WO2007007058A1 - Method and apparatus for regulating blood pressure - Google Patents

Abstract

Blood pressure is influenced in a human by stimulating a region of the brain in the human. In one embodiment, a dorsal region of the brain is stimulated. In another embodiment, a ventral region of the brain in stimulated. An apparatus is employed to stimulate a region of the brain in the human.

Description

METHOD AND APPARATUS FOR REGULATING BLOOD PRESSURE

RELATED APPLICATIONS

This application claims the benefit of U.S. Provisional Application No. 60/697,211, filed on July 7, 2005. The entire teachings of the above application are incorporated herein by reference.

BACKGROUND OF THE INVENTION

Blood pressure irregularities are a health concern for humans. Currently available methods to adjust irregularities in blood pressure (BP) in a human include diet, exercise, relaxation, lowering of salt intake and prescription drugs. However, such treatments can result in variable outcomes, may be effective for only relatively short periods of time, and maybe less effective in humans with severe blood pressure irregularities. Thus, there is a need to develop new, improved, and effective methods of treating blood pressure irregularities in humans and devices for use in methods of treating blood pressure irregularities in humans.

SUMMARY OF THE INVENTION

The present invention relates to a method and a device for influencing blood pressure in a human.

In one embodiment, the invention includes a method of influencing blood pressure in a human, comprising the step of applying a stimulation in a region of a brain in a human in a manner influencing blood pressure.

In another embodiment, the invention includes a method of influencing blood pressure in a human, comprising the step of applying a stimulation in a region of a brain in a human having a hypotension condition.

In yet another embodiment, the invention includes a method of influencing blood pressure in a human, comprising the step of applying a stimulation in a region in a brain in a human having a hypertension condition. In a further embodiment, the invention includes an apparatus for influencing blood pressure in a human subject, comprising a blood pressure sensor detecting blood pressure; a processor in communication with the blood pressure sensor and generating a control signal based on the blood pressure; a signal generator in communication with the processor generating a stimulation signal based on the control signal; and an electrode including at least two conductors in contact with a region of the brain that stimulates the region as a function of the stimulation signal in a manner influencing blood pressure in a human subject.

In still another embodiment, the invention is an apparatus for stimulating a region in a human brain, comprising a signal generator adapted to generate a signal; and at least one electrode disposed in a region of a brain in a human subject adapted to produce an output as a function of the signal to stimulate the region in a manner influencing blood pressure in the human subject.

Another embodiment of the invention is an apparatus for generating a signal for controlling stimulation of a region in a human brain in a human subject, comprising a blood pressure sensor adapted to provide a blood pressure measurement of blood pressure in a human subject; a microprocessor controller in communication with the blood pressure sensor adapted to convert the blood pressure measurement to a control signal that, when received by a signal generator in operative arrangement with electrodes deployed in a region of a brain of the human subject, causes the signal generator to generate a signal that stimulates the region in a manner that influences the blood pressure.

The invention described herein provides a method and a device for influencing blood pressure in a human. Advantages of the claimed invention include, for example, treatment of a human with a blood pressure irregularity with a relatively consistent outcome and for an extended period of time, especially in humans with severe blood pressure irregularities or where currently available treatments have failed. Thus, blood pressure can be influenced in a human by applying a stimulation in a region of the brain in the human in a manner to influence blood pressure to potentially reverse, adjust, or dimmish an irregularity in blood pressure in a human. BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing will be apparent from the following more particular description of example embodiments of the invention, as illustrated in the accompanying drawings in which like reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating the principles of the invention.

Figs. IA and IB are diagrams of a device, employing the principles of the present invention, using an electrode deployed in a brain of a human subject.

Fig. 2 A is a close-up diagram of an embodiment of an aspect of the device of Figs. IA and IB.

Fig. 2B is an alternative embodiment of the aspect of the device of Fig. 2 A.

Figs. 2C-2E are diagrams of various embodiments of the electrodes used by the aspect of the device of Figs. 2 A and 2B.

Fig. 3 is a block diagram of electronics used in the device of Figs. IA and IB.

Fig. 4 is a flow diagram of an example process performed by the electronics of Fig. 3.

Fig. 5 is a post-operative, axial, Tl -weighted, Magnetic Resonance (MR) scan showing the electrode of Fig. 2 A positioned in the brain. The electrode is in contact with the VPL of the left thalamus (lateral) as distinct from the wire of the more medial and deeper PVG electrode. Insert depicts the ACPC plane.

Fig. 6 A is an annotated scan indicating saggittal positions of the electrode of Fig. 2A in human subjects in whom there were changes in blood pressure during testing. Fig. 6B is an annotated scan indicating coronal positions of the electrode corresponding to the annotations of Fig. 6A. Patients #1-7 all had reduction in BP (blue contacts) and are the most ventral electrodes. Conversely, #8-11 and the upper 2 electrodes of #1 and #6 had a rise in BP. Gray contacts are those that, when stimulated, had no effect on BP. AC=anterior commissure, PC=posterior commissure, PVG=periventricular gray, PAG=periaqueductal gray, SC=superior colliculus (the level of which is depicted by the dotted circle in IB), RN=red nucleus, EGNhird ventricle, Aq=aqueduct. Inset depicts the slice position. Fig. 7 A is a plot of an intra-arterial blood pressure recording of a human subject responsive to stimulation by the electrode of Fig. 2A. Both systolic and diastolic blood pressure sustained a rise when the stimulation ("stim") was 10Hz, 120 μs, 4.Ov. Fig. 7B is a plot of finger arterial pressure measurements (Finapress^®) at

10Hz, 120 μs, 3.5v.

Fig. 8 A shows plots of multiple metrics related to blood pressure averaged across multiple human subjects resulting from stimulation of the electrode of Fig. 2A in a ventral area of a region in the brain. Mean changes in systolic blood pressure (SBP), diastolic blood pressure (DBP), pulse pressure (PP), R-R interval and maximum dP/dt for the seven patients who had a reduction in BP during stimulation of ventral PVG. Stimulation was started at 100s (the position of the lefthand dashed vertical line) and stopped at 400s (righthand line). The gray area denotes ± one standard error of the mean. Fig. 8B is an annotated scan of a brain indicating regions of deployment of the electrode of Fig. 2A corresponding to the plots of Fig. 8 A. The image shows the location of the active contacts (squares) and their proximity to the centre of the Red Nucleus (dot). Radial distances are 5, 8 and 1 lmm. SC=suρerior colliculus, AC=anterior commissure, PC=posterior commissure. Pink line=ACPC plane. Fig. 9 A shows plots of multiple metrics related to blood pressure averaged across multiple human subjects resulting from stimulation of the electrode of Fig. 2A in a dorsal area of a region in the brain. Mean changes in systolic blood pressure (SBP), diastolic blood pressure (DBP), pulse pressure (PP), R-R interval and maximum dP/dt for the six patients with a rise in BP during stimulation of dorsal PVG (note that two of these are the same patients who also had a reduction in blood pressure). Stimulation was started at 100s and stopped at 400s. The gray area is one standard error of the mean.

Fig. 9B is an annotated scan of a brain indicating regions of deployment of the electrode of Fig. 2A corresponding to the plots of Fig. 9A. The image shows the location of the active contacts (squares) and their proximity to the centre of the Red Nucleus (dot). Radial distances are 5, 8 and 1 lmm. SC=superior colliculus, AC=anterior commissure, PC=posterior commissure. Pink line=ACPC plane. Fig. 1OA is a plot of power spectra of Systolic Blood Pressure (SBP) for the human subject in response to stimulation in ventral and dorsal areas of the brain regions of Figs. 8B and 9B, respectively. Change in the power spectra of systolic blood pressure (SBP) for patient #6 during rest, increase and decrease in SBP induced by 10Hz stimulation between proximal two and deepest two contacts at 3.0v. The absence of a visible high frequency component in this patient is probably due to an irregular respiratory pattern.

Fig. 1OB is a logarithmic histogram indicating changes in SBP for the human subject caused by stimulation of the electrode in ventral and dorsal areas of the brain regions of Figs. 8B and 9B. Composite changes in the logarized low frequency (LF) and high frequency (HF) power of systolic blood pressure (SBP) during no stimulation (resting) versus episodes of reduced SBP (SBP drop) and increased SBP (SBP rise) for all patients. Error bars show one standard error of the mean. Note that changes in LF were significant for both increases and decreases in BP, but changes in HF were not. (p<0.05, paired t-test vs. resting).

Figs. 1 IA-I II are plots that illustrate changes in systolic blood pressure (SBP), heart rate (HR), and dP/dt on standing. A-C, mean changes in systolic blood pressure for Patient 1 (#1), MOI group (MOI), and non-MOI group, respectively. D- F: changes in heart rate for the same groups. G-I, changes in dP/dt for the same groups. All traces include the mean of three sessions, averaged every 10 seconds. Gray area, period when patient was sitting; yellow area, period of standing; black line, stimulation off; red line, stimulation on. Error bars show + one standard error of the mean.

Figs. 12A-12D include timing diagrams, a scan, and a diagram of an electrode implanted into the PVG and PAG as described in reference to Example 3. Fig. 13 examples of raw ontraoperative data for each patient. Patient #1 and #2 had increases and decreases in arterial blood pressure (ABP), depending on the electrode position. 0-100 and 200-300 seconds = resting and recovery periods. 100- 200 seconds = PAG stimulation on. Fig. 14 patient #1 - summary of changes. A. Decrease in SBP with stimulation (location 1) was accompanied by a decrease in pulse pressure (PP), diastolic BP (DBP) and dPdt. Although decrease in RR interval was significant, it was a small change. 100-200 seconds = stimulation period. B Power spectrum of heart rate showing a large increase in Meyer's wave (~ 0.1Hz) with increase in BP, and a decrease with fall in BP. C Post-operative MRI scan showing the electrode position (white arrow) equivalent to position #1. Fig. 15 patient #2 - summary of changes. A.In this patient, increase in SBP

(location 2) was accompanied by a significant fall in RR interval, representing an increase in heart rate, with little change in pulse pressure (PP). Yellow area = stimulation period. B. Power spectrum of heart rate showing an increase in Meyer's wave with stimulation, indicating an increase in sympathetic activity. C. Post- operative MRI scan showing electrode position (white arrow), equivalent to position 2.

Fig. 16 normalised changes in cardiovascular variables with stimulation. The Y axis is a ratio of the variable during 100s of stimulation, versus an equivalent resting period with stimulation off. The horizontal lines indicate a ratio of 1 i.e. no change. Error bars indicate ± one SD from the mean. Note that the upper limits of the following variables are not shown for scaling purposes; #3 (HF = 2.82), #6 (LF = 7.6. HF = 10.2).

Fig. 17 post-operative electrode location. A. Sagittal view of the midbrain showing the superimposed electrode positions in those patients in whom BP significantly changed with stimulation. The electrode contacts that provided optimum pain relief are coloured and are both those that were studied and those used for chronic stimulation. Contacts that reduced BP are shown in blue, those that increased BP are red. Inset shows the level of the AC-PC line. The orange line indicates the approximate level of the aqueduct and therefore the distinction between ventral and dorsal PAG. B Location of electrodes that had no effect on BP (the contacts used are coloured green). RN= red nucleus, SC= superior colliculus, AC = anterior commissure, PC= posterior commissure (with pink line joining the two), PAG = periaqueductal grey, PVG = periventricular grey. Patient numbers refer to those in table 1. Background reprinted from Mai et al ©1998, with permission from Elsevier.

Fig. 18 comparison of visual analogue scores with cardiovascular variables. A) Average changes in VAS (%) significantly correlated to systolic blood pressure (A₅ r2=0.62. p=0.01, n=16) and dP/dt changes (C, .4=0.62. p=0.01, n=16), but weakly associated with pulse pressure changes with stimulation (B, ^= 0.48, p=0.06, n=16). Black lines= linear regression, Outer grey lines = upper and lower 95% confidence intervals. VAS= visual analogue score,^' SBP= systolic blood pressure Fig. 19 power spectral analysis of systolic blood pressure A) in those patients with a significant reduction in systolic blood pressure with stimulation, there was a significant decrease (p<0.001) in the low frequency component (0.05 to 0.15Hz) of systolic blood pressure variability. B) Similar results for patients with increases in systolic blood pressure (pO.OOl). Error bars indicate one standard deviation of the mean. n=7 for drop and n=4 for rise group. *denotes statistical significance.

Fig. 20 long-term changes in McGiIl Pain Questionnaire (MPQ) scores compared to BP changes. A) Changes in MPQ scores at one year, compared to pre- operatively, depending on whether BP increased (BP?), decreased (BP ?) or remained unchanged (BP ? ) in the laboratory. Comparisons were made using the Wilcoxon signed ranks test. * indicates significance (pO.OOl). Error bars denote one standard error of the mean. B), C) similar results for the sensory and miscellaneous component of the MPQ. D) Similar results for Question 7 which gives a patient the choice of one out of four words that describe 'burning' sensation, in increasing severity (0-4). These results show that those patients who had a decrease in blood pressure obtained the best analgesia over the long-term.

Stimulation appears to work particularly well on the burning component of pain, which may have a vascular component.

Fig. 21 long-term changes in MPQ scores compared to BP changes. A) Changes in MPQ scores at one year, compared to pre-operatively, significantly correlated to changes in blood pressure (j?= 0.69, p= 0.002, n=l 6). Black lines= linear regression, Outer grey lines = upper and lower 95% confidence intervals. B), C) similar results for the 'sensory' and 'miscellaneous' components of the MPQ. (T²= 0.67, ρ=0.004, n=16 for sensory; I²== 0.68, ρ=0.003, n=16 for miscellaneous) These results show that decrease in blood pressure correlates to analgesia over the long-term.

Fig. 22 comparison of absolute BP and pain variables. None of the comparisons showed any significant correlation; A) Systolic Blood Pressure (SBP) versus VAS (r² = 0.22, ρ>0.25, n=27). B) SBP versus MPQ score (^=0.26, p>0.19, N=27). C) Pulse pressure versus VAS (?= 0.18, ρ> 0.35, n=27), D) Pulse pressure versus MPQ score Cr²R= 0.21, Pp> 0.29, n=27), E) dP/dt versus VAS (r² = 0.26, p>0.18, n=27), F) dP/dt versus MPQ score Cr²= 0.007, >0.97, n=27). Fig. 23 comparison of short-term VAS with long-term MPQ scores. There was a significant correlation (V²= 0.6) between initial improvement in VAS (% reduction in VAS) and percentage reduction in long-term MPQ score. Note that two patients had an increase in MPQ score, one of which is not shown as the increase was 300% (with reduction of VAS of 0%). \ = patient with reduction in BP on stimulation, * = no change in BP,^s = increase in BP.

DETAILED DESCRIPTION OF THE INVENTION

The features and other details of the invention, either as steps of the invention or as combinations of parts of the invention, will now be more particularly described and pointed out in the claims. It will be understood that the particular embodiments of the invention are shown by way of illustration and not as limitations of the invention. The principle features of this invention can be employed in various embodiments without departing from the scope of the invention.

The present invention relates to a method of influencing blood pressure in a human comprising the step of applying a stimulation in a region of a brain in the human in a manner influencing blood pressure and a device to influence the blood pressure in a human. In one embodiment, ventral stimulation of a region of the brain (e.g., PVG₅) can have a suppressor effect (e.g., decrease blood pressure). In another embodiment, dorsal stimulation of a region of the brain (e.g., PVG, PAG) can have a pressor effect (e.g., increase blood pressure). The stimulation is in a region of the brain (internal to the cranium), not to the skin or cranium.

"Influencing blood pressure," as used herein, refers to a change (e.g., increase, decrease) in blood pressure in a human following stimulation in a region of the brain compared to the blood pressure in the human before stimulation in a region of the brain. For example, by applying a stimulation in a region of the brain of the human, blood pressure in the human can be influenced to increase or decrease compared to the blood pressure in the human before application of the stimulation. A human is also referred to herein as a subject or a patient.

In one embodiment, the region of the brain that is stimulated includes the periventricular gray (PVG) region of the brain. In another embodiment, the region of the brain that is stimulated includes the periaqueductal gray (PAG) region of the brain.

The human treated by the methods described herein can have hypotension (e.g., a hypotension condition such as orthostatic hypotension) or hypertension. The human having hypertension or hypotension can further have a pain condition, such as neuropathic pain. Li a further embodiment, the human treated by methods described herein can have a pain condition (e.g., neuropathic pain). Stimulation in a region of the brain includes applying stimulation to a dorsal region of the brain. Ih a particular embodiment, the dorsal region of the PVG region of the brain is stimulated. Stimulation of the dorsal region of the brain (e.g., PVG, PAG) can be associated with a pressor response in the blood pressure of the human. A pressor effect can increase blood pressure in a human. hi another embodiment, stimulation in a region of the brain includes stimulation of a ventral region of the brain (e.g., PVG, PAG). Li a particular embodiment, the ventral region of the PVG region of the brain is stimulated. Stimulation of the ventral region of the brain can be associated with a suppressor response in the blood pressure in the human. A suppressor response (also referenced to herein as a "depressor" response) can decrease blood pressure hi the human.

The method of influencing blood pressure in a human, comprising stimulation in a region of the brain in a human in a manner mfluencing blood pressure, can further include selecting a human having a blood pressure irregularity. A "blood pressure irregularity," as used herein, refers to a blood pressure measurement or assessment hi the human that is not within a normal range for that human. Normal ranges for the human can be determined by one of skill in the art and can depend, for example, on the gender, age, environmental factors, general health and other medications and/or treatments the human may be undergoing.

When stimulation is applied in a region of the brain in a human having a blood pressure irregularity, the stimulation includes correcting the blood pressure irregularity. "Correcting the blood pressure irregularity," as used herein, means that the blood pressure of the human is changed (e.g., increased, decreased) from a blood pressure measurement prior to stimulation to a blood pressure measurement after stimulation that deviates from the pre-stimulation value and approaches or is within a blood pressure measurement observed in a human without a blood pressure irregularity.

Correction in the blood pressure irregularity in the human can include normalizing the blood pressure in the human. "Normalizing the blood pressure" in the human, as used herein, refers to an alteration in blood pressure in the human to approach levels observed or expected in a human of, for example, similar age, weight, gender, health history and medication regimen that does not have a blood pressure irregularity. For example, a human having an elevated blood pressure (e.g., hypertension) can have a decrease in blood pressure (e.g., to within normal levels, levels observed in a human without a blood pressure irregularity) following stimulation in a region of the brain in a manner influencing blood pressure (e.g., stimulation of the ventral region of the PVG, PAG or both PVG and PAG), thereby correcting the blood pressure irregularity. Likewise, for example, a human having a low blood pressure (e.g., hypotension, including orthostatic hypotension) can have an increase in blood pressure (e.g., to within normal levels) following stimulation in a region of the brain in the human in a manner influencing blood pressure (e.g., stimulation of the dorsal region of the PVG, PAG or both PVG and PAG), thereby correcting the blood pressure irregularity.

The method of influencing blood pressure in a human described herein can further include feeding back a metric representative of blood pressure in an automated manner and responsively adjusting the stimulation based on the metric. Alternatively, or additionally, the method can further include enabling the feedback of a metric representative of a blood pressure in a manual manner and adjusting the stimulation in response to the metric

The stimulation employed in a method of stimulation in a region of the brain in the human in a manner to influence blood pressure can include at least one member selected from the group consisting of an electrical stimulation, a magnetic stimulation, an electromagnetic stimulation, a thermal stimulation, and a mechanical stimulation. The method of influencing blood pressure in the human can further include inductively communicating stimulation parameters used for applying the stimulation. The method can also further include powering a system used for applying the stimulation based on the inductive communications. The method of influencing blood pressure in a human comprising applying a stimulation in a region of the brain to influence blood pressure can further include communicating stimulation parameters used for applying the stimulation by at least one member selected from the group consisting of a radio frequency, an electrical signal, and an optical signal. The method of influencing blood pressure in a human can further include disabling the stimulation in a fail-safe manner. The disabling can be performed in an automated manner based on a metric associated with blood pressure or a detected problem applying a stimulation. The disabling of the stimulation can be activated by the human whose blood pressure is being influenced or by another human (e.g., a physician, a caretaker).

The stimulation in the region of the human brain can include selectively applying the stimulation to at least one region of the brain. "Selectively applying the stimulation," as used herein, refers to the stimulation of a particular region of the brain. The particular region of the brain can be one region (e.g., PAG alone, PVG alone) or more than one region (e.g., PAG and PVG) of the brain. The selective application of the stimulation can be employed for a particular effect in influencing blood pressure. For example, stimulation of the ventral region of the brain (e.g., PVG, PAG) can decrease (also refered to herein as suppress, depress) blood pressure in a human having elevated (e.g., hypertension) blood pressure. Likewise, stimulation of the dorsal region of the brain (e.g., PVG, PAG) can increase blood pressure (also referred to herein as pressor effect in blood pressor) in the human.

The method of influencing blood pressure in a human can include applying stimulation to a region of the brain that includes selectively energizing at least one multiple conductor of at least one electrode disbursed within the region. The method can also include applying a stimulation to the region of the brain including selectively energizing multiple conductors of a single electrode having at least two distal conductors positioned in a ventral region of a nuclei and at least two proximal conductors positioned in a ventral region of a nuclei. The method of influencing blood pressure in the human can include applying a stimulation that includes generating a voltage differential between at least two electrodes of between about - 10V and about +10V with a frequency between about 0.1 Hz and about 1 kHz. As described herein, in awake humans electrical stimulation of the PVG and/or PAG can influence arterial blood pressure (ABP). Ventral stimulation has a supressor effect and dorsal stimulation has a pressor effect. Furthermore, the electrically induced BP modulation appears to be frequency-dependent, working most often at 10Hz and sometimes at 50Hz. Electrical stimulation of different regions of the human P VG/P AG can selectively modulate blood pressure.

Hypertension or postural hypotension may be controlled by manipulation of the stimulation of the PVG, PAG or both PVG/PAG.

The following is a detailed description of an apparatus that maybe employed to influence blood pressure in a human by applying a stimulation in a region of the brain (e.g. PVG, PAG) in the human in a manner influencing blood pressure. The apparatus may be used in the methods described herein to influence blood pressure, for example, in a human.

Figs. IA and IB are macro level views of a human subject 50 in which a device 100 with an electrode 110 is deployed in operative arrangement with a human brain 101 according to the principles of the present invention. The device 100 is referred to herein as a blood pressure regulator 100 when used in connection with treating or otherwise influencing blood pressure 104. Other components associated with the device 100 are similarly referenced, but may be referred to as simply "device" when used for applications other than influencing blood pressure. In the embodiment of Fig. IA, the human subject 50 may have a blood pressure irregularity, such as hypertension or hypotension. The blood pressure regulator 100 may include brain stimulation components, such as a signal generator 105, electrode 110, and blood pressure sensor 115, distributed on or within the human subject's cranium 102 or body 103 in operative communication with each other.

In one embodiment, a wrist band 112 that includes the blood pressure sensor 115 with a transmitter 120 may be positioned on an arm 106 of the human subject 50. In some embodiments, the transmitter 120 generates inductive communications signals 125 that use conductivity of the human subject's body 103 to conduct the inductive communications signal 125 to a receiver 122 that receives the inductive communications signal 125 for the signal generator 105. The blood pressure regulator 100 may also include a microprocessor controller (not shown), which may allow the blood pressure regulator 100 to operate in an automated manner. In some embodiments, the microprocessor controller may be positioned at the signal generator 105 or the blood pressure sensor 115. Further description of such embodiments is provided below in reference to Fig. 3. Continuing to refer to the embodiment of Fig. IA, the signal generator 105 may be positioned at or in the human subject's cranium 102. Based on metrics or other information related to the blood pressure 104 that may be provided by the blood pressure sensor 115 or microprocessor controller by way of a wireless communications transmitter 120 and receiver 122, the signal generator 105 may cause the electrode 110 to stimulate a selected region or nuclei in the brain or brain stem to influence the blood pressure 104 in a manner (i) treating a blood pressure irregularity or (ii) influencing the blood pressure to change in a predictable manner. The blood pressure regulator 100 may, for example, be used to influence a change in blood pressure 104 for purposes other than for treating blood pressure, such as for testing a drug at an elevated or reduced blood pressure level.

The signal generator 105 may generate and output a wide range of electrical signals to the electrode 110. For example, the electrical signals may be a combination of some of the following parameters: -10 to +10 volts, 0.1Hz to IKHz, and have a pulsewidth of 1 nsec to 60 minutes. Specific examples of combinations used during testing are presented below in the Exemplifications section. It should be understood that the electrode 110 may stimulate the region or nuclei with at least one of the following outputs in response to the electrical signal: voltage, current, induced magnetic or electromagnetic field, thermal temperature, and so forth.

Other embodiments of the signal generator 105 may transmit signals other than electrical signals to the electrodes 110, such as acoustical, Radio Frequency (RF), or optical signals, in which case, the electrodes 110 may serve as a transducer to convert the signals to a form that can stimulate a region or nuclei in the brain. Fig. IB is a diagram of another embodiment of the blood pressure regulator 100 in which the blood pressure sensor 115 is permanently implanted within the human subject's arm 106. In some embodiments, a wire or optical fiber 135 is "threaded" through the human subject's body 103 to form a communications path from the sensor 115 to the signal generator 105. In other embodiments, the blood pressure sensor 115 may use the wireless transmitter 120 to communicate with the wireless receiver 122 via inductive or RF communications, as described in reference to Fig. IA.

Fig. 2A is a side view of the human subject 50 having the device (e.g., blood pressure regulator) 100 influencing at least one physiological condition of the human body 103 or mind through stimulation of a region or nuclei of the brain 101 according to the principles of the present invention. In a blood pressure application, the blood pressure regulator 100 and electrode 110 are, in one embodiment, disposed in the brain 101 or brain stem 202 in a manner adapted to apply stimulation to a region or nuclei in the brain 101 or brain stem 202 associated with influencing the blood pressure 104.

The electrode 110 may include two pairs of conductors 215a, 215b (collectively, conductors 215) in this embodiment. A first pair of conductors 215a may be positioned in a ventral area of the PVG 205, and a second pair of conductors 215b may be positioned in a dorsal area of the PVG 205. The conductor pairs 215a, 215b may be energized by the signal generator 110 via a wire 107 with an electrical signal 220 that produces a voltage differential across the conductors of one or both pairs of conductors 215a or 215b. In turn, the voltage differential stimulates the PVG 205 in a manner influencing the blood pressure 104, as described in more detail below in reference to Fig. 2B.

Fig. 2B is a side view of another embodiment of the device 100 deployed in the human subject 50. In this embodiment, the electrode 110 maybe positioned in the PVG 205, PAG 210, or both. The signal generator 105 is illustrated as being disposed in the brain 101; however, it should be understood that the signal generator 105 may be positioned anywhere within the cranium 102 or external from the cranium 102. For example, the signal generator 105 may be positioned beneath skin (not shown) covering the cranium 102. Wire(s) 107 extending into the cranium 102 may connect the signal generator 105 to the electrode(s) 110 to conduct the electrical signal 220 to the electrode(s) 110.

The blood pressure regulator 110 may include one, two, or more electrodes 110, which may be determined on a case-by-case basis. Example embodiments include one electrode 110 that is positioned to have its conductors 215 positioned entirely within the PVG 205. Alternatively, a single electrode 110 may be positioned such that the conductors 215 are positioned entirely within the PAG 210. In yet another embodiment, a single electrode 110 may be positioned so that the electrode 110 extends into both the PVG 205 and PAG 210 with its conductors 215a, 215b mechanically disposed in a respective region or nuclei 205, 210 to selectively stimulate one or both regions or nuclei at a time.

In the embodiment illustrated in Fig. 2B, the electrode(s) 110 may be fixedly positioned ventrally or dorsally within one or both of the nuclei 205, 210. For example, if a diagnosis has been made that only a pressor response of the blood pressure 104 is required, the electrode(s) 110 may be fixedly positioned dorsally in the PVG 205 or PAG 210. Alternatively, if the human subject 50 is known to only require a suppressor response of the blood pressure 104, the electrode(s) 110 maybe fixedly positioned ventrally in one or both of the nuclei 205, 210.

The conductors 215 on the electrode(s) 110 may be activated in a variety of ways through control by the signal generator 105. For example, a first pair of conductors 215a may be positioned in the PVG 205, and a second pair of electrodes 215b maybe positioned in the PAG 210. Depending upon the human subject's 50 response to stimulation of the PVG 205 or PAG 210, a microprocessor controller (discussed below in reference to Fig. 3) may cause the signal generator 105 to activate one or both pairs of conductors 215a, 215b in a manner adjusting the stimulation of the PVG 205 or PAG 210 to optimize blood pressure response to the stimulation. At least one mechanical or electrical switch (not shown) may be used to direct the electrical stimulation to the selected electrode(s) 110.

In another embodiment, the electrode(s) 110 may be telescoping electrodes such that the conductors 215 on the electrodes 110 may be positioned in the PVG 205 or PAG 210 in a selectable manner by selectively lengthening or shortening the electrode(s) 110, preferably in a remotely controlled manner. For telescoping operations, a stint (not shown) or other permanent "tunnel" that allows the electrode(s) 110 to telescope and the conductors 215 to provide stimulation to tissues of the PVG 205 or PAG 210 maybe provided to ensure long-term operability of the telescoping feature. Fig. 2C is another embodiment of a deployment of the electrode(s) 110 in a region of the brain 101. In this embodiment, one electrode 110 is positioned in the PVG 205 in a ventral-to-dorsal orientation, and a second electrode 110 is positioned in the PAG 210 in a similar orientation. In each of these electrodes 110, a first pair of conductors 215a is positioned ventrally in their respective nuclei, and a second pair of conductors 215b are positioned dorsally in their respective nuclei. It should be understood that fewer or more conductors in the "pairs" of conductors 215a, 215b may be provided along the length of the electrode 110.

Fig. 2D is an embodiment having a single signal conductor 215c positioned centrally between two reference conductors 215d, 215e positioned at opposite ends of the electrode 110 for stimulating a ventral area or dorsal area, respectively, of the PVG 205. To cause stimulation, the signal generator 105 may activate one of the reference conductors 215d or 215e by switching electrical or mechanical switches (not shown) to connect the selected reference conductor 215d or 215e to a "ground" reference potential. When connected to a ground reference potential, the signal conductor 215c and selected, grounded, reference conductor 215d or 215e establishes a voltage differential between them, which causes current to flow between the signal conductor 215c and selected, grounded, reference conductor 215d or 215e. The voltage differential/current flow causes stimulation in the ventral or dorsal region of the nuclei. The ungrounded reference conductor (i.e., electrically "floating" conductor) is not involved in the stimulation. It should be understood that any other number of conductor configurations are possible to optimize efficiency, minimize blood pressure response times, minimize side effects, or otherwise favorably enhance experience for the human subject 50. hi another embodiment, the signal conductor 215c may be grounded, and the outer two conductors 2l5d, 215e may be selectively energized. Again, it is a voltage differential (or other potential differential) that causes stimulation. Further, in such an embodiment, one conductor 215d maybe energized at a first voltage level, and the other conductor 215e may be energized at a different voltage level to stimulate ventral and dorsal areas of the PVG 205 or PAG 210 at different levels. It should be understood that the "voltage" level(s) maybe steady state, oscillatory, pulsewidth modulated, duty cycle controlled, or employ other non-steady state modulation. Fig. 2E is a rear view of conductor 110 deployment in a region or nuclei

(e.g., PVG 205) of the brain 101. This embodiment includes multiple electrodes 110 that may be positioned vertically, horizontally, or otherwise apart from one another. In this embodiment, the conductors 215 maybe activated to cause stimulation in tissue between the electrodes 110 in a ventral, dorsal, or other area of the PVG 205 or other region. Further, a multi-dimensional array of electrodes 110 may be formed within one or both of the nuclei 205, 210, subject to physical limitations. It should be understood that deployment of the electrodes 110 can be done in other nuclei or regions of interest.

Fig. 3 is a block diagram of the device 100 according to the principles of the present invention. The device 100 or blood pressure regulator 100 may include a microprocessor controller 305, fail-safe unit 315, signal generator 105, and at least one stimulus electrode 110. The blood pressure regulator 100 may also include a blood pressure sensor 115 and, optionally, a human-controlled feedback interface 310. The components 305, 315, 105, 115, 310 maybe analog, digital, or hybrid circuits, and clear distinctions among these components may or may not be clear depending on implementation.

The device 100 may be deployed in the human subject 50 in various ways. For example, dashed line zones 305a, 305b, and 305c illustrate different example deployment configurations on or within the human subject 50. In one embodiment, the microprocessor controller 305, fail-safe unit 315, signal generator 105, and stimulus electrode(s) 110 (i.e., zones 305a and 305b) are disposed inside the human subject 50. In this same embodiment, the blood pressure sensor 115 and human- controlled feedback interface 310 are disposed external from the human subject 50. In another embodiment, the signal generator 105 and stimμlus electrode 110 (i.e., zone 305a) are the only components of the device 100 that are disposed inside the human subject 50. In such a case, the signal generator 105 receives command signals from the microprocessor controller 305 and optionally fail-safe unit 315 via wireless, wired, or optical signal path(s). In another embodiment, all components 105, 110, 115, 305, and 315 (i.e., zones 305a, 305b, and 305c) are deployed in the human subject 50.

In operation, the blood pressure sensor 100 is adapted to automatically detect the blood pressure 104 in the human subject 50. The blood pressure sensor 115 includes a transmitter (Tx) 120 that transmits inductive communications signals 320 to a receiver (Rx) 122 used by the microprocessor controller 305 to receive the inductive communications signals 320. In other embodiments, the inductive communications signals 320 may instead be Radio Frequency (RF) signals. In other embodiments, the blood pressure sensor 115 communicates with the microprocessor controller 305 via an electrical or fiber optic transport medium 325 via electrical or optical signals 330.

The microprocessor controller 305 uses information in the communications 320 or 330 to determine or calculate control signals for applying a stimulation to a region or nuclei (e.g., PVG 205 or PAG 210) of the brain 101. The microprocessor controller 305 may communicate the determined or calculated control signals to the signal generator 105 via a transmitter 120 and receiver 122, respectively. The control signals may be physically communicated between the microprocessor 305 and signal generator 105 via inductive communications path signals 320 through conduction via the body 103 or electrical or fiber optic path 325.

Responsive to receiving the control signals from the microprocessor controller 305, the signal generator 105 produces an electrical signal 220 and transmits the electrical signal 220 via a wire or other electrically conductive medium 220 to at least one stimulus electrode 110 positioned in the brain 101 in a region or nuclei that influences the blood pressure 104 in the human subject 50 in a manner treating a blood pressure irregularity, for example, or for some other reason.

The human-controlled feedback interface 310 is optionally provided as a means for the human subject 50, doctor, technician, or lay person to "dial in" a blood pressure reading to the microprocessor controller 305, fail-safe unit 315, or other component having a use for blood pressure information or other information the human controlled interface 310 is adapted to provide. The interface 310 may optionally allow simple 'increase' or 'decrease' commands to be entered. In one embodiment, the only components in the device 100 are the interface 310, signal generator 305, and electrode(s) 110. In such an embodiment, the interface 310 and signal generator 105 may be an integrated unit accessible externally from the cranium 102 or body 103. The fail-safe unit 315 may receive signals from any number of components of the device 100. The fail-safe unit 315 may include analog or digital circuitry and optionally circuitry for converting between analog and digital formats. The fail-safe unit may have preset or adaptive protocols used to determine when or how to disable or shut-down the device 100 in a manner safest for the human subject 50. Fig. 4 is a flow diagram of a process 400 employed by the blood pressure regulator 100 according to the principles of the present invention. Some steps in the process 400 may be executed in the microprocessor controller 305, and other steps maybe performed by other components or combinations of components, including the microprocessor controller 305, of the blood pressure regulator 100, as suggested above and described below.

The process 400 starts (step 405) and initializes (step 410) to begin operation. Initialization can include any number of initialization sequences, such as power-up sequences, verifying processor operational readiness, verifying transmitters and receivers are using the same communications protocol, and so forth. The process 400 continues by checking whether a 'disable' of the blood pressure regulator 100 has been requested (e.g., manually) or a blood pressure regulator 100 failure has been detected (step 415). An example of a failure detection maybe detection of a low power condition, loss of communications, software error, or other error that may interfere with operations of the blood pressure regulator 100. If disable has not been requested and failure has not been detected (step 415), the process 400 measures and feeds back blood pressure (step 420). hi one embodiment, described in reference to Fig. 3, blood pressure measurement and feedback is performed in an automated manner. In another embodiment, the blood pressure measurement and feedback is performed in a manual manner through use of the human-controlled feedback interface 310.

The process 400 continues and determines whether the blood pressure is within a safe operating range (step 425), meaning that a determination is made as to whether it is safe to continue operating the blood pressure regulator 100. For example, if the blood pressure is observed to be outside a given positive or negative threshold from a nominal or normal operating pressure, the blood pressure regulator 100 may determine that it is itself a cause of a blood pressure irregularity due to, for example, a failure or "runaway" condition.

If the process 400 determines it is safe to continue operating, the process 400 may determine whether the blood pressure is at a desired pressure (step 430). If the blood pressure is nominal or normal (step 435), the process 400 returns to a step of checking whether a 'disable' has been requested or a blood pressure regulator failure has been detected (step 415). If the process 400 determines that the blood pressure is low, the process 400 stimulates a dorsal region of a nuclei (e.g., PVG 205 or PAG 210) in the brain 101 or brain stem 202 (step 440) to influence a pressor response of the blood pressure in the human subject's body 103. The process 400 thereafter continues operations (step 415). If the blood pressure 104 is determined to be high, the process 400 stimulates a ventral region of the nuclei (step 445) to influence a suppressor response of the blood pressure 104. Thereafter, the process 400 continues operations (step 415).

If a 'disable' has been requested or a failure has been detected in the blood pressure regulator (step 415), the process 400 disables the blood pressure regulator (step 450). Similarly, if the blood pressure is outside a safe operating range (step 425) as described above, the process 400 disables the blood pressure regulator (step 450). Thereafter, the process 400 determines whether to suspend operations (step 455), optionally based on a number of criteria or as a result of the human subject's triggering of a fail-safe signal (i.e., 'disable'). If operation is not to be suspended, the process 400 initializes the blood pressure regulator 100 (step 400) as a matter of precaution in one embodiment. If operation is to be suspended, the process 400 ends (step 460), and the blood pressure regulator 100 is set into a safe operating mode by, for example, disabling the electrodes 110, powering down, or entering a 'safe mode.' It should be understood that the process 400 is an example embodiment used for illustration purposes only. Other embodiments within the context of regulating blood pressure may be employed. Some or all of the steps in the process 400 maybe implemented in hardware, firmware, or software. If implemented in software, the software may be (i) stored locally with the microprocessor controller 305 or (ii) stored remotely and downloaded to the microprocessor controller 305 during initialization (step 410). To begin operations in a software implementation, the microprocessor controller 305 loads and executes the software in any manner known in the art.

It should be understood that any form of communications protocol(s) maybe employed to provide communications between or among the several components of the blood pressure regulator 100. For example, wireless communications signals 320 may include inductive communications signals, Radio Frequency (RF) communications signals, Bluetooth® communications signals, or other forms of wireless communications signals. For any of such wireless communications signals, various protocols can be employed, such as coding, encryption, or other protocols known to improve communications and make the device 100 resistant to communications errors. As known in the art, communications errors may be caused by internal noise sources (e.g., low battery power, noisy amplifiers, poor analog or digital signal(s) isolation, etc.) or external noise sources, such as large electromagnetic fields (e.g., airport metal detectors, car electronics, etc.).

The stimulation that may be applied to the region or nuclei may be electrical stimulation, magnetic stimulation, electromagnetic stimulation, mechanical stimulation, thermal stimulation, or combination thereof. Although testing is described herein in reference to electrical stimulation, further studies maybe conducted to ascertain effectivity and operating parameters of these example other forms of stimulation. The blood pressure regulator 100 is described hereinabove as using a manual feedback system or an automatic feedback system. In the case of the manual feedback system, it should be understood that the blood pressure regulator 100 may operate in a "set and forget" mode or provide temporary stimulation while the human subject 50 is operating the human controlled feedback interface 310. In the case of being an automatic feedback system, the microprocessor controller 305 may calibrate a target 'set point' or nominal blood pressure level for the signal generator electrodes 110 on a one-time, periodic, aperiodic, as-needed basis, or as-requested basis.

In some embodiments, power is provided to portions of the blood pressure regulator 100 that are disposed inside the human subject 50 through inductive communications. In other words, inductive signals are received by the signal generator 105, for example, via the receiver 122 and converted into power that operates the microprocessor 305, memory (not shown), signal generator 105, failsafe unit 315 or other electronics. A power receiver circuit (not shown) may include a coil, capacitor(s), power regulator, power collection controller, filters, transformers, DC-to-DC converter controller, or other circuits commonly used to convert inductively coupled signals into usable energy for operating circuits and generating signals to cause the electrode(s) 110 to stimulate the regions or nuclei of the brain 102, as described herein.

In other embodiments, portions of the blood pressure regulator 100 disposed inside the human subject 50 are powered by way of battery or other power generator source, including, for example, fuel cells or the like. In such embodiments, access to the power source must be made available for changing or replenishing the power source on a regular or as-needed basis.

Although referred to as a "blood pressure regulator," it should be understood that the device 100 is operable to influence the blood pressure 104 as a blood pressure controller. The difference is that a blood pressure regulator generally maintains a preselected blood pressure and a blood pressure controller can control the blood pressure to be at a selectable level or follow a command trajectory in a dynamic manner. Although the principles of the present invention are described herein as being used to influence, regulate or control blood pressure, the principles of the present invention can be applied to other physiological applications, such as pain, anxiety, body temperature 'set point, 'and so forth.

In addition to some embodiments described herein, there are other embodiments that may be implemented. For example, the electrode(s) can be positioned within other regions or nuclei other than the PVG 205 or PAG 210. The electrodes 110 maybe partially disposed inside the cranium 102 (with conductors 215 positioned as described above) and extending into or through the cranium 102 in a permanent or temporary configuration. Other voltages/frequencies/pulsewidths can be used to stimulate other regions or nuclei. Other electrode embodiments may be used. Other communications techniques between multiple components may be employed. Positioning the electrode(s) or dorsal areas in the selected region(s) or nuclei may be changed to be in upper or lower areas in the selected region(s) or nuclei when applied for treating or influencing physiological functions other than blood pressure or blood pressure but with lesser efficacy.

The device 100 maybe used in the human subject 50 or animals for testing, treatment, or research purposes. PAG stimulation influences BP in the cat (see Kabat H, et α/.Magoun HW, Ranson JW. Electrical stimulation of points in the forebrain and midbrain. The resultant alteration in blood pressure. Archs Neurol 1935; 34:931-955).

The PAG is organized into four longitudinal columns (Carrive P, Bandler R, Dampney RA: Somatic and autonomic integration in the midbrain of the unanesthetized decerebrate cat: a distinctive pattern evoked by excitation of neurones in the subtentorial portion of the midbrain periaqueductal grey. Brain Res 1989;483:251-8; Carrive P, Bandler R: Viscerotopic organization of neurons subserving hypotensive reactions within the midbrain periaqueductal grey: a correlative functional and anatomical study. Brain Res 1991;541:206-15; Carrive P, Bandler R: Control of extracranial and hindlirrib blood flow by the midbrain periaqueductal grey of the cat. Exp Brain Res 1991 ;84:599-606).

Stimulation of the dorsomedial and dorsolateral columns produces an increase in BP whereas stimulation of the lateral and ventrolateral columns produces hypotension and freezing behavior (Abrahams VC, Hilton SM, Zbrozyna A: Active muscle vasodilatation produced by stimulation of the brain stem: its significance in the defense reaction. J Physiol 1960;154:491-513; see also Duggan AW, Morton CR: Periaqueductal grey stimulation: an association between selective inhibition of dorsal horn neurones and changes in peripheral circulation. Pain 1983; 15:237-48, Lovick TA: Inhibitory modulation of the cardiovascular defense response by the ventrolateral periaqueductal grey matter in rats. Exp Brain Res l992;89:133-9; Carrive P, Bandler R: Control of extracranial and hindlimb blood flow by the midbrain periaqueductal grey of the cat. Exp Brain Res 1991;84:599-606; Lovick TA: Ventrolateral medullary lesions block the antinociceptive and cardiovascular responses elicited by stimulating the dorsal periaqueductal grey matter in rats. Pain 1985;21:241-52). In the human, there is some evidence that stimulation of deep brain nuclei, such as the subthalamic nucleus, can influence the cardiovascular system (Thornton JM, Aziz T, Schlugman D, Paterson DJ: Electrical stimulation of the midbrain increases heart rate and arterial blood pressure in awake humans. J Physiol 2002;539:615-21; Priori A, Cinnante C, Genitrini S, Pesenti A, Tortora G, Bencini C, Barelli MV, Buonamici V, Carella F, Girotti F, Soliveri P, Magrini F, Morganti A, Albanese A, Broggi S, Scarlato G, Barbieri S: Non-motor effects of deep brain stimulation of the subthalamic nucleus in Parkinson's disease: preliminary physiological results. Neurol Sd 2001 ;22:85-6).

However, to date there have been no such reports related to the PVG/PAG. As described herein, the role of this region in the human is investigated. This study has found that ventral stimulation at 10Hz can have a consistent depressor (also referred to herein as "suppressed") response, whereas dorsal stimulation can have a pressor response. In addition to changes in SBP, this study has found corresponding analogous changes in DBP, pulse pressure, and maximum dP/dt (although change in the latter was only weakly significant with respect to fall in BP), but no change in RR interval. The changes may be elicited by a mixture of increased/decreased myocardial contractility (change in dP/dt) and a change in total peripheral resistance (changes in pulse pressure). The changes may be due to an altered sympathetic activity, with little or no change in parasympathetic activity. This is further corroborated by the changes in the power spectra of systolic blood pressure that show a change in the low frequency component implying a change in sympathetic activity, i.e., the spectral estimate of Mayer's wave (Pagani M, Montano N, Porta A, Malliani A, Abboud FM, Birkett C, Somers VK: Relationship between spectral components of cardiovascular variabilities and direct measures of muscle sympathetic nerve activity in humans. Circulation 1997;95:1441-8; Pagani M,

Lombardi F, Guzzetti S, Rimoldi O, Furlan R, Pizzinelli P, Sandrone G, Malfatto G, Dell'Orto S, Piccaluga E, et al: Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res 1986;59:178-93; see Furlan R, Guzzetti S, Crivellaro W, Dassi S, Tinelli M, Baselli G, Cerutti S, Lombardi F, Pagani M, Malliani A: Continuous 24-hour assessment of the neural regulation of systemic arterial pressure and RR variabilities in ambulant subjects. Circulation 1990;81 :537-47).

There are some important differences between the findings disclosed herein related to humans and findings related to animals. For example, the latency between stimulus and peak response in animals has been consistently reported at around 5-20 seconds (Kabat H, Magoun HW, Ranson JW. Electrical stimulation of points in the forebrain and midbrain. The resultant alteration in blood pressure. Archs Neurol 1935; 34:931-955; Abrahams VC, Hilton SM, Zbrozyna A: Active muscle vasodilatation produced by stimulation of the brain stem: its significance in the defence reaction. J Physiol 1960; 154:491-513; Hilton SM₅ Redfern WS. A Search for Brain Stem Cell Groups Integrating the Defence Reaction in the Rat. J Physiol 1986; 378: 213-228; Lindgren P, Uvnas B. Activation of sympathetic vasodilator and vasoconstrictor neurons by electrical stimulation in the medulla of the dog and cat. Circulation Res 1953; 1:479-485).

This study found much longer latencies to peak effect in the order of minutes, rather than seconds. A second difference is that it was found no significant change in RR interval with stimulation. This may, in fact, be the reason why the latencies are so much longer. It would appear that the stimulation can alter blood pressure by a non-vagal (i.e., a sympathetic pathway) that takes longer to exert its effect. Anatomical substrates for this sympathetic pathway have been demonstrated. For example, the ventrolateral PAG projects to the Nucleus Raphe Magnus (NRM) (Gallager DW, Pert A. Afferents to brain stem nuclei (brain stem raphe, nucleus reticularis pontis caudalis and nucleus gigantocellularis) in the rat as demonstrated by microiontophoretically applied horseradish peroxidase. Brain Res 1978; 144: 257-275) and PAG neurones are excited antidromically by NRM stimulation (Shah Y, Dostrovsky JO. Electrophysiological evidence for a projection of the periaqueductal gray matter to nucleus raphe magnus in cat and rat. Brain Res 1980; 193: 534-538). PAG also sends collaterals to the rostroventrolateral medulla (RVLM) (Hudson PM, Lumb BM. Neurons in the midbrain periaqueductal grey send collateral projections to nucleus raphe magnus and the rostral ventrolateral medulla in the rat. Brain Res. 1996 ;733(1):138-41) and serotonin receptor agonists applied to the RVLM produce hypotension (Mandal AK, Zhong P, Kellar KJ, GiIHs R. Ventrolateral medulla: an important site of action for the hypotensive effect of drugs that activate serotonin-1 A receptors. J Cardiovasc Pharmacol 1990; 15: S49- 60).

Other ways in which this study differs from animal experiments, which might explain the latency of response, include the fact that our subjects were awake, as opposed to the anaesthetized, decerebrate animals that would not have had the influence of higher brain centers such as prefrontal cortex that has been shown to inhibit cardiovascular responses in animals (Zbrozyna AW, Westwood DM. Stimulation in prefrontal cortex inhibits conditioned increase in blood pressure and avoidance bar pressing in rats. Physiol Behav. 1991; 49(4):705-8). Also, the methodology differs in that stimulation is applied continuously for several minutes, rather than giving a 'pulse' of stimulation for 5-10 seconds.

Limitations of this study include the use of a finger plethysmograph that has been shown to have a bias of approximately 2-4mmHg (Constant I, Laude D, Elghozi JL, Murat I: Assessment of short-term blood pressure variability in anesthetized children: a comparative study between intraarterial and finger blood pressure. J Clin Monit Comput 1999;15:205-14). However, the Finapres® has been shown to be reliable when looking at BP variability (Imholz BP, Wieling W, van Montfraus GA, Wesseling KH. Fifteen, years experience with finger arterial pressure monitoring: assessment of the technology. Cardiovasc Res 1998; 38:605- 16). Also, with changes of greater than 14mmHg, as well as intra-arterial recordings that fit exactly with lab-based recordings, it is believed that this problem is circumvented. Another limitation is that it has been shown BP changes are sustained only for short periods of time (up to 6 minutes). The assessment of longer-term changes may necessitate the use of an entirely different protocol.

EXEMPLIFICATION EXAMPLE 1

Methods and Results Cardiovascular responses to electrical stimulation of the periventricular/periaqueductal gray (PVG/PAG) were measured in fifteen awake human subjects following routine implantation of deep brain stimulating electrodes for treatment of chronic pain. Stimulation parameters were manipulated under controlled conditions and blood pressure measurements were made with a finger plethysmograph and confirmed with an intraoperative recording. Six controls were also investigated. Heart rate, rate of change of blood pressure and power spectra were calculated.

It was found that stimulation of the ventral PVG/PAG caused a mean reduction in systolic blood pressure of 14.2mmHg (ρ<0.001 ) in 7 patients and stimulation of the dorsal PVG/PAG caused a mean increase of 16.7rnmHg (p<0.001) in 6 patients. These changes were accompanied by analogous changes in diastolic blood pressure, pulse pressure, maximum dP/dt but not R-R interval. Not all patients showed a response. Power spectral analysis of the systolic blood pressure curve suggests a change in sympathetic outflow as the underlying mechanism. Conclusions

Deep Brain Stimulation of the human PVG/PAG can modulate blood pressure in awake humans. The effect is site specific and related to frequency of stimulation. Introduction

Control of arterial blood pressure is a complex process that is influenced by both hormonal and neural pathways from the forebrain down to each individual cardiac and vascular myocyte. In the midbrain, the periaqueductal gray (PAG) projects to all medullary regions that control blood pressure (BP) and heart rate, as well as having reciprocal connections with higher centers (Shipley, M. T., Ennis M. Rivzi T. A., Behbehani M. M. Topographical specificity of forebrain in the periaqueductal grey and inputs to the midbrain periaqueductal grey: Evidence for discrete longitudinally organised input columns. In: Depaulis, A. BandlerR. The Midbrain Periaqueductal Gray Matter: Functional, Anatomical and Neurochemical Organization. New York, USA, Plenum Press. 1991, 417-448; see also Beitz, A. J. Central gray. In: Paxinos, G. The Human Nervous System. San Diego, USA, Academic Press. 1990; see also M'hamed SB, Sequeira H, Poulain P, Bennis M, Roy JC): Sensorimotor cortex projections to the ventrolateral and the dorsomedial medulla oblongata in the rat. Neurosci Lett 1993;164:195-8, Guyenet PG, Haselton JR, Sun MK: Sympathoexcitatory neurons of the rostroventrolateral medulla and the origin of the sympathetic vasomotor tone. Prog Brain Res 1989;81 : 105- 16, Barman SM: Descending projections of hypothalamic neurons with sympathetic nerve- related activity. J Neurophysiol 1990;64: 1019-32.

Thus, the neurocircuitry of the PAG may play a pivotal role in cardiovascular control, probably via the medulla.

The periventricular gray (PVG) is the most medial of the three regions of the hypothalamus, located adjacent to the third ventricle. This is continuous with the PAG that encircles the cerebral aqueduct. These nuclei have long been known to have an important role in the modulation of pain (Magoun, H. W. Atlas D. Ingersoll E. H. Ranson S. W. Associated facial, vocal and respiratory components of emotional expression: An experimental study. J Neurol Psychopath 17, 241-155; Melzack R, Stotler WA, Livingston WK. Effects of discrete brainstem lesions in cats on perception of noxious stimulation. J Neurophysiol 1958;21:353-67). In addition, stimulation of this area in humans can produce fear (Nashold BS Jr, Wilson WP, Slaughter DG. Sensations evoked by stimulation in the midbrain of man. J Neurosurg 1969;30: 14-24) and in animals may increase or decrease blood pressure (Kabat H, Magoun HW, Ranson JW. Electrical stimulation of points in the forebrain and midbrain. The resultant alteration in blood pressure. Archs Neurol 1935; 34:931-955; Abrahams VC, Hilton SM, Zbrozyna A: Active muscle vasodilatation produced by stimulation of the brain stem: its significance in the defense reaction. J Physiol 1960;154:491-513; Duggan AW, Morton CR: Periaqueductal grey stimulation: an association between selective inhibition of dorsal horn neurones and changes in peripheral circulation. Pain 1983; 15:237-48; Lovick TA: Inhibitory modulation of the cardiovascular defense response by the ventrolateral periaqueductal grey matter in rats. Exp Brain Res 1992;89:133-9). As described herein, in awake humans, deep brain stimulation in this area can increase or decrease arterial blood pressure (ABP). This effect was dependent on both the ventral/dorsal location of the electrode as well as the frequency of stimulation, and highlights a potential neuro-therapautic target to regulate ABP. Methods

The patient demographics are summarized in Table I. Fifteen patients (twelve male, three female) were referred for deep brain stimulation for neuropathic pain. Mean age was 51.3 years (range 30-74 years). Four patients acted as their own controls as they had both PVG/PAG and thalamic stimulators. Two controls were patients with non-pain conditions - one with a thalamic deep brain stimulator, the other with a spinal cord stimulator. Informed consent for participation in the study was obtained from each patient, and the study was approved by the local ethics committee.

TABLEl

Surgical Technique

Details of an example surgical technique for targeting the thalamus has been described previously (Papanastassiou V, Rowe J., Scott R., Silburn P., Davies L., Aziz T.: Use of the Radionics Image Fusion™ and Stereoplan™ programs for target localisation in functional neurosurgery. J CHn Neurosci 1998;5:28-32). An embodiment of an image-guided technique of targeting PVG/PAG according to the principles of the present invention has been tested as follows. The targeting was carried out on fused stereotactic Magnetic Resonance Imaging (MRI)/computed tomography scans of 2mm in thickness using the Radionics Image Fusion™ and Stereoplan™. The anterior and posterior commissures were identified on the axial images. The intended target for placing the deepest electrode contact was marked at the PAG at a level of < 10mm below the AC-PC line; between the dorsal part of the red nucleus and the superior colliculus in the AP plane; and approximately 5mm lateral to the lateral boundary of the aqueduct and the third ventricle. Next, the electrode trajectory was selected to avoid possible penetration of the surface vessels on the cortex and the lateral ventricle. This leads to some adjustment of the target localization for each individual patient, and likely contributes to inter-patient variation in electrode placement.

In patients with post-stroke pain and severely deformed hemispheres, targeting can be quite difficult. For these patients, relative anatomic landmarks such as the third ventricle, the aqueduct, the red nucleus and the superior colliculus are more reliable than AC-PC measurements. After making a 2.7mm twist-drill hole in the skull, a Radionics™ electrode of 1.8 mm diameter and 2.0 mm exposed tip was slowly passed towards the target while the impedance values were monitored for a sudden drop in impedance value from 500 - 600? to under a few tens of ohms, suggesting possible penetration of a ventricle. The Radionics™ electrode was replaced by a Medtronic 3387® electrode (Medtronic Inc., Minneapolis, USA).

Test stimuli were applied at <3.0V in amplitude, 120μs in pulse width and 10 - 50Hz in frequency to check for a warm feeling or paraesthesiae in the area of pain or pain suppression and abnormal eye movements. Once the pain suppression area was identified functionally, the DBS electrode was fixed onto the skull and externalized for further investigation. The whole stimulation system was then internalized a few days later in a second procedure.

Post-operative Localization of Electrodes The electrode positions were plotted on a brain atlas (Mai J. K, Assheuer J,

Paxinos G. Atlas of the Human Brain , San Diego, USA. Academic Press. 1998; p 79, 118) using the post-operative MRI and a manipulation program (MRIcro version 1.38 build 1, Chris Rorden). First, the scan was rotated such that the anterior and posterior commissures (AC and PC respectively) were on the same slice. The mid- commissural point was then calculated, followed by the position, in Talairach space, of the electrode contacts. The contacts are visible, circular thickenings in the low signal on the axial scan (Fig. 5). The center of each contact was taken as the position of the electrode, and this corresponds to the center of the contacts in Figs. 6 A and 6B. Using the coronal and sagittal scans, the angles of the electrode to the midline and AC-PC line, respectively, were calculated. Once plotted on the brain atlas, the relative position of the lowest contact to the posterior wall of the superior colliculus was verified, as was the relative position of the upper electrode to the mid- commissural point. As a further verification, the relative positions of the electrodes from all patients were compared, to rule out inconsistencies among the groups.

Measurements

During lab-based recordings, the non-invasive continuous finger arterial pressure was measured with an Ohmeda Finapres 2300 (FinapresTM, BOC Healthcare, USA). The blood pressure was calibrated using a sphygmomanometer, and the pressure transducer and finger cuff were positioned at heart level during the experiment. Automatic BP measurements from the upper arm were also made every three minutes or when a change in BP was observed (Omron 705CP-II Automatic Blood Pressure Monitor, Omron® Healthcare Europe B.V, Hoofdorp, Netherlands) in order to corroborate the finger arterial pressure measurements. Lead H electrocardiogram (ECG) was recorded using disposable adhesive

Ag/AgCl electrodes (H27P, Kendall-LTP, MA, USA), amplified* 1,000 (CED 1902, Cambridge Electronic Design, Cambridge, UK). The finger pressure and ECG were digitized at 4kHz with 16-bit resolution (CED 1401 Mark II, Cambridge Electronic Design, Cambridge, UK) using Spike II software® (version 5.0, Cambridge Electronic Design, Cambridge, UK).

In order to validate the lab-based (Finapres®) recordings, in one patient (#8), an intra-arterial catheter (BD Angiocath™, Infusion Therapy Systems Inc, Sandy, Utah) was inserted into the radial artery during general anesthesia. The patient was induced with a sleep dose of propofol, and anesthesia was maintained with nitrous oxide in oxygen, midazolam, end tidal concentration sevoflurane 0.5-1% and fentanyl. The arterial pressure signal was transduced using a Medex Medical® transducer and recorded from the anesthetic machine (AS/3® Datex-Ohmeda inc., Tewksbury, MA), sampled at 500Hz with 12-bitresolution (MP 100®, Biopac Systems, Santa Barbara, Ca) using Acqknowledge® software (version 3.7.3, Biopac systems). The dorsal PVG was stimulated at 10Hz at 4.Ov (pulse width 120μs) for 3 minutes. This was repeated three times with five minutes rest in-between to confirm the effect.

Study Design

Experiments were performed more than 2 hours after any meal. On the day of the experiment, patients were asked not to take any analgesics, such as morphine or its derivatives, that might affect BP. They also abstained from coffee and tea. Hypertensive patients took their usual antihypertensive medication (as listed in Table I). The deep brain stimulator was initially turned off for at least 10 minutes prior to experiments.

The experiments were started with the patient sitting for 5 minutes. The first session consisted of a 12-minute rest period (while recording cardiovascular variables) with the stimulator turned off. The same procedure was then repeated with the stimulator turned on and randomly set at different settings. If there was a change in BP, stimulation was stopped 300s after it had been started to look at the recovery phase. If there was no change, stimulation was continued for 12 minutes to confirm that there was no effect. There was a 9-minute rest period with the stimulation off in-between each session. This rest period was extended if blood pressure had not yet returned to the baseline value of session 1. The Medtronic 3387® electrode consists of four circumferential contacts measuring 1.5mm, spaced by lmm. During each session with stimulation on, bipolar stimulation was used between the two deepest or the two most proximal contacts, at either 10 or 50Hz. The pulse width was 120μs and amplitude was increased to the maximum tolerated by the patient, without side effects, up to 3 volts (equivalent to a current magnitude up to 3mA). Thus, a typical experiment would consist of at least eight sessions including the rest periods. To confirm consistency of responses to stimulation, if a change in BP was observed for any particular session, the session was repeated two further times to confirm the response. Both the patient and the person recording the data were blinded from the actual stimulation parameters. The pain was quantified with a visual analogue score at the beginning and end of each session.

Signal processing and statistical analysis In order to avoid change of BP caused by pain, the data segment was excluded when visual analogue score changed. One-way analysis of variance of BP with time was performed on all raw data segments for each session to determine significant change (p<0.05). Data segments from sessions with significant changes (on at least three occasions) were then averaged every 30 seconds, and the mean of each of these values was plotted sequentially to provide the overall changes over time. One-way analysis of variance with time was performed on the overall figures to determine significance. The blood pressure changing rate, dP/dt, was derived by differentiating the blood pressure. The maximum dP/dt (maximum slope of the blood pressure curve) was then extracted. All results are expressed with ± one standard error of the mean.

Auto-regressive power spectra analysis of SBP was performed on all data segments that had been chosen for analysis (i.e., those with significant changes in BP). Frequencies below 0.02 Hz were filtered out to remove the trend in the signal (Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation 1996;93: 1043-65) for methodology). The power of the low and high frequency components were computed as the integral of the power spectra between 0.05 and 0.15 Hz and between 0.15 and 0.4 Hz. The logarized low and high frequency power between recording sessions was analyzed using a paired t-test.

AU signal processing was performed in Matlab® (Version 6.1, MathWorks Inc., Natick, Ma., USA, and statistical analysis was performed in SPSS (Version 11, SPSS Inc, IL, USA).

Results

Of the fifteen patients with PVG/PAG electrodes (two bilateral), five had episodes of significant decreases in BP (five electrodes), four had significantly increased BP (four electrodes), and two had episodes of both (two electrodes). Stimulation of the four remaining PVG patients (six electrodes) and six control electrodes caused no significant changes.

Intra-arterial blood pressure recordings

To verify the non-invasive ABP measurements (Finapres®), testing was conducted to show, under general anesthetic, that stimulation of the dorsal PVG caused an increase in ABP when measured via an arterial line (Fig. 7A). This was reproduced three times for at least one minute. Thus, the intra-arterial measurements confirmed the changes in BP as measured with the Finapres® in this patient (compare with Fig. 7B).

Reduction of Arterial Blood Pressure with Stimulation of Ventral Periventricular Gray Fig. 8A shows the composite data from all seven patients in whom SBP dropped significantly after the onset of stimulation, without significant changes in pain severity (#1 to #7, lower two contacts only in #1 and #6). It is striking that the contacts that reduced BP were the most ventral electrodes (Figs. 6A and 6B, blue electrodes: electrodes 1-6, lower two conductors; electrode 7, upper two conductors). Thus, it appears that stimulation of the ventral PVG/PAG is required to reduce BP. There is considerable variation in the lateral location of electrode placement, which is due to variations in ventricular size that determines intraoperative electrode trajectory, but there does not appear to be a relation to decreased BP.

The average reduction in systolic BP was 14.2 mmHg ±3.6 (p<0.001, single factor ANOVA, n=7, range of reduction = 7-25mmHg), equivalent to 13.9%, at the end of a 400 second period where stimulation was started at 100s. The mean latency (i.e., the time from initiation of stimulation to the maximum fall in SBP) was 160 ± 29s, although there was a considerable range between subjects (34 to 214 seconds). It is also worth noting that there was a much shorter time between stimulation onset and the initial change in SBP (mean = 24±8 seconds). Fig. 8A shows that when stimulation was turned off, the latency was much shorter (mean = 48 ± 23 seconds), and the time between turning off the stimulus and initial change in SBP was even shorter (mean = 6±4 seconds).

In all seven patients, the stimulation parameters required to drop BP were 10Hz with a pulse width of 120μs and a voltage range from 0.6v to 3.Ov (equivalent to a current of 0.6 to 3mA). In three patients, 50Hz had a similar effect. Fig. 8A shows that the drop in SBP is accompanied by a fall in diastolic BP (DBP) of 4.9mmHg± 2.9 (ρ=0.03, single factor ANOVA, n=7, range 1.5-9.3), equivalent to 6%. As the systolic drops more than the diastolic BP, there is a mean decrease in pulse pressure of 9.3mmHg ±3.16 (p=0.04, single factor ANOVA). Analysis of the change of SBP with time (maximum dP/dt, i.e., the slope of the blood pressure curve) showed a mean reduction of 222 mmHg/s ±126 (19.8%, p=0.06). This is suggestive, but^"not absolute proof that the contractility of the myocardium was reduced. On the other hand, R-R interval did not change significantly throughout the stimulation period (mean change = 0.01s ±0.04, range 0-0.08). Frequency analysis of SBP revealed that the average reduction in the logarithm of the low frequency component was 20.2±14% (p=0.03, paired t-test, n=7).

Fig. 1OA shows raw data from one patient; Fig. 1OB shows the group data from all seven patients. In contrast, there was no significant change in the high frequency component. This implies a reduction in sympathetic outflow (Cohen MA, Taylor JA: Short-term cardiovascular oscillations in man: measuring and modeling the physiologies. J Physiol 2002;542:669-83). Increase of ABP with Stimulation of Dorsal Periventricular Gray

Six patients (#8- #11 and the upper 2 contacts in patients #1 and #6) had episodes of a sustained increase in BP at or shortly after the onset of stimulation (Fig. 9A). One of these patients (#10) was hypertensive. Two of these patients (#1, #6) also had episodes of decreased BP when the lower contacts were stimulated, supra). Analysis of the image data (Figs. 6A and 6B) showed that the electrodes in this group were placed dorsally (#8-#l 1), with the exception of patients #1 and #6 in whom the distal part of the electrodes were ventral and the proximal parts were dorsal (the episodes of raised BP occurring when the most proximal two contacts were stimulated, i.e., those located dorsally).

The mean rise in SBP was 16.73 mmHg ±5.9 (pO.001, single factor ANOVA, n=6, range 16-31mmHg), equivalent to 16.4% at the end of a 400s period where stimulation was started at 100s; however, the maximum rise of 22.23mmHg occurred just before this - see Fig. 9 A. The mean latency was 230±44s (with a range of 48 to 289s). As with reduction in BP₅ there was also a much shorter time between stimulation and initial rise in blood pressure of 8±4s.

Stimulation parameters required to raise BP were the same as with the episodes of reduced BP (i.e., 10Hz, 120μs and up to 3.0v), except that 50Hz did not have the same effect in any patient. As with BP reduction, increases were accompanied by a smaller rise in DBP of 4.9mmHg±2.8 or 6.4% (p=0.04, single factor ANOVA, n=6, range = 2.4 to 12.1mrαHg). There was also an increase in mean pulse pressure of 11.83±5.4mmHg or 14.5% (pO.Ol, single factor ANOVA), and again, the maximum rise of 17.33mmHg occurred just before 400s. Maximum dP/dt increased by 212±97 mrnHg/s (ρ<0.03, single factor ANOVA). As with reduction in BP₅ there was no significant change in R-R interval. Thus, it appears that increasing BP is accompanied by a mirror of the changes that occur during reduction in BP.

Frequency analysis shows a significant increase in the logarithm of the low frequency range of 30.4±12.1% ± (p=0.03₅ paired t-test, n=6), implying an increase in sympathetic outflow. Again, there was not a significant change in the high frequency component (Fig. 10B). Controls and electrodes that had no effect

Six control patients were investigated (six thalamic electrodes, one spinal cord stimulator). Stimulation was unable to modulate the BP in any of these patients despite extensive investigation using a variety of frequencies and voltages, as well as a variety of electrode contact configurations,.

In addition to the control electrodes that had no effect on BP, four patients with PVG electrodes (six electrodes in total) also had no effect. Electrode positions was plotted for five of these six electrodes (one had not had a post-operative scan). Four of the five electrodes were dorsal to the group that raised BP and were therefore probably outside the PAG/PVG. The remaining electrode (#15) was in mid-PVG (similar to #2), and it maybe of note that he was being actively treated for hypertension.

EXAMPLE 2: TREATMENT OF ORTHOSTATIC HYPOTENSION Methods

Patients (n=l 1) who had chronic neuropathic pain and who had undergone implantation of a deep brain stimulator in the PVG/PAG were employed. Patients were divided into three groups depending on whether they had orthostatic hypotension (one patient), 'mild orthostatic intolerance' (five patients) or no orthostatic intolerance (five patients). Post-operatively, we continuously recorded blood pressure and heart rate with stimulation off and on and in both sitting and standing positions. From these we derived the blood pressure changing rate (dP/dt). Using autoregressive modelling techniques, we calculated changes in low and high frequency power spectra of heart rate and barorefiex sensitivity. Results

Electrical stimulation reduced the fall in SBP on standing from 28.2% to 11.1% in one patient with orthostatic hypotension (pO.001). In the mild orthostatic intolerance group, an initial drop in SBP of 15.4% was completely reversed (pO.001). There were no side-effects in the remaining group. These changes were accompanied by increases in dP/dt, baroreflex sensitivity, and baseline (sitting) low frequency power of RR interval but not high frequency power. Conclusions

Electrical stimulation of the human PVG/PAG can reverse orthostatic hypotension. The cause appears to be an increase in sympathetic outflow and in baroreflex sensitivity. This has important implications for future therapies. Introduction

Orthostatic Hypotension is a significant clinical problem that affects a large number of people, particularly the elderly (Rutan GH, Hermanson B, BiId DE, Kittner SJ, LaBaw F, Tell GS. Orthostatic hypotension in older adults. The

Cardiovascular Health Study. CHS Collaborative Research Group. Hypertension 19: 508-19, 1992) and, in extreme cases, has even led some clinicians to implant subcutaneous norepinephrine pumps (Oldenburg O, Mitchell A, Nurnberger J, Koeppen S, Erbel R, Philipp T, Kribben A. Ambulatory norepinephrine treatment of severe autonomic orthostatic hypotension. J Am Coll Cardiol 37: 219-23, 2001.) Ascending projections of barosensitive adrenergic cells in the rostro ventrolateral medulla project to PAG (Haselton JR, Guyenet PG. Ascending collaterals of medullary barosensitive neurons and Cl cells in rats. Am J Physiol 258: R1051-63, 1990.) PAG may project to preganglionic cardiac vagal neurones in the nucleus ambiguus and chemical stimulation of the PAG inhibits baroreflex vagal bradycardia in rats (Inui K, Nosaka S. Target site of inhibition mediated by midbrain periaqueductal gray matter of baroreflex vagal bradycardia. J Neurophysiol 70: 2205-14, 1993.) Stimulation of this area in the human may affect the baroreceptor reflex, hi this study, we investigate the effect of electrical stimulation of this area on postural changes in blood pressure in patients treated for neuropathic pain. A secondary aim was to elucidate the mechanisms of any changes, including autonomic nervous activity and baroreflex sensitivity. These findings will have important implications for the potential use of chronic PVG/PAG stimulation in the treatment of orthostatic hypotension.

Methods Subjects

Patients with chronic neuropathic pain undergoing deep brain stimulation of the PVG/PAG areas were recruited to this study (see Table 2 for details). Patients were excluded if they i) were unable to stand for 280 seconds (due to paraplegia or hemiplegia) ii) had an irregular heart rhythm, iii) had heart disease, including previous myocardial infarction or valvular disease, iv) were taking medication that might affect autonomic response e.g. betablockers, antidepressants, v) had any other disease such as Parkinson's Disease or alcoholism that might affect the autonomic response. The study was approved by the local ethics committee and informed consent was obtained from all patients. Eleven subjects (10 male) were recruited. Mean age was 53 years (range 34-74 years). One subject (henceforth referred to as 'subject #1') had a past clinical history of orthostatic hypotension that resolved after insertion of the stimulator two years previously (the stimulator was constantly on). Etiology of neuropathic pain was as follows; six had suffered thalamic hemorrhage, one a pontine hemorrhage; one brachial plexus trauma; one anaesthesia dolorosa; one post-traumatic headache; one post-craniotomy facial pain. Subjects were divided into three groups depending on the initial change in systolic blood pressure on standing (Table 3); orthostatic hypotension i.e. a fall of >20mmHg at three minutes (one patient -subject #1); mild orthostatic intolerance i.e. a fall of >20mmHg but recovery to less than 20mrnHg before three minutes (MOI group - 5 patients); A fall of <20mmHg on standing (non-OI group - 5 patients). TABLE 2

TABLE 3

Surgical Technique

Details of our surgical technique for targeting the thalamus has been described previously (Papanastassiou V, Rowe J., Scott R._} Silburn P., Davies L., Aziz T. Use of the Radionics Image Fusion™ and Stereoplan™ programs for target localisation in functional neurosurgery. J Clin Neurosci 5:28-32, 1998.) Our image- guided technique of targeting PVG/PAG is as follows. The targeting was carried out on fused stereotactic magnetic resonance (MRT)/computed tomography scans of 2mm in thickness using the Radionics Image Fusion™ and Stereoplan™. The anterior and posterior commissures were identified on the axial images. The intended target for placing the deepest electrode contact was marked at the PAG at a level of <10mm below the AC-PC line; between the dorsal part of the red nucleus and the superior colliculus in the AP plane; and approximately 5mm lateral to the lateral boundary of the aqueduct and the third ventricle. Next, the electrode trajectory was selected to avoid possible penetration of the surface vessels on the cortex and the lateral ventricle. This leads to some adjustment of the target localisation for each individual patient, and likely contributes to inter-patient variation in-electrode placement. In patients with post-stroke pain and severely deformed hemispheres, targeting can be quite difficult. For these patients, relative anatomic landmarks such as the third ventricle, the aqueduct, the red nucleus and the superior colliculus are more reliable than AC-PC measurements. After making a 2.7mm twist-drill hole in the skull, a Radionics™ electrode of 1.8 mm diameter and 2.0 mm exposed tip was slowly passed towards the target while the impedance values were monitored for a sudden drop in impedance value from 500 - 600? to under a few tens of? , suggesting possible penetration of a ventricle. The Radionics™ electrode was replaced by a Medtronic 3387® electrode (Medtronic Inc., Minneapolis, USA). Test stimuli were applied at <3.0V in amplitude, 120μs in pulse width and 10 - 50Hz in frequency to check for a warm feeling or paraesthesiae in the area of pain or pain suppression, and abnormal eye movements. Once the pain suppression area was identified functionally, the DBS electrode was fixed onto the skull and externalised for further investigation. The whole stimulation system was then internalised a few days later in a second procedure.

Study Design

Experiments were performed in the morning, more than 2 hours after any meal. On the day of the experiment, patients abstained from analgesics such as morphine or its derivatives that might affect blood pressure as well as caffeine- containing drinks. None of the subjects smoked. Hypertensive patients delayed their usual antihypertensive medication until after the experiment (as listed in table I). AU experiments were carried out in an ambient temperature of 22⁰C in a quiet room. Patients were asked not to sleep or talk during the study.

The experiments were started with the patient sitting for 10 minutes with the stimulator 'off. Then, while cardiovascular variables were recorded, the first 'session' was conducted with the stimulator turned off. During this, and all subsequent sessions, the patient was sitting for three minutes, followed by standing for 280 seconds, followed by sitting. A sitting 'rest' period of at least nine-minutes followed each 'session' to allow the blood pressure to recover to baseline. Stimulation status ('on' or Off) was randomly ordered and both patients and researchers were blinded to it. The Medtronic 3387® electrode consists of 4 circumferential contacts measuring 1.5mm, spaced by lmm. During each session with stimulation on, bipolar stimulation was used between the middle two contacts, at 10Hz. The pulse width was 120μs and amplitude was increased to the maximum tolerated by the patient, without side effects, up to 3 volts. As the settings used were different to those generally used for pain control, patients were unable to discern whether stimulation was on or off. For each stimulator setting, including stimulation 'off, the session was repeated three times and the average of the three sessions was used. Thus, an experiment consisted of 6 sessions including the rest periods.

Measurements

During lab-based recordings, the non-invasive continuous finger arterial pressure was measured with an Ohmeda Finapres 2300 (Finapres™, BOC Healthcare, USA). The blood pressure was calibrated using a sphygmomanometer and the pressure transducer and finger cuff (placed on the middle finger) were positioned at heart level during the experiment. Automatic blood pressure measurements from the upper arm were also made every three minutes (Omron 705CP-II Automatic Blood Pressure Monitor, Omron® Healthcare Europe B.V, Hoofdorp, Netherlands) in order to corroborate the finger arterial pressure measurements. Lead II electrocardiogram (ECG) was simultaneously recorded using disposable adhesive Ag/AgCl electrodes (H27P, Kendall-LTP, MA, USA), amplifiedx 1,000 (CED 1902, Cambridge Electronic Design, Cambridge, UK). The finger pressure and ECG were digitized at 4kHz with 16-bit resolution (CED 1401 Mark II, Cambridge Electronic Design, Cambridge, UK) using Spike II software^® (version 5.0, Cambridge Electronic Design, Cambridge, UK).

Data Analysis and Signal Processing

For each session, data were averaged every 10 seconds and the mean values over three sessions were taken. For each group (subject #1 , MOI or non-MOI), all patient data were then averaged at each session. From these data, the mean peak change in systolic blood pressure and heart rate were calculated — this was defined as the maximum change from baseline within 30s of standing. Also derived was the percentage change in systolic and diastolic blood pressure, heart rate, pulse pressure, and blood pressure changing rate (dP/dt) at plateau which was defined as the period from 220-28Os. The dP/dt was derived by differentiating the blood pressure. The maximum dP/dt value is the measure of maximum changing of the blood pressure which reflects the contractility of the myocardium.

For each patient, auto-regressive power spectra analysis of R-R interval was performed from 3 -minute data segments of each session. This mathematical technique allows us to look at heart rate variability (HRV) and provides a quantitative measure of the low and high frequencies in the HRV. In turn these low and high frequencies are associated with different aspects of the sympatho-vagal components of heart rate and provide an estimate of sympathetic/ parasympathetic activity (Pagani M, Lombardi F, Guzzetti S, Rimoldi O, Furlan R, Pizzinelli P,

Sandrone G, Malfatto G, Dell'Orto S, Piccaluga E, et al. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res 59: 178-93, 1986). Frequencies below 0.02 Hz were filtered out to remove the trend in the signal (see Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation 93: 1043-65, 1996 for methodology). The power of the low and high frequency components were computed as the integral of the power spectra between 0.05 and 0.15Hz and between 0.15 and 0.4Hz respectively. The baroreflex sensitivity index was calculated from the transfer function of systolic blood pressure and RR interval signals using bivariate autoregressive modeling (Barbieri R, Bianchi AM, Triedman JK, Mainardi LT, Cerutti S, Saul JP. Model dependency of multivariate autoregressive spectral analysis. IEEE Eng Med Biol Mag 16: 74-85, 1997, and Zhang Y, Critchley LA, Tarn YH, Tomlinson B. Short-term postural reflexes in diabetic patients with autonomic dysfunction.

Diabetologia 47: 304-11, 2004) RR(ή (k)RR(n -k) (k)SBP{n-k)+w{ή)-

After checking that the squared coherency was greater than 0.3 (Lefrandt JD, Hoogenberg K, van Roon AM, DuUaart RP, Gans RO, Smit AJ. Baroreflex sensitivity is depressed in microalbuminuric Type I diabetic patients at rest and during sympathetic manoeuvres. Diabetologia 42: 1345-9, 1999), the baroreflex sensitivity was calculated as the average of low and high frequency gains of the transfer function from systolic blood pressure to R-R interval for both sitting and standing positions with stimulation off and on for each patient. This mathematical model essentially compares the effects of changes in systolic blood pressure to changes in heart rate (for both high and low frequencies). For example, if a change in systolic blood pressure leads to a large change in RR interval i.e. a large gain in the transfer function, the baroreflex sensitivity index would be high i.e. because the baroreflex change is greater, this implies a greater sensitivity of the baroreflex.

All signal processing was performed in Matlab® (Version 6.1, MathWorks Inc., Natick, Ma., USA).

Statistical Analysis

Data were analyzed using the Statistical Package for Social Sciences (SPSS® version 11, SPSS Lie, IL, USA). The changing of systolic and diastolic blood pressure, heart rate, pulse pressure and dP/dt were tested using one way analysis of variance with time (ANOVA) when the stimulation was on or off. A paired t-test was used for testing the difference between resting and initial standing conditions when the stimulation was on or off, the difference between resting and plateau standing conditions when the stimulation was on or off, the difference between stimulation on and off at resting or plateau standing conditions, and the changes in the low and high frequency power and baroreflex sensitivity changes between stimulation on and off. Significance was taken as p<0.05 (two-tailed). All results are expressed with mean ± one standard error of the mean or ranges (in brackets).

Results

In subject #1, who was found to fulfil the clinical definition of orthostatic hypotension (Schatz J B Real. Consensus statement on the definition of orthostatic hypotension, pure autonomic failure, and multiple system atrophy. J Neurol Sci

144: 218-9, 1996) his postural effects on systolic blood pressure were reversed with stimulation at 10Hz, 120μs pulse width, and 3.Ov (Fig. 1 IA). On three occasions, while sitting with the stimulator off, his mean systolic blood pressure was 140.8±2.9mmHg. On standing, this gradually fell to a mean of 96.9±1.4mmHg after 280 seconds, a 28.2% reduction (ρ<0.001, ANOVA). Electrical stimulation of the PVG/PAG (Fig. 1 IA- upper line) reduced the postural reduction to 11.1% at 280 seconds (p<0.001 ANOVA). This phenomenon of a lessening in the postural reduction of systolic blood pressure with stimulation is concordant with the patient's clinical history that his symptoms of orthostatic hypotension resolved after the stimulator was originally inserted. Statistical analysis showed that the reduction of systolic blood pressure after standing with stimulation 'on' was significantly less than that when stimulation was off (t-test, p<0.001). hi the mild orthostatic intolerance group (MOI), the mean initial peak postural reduction of 22.3±2.4mmHg or 15.4% in systolic blood pressure on standing (p<0.001, ANOVA) was prevented with PVG/PAG stimulation (Fig. 1 IB). The reduction was significantly different between stimulation 'off and 'on' conditions (t-test, p<0.001). In the non-01 group, systolic blood pressure increased significantly on standing in either 'on' or OfP conditions (p<0.001). However, a t-test revealed no significant difference between stimulation 'on' or 'off conditions (p>0.5) (Fig. HC).

Investigation of heart rate revealed that subject #1 had a less than expected peak increase on standing with stimulation off, from 64.7±2.0bpm to 76.2±1 ,3bpm or 17.8% (ANOVA, pO.OOl, Fig. 1 ID). With stimulation, the peak increase on standing almost doubled (from 58.4±1.9bpm to 77.3±2.2bρm (pO.OOl, ANOVA)), however, a t-test revealed no significant difference between the 'on' and 'off conditions (p=0.6). Although heart rate increased significantly (pO.OOl, ANOVA) on standing for both on and off conditions in the MOI (Fig. 1 E) or non-01 groups (Fig. IF), there was no significant difference in heart rate response between 'on' and 'off in each group (p>0.1, t-test).

Systolic blood pressure changing rate (dP/dt) was not significantly altered by stimulation in the non-01 group (Fig. II). However, in subject #1, dP/dt dropped synchronously with the fall in blood pressure in the 'off state. In the 'on' state, the magnitude of the fall in dP/dt was markedly decreased (p=0.022, t-test). The MOI group, on the other hand, had a significantly raised baseline value of dP/dt (p=0.006, t-test) with stimulation, and did not change significantly on standing.

In the sitting position, stimulation did not significantly alter resting diastolic blood pressure or pulse pressure except in subject #1 , whose resting pulse pressure^• was higher (Table 3). However, in subject #1, the magnitude of the reduction in both of these variables associated with standing was significantly reduced with PVG/PAG stimulation (ρ=0.020, ρ=0.011 respectively, t-test). In the MOI group, the reduction was actually reversed (p=0.008, t-test).

Power Spectral Analysis and Baroreflex Sensitivity hi the MOI and Non-OI groups, baseline low frequency power of RR interval significantly increased with stimulation (t-test, p=0.021 and pO.OOl respectively, Table 4). However, baseline high frequency power in these groups was not significantly altered by the stimulation (p>0.1 , t-test). In the MOI group and subject #1 , the reduction in both low and high frequency power associated with standing . was prevented with stimulation (p=0.008, t-tests, 'on' vs. Off).

Baroreflex sensitivity reduced significantly on standing with stimulation off, in all three groups (p=0.004, Table 4). It was significantly raised by stimulation in the sitting position (t-test, ρ=0.018, 0.001 and 0.002 respectively) in all three groups. Although, with stimulation, baroreflex sensitivity reduced on standing, the magnitude of this reduction was significantly less in subject #1 (p=0.024 t-test) and the MOI group (pO.OOl, t-test). In the non-OI group, there was no difference between sitting and standing (p>0.1, t-test).

TABLE 4

Discussion

Orthostatic hypotension is a clinical sign that is defined as a fall in systolic blood pressure of =20mmHg, a fall in diastolic blood pressure of lOmmHg, or symptoms of cerebral hypoperfusion within 3 minutes of standing (Schatz IJ BReal. Consensus statement on the definition of orthostatic hypotension, pure autonomic failure, and multiple system atrophy. J Neurol Sci 144: 218-9, 1996). It maybe idiopathic or related to diseases with autonomic failure such as multiple system atrophy or diabetes. It is present in up to 20% of patients over 65 and its treatment may lead to troublesome hypertension (Kaplan NM. The promises and perils of treating the elderly hypertensive. Am J Med Sci 305: 183-97, 1993., and Rutan GH, Hermanson B₅ BiId DE, Kittner SJ, LaBaw F, Tell GS. Orthostatic hypotension in older adults. The Cardiovascular Health Study. CHS Collaborative Research Group. Hypertension 19: 508-19, 1992). In the normal subject, assumption of an upright posture leads to pooling of venous blood in the lower extremities and splanchnic circulation (Garas Z₅ Komor K. [Changes in the plasma volume in the erect position in hypertension. Comparative studies in the normotensive and hypertensive disease phase] . Z Gesamte Inn Med 26: 199-202, 1971.). The resulting decrease in venous return to the heart leads to a compensatory, centrally mediated increase in sympathetic and decrease in parasympathetic activity (i.e., the baroreceptor reflex). This activity usually results in a transient fall in systolic blood pressure of 5 to lOrnmHg, a small rise in diastolic blood pressure (5 to lOmmHg) and a rise in heart rate of 10-25bpm. We have shown that electrical stimulation of the P VG/P AG reverses or attenuates this fall in blood pressure, and increases the heart rate response to standing in subjects with orthostatic hypotension or intolerance. In a control group of subjects without orthostatic intolerance, the blood pressure changes on standing are unchanged with stimulation.

Changes in Heart Rate Variability and Baroreflex Sensitivity

The power of RR interval spectra in the high frequency band (0.15-0.4Hz) has been shown to be a marker of cardiac vagal control (Pagani M, Lombardi F, Guzzetti S, Rimoldi O, Furlan R, Pizzinelli P, Sandrone G, Malfatto G, Dell'Orto S, Piccaluga E, et al. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res 59: 178-93, 1986, and Saul JP, Berger RD, Albrecht P, Stein SP, Chen MH, Cohen RJ. Transfer function analysis of the circulation: unique insights into cardiovascular regulation. Am J Physiol 261: H1231-45, 1991). The low frequency band (0.04-0.15Hz) has been associated with cardiac sympathetic activity, although it has been shown to be affected by both vagal and sympathetic nerves (Berger RD, Saul JP, Cohen RJ. Transfer function analysis of autonomic regulation. I. Canine atrial rate response. Am J Physiol 256: H142-52, 1989, and Saul JP, Berger RD, Albrecht P, Stein SP, Chen MH, Cohen RJ. Transfer function analysis of the circulation: unique insights into cardiovascular regulation. Am J Physiol 261 : H1231-45, 1991). Previous research has shown a reduction in both these components of heart rate variability power with head up tilt in patients with autonomic neuropathy (Zhang Y, Critchley LA, Tarn YH₅ Tomlinson B. Short-term postural reflexes in diabetic patients with autonomic dysfunction. Diabetologia 47: 304-11, 2004) compared to the increase in low frequency power seen on standing in normal subjects (Sanderson JE₅ Yeung LY₅ Yeung DT, Kay RL, Tomlinson B₅ Critchley JA, Woo KS, Bernardi L. Impact of changes in respiratory frequency and posture on power spectral analysis of heart rate and systolic blood pressure variability in normal subjects and patients with heart failure. Clin Sci (Lond) 91: 35- 43, 1996).' We have shown that stimulation of the PVG/PAG significantly increases the baseline low frequency power of RR interval in the MOI and non-OI groups but not the high frequency power (table 3). Moreover, in the MOI group and subject #1 , stimulation prevents the reduction in both low and high frequency power that is usually associated with standing in these patients. These results suggest that stimulation may increase the cardiac sympathetic activity and enhance its response to standing. Li young and middle-aged healthy subjects, baroreflex sensitivity decreases on standing (Bahjaoui-Bouhaddi M, Henriet MT, Cappelle S, Dumoulin G, Regnard J. Active standing and passive tilting similarly reduce the slope of spontaneous baroreflex in healthy subjects. Physiol Res 47: 227-35, 1998, and Cooper VL, Hainsworth R. Carotid baroreceptor reflexes in humans during orthostatic stress. Exp Physiol 86: 677-81, 2001, and Wang, PhD thesis 2000). In autonomic neuropathy, such as that of diabetes, it has been shown that it is lower in the supine position and there is less further reduction on standing than in normal subjects (Zhang Y, Critchley LA, Tarn YH, Tomlinson B. Short-term postural reflexes in diabetic patients with autonomic dysfunction. Diabetologia 47: 304-11, 2004.) hi this study, baroreflex sensitivity in orthostatic hypotension and MOI groups were similar to those with a mild autonomic neuropathy. We have shown that stimulation significantly raises sensitivity in the sitting position and reduces the magnitude of reduction on standing. This suggests that tiie reversal of postural changes in blood pressure may be associated with increased sensitivity. Changes in dP/dt and Pulse Pressure

Systolic blood pressure changing rate (dP/dt) is a validated measure of cardiac contractility (Brinton TJ, Cotter B, Kailasam MT, Brown DL, Chio SS, O'Connor DT, DeMaria AN. Development and validation of a noninvasive method to determine arterial pressure and vascular compliance. Am J Cardiol 80: 323-30, 1997, and Germano G, Angotti S, Muscolo M, D'Auria F, Giordano M. The (dP/dt)max derived from arterial pulse waveforms during 24 h blood pressure oscillometry recording. Blood Press Monit 3: 213-216, 1998). In subject #1, the synchronous fall in dP/dt with systolic blood pressure in the 'off state, reflects the fact that he did not exhibit a normal increased cardiac contractility on standing (Lewis BS, Lewis N, Gotsman MS. Effect of standing and squatting on^" echocardiographic left ventricular function. Eur J Cardiol 11: 405-12, 1980). This was altered with stimulation. In the MOI group, the raised baseline value of dP/dt with stimulation that did not change significantly on standing implies that cardiac function has been improved in this group.

Pulse pressure is related to peripheral vasoconstriction (Laskey WK, Parker HG Ferrari VA Kussmaul WG Noordergraaf A. Estimation of total systemic arterial compliance in humans. J Appl Physiol 69 (1): 112-9. 1990). Our findings of increased pulse pressure (or reduction in the fall on standing) suggest that stimulation increases peripheral vasogenic tone. As this is a sympathetic effect, this adds further evidence that stimulation is acting via an up regulation of the peripheral sympathetic nervous system.

PAG connections to other baroregulatory centers In animals, the dorsal PAG is known to play a role in regulating the arterial baroreflex (Inui K, Nosaka S. Target site of inhibition mediated by midbrain periaqueductal gray matter of baroreflex vagal bradycardia. J Neurophysiol 70: 2205-14, 1993). The PAG is also linked to the 'exercise pressor reflex' which involves the cardiovascular response to exercise (Iwamoto GA, Wappel SM, Fox GM, Buetow KA, Waldrop TG. Identification of diencephalic and brainstem, cardiorespiratory areas activated during exercise. Brain Res 726: 109-122, 1996, and Kramer JM, Jarboe MO, Waldrop TG. Periaqueductal gray neuronal responses to hindlimb muscle contraction in the cat. Soc Neurosci Abstr 22: 89, 1996). It has therefore been suggested that the PAG acts as a site of integration of these reflexes. It is likely that the changes we have seen are due to the influence of the PAG on other baroregulatory centers in the brainstem. These include the nucleus of the tractus solitarius (NTS) and the rostroventerolateral medulla (RVLM). For example, it has been shown that the NTS receives excitatory convergence of PAG and somatic afferents involved in the baroreceptor reflex (Boscan P, Paton JFR. Excitatory convergence of periaqueductal gray and somatic afferents in the solitary tract nucleus: role for neurokinin 1 receptors. Am J Physiol Integr Comp Physiol 288: R262- R269, 2005). It has also been shown that baroreceptor sensitive neurons in the RVLM can be activated by PAG stimulation (Van Der Plas J, Maes FW, Bohus B. Electrophysiological Analysis of Midbrain Periaqueductal Gray Influence on Cardiovascular Neurons in the Ventrolateral Medulla Oblongata. Brain Res Bull 38 (5): 447-456, 1995). Conversely, barosensitive RVLM adrenergic neurons project to PAG and therefore do not solely provide an excitatory drive to sympathetic preganglionic neurons (Haselton JR, Guyenet PG. Ascending collaterals of medullary barosensitive neurons and Cl cells in rats. Am J Physiol 258: Rl 051-63, 1990). Electrical stimulation of PAG in this study is therefore likely to be influencing one stage in this complex circuit.

Summary

Li summary, we have found that stimulation of the PVG/PAG can prevent the postural drop in blood pressure in humans. It appears to work via an up regulation of both the cardiac and the peripheral sympathetic nervous systems and an increase in baroreceptor sensitivity. Furthermore, its effects on postural hypotension are not accompanied by any adverse effects or dangerous increases in resting blood pressure. We did notice autonomic side-effects such as sweating in two patients when higher voltages were used, but not at the amplitudes used to influence postural changes in blood pressure. Other effects such as changes in bladder activity were not noted by the patients, although these were not formally tested. Deep brain stimulation has been used largely in the treatment of movement disorders, and more recently for depression (Abelson JL, Curtis GC, Sagher O, Albucher RC, Harrigan M, Taylor SF, Martis B, Giordani B. Deep brain stimulation for refractory obsessive-compulsive disorder. Biol Psychiatry 57: 510-6, 2005, Bittar RG, Yianni J, Wang S, Liu X, Nandi D, Joint C, Scott R, Bain PG, Gregory R, Stein J, Aziz TZ. Deep brain stimulation for generalised dystonia and spasmodic torticollis. J Clin Neurosci 12(1): 12-16, 2005, and Parkin S, Nandi D, Giladi N, Joint C, Gregory R, Bain P, Scott R, Aziz TZ. Lesioning the subthalamic nucleus in the treatment of Parkinson's disease. Stereotact Funct Neurosurg 77: 68-72, 2001). Although we have not specifically used deep brain stimulation for the treatment of postural hypotension to date, a great advantage would be that the stimulation could be stopped when in the supine position, preventing the nocturnal hypertension that these patients suffer from (Kaplan NM. The promises and perils of treating the elderly hypertensive. Am J Med Sci 305: 183-97, 1993).

EXAMPLE 3: TREATMENT FOR HYPERTENSION We report a case of a 61 year old hypertensive man who underwent deep brain stimulation of the periventricular/ periaqueductal grey area for the relief of chronic neuropathic pain affecting his oral cavity and soft palate. During intraoperative stimulation, we were able to modulate his blood pressure up or down, depending on electrode location.

Introduction

The periaqueductal gray projects to all medullary regions that control blood pressure, as well as having reciprocal connections with higher centers (Shipley, M. T. Ennis M. Rivzi T. A. Behbehani M. M. Topographical specificity of forebrain in the periaqueductal grey and inputs to the midbrain periaqueductal grey: Evidence for discrete longitudinally organised input columns. In: Depaulis, A. Bandler R., eds. The Midbrain Periaqueductal Gray Matter: Functional, Anatomical and Neurochemical Organization. New York, Plenum Press, 1991 :417-448, M'hamed SB, Sequeira H, Poulain P, Bennis M, Roy JC. Sensorimotor cortex projections to the ventrolateral and the dorsomedial medulla oblongata in the rat. Neurosci Lett 1993;164:195-8, Barman SM. Descending projections of hypothalamic neurons with sympathetic nerve-related activity. J Neurophysiol 1990;64:1019-32, Guyenet PG, Haselton JR, Sun MK. Sympathoexcitatory neurons of the rostroventrolateral medulla and the origin of the sympathetic vasomotor tone. Prog Brain Res 1989;81 :105-16). Here we show reversal of hypertension with electrical stimulation of this area in an awake patient.

Case Report

A 61 year old gentleman presented with a five-year history of intractable neuropathic pain affecting the right side of his soft palate, oral cavity and lateral side of the tongue. He described this as a constant 'searing' pain that gradually worsened throughout the day. It was unrelated to any dental problems, trauma, or infection, and was unaffected by swallowing or movement of his jaw. A range of investigations, including Magnetic Resonance scans of his neck and brain, proved normal. He had tried numerous medications including acetaminophen, opioids, antiepileptics and gabapentin, all to no avail. He had also undergone local anaesthetic injections to his inferior alveolar nerve and two radiofrequency lesions of his right Gasserian ganglion. After these failed to cure his pain, he had a motor cortex stimulator inserted (and re-adjusted after it did not work) and this also failed. After considering his case, we decided to perform deep brain stimulation of the periaqueductal gray area and the ventroposteromedial (VPM) nucleus of the thalamus, areas that have both been successfully used in the treatment of facial pain. Green A.L. Owen S.L.F., Davies P, Moir L, Aziz TZ. Deep Brain Stimulation for Neuropathic Cephalalgia. Cephalalgia 2006; In press (doi: 10.111 l/j.1468- 2982.2005.01068.x) Despite medication, on admission, the patient's mean blood pressure was

162/93 mmHg over three days. His previous mean blood pressure off-medication was 205/110 mmHg. On the morning of operation, he took Atenolol 50 mg, Perindopril 8 mg and Felodipine 5 mg as usual.

Our surgery for deep brain stimulation is performed with the patient awake and has been described in detail elsewhere (Bittar RG, Burn SC, Bain PG, et al., Deep brain stimulation for movement disorders and pain. J Clin Neurosci 2005 ; 12:457-463). After permission from the local ethics committee and patient' s consent, we intra-operatively recorded the patient's arterial blood pressure during the procedure. The arterial pressure signal was transduced using a Medex Medical® transducer and recorded from the anesthetic machine (AS/3® Datex-Ohmeda inc., Tewksbury, MA), sampled at 500 Hz with 12-bit resolution (MP100®, Biopac Systems, Santa Barbara, Ca) using Acqknowledge® software (version 3.7.3, Biopac systems).

Results

With the patient awake and resting on the operating table (his head secured to the table in a stereotactic frame), his mean baseline blood pressure was 157.4/87.6 mmHg. Insertion of the electrode to target had no effect on the blood pressure. However, when the periaqueductal gray was electrically stimulated at 30 Hz and at an amplitude of 2 v, the arterial blood pressure fell to a mean of 132.4/79.2 mmHg at the end of a 110 second period of stimulation (Fig. 12A). The latency between stimulation and initial reduction was less than five seconds but it took approximately thirty seconds for full effect. When stimulation was turned off, the blood pressure promptly increased, returning to its original value after twenty seconds. This stimulation was repeated three times to confirm the effect. The electrode was then advanced 3 mm in order to find the position for optimal pain relief. Stimulation at the same parameters had the opposite effect (Fig. 12B) i.e. the blood pressure increased (consistently on three occasions) to a mean of 179 mmHg. Interestingly, in this position, pain relief was not so good. The electrode was therefore withdrawn 3mm before being secured in its final position (Figs 12C, 12D). Stimulation of the thalamic electrode had no effect on blood pressure, although it did provide pain relief.

Discussion

Arterial blood pressure can be increased or decreased by electrically stimulating the periaqueductal gray area in the human[9]. Moreover, the direction of change of blood pressure depends on whether the electrode is in the ventral or dorsal part of the nucleus. Young found that intraoperative blood pressure changes were common with PAG stimulation [10]. This case illustrates that it is possible to reduce blood pressure with electrical stimulation of this area in a hypertensive patient.

The medulla contains the major nuclei that control heart rate, blood pressure and respiration. Information from the peripheral arterial and cardiopulmonary baroreceptors and chemoreceptors passes, via the glossopharyngeal and vagus nerves, to the caudal nucleus of the tractus solitarius (NTS)[I I]. From here, impulses are transmitted to the nucleus ambiguous (NA), the dorsal motor nucleus of the vagus nerve (DMNV) and then to the rostral and caudal ventrolateral medulla (RVLM/ CVLM)[12;13]. The parasympathetic system is activated by the NA and DMNV which contain parasympathetic preganglionic neurones that alter contractility of the heart and heart rate via the vagus. The sympathetic system is activated by the CVLM whose neurones synapse on the sympathetic pre-ganglionic intermediate lateral (IML) neurones that innervate blood vessels and the adrenal medulla and, via catecholamine release, alter the basal tone of blood vessels (Smith JE, Jansen AS, Gilbey MP, Loewy AD. CNS cell groups projecting to sympathetic outflow of tail artery: neural circuits involved in heat loss in the rat. Brain Res 1998; 786: 153-64., and Malhotra V, Kachroo A, Sapru HN. Role of alpha 1 -adrenergic receptors in the intermediolateral column in mediating the pressor responses elicited by the stimulation of ventrolateral medullary pressor area. Brain Res 1993; 626: 278-86.) Although this is an oversimplification and there is considerable interaction between these autonomic medullary centres, this mechanism is believed to be a reflex to maintain cardiovascular homeostasis.

In 1953, Kabat showed that PAG stimulation can alter blood pressure (Kabat H, Magoun HW Ranson SW. Electrical stimulation of Points in the forebrain and midbrain. The resultant alteration in blood pressure. Arch Neurol Psych 34; 931- 955). Evidence for an anatomical substrate for this 'higher' control is abundant. For example, serotonergic and adrenergic sympathetic pathways project to the rostroventromedial medulla (Farkas E, Jansen AS, Loewy AD. Periaqueductal gray matter input to cardiac-related sympathetic premotor neurons. Brain Res 1998; 792: 179-92, and Cameron AA, Khan IA, Westlund KN, Cliffer KD, Willis WD. The efferent projections of the periaqueductal gray in the rat: a Phaseolus vulgaris- leucoagglutinin study. I. Ascending projections. J Comp Neurol 1995; 351 : 568-84), the rostroventrolateral medulla, locus coeruleus (Farkas E, Jansen AS, Loewy AD. Periaqueductal gray matter projection to vagal preganglionic neurons and the nucleus tractus solitarius. Brain Res 1997; 764: 257-61) and pontobulbar reticular formation (Odeh F AM. The projections of the midbrain periaqueductal grey to the pons and medulla oblongata in rats. Eur J Neurosci 2001; 14: 1275-1286), amongst others. PAG neurones also project to cardiac vagal preganglionic neurones in the nucleus ambiguus, dorsal motor vagal nucleus and the nucleus of the tractus solitarius Farkas E, Jansen AS, Loewy AD. Periaqueductal gray matter projection to vagal preganglionic neurons and the nucleus tractus solitarius. Brain Res 1997; 764: 257-61.

The PAG also has reciprocal connections with a number of higher centres. Many of these centres are involved in cardiovascular regulation. Of particular note is the hypothalamus as nearly all its nuclei influence blood pressure and heart rate. Many of the descending connections from hypothalamus pass through and are influenced by PAG. For example, the hypotensive response elicited by lateral hypothalamus stimulation can be attenuated by lidocaine injection into PAG (Pajolla GP, Tavares RF, Pelosi GG, Correa FM. Involvement of the periaqueductal gray in the hypotensive response evoked by L-glutamate microinjection in the lateral hypothalamus of unanesthetized rats. Auton Neurosci 2005; 122: 84-93). In addition, the PAG has direct reciprocal connection with amygdala, prefrontal cortex and insular cortex (Rizvi TA, Ennis M, Behbehani MM, Shipley MT. Connections between the central nucleus of the amygdala and the midbrain periaqueductal gray: topography and reciprocity. J Comp Neurol 1991; 303: 121-31, and Bandler R, Keay KA, Floyd N, Price J. Central circuits mediating patterned autonomic activity during active vs. passive emotional coping. Brain Res Bull 2000; 53: 95-104). Thus, the role of PAG in autonomic regulation is probably as an 'integrator' of emotional/ higher influence on cardiovascular control, similar to its role in pain control.

This treatment of hypertension may have important implications for possible future therapies. Refractory hypertension i.e. hypertension that persists despite all available medical therapy, affects approximately 3 % of hypertensive patients (Alderman MH, Budner N, Cohen H, Lamport B, Ooi WL. Prevalence of drug resistant hypertension. Hypertension 1988;11 :II71-5) and can lead to stroke (Almgren T, Persson B, Wilhelmsen L, Rosengren A, Andersson OK. Stroke and coronary heart disease in treated hypertension — a prospective cohort study over three decades. J Intern Med 2005; 257:496-502) and cardiovascular disease such as myocardial infarction (Wang JG, Staessen JA, Franklin SS, Fagard R, Gueyffier F. Systolic and diastolic blood pressure lowering as determinants of cardiovascular outcome. Hypertension 2005;45:907-13). Some estimates have suggested that as many as 27 % of hypertensives do not have their disease well controlled (McBride W, Ferrario C, LyIe PA. Hypertension and medical informatics. J Natl Med Assoc 2003; 95:1048-56). However, there are a number of issues that need to be resolved regarding effects of periaqueductal gray stimulation on blood pressure. For example, the changes we have shown are acute changes. For an antihypertensive treatment to be beneficial, we need to demonstrate long-term effect. Secondly, in our experience, deep brain stimulation presents a 0.3 % risk of stroke as a direct result of the procedure (this compares to the experience of other groups (Lyons KE, Wilkinson SB, Overman J,Pahwa R. Surgical and hardware complications of subthalamic stimulation: a series of 160 procedures. Neurology 2004;63:612-6). If we are considering this as a treatment to prevent stroke, these risks need to be reduced. Thirdly, deep brain stimulation is an expensive treatment (Yianni J., Green AX, McMosh E,et al. The Costs and Benefits of Deep Brain Stimulation Surgery for Patients with Dystonia: An Initial Exploration. Neuromodulation 2005;8:155-161), particularly in comparison to antihypertensive medication.

EXAMPLE 4: BLOOD PRESSURE IN HUMANS Introduction

The defence reaction in the rat is an integrated response that is associated with survival in the wild. For example, if escape from danger is possible, the response involves a 'fight or flight' reaction that includes raised blood pressure and heart rate, non-opioid mediated analgesia and emotional effects such as fear McGaraughty, S., Fair, D. A., and Heinricher, M.M. Lesions of the periaqueductal gray disrupt input to the rostral ventromedial medulla following microinjections of morphine into the medial or basolateral nuclei of the amygdala. Brain Res 1009, 223-7 (2004), and Carrive, P. and Bandler, R. Control of extracranial and hindlimb blood flow by the midbrain periaqueductal grey of the cat. Exp Brain Res 84, 599- 606 (1991). Conversely, if escape is not possible and it is safer to remain undetected, the reaction consists of lowered blood pressure, opioid-mediated analgesia and 'freezing' behaviour as well as fear (Johnson, P.L., Lightman, S.L., and Lowry, CA. A functional subset of serotonergic neurons in the rat ventrolateral periaqueductal gray implicated in the inhibition of sympathoexcitation and panic. Ann N Y Acad Sci 1018, 58-64 (2004), and Finnegan, T.F., Chen, S.R., and Pan, HX. Effect of the {mu} opioid on excitatory and inhibitory synaptic inputs to periaqueductal gray-projecting neurons in the amygdala. J Pharmacol Exp Ther 312, 441-8 (2005). Other components of the defense reaction include vocalisation, pupillary changes, micturition and changes in skeletal blood flow (Bittencourt, A.S., Carobrez, A.P., Zamprogno, L.P., Tufik, S.^and Schenberg, L.C. Organization of single components of defensive behaviors within distinct columns of periaqueductal gray matter of the rat: role of N-methyl-D-aspartic acid glutamate receptors. Neuroscience 125, 71-89 (2004).) An important area involved in the defense reaction is the periaqueductal gray matter (PAG). This area is organised into longitudinal columns that are functionally distinct and opposite (Carrive, P., Bandler, R., and Dampney, R. A. Viscerotopic control of regional vascular beds by discrete groups of neurons within the midbrain periaqueductal gray. Brain Res 493, 385-90 (1989). Activation of the dorsomedial and dorsolateral columns evokes the 'fight or flight' response and activation of the lateral and ventrolateral columns produces the passive coping responses described above (Schenberg, L.C. et al. Functional specializations within the tectum defense systems of the rat. Neurosci Biobehav Rev (2005)).

Serotonergic and adrenergic sympathetic pathways project to the rostroventromedial medulla (Farkas, E., Jansen, A.S., and Loewy, A.D.

Periaqueductal gray matter input to cardiac-related sympathetic premotor neurons. Brain Res 792, 179-92 (1998), Cameron, A.A., Khan, LA., Westlund, K.N., and Willis, W.D. The efferent projections of the periaqueductal gray in the rat: a Phaseolus vulgaris-leucoagglutinin study. IL Descending projections. J Comp Neurol 351, 585-601 (1995), and Cameron, A. A., Khan, I.A., Westlund, K.N., Cliffer, K.D., and Willis, W.D. The efferent projections of the periaqueductal gray in the rat: a Phaseolus vulgaris-leucoagglutinin study. I. Ascending projections. J Comp Neurol 351, 568-84 (1995)) the rostroventrolateral medulla, locus coeruleus (Farkas, E., Jansen, A.S., and Loewy, A.D. Periaqueductal gray matter input to cardiac-related sympathetic premotor neurons. Brain Res 792, 179-92 (1998)) and pontobulbar reticular formation (Odeh F, A.M. The projections of the midbrain periaqueductal grey to the pons and medulla oblongata in rats. Eur J Neurosci 14, 1275-1286 (2001)) amongst others. PAG neurones also project to cardiac vagal preganglionic neurones in the nucleus ambiguus, dorsal motor vagal nucleus and the nucleus of the tractus solitarius (Farkas, E., Jansen, A.S., and Loewy, A.D. Periaqueductal gray matter projection to vagal preganglionic neurons and the nucleus tractus solitarius. Brain Res 764, 257-61 (1997).

In this study, we show, using intraoperative arterial pressure recordings in awake humans, that electrical stimulation of the PVG/PAG can alter blood pressure, even in a hypertensive patient. In contrast to our previous studies, we have found that it is possible to alter heart rate as well as other cardiovascular parameters. We have performed power spectral analysis of heart .rate in order to elucidate the underlying sympathetic/ parasympathetic components of these changes.

Methods Six consecutive patients (all male) undergoing deep brain stimulation of the

PAG/ PVG were recruited to the study. Patient demographics are shown in Table 5. AU suffered from chronic neuropathic pain ranging from phantom limb pain to severe orofacial pain of unknown aetiology. Mean age was 49.7 years. One patient (#1) was hypertensive, although poorly controlled, despite polypharmacy. The remaining five patients were normotensive. The study was approved by the local regional ethics committee and informed consent for participation was obtained from all patients. Four patients entered the 'awake' arm of the study, and two entered the 'general anaesthetic' arm (see below).

TABLE 5

Surgical Technique

Patients had a Tl -weighted axial Magnetic Resonance Image (MRI) brain scan prior to the day of surgery. Slice thickness was 2.9mm. On the day of surgery, the Cosman-Roberts- Wells™ stereotactic frame was applied to the patients' scalp under local anaesthetic (1 OmIs 0.25% bupivicaine mixed with 1 OmIs 2% lidocaine with 1 :200000 adrenaline). A stereotactic Computed Tomography (CT) scan was then performed (1.9mm slice thickness, zero gantry tilt,). The pre-operative MRI was fused to the stereotactic CT using the Radionics Image Fusion™ and Stereoplan™ software (Radionics, Burlington, MA). The anterior and posterior commissures were identified on the axial images. The intended target for placing the deepest electrode contact was marked at the PAG at a level of < 10mm below the AC-PC line; between the dorsal part of the red nucleus and the superior colliculus in the AP plane; and approximately 3mm lateral to the lateral boundary of the aqueduct and the third ventricle. Next, the electrode trajectory was selected to avoid possible penetration of the surface vessels on the cortex and the lateral ventricle. This leads to some adjustment of the target localisation for each individual patient, and likely contributes to inter-patient variation in electrode placement. In patients with post- stroke pain and severely deformed hemispheres, targeting can be quite difficult. For these patients, relative anatomic landmarks such as the third ventricle, the aqueduct, the red nucleus and the superior colliculus are more reliable than AC-PC measurements.

After preparing the patient's scalp with cetrimide® and alcoholic chlorhexidine® solution, 2OmIs of local anaesthetic (as above) was injected into the scalp on the side of targeting. A curved incision was made over the coronal suture, with the base of the flap posteriorly. Next, a 2.7mm twist-drill craniostomy was performed, passing through skull and dura. A Radionics™ electrode of 1.8 mm diameter and 2.0 mm exposed tip was slowly passed towards the target while the impedance values were monitored for a sudden drop in impedance value from 500 - 600? to under a few tens of? , suggesting possible penetration of a ventricle. The Radionics™ electrode was replaced by a Medtronic 3387® electrode (Medtronic Inc., Minneapolis, USA). In our usual procedure, test stimuli (<5.0V in amplitude, 120μs in pulse width and 10 - 50Hz in frequency) are applied 2-3mm above target (i.e. rostral to target), then at target, and beyond, if necessary, until we have achieved the desired clinical effect. This is usually a warm feeling or paraesthesiae in the area of pain or pain suppression, and sometimes abnormal eye movements.

In this study, our experimental protocol (see below) was carried out at each 'target' site that was clinically tested i.e. the number of sites tested was dependent on the procedure that was carried out. Once clinical efficacy (and the experimental recordings) had been established, the DBS electrode was fixed onto the skull with a titanium bioplate™ and externalised for further investigation. The wounds were then sutured. If the patient had experienced good clinical effect, the whole stimulation system was internalised one week later in a second procedure under general anaesthetic as follows; the patient was induced with a sleep dose of propofol, and anesthesia was maintained with nitrous oxide in oxygen, midazolam, end tidal concentration sevoflurane 0.5-1% and fentanyl. The extension leads (Medtronic) and an implantable pulse generator (IPG - kinetra™, Medtronic) that is placed in a subcutaneous pocket in the chest wall or abdomen were inserted. The GA recordings (two patients) were performed at the end of this second procedure. As the electrode is fixed, these only involved one target site.

Experimental Protocol

The experimental paradigm was as follows; 300s recordings were made; 100s before stimulation, 100s of stimulation at the highest amplitude tolerated (3- 4.5v) followed by 100s after stimulation. This was repeated two more times at each 'target' with 3 minutes of rest (no stimulation) in between each recording. The mean changes in all cardiovascular parameters was taken as the mean of each parameter for the last 30 seconds of stimulation across all three recordings, to take into account any latency of effect. For the GA recordings, after 10 minutes of 'rest' at the end of the surgical procedure, three separate three-minute recordings were made with stimulation on, separated by three minutes of stimulation 'off in between.

For intraoperative recording of arterial blood pressure, an intra-arterial catheter (BD Angiocath™, Infusion Therapy Systems Inc, Sandy, Utah) was inserted into the radial artery prior to either the DBS insertion or the implantable pulse generator (IPG) insertion. The arterial pressure signal was transduced using a Medex Medical® transducer and recorded from the anaesthetic machine (AS/3® Datex-

Ohmeda inc., Tewksbury, MA), sampled at 500Hz with 12-bit resolution (MP 100®, Biopac Systems, Santa Barbara, Ca) using Acqknowledge® software (version 3.7.3, Biopac systems).

The pulse pressure and RR interval were calculated from the measurements of SBP/DBP and ECG respectively. The blood pressure changing rate, dP/dt, was derived by differentiating the blood pressure. The maximum dP/dt (maximum slope of the blood pressure curve) was then extracted.

Signals Processing and Statistical Analysis Cardiovascular variables during stimulation were compared to resting periods using one way analysis of variance of BP etc with time (ANOVA). The results represent the average of the changes over three stimulation periods (at each target) for each period. Significance was taken as p<0.05. All changes quoted in the results are expressed ± one standard deviation of the mean.

For all data segments, auto-regressive power spectra analysis of SBP was performed on both the resting and stimulation-on components. Frequencies below 0.02 Hz were filtered out to remove the trend in the signal (see Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation 93, 1043-65 (1996) for methodology). The power of the low and high frequency components were computed as the integral of the power spectra between 0.05 and 0.15Hz and between 0.15 and 0.4Hz. Again, these changes were averaged across three sessions and significance between stimulation and resting periods was determined using a paired t-test. AU signal processing was performed in Matlab^® (Version 6.1, MathWorks Inc., Natick, Ma., USA) and statistical analysis was performed in SPSS (Version 11, SPSS Inc, IL, USA).

Results

Subject #1 and #2 had a significant decrease in ABP when the first (most proximal) location was stimulated, followed by increase in ABP when the second (more caudal) target was stimulated. Subjects #3 and #4 had decrease or increase in

ABP only. Regarding the two patients tested under GA, one had a significant increase in ABP, and one had no change. Examples of raw data from each patient are shown in Fig. 13. To best illustrate changes in cardiovascular variables, two case examples are shown in more detail;

Patient #1 (Chronic oral/facial pain)

The effect of stimulation at 2 locations was examined: the 'planned' target and 3mm below;

At 'target', the mean change in SBP was a decrease with stimulation from 159.3±2.6 mmHg to 138.9±4.5 mniHg, a decrease of 12.8% (PO.01). This reversed after stimulation, rising to 157.8±4.7 mmHg. DBP decreased with stimulation, falling from 87.2±1.8 mmHg to 82.2±3.6 mmHg, a decrease of 5.7%. It did not recover, however (P=0.32). The RR interval fell with stimulation from 1.14 ± 0.06s to 1.06 ± 0.02s-a fall of 7%-rising again after stimulation to 1.15 ± 0.12s (ρ<0.05).

2mm beyond the planned target, the mean change in SBP was an increase of 7.8% with stimulation from 168.7 ± 1.69 mmHg to 181.9 ± 5.28 mmHg (p<0.01). After stimulation it fell, but did not attain the pre-stimulation value (mean 176.6 ± 2.54 mmHg). DBP increased to a lesser extent with stimulation, by 2.9%, from 96.0 ± 1.55 mmHg to 98.9 ± 2.13 mmHg (ρ<0.01). After stimulation DBP fell back to the pre-stimulation value: 95.6 ± 1.76 mmHg. RR interval did not change (1.07 to 1.068s). PP and dP/dt both increased significantly with stimulation (15.4mmHg and 21 OmmHg^s respectively).

An explanation for these differential effects on blood pressure is likely to be that the advancement of the electrode changed the stimulation to a different part of the PAG. Analysis of the Power Spectrum of RR interval showed that the low frequency greatly increased with a rise in ABP, consistent with an increase in Meyer's wave (see discussion). These results are summarized in Fig. 14.

Patient #2 (post-stroke pain)

As with patient #1, two electrode locations were stimulated. At the proximal location, SBP significantly reduced from 136.4±7 mmHg to 128±8.8 mmHg, a fall of 5.6% (p<0.05). DBP reduced from 74.5±2.4 mmHg to 72.4+3.6 mmHg (2.8%) although this did not reach significance. RR interval did not significantly change. dP/dt reduced from 1018 mmHgVs to 908 mmHg^s (p=0.06).

When the distal location was stimulated (3mm below target), BP consistently increased. SBP increased from 136.3±8.9 mmHg to 162.3±5.6 mmHg, an increase of 19.1% (ρθ.001). DBP and dP/dt also significantly increased (35.0%, 45% respectively), but PP did not significantly change. In contrast to most other patients, RR interval decreased significantly with SBP rise, from 0.73±0.01 s to 0.48±0.03 s, representing an increase in pulse rate from 82 bpm to 125 bpm. Power spectral analysis of heart rate showed an increase in Meyer's wave accompanying the rise in ABP. A selection of these results are summarized in Fig. 15.

Overall Results Fig. 16 shows the results (normalised) for the four awake and two anaesthetized patients. Generally, changes in SBP were either accompanied by large changes in pulse pressure or RR interval. For reduction in BP, two out of three patients had large reductions in pulse pressure and reduction in dPdt, with a corresponding decrease in the low frequency component of the power spectrum of heart rate variability (#1 and #2). The third patient (#3) had an increase in RR interval and no change in pulse pressure. The low frequency component reduced, but the high frequency component increased (by 2.82 times - not shown in graph). This increase in the HF:LF ratio is consistent with an increase in parasympathetic activity and is consistent with the increase in RR interval (see discussion).

Similar to the decreases in BP, increases were associated with increased pulse pressure and dPdt, or decreased RR interval (two out of four). These sympathetically mediated effects were all associated with an increase in the overall power of HRV, but particularly the LF component. As expected, patient #5, who had no change in BP, had no significant change in any of the variables with stimulation.

Discussion

PAG is organised into four longitudinal columns (Carrive, P. and Bandler, R. Control of extracranial and hindlimb blood flow by the midbrain periaqueductal grey of the cat. Exp Brain Res 84, 599-606 (1991), Carrive, P., Bandler, R., and Dampney, R. A. Somatic and autonomic integration in the midbrain of the unanesthetized decerebrate cat: a distinctive pattern evoked by excitation of neurones in the subtentorial portion of the midbrain periaqueductal grey. Brain Res 483, 251-8 (1989), Carrive, P. and Bandler, R. Viscerotopic organization of neurons subserving hypotensive reactions within the midbrain periaqueductal grey: a correlative functional and anatomical study. Brain Res 541, 206-15 (1991)). Stimulation of the dorsomedial and dorsolateral columns produces an increase in BP whereas stimulation of the lateral and ventrolateral columns produces hypotension and freezing behaviour (Carrive, P. and Bandler, R. Control of extracranial and hindlimb blood flow by the midbrain periaqueductal grey of the cat. Exp Brain Res 84, 599-606 (1991)., ABRAHAMS, V.C., HILTON, S.M., and ZBROZYNA, A. Active muscle vasodilatation produced by stimulation of the brain stem: its signifϊcance in the defence reaction. J Physiol 154, 491-513 (1960), Duggan, A. W. and Morton, CR. Periaqueductal grey stimulation: an association between selective inhibition of dorsal horn neurones and changes in peripheral circulation. Pain 15, . 237-48 (1983), Lovick, T.A. Inhibitory modulation of the cardiovascular defence response by the ventrolateral periaqueductal grey matter in rats. Exp Brain Res 89, 133-9 (1992), and Lovick, T.A. Ventrolateral medullary lesions block the antinociceptive and cardiovascular responses elicited by stimulating the dorsal periaqueductal grey matter in rats. Pain 21, 241-52 (1985)). These cardiovascular changes are part of the 'defence reaction' that an animal uses to improve survival (HUNSPERGER, R. W. [Affective reaction from electric stimulation of brain stem in cats.]. HeIv Physiol Pharmacol Acta 14, 70-92 (1956)). Non-invasive recordings, that PAG stimulation can influence blood pressure in humans, see supra. As shown in the results described herein, intra-arterial recordings show that stimulation can, in some cases, alter RR interval. This may be due to the greater changes in ABP in this study, and also the 'cleaner' traces that are provided by a direct ABP recording. As well as changes in SBP, we have found corresponding analogous changes in DBP, pulse pressure and maximum dP/dt. Where RR interval changes did occur, changes in PP tended to be less (see Fig. 16). In general, these changes suggest that the mechanism of BP change is a mixture of increased/decreased myocardial contractility (change in dP/dt) (Brinton, TJ. et al. Development and validation of a noninvasive method to determine arterial pressure and vascular compliance. Am J Cardiol 80, 323-30 (1997)) change in total peripheral resistance (changes in pulse pressure) (Laskey WK, Parker HG Ferrari VA Kussmaul WG Noordergraaf A. Estimation of total systemic arterial compliance in humans. J Appl Physiol 69 (1), 112-9. 90.) and, in some cases, change in RR interval. The first two of these are sympathetically mediated, whereas the latter could be due to a change in parasympathetic activity. Our finding that the low frequency component (0.023 to 0.15 Hz) of HRV significantly changed in the direction of BP change, provides strong evidence for an underlying sympathetic mechanism, as there is much evidence that this part of the power spectrum is related to sympathetic activity (Berger, R.D., Saul, J.P., and Cohen, RJ. Transfer function analysis of autonomic regulation. I. Canine atrial rate response. Am J Physiol 256, H142-52 (1989), Saul, J.P. et al. Transfer function analysis of the circulation: unique insights into cardiovascular regulation. Am J Physiol 261, H1231-45 (1991)).

Also, the ratio of low frequency: high frequency (0.15 to 0.4Hz) is important, as this has been shown to be an indicator of the balance between the sympathetic and parasympathetic nervous systems (Accurso V, S. A. S. V. Rhythms, rhymes and reasons - spectral oscillations in neural cardiovascular control. Autonomic Neuroscience: Basic and Clinical 90, 41-46 (2001)). In the majority of patients, the LF:HF ratio reduced with reduction in blood pressure, implying a greater fall in sympathetic activity. However, in one patient (#3), there was actually an increase in HF power, despite a fall in LF and fall in SBP. Notably, this was the one patient with a significant increase in RR interval, a change associated with increased parasympathetic activity. In summary, it appears that in the majority of cases we are altering sympathetic activity, but it may also be possible to alter parasympathetic activity. There are some important differences between our findings and those in animals. For example, the latency between stimulus and peak response in animals has been consistently reported at around 5-20 seconds (Kabat H, Magoun HW Ranson SW. Electrical stimulation of Points in the forebrain and midbrain. The resultant alteration in blood pressure. Arch Neurol Psych 34, 931-955, ABRAHAMS, V.C., HILTON, S.M., and ZBROZYNA, A. Active muscle vasodilatation produced by stimulation of the brain stem: its significance in the defence reaction. J Physiol 154, 491-513 (1960), LINDGREN, P. and UVNAS, B. Activation of sympathetic vasodilator and vasoconstrictor neurons by electric stimulation in the medulla of the dog and cat. Circ Res 1, 479-85 (1953)). One possible reason is that we have activated a hormonal mechanism such as that shown by Bamshad (Bamshad, M. and Albers, H.E. Neural circuitry controlling vasopressin-stimulated scent marking in Syrian hamsters (Mesocricetus auratus). J Comp Neurol 369, 252-63 (1996)) in which PAG stimulation caused a release in arginine vasopressin via a rostral projection to the hypothalamus. Other ways in which this study differs from animal experiments, that might explain the latency of response include the fact that our subjects were awake, as opposed to the anaesthetised, decerebrate animals that would not have had the influence of higher brain centres such as prefrontal cortex that has been shown to inhibit cardiovascular responses in animals (Zbrozyna AW, W.D. Stimulation in prefrontal cortex inhibits conditioned increase in blood pressure and avoidance bar pressing in rats. Physiol Behav 49, 705-8 (1991)). Also, our methodology differs in that we stimulated continuously for several minutes, rather than giving a 'pulse' of stimulation for 5-10 seconds.

In summary, we have shown that electrical stimulation of different regions of the human PVG/PAG can selectively modulate blood pressure, mainly via a sympathetic effect. These findings are important because demonstrating these changes in humans provides the gold standard by which future therapies can be targeted. We have shown that we can reduce blood pressure, even in hypertensive patients (patient #1). m the future, perhaps hypertension or indeed postural hypotension may be controlled by manipulation of this area.

EXAMPLE 5: STIMULATING THE HUMAN MIDBRAIN Introduction hi 1884 William James suggested that pain sensations are at least partly due to autonomic reactions changing local blood flow and blood pressure (James W, What is an Emotion? Mind 1884; 9:188-205). Craig (Craig AD. How do you feel? Interoception: the sense of the physiological condition of the body. Nat Rev Neurosci 2002;3:655-66) has argued that 'interoception' (sensation of the physiological condition of the body) should be defined as the sense of the physiological condition of the entire body, not just the viscera. This hypothesis questions the traditional notion that pain and temperature are simply somatosensory aspects of touch but are part of a wider homeostatic mechanism that integrates pain with emotions to provide a sense of 'how you feel'. The periaqueductal gray area (PAG) in the midbrain is an important area for both cardiovascular control and modulation of pain (Vanegas H, Schaible HG. Descending control of persistent pain: inhibitory or facilitatory? Brain Res Rev 2004;46:295-309, Green AL, Wang S, Owen SLF, Xie K, Liu X, Paterson DJ, Stein JF, Bain PG, Aziz TZ. Deep Brain Stimulation Can Regulate Arterial Blood Pressure in Awake Humans. Neuroreport,2005;16:1741-5, Young RF, Rinaldi PC. Brain Stimulation. In: North RB, Levy RM, editors. Neurosurgical Management of Pain, New York: Springer- Verlag, 1997. pp. 288-90, Fields HL, Vanegas H, Hentall ID, Zorman G. Evidence that disinhibition of brain stem neurones contributes to morphine analgesia. Nature 1983;306:684-6, and Hosobuchi Y. Subcortical electrical stimulation for control of intractable pain in humans. Report of 122 cases (1970-1984). J Neurosurg 1986;64:543-53.) In rats, stimulation of the ventral PAG is associated with reduction in arterial blood pressure (ABP) and analgesia (Johnson PL, Lightman SL, Lowry CA. A functional subset of serotonergic neurons in the rat ventrolateral periaqueductal gray implicated in the inhibition of sympathoexcitation and panic. Ann N Y Acad Sci 2004;1018:58-64) whereas dorsal stimulation is associated with increased ABP (Carrive P, Bandler R. Control of extracranial and hindlimb blood flow by the midbrain periaqueductal grey of the cat. Exp Brain Res 1991 ;84:599- 606). These midbrain sites are inextricably coupled and are part of the 'defence' reaction that an animal uses to combat danger (Reis DJ, Miura M, Weinbren M, Gunne LM. Brain catecholamines: relation to defense reaction evoked by acute brainstem transection in cat. Science; 156: 1768-70). Similarly, in the human, we have recently shown that blood pressure can be modulated (up or down) by stimulating the rostral PAG in patients with chronic pain (Green AL, Wang S, Owen SLF, Xie K, Liu X, Paterson DJ, Stein JF, Bain PG, Aziz TZ. Deep Brain Stimulation Can Regulate Arterial Blood Pressure in Awake Humans. Neuroreport, 2005;16:1741-5). Moreover, the direction of the change in blood pressure depends on whether the electrode is in dorsal or ventral PAG.

Although the precise relationship between pain and blood pressure has not been fully elucidated, that it is a complex relationship and is altered in chronic pain patients is well accepted. For example, in these patients, pain sensitivity (defined as the level of stimulus required to produce a subjective feeling of pain) positively correlates with blood pressure, the opposite of the relationship in pain-free individuals (Bruehl S, Chung OY, Ward P, Johnson B, McCubbin JA, The relationship between resting blood pressure and acute pain sensitivity in healthy normotensives and chronic back pain sufferers: the effects of opioid blockade. Pain 2002;100: 191-201). The aim of this study was to correlate blood pressure changes following electrical stimulation of the rostral PAG with changes in the patients' resulting sensations of pain. We have shown that the degree of analgesia induced by deep brain stimulation of the rostral PAG in man is related to the magnitude of reduction in ABP. We found that this relationship is linear and is due to reduced activity of the sympathetic nervous system. Thus stimulation of the PAG may partly control pain by reducing sympathetic activity as predicted by William James over a century ago (James W, What is an Emotion? Mind 1884; 9:188-205).

In this study, we used both visual analogue scores (VAS) and the McGiIl Pain Questionnaire (MPQ)TtO quantify and to give a qualitative measuϊ§ of the pain. The VAS method has been extensively used and validated (McCarthy et al., 2005). The MPQ is a widely used pain scoring method (Melzack R. The McGiIl Pain Questionnaire: major properties and scoring methods. Pain 1975; 1:277-99) that has been validated by other authors (Cohen MM, Tate RB. Using the McGiIl Pain Questionnaire to study common postoperative complications. Pain 1989;39(3):275- 9, and Mystakidou K, Parpa E, Tsilika E, Kalaidopoulou O, Georgaki S, Galanos A, Vlahos L. Greek McGiIl Pain Questionnaire: validation and utility in cancer patients. J Pain Symptom Manage 2002;24(4):379-87). It also allows us to look specifically at the 'burning' component of the pain that is common in these patients.

Methods

Sixteen patients undergoing deep brain stimulation for pain were entered prospectively into the study after informed consent and approval from the local ethics committee. AU patients suffered from chronic neuropathic pain. Table 1 shows the demographics of the patients and the aetiology of the pain. There were 13 male and 3 female patients. Mean age was 52 years (median 54.5). All patients had a PAG deep brain stimulator and 4 had a second electrode inserted into the ventroposterolateral (VPL) nucleus of the thalamus (turned off during this study). None of the six thalamic or one pontine stroke (see table 6) were in the vicinity of the electrodes and in fact the closest thalamic infarct was 4mm from the wire of a passing electrode (and further from the electrical contacts). TABLE 6. Demographics. PAG= periaqueductal grey, VPL = ventroposterolateral

nucleus.

Surgical Technique

Details of our surgical technique have been described elsewhere (Bittar RG, Kar-Purkayastha I, Owen SL, Bear RE, Green A, Wang S, Aziz TZ. Deep brain stimulation for pain relief: A meta-analysis. J Clin Neurosci 2005;12:515-519). In brief, Medtronic 3387® electrodes were stereotactically implanted under local anaesthetic. Intra-operative electrode localisation was aided by a feeling of warmth in the area of pain, during PAG stimulation, and parasthesia during thalamic stimulation, as described by others (Hosobuchi et al., 1986 and Young and Rinaldi, 1997). We did not elicit fear or anxiety in any patient, as previously reported (Nashold, 1969).

Measurements

During lab-based recordings, non-invasive continuous finger arterial pressure was measured with an Ohmeda Finapres 2300 (BOC Healthcare, USA). The blood pressure was calibrated using a sphygmomanometer and the pressure transducer and finger cuff were positioned at heart level. The finger pressure was digitised at 4kHz with 16-bit resolution (CED 1401 Mark II, Cambridge Electronic Design,

Cambridge, UK) using Spike II software^® (version 5.0, Cambridge Electronic Design, Cambridge, UK).

Study Design

Experiments to determine blood pressure change and VAS were performed in the week after the initial placement of the electrode. All sessions were carried out on a single day. Each experiment was performed more than 2 hours after any meal and room temperature was kept constant at 22°C. Patients delayed opioid (one patient) or antihypertensive medication (three patients) until after the experiment. They also abstained from caffeine. The deep brain stimulator was initially turned off for at least 10 minutes prior to experiments.

Experiments were started with the patient sitting for 5 minutes. The first session consisted of a 12-minute rest period with the stimulator turned off, while recording cardiovascular variables. Subsequent sessions consisted of six randomly ordered 'on' or 'off periods lasting for 9 minutes per session. Between each of these 'on' or 'off sessions, there was a 9 minute rest period with stimulation off to allow cardiovascular variables to return to baseline. The person recording the data and the patient were blinded as to whether the stimulator was on or off. The patient was asked to provide a visual analogue score (0-100, 0= no pain, 100= worst pain ever experienced) to estimate pain level, during each session. The stimulator settings used were those that provided the optimal analgesia for each individual patient (see below). Frequency ranged from 5 to 80Hz, amplitude from 2.5 to 5 volts, and pulse width from 150 to 450μs.

Determining Optimal Stimulator Parameters

This occurred prior to each experiment and usually lasts for 2-3 days. A full range of combinations of contacts, frequencies (5 to 80Hz), pulse width (120 to 450μs) and amplitude (0.5 to 5 volts) are trialled. The final parameters chosen are those that provide best subjective analgesia, often helped by looking at VAS. hi this study, only the contacts that provided best analgesia were stimulated during the cardiovascular measurements. In all cases, these were the chronic settings that were being used at one year, except that voltage was subsequently increased hi three cases. The effect of location of stimulation and blood pressure is the subject of another study (Green at al, 2005) and is therefore not presented here.

Recording of McGill's Pain Questionnaire Scores

Pre-operatively, and at one year, each patient completed a McGiIl pain questionnaire (MPQ) on both occasions, in the presence of a specialist nurse. For analysis of data, we used the Ranked Pain Rating Index (PRI(R)) as described by Melzack (Melzack R. The McGiIl Pain Questionnaire: major properties and scoring methods. Pain 1975;l:277-99). In this method of scoring, each word in a category is assigned a number, depending on'its severity. Using this method, we calculated overall pain rating index -PRI (R) - (out of 78), sensory PRI(R) (out of 42), affective PRI(R) (out of 14), evaluative PRI(R) (out of 5) and miscellaneous PRI(R) (out of 17). In addition, as we were particularly interested in the burning component of pain, we scored question 7 separately. Question 7 gives the patient the opportunity to select one of four words of increasing severity (hot, burning, scalding, and searing) and they were thus assigned a score of 0 to 4.

Signal processing and statistical analysis

One-way analysis of variance of BP with time was performed on all raw data segments for each session to determine significant change from baseline BP (p<0.05). For each session, data were averaged every 30 seconds and the mean values were taken to give the average across all three of the 'stimulation on' or three 'stimulation off sessions in each patient. These were then compared to the mean change in visual analogue score for each patient. To assess the contractility of the heart, and therefore to provide a measure of central sympathetic drive, we calculated the blood pressure changing rate, dP/dt (see discussion). This was derived by differentiating the continuous blood pressure waveform. The maximum dP/dt (maximum slope of the blood pressure curve) was then extracted.

To assess whether blood pressure, dP/dt, and pulse pressure were related to changes in pain scores, linear regression analysis was performed. For long-term changes, using the McGill's Pain Questionnaire, pre- operative and one-year MPQ scores for each group were compared and linear regression analysis was applied to test any relationship between reduction in blood pressure and degree of analgesia. As 'question 7' or the burning component of pain consists of a scale of one to four, we did not consider linear regression analysis to be valid. We therefore divided the patients into three groups depending on whether their blood pressure significantly changed with stimulation (significant change was regarded as two-tailed p<0.05 using analysis of variance of blood pressure with time). We then compared the mean pre- operative and one-year scores for each group using the Wilcoxon Signed Ranks Test for non-parametric data with a two-tailed p-value of <0.05 taken as significant.

To assess changes in the sympathetic and parasympathetic components of

SBP, auto-regressive power spectral analysis of SBP was performed (see Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use, Circulation 1996;93: 1043-65, for methodology) after filtering out frequencies below 0.023 Hz to detrend the signal. The power of the low and high frequency components were computed as the integral of the bands from 0.05 to 0.15Hz and 0.15 to 0.4Hz. The logarithm of the LF and HF measures was further analysed using a t-test.

Signal processing was performed in Matlab^® (vβ.l, MathWorks Inc., Natick, Ma., USA) and statistical analysis was performed in SPSS (Version 11, SPSS Inc, IL, USA). Graphs were plotted using Origin® (v7. 0300, Northampton, MA, USA). All p values are two-tailed.

Post-operative Localisation of Electrodes

Electrode positions were plotted on a brain atlas (Mai et al, 1998) using the post-operative MRI and a manipulation program (MRIcro version 1.38, Chris Rorden). First, the scan was rotated such that the Anterior and Posterior commissures (AC and PC respectively) were on the same slice. The mid- commissural point was then calculated, followed by the position, in Talairach space, of the electrode contacts. The contacts are visible, circular thickenings in the low signal on the axial scan. The centre of each contact was taken as the position of the electrode and this corresponds to the centre of the contacts in Fig. 17. Using the coronal and sagittal scans, the angles of the electrode to the midline and AC-PC line, respectively, were calculated. Once plotted on the brain atlas, the relative position of the lowest contact to the posterior wall of the superior colliculus was verified, as was the relative position of the upper electrode to the mid-commissural point. As a further verification, the relative positions of the electrodes from all patients were compared, to rule out inconsistencies among the groups.

Results

Blood Pressure Changes

Seven out of sixteen patients had a significant decrease in systolic blood pressure with stimulation (mean reduction = 13.1 ± 6.1 mmHg). Four patients had a significant increase (mean = 11.7 ± 6.0 mmHg). The remaining five patients had no significant changes in BP .

Visual Analogue Scores (Short-term changes)

Using one-way ANOVA of VAS versus change in BP, mean pre-operative VAS scores were not found to be significantly different in the three groups of decrease, increase or no change in BP with stimulation (p>0.5 in all comparisons). VAS (within one week of surgery) were compared to changes in blood pressure during six 9-minute stimulation periods for each patient (three 'on' and three Off). Linear regression analysis of change in VAS (%) versus change in systolic blood pressure showed that there is a significant linear relationship between reduction in blood pressure and reduction in VAS (^=0.62, p=0.0105, n=l 6 - see Fig. 18A). Similar analysis revealed that fall in dP/dt was also strongly correlated with reduction in VAS (^= 0.62, p<0.008, n=16 - Fig. 18C) but that changes in pulse pressure were only weakly correlated (^=0.48, n=16 - Fig 18B). Autoregressive power spectral analysis of systolic blood pressure variability revealed that in the group with decreased blood pressure, there was a significant reduction (pO.OOl , n=7) in the logarithm of the low frequency component (Fig. 19A). The converse was true for increases in blood pressure (pO.OOl, n=4, Fig. 19B). This implies that blood pressure changes are related to changes in sympathetic activity (see discussion). McGiIl Pain Questionnaire - long-term results

To test whether this association between blood pressure changes and pain relief lasts for a long time, we compared blood pressure changes (taken within the 5 first week post-operatively) to improvements in pain scores at one year, using the McGiIl Pain Questionnaire (Figs. 20A-20D). Pre-operatively, one-way ANOVA of all MPQ categories versus change in BP (decrease, increase or no change) revealed that there was no significant difference in starting values between each group for each category (p>0.2).

1.0 At one year, we found that total pain remained lowest, having reduced from a mean pain score of 34 (SD=I 3) pre-operatively to a mean of 12.2 (SD=9.5) 1 yr post-operatively (p=0.003 Wilcoxon Signed Ranks Test, n=7) in those whose BP had decreased on stimulation (seven patients). The no change in BP group reduced from a mean of 28.5 (SD=5.1) pre-operatively to 14 (SD7.0) post-operatively

15 (p=0.055, n=5). Although there was an improvement in total pain score in the increased BP group from a mean of 41.5 pre-operatively to 36.8 post-operatively, this was not significant (Fig. 20A).

Analysis of the sensory component of the MPQ (scored out of 42) shows a similar decrease in the total scores (Fig. 20B). In the decreased BP group, mean 0 reduced from 19.4 (SD=7.9) pre-operatively to 6.2 (SD=5.5) post-operatively (p<0.05, Wilcoxon Signed Ranks test). In the no change group, mean decreased from 15.3 (SD=3.6) pre-operatively to 7.7 (SD=0.6) post-operatively (p<0.05, Wilcoxon Signed Ranks test). Again, there was a reduction from 22.5 (SD=9.5) to 18.0 (SD=I 0.7) in the increased BP group but this was not significant. Although 5 there was a reduction in the affective and evaluative components in all groups, due to a wide variation between patients, none of these were found to be significant.

In the miscellaneous section, pre-operative scores reduced with weak significance (p=0.066) only in the decreased BP group (6.1 (SD=2.5) pre- operatively versus 1.6 (SD=I .8) post-operatively) but not in the other two groups. 0 See Fig. 2OC for a representation of these results. Analysis of question 7 of the MPQ (related to 'burning' sensation) showed that all but two patients studied had burning pain pre-operatively (one in the 'no change in BP' group, the other in the 'decreased BP' group did not have burning). Post-operatively, in contrast, only one patient had burning pain in the decreased BP group, one in the no change in BP group, and all of the patients in the increased BP group. Thus, the burning component of pain reduced in the decreased BP group but not in the increased BP group. Fig. 2OD shows that, using a quantitative score of 1 to 4 for the burning component, there was only a significant reduction in the burning component of pain in the decreased BP group, but not the other two. The mean score dropped from 2.3 pre-operatively to 0.4 post-operatively. It therefore appears that there is a relationship between reduced BP and improvement of burning pain.

As with short-term changes there was a linear correlation between reduction in blood pressure and percentage reduction in MPQ score at one year (r¹- 0.69, p= 0.002, n=16; Fig. 21 A). There were also strong linear correlations with both the sensory and miscellaneous subsets of the MPQ score at one year, that describe the qualitative aspects of a patient's pain (I²= 0.67, p=0.004, n=16 for sensory; Fig. 21B, r²^ 0.68, p=0.003, n=16 for miscellaneous; Fig. 21C). Thus, it appears that reduction in blood pressure on deep brain stimulation is associated with prolonged pain relief for at least a year.

Electrode Location and Analgesic Effect

Fig. 17 shows the electrode positions, as determined by the post-operative MRI scans. Note that these are an approximation as every brain is slightly different and there are errors inherent in plotting all electrodes onto one image. We estimate an error of 2-3mm using this technique. We have previously shown that electrodes in ventral PAG reduce blood pressure and those in dorsal PAG increase blood pressure (Green et al, 2005). Comparison of changes in VAS between the 8 most ventral and 8 most dorsal electrodes (as in Fig. 17) revealed a mean reduction in VAS of 56.6% in the ventral group and 33.6% in the dorsal group. This difference was found to be significant (p=0.036, Wilcoxon, n=16). Comparison of Absolute Values of Pain and Blood Pressure

Linear regression analysis was performed on the absolute values of pain and blood pressure in each patient (Figs. 22A-22F). For each patient with a change in blood pressure during stimulation there are therefore two data points - stimulation on versus stimulation off, with corresponding VAS. For those with no significant change in BP, there is only one data point for each pain score. The results show that there is no significant correlation between absolute systolic blood pressure, pulse pressure or dP/dt and pain scores for any of these variables (see legend for r² values).

Comparison of short-term versus long-term pain scores

Short term improvements in VAS (i.e. within one week of surgery) were compared to one-year MPQ scores (Fig. 23). Regression analysis showed that there is a significant correlation between early efficacy and improvement in long-term pain score.

Discussion

Pain processing in the central nervous system is complex and involves many areas, but there are some areas that have been shown to be important both in pain processing and blood pressure control. These include the nucleus tractus solitarius (this is the first relay station of the baroreceptor afferents), the locus coeruleus and the periaqueductal grey region (Ghione S. Hypertension-associated hypalgesia. Evidence in experimental animals and humans, pathophysiological mechanisms, and potential clinical consequences. Hypertension 1996;28:494-504.) That there is a relationship between acute pain and raised blood pressure is well established, the mechanism being a combination of increased arousal and increased sympatlietic activity (Maixner W, Gracely RH, Zuniga JR, Humphrey CB₅ Bloodworth GR. Cardiovascular and sensory responses to forearm ischemia and dynamic hand exercise. Am J Physiol 199O;259:R1156-63, and Nordin M, Fagius J. Effect of noxious stimulation on sympathetic vasoconstrictor outflow to human muscles, J Physiol 1995;489(3):885-94). The relationship between chronic pain and blood pressure, however, is much more complicated, hi pain-free individuals, there appears to be an inverse relationship between resting BP and pain sensitivity. For example, hypertensive subjects do not feel pain as intensely as normotensives

(Ghione S, Rosa C, Mezzasalma L, Panattoni E. Arterial hypertension is associated with hypalgesia in humans. Hypertension 1988;12(5):491-497, Sheps DS, Bragdon EE, Gray TF 3rd, Ballenger M, Usedom JE, Maixner W. Relation between systemic hypertension and pain perception. Am J Cardiol 1992;70:3F-5F, Ghione S. Hypertension-associated hypalgesia. Evidence in experimental animals and humans, pathophysiological mechanisms, and potential clinical consequences. Hypertension 1996;28:494-504, Bradley KM, O'Sullivan VT, Soper ND, Nagy Z, King EM, Smith AD, Shepstone BJ. Cerebral perfusion SPET correlated with Braak pathological stage in Alzheimer's disease. Brain 2002; 125: 1772-81).

Dworkin et al (Dworkin BR, Filewich RJ, Miller NE, Craigmyle N,

Pickering TG. Baroreceptor activation reduces reactivity to noxious stimulation: implications for hypertension. Science 1979;205:1299-301) showed, in rats, that this effect is mediated by the baroreceptors and proposed that, not only could the stimulation of baroreceptors explain the hypalgesia of hypertension, but that hypertension may be a learned behaviour to minimise the effects of stress. In contrast to this inverse relationship of BP and pain sensitivity, chronic pain patients have been shown to lose, or even reverse this relationship i.e. pain sensitivity positively correlates with BP (Maixner W, Fillingim R, Kincaid S, Sigurdsson A, Harris MB. Relationship between pain sensitivity and resting arterial blood pressure in patients with painful temporomandibular disorders, Psychosom Med

1997;59:503-l 1., Bruehl S, Chung OY, Ward P, Johnson B, McCubbin JA, The relationship between resting blood pressure and acute pain sensitivity in healthy normotensives and chronic back pain sufferers: the effects of opioid blockade. Pain 2002;100:191-201, Bragdon EE, Light KC, Costello NL, Sigurdsson A, Bunting S, Bhalang K, Maixner W. Group differences in pain modulation: pain-free women compared to pain-free men and to women with TMD. Pain 2002;96:227-37). Our results show that the degree of pain relief achieved by stimulating the rostral PAG, correlates with the magnitude of blood pressure changes; the greater the reduction in blood pressure during stimulation, the greater the analgesic effect (Fig.20A). The fact that absolute BP variables did not correlate with pain scores suggests that it is the reduction in BP rather than the absolute blood pressure that is important. These findings could not be deduced from animal experiments, because they cannot report their feeling of pain. Young and Rinaldi (Young RF, Rinaldi PC. Brain Stimulation. In: North RB, Levy RM, editors. Neurosurgical Management of Pain, New York: Springer- Verlag, 1997. pp. 288-90) previously reported elevation of heart rate with intraoperative PVG stimulation (32 ± 12 beats per minute) as well as elevation of SBP and DBP (71 ± 21 and 47 ± 10 rnmHg respectively) with stimulation at a threshold generally higher than that used to produce analgesia. As this effect was consistent, it was used to aid electrode localisation in over 120 patients. As their electrode localisation appears similar to ours, it is likely that their high amplitude of stimulation was exciting dorsal PAG in preference to ventral

PAG. Their improved results in patients with cardiovascular changes is in agreement with our results, in that if the electrode is in the 'cardiovascular area', it can be used to reduce BP.

Does the reduction in blood pressure itself lead to analgesia; is it another consequence of the change in sympathetic nervous system activity; or is it due to another output of the PAG? Peripheral blood flow is increased in neuropathic pain conditions (Archer AG, Roberts VC, Watkins PJ. Blood flow patterns in painful diabetic neuropathy. Diabetologia 1984;27:563-7) and reduction of cutaneous blood flow is associated with pain relief(Tanaka S, Komori N, Barron KW, Chandler MJ, Linderoth B, Foreman RD. Mechanisms of sustained cutaneous vasodilation induced by spinal cord stimulation. Auton Neurosci 2004; 114:55-60). Thus reduction in ABP is not itself likely to reduce pain directly, but its effect of reducing local blood flow might. To test this, we compared changes in pain scores to changes in pulse pressure, a validated marker of peripheral vasodilatation (Laskey WK, Parker HG, Ferrari VA, Kussmaul WG, Noordergraaf A. Estimation of total systemic arterial compliance in humans. J Appl Physiol 1990;69(l):l 12-9). Our results show that analgesic effects were only weakly correlated to reductions in pulse pressure (Fig. 20B) i.e. that pain relief was not strongly associated with local vasoconstriction. More importantly, correlation of the rate of change of systolic blood pressure (dP/dt), a good index of cardiac contractility, implies that the blood pressure changes were likely to have been centrally mediated (Brinton TJ, Cotter B,

Kailasam, MT, Brown DL, Chio SS, O'Connor DT, DeMaria AN. Development and validation of a noninvasive method to determine arterial pressure and vascular compliance. Am J Cardiol 1997; 80:323-30). The reduction in dP/dt correlated significantly with the patients' relief of pain (Fig. 20C). Thus the reduction in blood pressure was probably largely cardiac in origin, and possibly mediated by a reduction in central sympathetic drive.

To elucidate the underlying mechanisms of these blood pressure changes further, we analysed blood pressure variability (BPV) by autoregressive power spectral analysis in the patients with increases and in those with decreases in blood pressure. The power of Mayer's wave c. 0.1 Hz on the BPV spectra provides another, albeit crude, index of sympathetic nervous activity (Pagani M, Montano N, Porta A, Malliani A, Abboud FM, Birkett C, Somers VK. Relationship. between spectral components of cardiovascular variabilities and direct measures of muscle sympathetic nerve activity in humans. Circulation 1997;95:1441-8). We found that there was a significant reduction in power at this frequency in those with reduced blood pressure, and an increase in those with increased blood pressure (Fig. 20D). This again suggests that our patients' blood pressure changes were accompanied by changes in sympathetic nervous system activity. This is consistent with studies that have shown that sympathetic ganglion blockade can often reverse neuropathic pain (Treede RD, Davis KD, Campbell JN, Raja SN. The plasticity of cutaneous hyperalgesia during sympathetic ganglion blockade in patients with neuropathic pain. Brain 1992;115(2):607-21).

Using the McGill's Pain Questionnaire, we have also shown that those patients with the greatest magnitude of blood pressure reduction in the laboratory had the best results from deep brain stimulation in the long-term. This may be because of the long-term blood pressure changes (although these were not measured in this study) or it could be that acute blood pressure changes imply that the electrode is in the optimal position. Our comparison of short-term VAS and long- term MPQ scores suggest that long-term improvements are a reflection of early success, although there is still room for minor parameter adjustments leading to further improvements in some cases.

^■ The 'defence' reaction in the rat is part of an integrated response that aids survival in the wild (Hunsperger, 1956). There are two components to this reaction. Firstly, 'freezing' behaviour that is associated with hypotension, bradycardia, and non-opioid related analgesia and can be induced by ventral PAG stimulation

(Abrahams et al., 1960, Duggan and Morton, 1983 and Lovick, 1992). Secondly, a 'fight or flight' response that is associated with hypertension, tachycardia and opioid related analgesia and can be induced by dorsal PAG stimulation (Carrive and Bandler 1991 and Lovick 1985). Thus this basic survival response in animals couples the cardiovascular system to the modulation of pain. Our results provide evidence, for the first time, that similar responses may exist in the human.

The teachings of all of the references cited herein are hereby incorporated by reference in their entirety.

EQUIVALENTS

While this invention has been particularly shown and described with references to preferred embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.

Claims

CLAMS What is claimed is:

1. A method of influencing blood pressure in a human, comprising the step of applying a stimulation in a region of a brain in the human in a manner influencing blood pressure.

2. The method according to claim 1 wherein the region includes the periventricular gray region of the brain.

3. The method according to claim 1 wherein the region includes the periaqueductal gray region of the brain.

4. The method according to claim 1 wherein applying the stimulation in the region of the brain includes applying the stimulation to a dorsal area of the region.

5. The method of claim 4, wherein the stimulation is associated with a pressor response in the blood pressure in the human.

6. The method according to claim 1 wherein applying the stimulation in the region of the brain includes applying the stimulation to a ventral area of the region.

7. The method of claim 6, wherein the stimulation is associated with a suppressor response in the blood pressure of the human.

8. The method according to claim 1 further including selecting a human having a blood pressure irregularity.

9. The method of claim 8, wherein the blood pressure irregularity is normalized.

10. The method according to claim 8 wherein the irregularity is selected from the group consisting of hypertension and hypotension.

11. The method according to claim 1 further including feeding back a metric representative of blood pressure in an automated manner and responsively adjusting the stimulation based on the metric.

12. The method according to claim 1 further including enabling feedback of a metric representative of blood pressure in a manual manner and adjusting the stimulation in response to the metric.

13. The method according to claim 1 wherein the stimulation includes at least one member selected from the group consisting of an electrical stimulation, a magnetic stimulation, an electromagnetic stimulation, a thermal stimulation, and a mechanical stimulation.

14. The method according to claim 1 further including inductively communicating stimulation parameters used for applying the stimulation.

15. The method according to claim 14 further including powering a system used for applying the stimulation based on the inductive communication.

16. The method according to claim 1 further including communicating stimulation parameters used for applying the stimulation by at least one member selected from the group consisting of a radio frequency signal, an electrical signal, and an optical signal.

17. The method according to claim 1 further including disabling the stimulation in a fail-safe manner.

18. The method according to claim 17 wherein the disabling is performed in an automated manner based on a metric associated with a blood pressure or a detected problem in applying the stimulation.

19. The method according to claim 17 wherein the disabling is activated by the human or another human.

20. The method according to claim 1 wherein applying the stimulation includes selectively applying the stimulation to at least one area of the region of the brain.

21. The method according to claim 1 wherein applying the stimulation to the region of the brain includes selectively energizing at least one multiple conductor of at least one electrode disposed within the region.

22. The method according to claim 1 wherein applying a stimulation to the region of the brain includes selectively energizing multiple conductors of a single electrode having at least two distal conductors positioned in a ventral region of a nuclei of the brain and at least two proximal conductors positioned in a dorsal region of a nuclei of the brain.

23. The method of Claim 1 , wherein the human has a hypotension condition.

24. The method of Claim 23, wherein the hypotension condition is orthostatic hypotension.

25. The method of claim 23, wherein the human further has a pain condition.

26. The method of Claim 1 , wherein the human has a hypertension condition.

27. The method of Claim 26, wherein the human further has a pain condition.

28. A method of influencing blood pressure in a human, comprising the step of applying a stimulation in a region in a brain in a human having a hypertension condition.

29. A method of influencing blood pressure in a human, comprising the step of applying a stimulation in a region of a brain in a human having a hypotension condition.

30. The method according to claim 1 wherein applying the stimulation includes generating a voltage differential between at least two electrodes of between about -10V and about +10V with a frequency of between about 0.1 Hz and about 1 kHz.

31. An apparatus for influencing blood pressure in a human subject, comprising: a blood pressure sensor detecting blood pressure; a processor in communication with the blood pressure sensor and generating a control signal based on the blood pressure; a signal generator in communication with the processor generating a stimulation signal based on the control signal; and an electrode including at least two conductors in contact with a region of the brain that stimulates the region as a function of the stimulation signal in a manner influencing blood pressure in a human subject.

32. The apparatus according to claim 31 wherein the blood pressure sensor includes a transmitter and the processor includes a receiver configured to receive a metric from the transmitter representative of blood pressure in an automated feedback manner and wherein the processor, responsive to the metric, is configured to adjust the stimulation based on the metric.

33. The apparatus according to claim 31 further including a human-controlled feedback interface that enables feedback of a metric representative of blood pressure in a manual manner to allow an operator to adjust the stimulation.

34. The apparatus according to claim 31 wherein the signal generator generates a stimulation signal that causes the electrode to generate a stimulation selected from the group consisting of an electrical stimulation, magnetic stimulation, electromagnetic stimulation, thermal stimulation, and mechanical stimulation.

35. The apparatus according to claim 31 wherein the processor is coupled to a transmitter configured to transmit stimulation parameters via wireless communications to a receiver coupled to the signal generator to cause the signal generator to generate a stimulation signal as a function of the stimulation parameters.

36. The apparatus according to claim 35 further including a power source to power the signal generator to apply the stimulation based on the wireless communications.

37. The apparatus according to claim 31 wherein the signal generator is coupled to a receiver configured to receive stimulation parameters used for applying the stimulation by at least one member selected from the group consisting of a radio frequency signal, electrical signal, and optical signal.

38. The apparatus according to claim 31 further including a fail-safe unit coupled to the signal generator to disable the signal generator in a fail-safe manner.

39. The apparatus according to claim 38 wherein the fail-safe unit is coupled to the processor or blood pressure sensor and is configured to cause the signal generator to be disabled based on the metric associated with blood pressure or a detected problem in applying the stimulation.

40. The apparatus according to claim 38 wherein the fail-safe unit is activated by the human subject or another human.

41. The apparatus according to claim 31 wherein the signal generator and electrode are configured to selectively apply the stimulation to at least one area of the region of the brain.

42. The apparatus according to claim 31 wherein the signal generator causes at least one multiple conductor of the electrode to apply a stimulation to the region of the brain.

43. The apparatus according to claim 31 wherein the signal generator applies a stimulation to the region of the brain by selectively energizing multiple conductors of a single electrode having at least two distal conductors positioned in a ventral region of a nuclei of the brain and at least two proximal conductors positioned in a dorsal region of a nuclei of the brain.

44. The method according to claim 31 wherein the signal generator is configured to generate a voltage differential between at least two conductors of the electrode between about^"10V and about⁺10V with a frequency of between about 0.1Hz and about IkHz.

45. An apparatus for stimulating a region in a human brain, comprising: a signal generator adapted to generate a signal; and at least one electrode disposed in a region of a brain in a human subject adapted to produce an output as a function of the signal to stimulate the region in a manner influencing blood pressure in the human subject.

46. An apparatus according to claim 45 wherein the signal generator is coupled to a receiver configured to receive stimulation parameters used for applying the stimulation by at least one member selected from the group consisting of a radio frequency signal, electrical signal, and optical signal.

47. An apparatus for generating a signal for controlling stimulation of a region in a human brain in a human subject, comprising: a blood pressure sensor adapted to provide a blood pressure measurement of blood pressure in a human subject; a microprocessor controller in communication with the blood pressure sensor adapted to convert the blood pressure measurement to a control signal that, when received by a signal generator in operative arrangement with electrodes deployed in a region of a brain of the human subject, causes the signal generator to generate a signal that stimulates the region in a manner that influences the blood pressure.