Comparing the cost-effectiveness of rituximab maintenance and radio-immunotherapy consolidation versus observation following first-line therapy in follicular lymphoma patients
Qiushi Chen
Turgay Ayer,PhD
Adam C Rose
Loretta J Nastoupil,MD
Christopher R Flowers,MD, MS
Corresponding author: Christopher R. Flowers, MD, MS., 1365 Clifton Road, N.E. Building B, Suite 4302, Emory University, Atlanta, GA 30322,crflowe@emory.edu, Telephone: 404-778-5554, Fax: 404-778-3366
Abstract
OBJECTIVES
Phase 3 randomized trials have shown that rituximab maintenance (MR) therapy or radio-immunotherapy (RIT) consolidation following frontline therapy can improve progression-free survival for patients with follicular lymphoma (FL), but the cost-effectiveness of these approaches with respect to observation has not been examined using a common modeling framework.
METHODS
We developed Markov models to estimate patients’ lifetime costs, quality-adjusted life years (QALYs), and life years (LYs) after MR, RIT, and observation following frontline FL treatment from the US payer’s perspective. Progression risks, adverse event probabilities, costs, and utilities were estimated from clinical data of PRIMA, ECOG trial (for MR) and FIT trial (for RIT), and the published literature. We evaluated the incremental cost-effectiveness ratio for direct comparisons between MR/RIT and observation. Model robustness was addressed by one-way and probabilistic sensitivity analyses.
RESULTS
Compared with observation, MR provided additional 1.089 QALYs (1.099 LYs) and 1.399 QALYs (1.391 LYs) based on PRIMA and ECOG trials respectively, and RIT provided additional 1.026 QALYs (1.034 LYs). The incremental cost per QALY-gained was $40,335 (PRIMA) or $37,412 (ECOG) for MR, and $40,851 for RIT. MR and RIT had comparable incremental QALYs before first progression, while RIT had higher incremental costs of adverse events due to higher incidences of cytopenias.
CONCLUSIONS
MR and RIT following frontline FL therapy demonstrated favorable and similar cost-effectiveness profiles. The model results should be interpreted within the specific clinical settings of each trial. Selection of MR, RIT, or observation should be based on patient characteristics and expected tradeoffs for these alternatives.
Keywords: Cost-effectiveness, follicular lymphoma, maintenance, rituximab, radio-immunotherapy, lymphoma, Non-Hodgkin lymphoma
INTRODUCTION
Follicular lymphoma (FL) is the most common subtype of indolent non-Hodgkin’s lymphoma (NHL) in the United States (US) (1,2), accounting for approximately 20% of 580,000 prevalent NHL cases in 2011 (1,3). Although FL in limited stage is curable with standard radiation therapy (4), the majority of FL patients are diagnosed with advanced-stage disease (5,6), which remains incurable. FL management also produces an economic burden to patients and US society, with an annual cost ranging from $20,000 to $36,000 per patient (7).
This cost is associated with substantial patient benefit. In the past few decades, the median overall survival (OS) of FL patients significantly improved from 11 years to 18 years, following advances in effective therapies and supportive care (8). In the modern era, chemotherapy and rituximab (R) plus chemotherapy commonly have been used for previously untreated patients with advanced-staged FL. In current practice, however, there is no single approach that has become the standard for first-line treatment (9). Advanced-stage FL typically produces a course of recurrent remissions and relapses with reducing response rate, remission duration, and health-related quality of life along with the subsequent treatments. As a result, in the absence of curative therapies, many efforts have focused on extending the duration of the first remission to postpone subsequent treatment and to help patients maintain a higher health-related quality of life (HRQL).
Maintenance with rituximab (MR) and radio-immunotherapy consolidation (RIT) are two approaches aiming at such improvement. Rituximab, an anti-CD20 monoclonal antibody with favorable toxicity profile, has been a major therapeutic advance for FL treatment in the last several decades. It has been utilized as a single agent, in combination with chemotherapy, or as maintenance therapy in newly diagnosed and relapsed patients (10). Patients undergoing MR following the induction therapy continue to receive rituximab for additional two years. Radio-immunotherapy uses radiation-labeled anti-CD20 antibody to deliver radiation to malignant cells. It first showed high response rate in patients with relapsed FL (11) and was later applied as a consolidation strategy following first-line treatment.
For untreated patients with FL, MR and RIT consolidation also have demonstrated clinical benefit. MR for two years has been shown to significantly improve the progression-free survival (PFS, i.e., time from randomization to disease progression or death) and rate of complete response (CR, i.e., complete disappearance of all evidence of disease (12)) in the randomized Primary RItuximab and MAintenance (PRIMA) and Eastern Cooperative Oncology Group (ECOG) trials (13,14). RIT consolidation following induction chemotherapy or R-chemotherapy also showed similar efficacy results in the randomized First-line Indolent Trial (FIT) (15,16). Each approach demonstrated improvement in PFS over observation without producing significant differences in patients’ HRQL (14,17). As a result, MR and RIT consolidation have been approved for use in the front-line setting since 2011 and 2009, respectively. A randomized phase 2 trial, ZAR2007, will provide a head-to-head comparison between MR and RIT following first-line induction with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone) therapy (18). Preliminary results from this trial showed similar partial response (PR) to CR conversion rate, no significant difference in OS, and a superior 3-year PFS for MR. As indicated above, however, studies comparing front-line strategies in FL may require more than a decade of follow-up to demonstrate difference in OS. These data remain immature and longer follow-up is awaited for a more comprehensive comparison of survival benefits for the two approaches.
In clinical practice, MR is commonly used. An analysis based on the largest prospective study of FL in the US, the National LymphoCare Study, showed that among 1186 patients who received front-line rituximab-based induction therapy, 46% received MR (19). On the other hand, since a single dose of RIT consolidation may provide comparable efficacy to MR for two years, RIT could be preferred in some circumstances, although it is much less commonly used (20). For MR and RIT consolidation it is unclear whether the additional costs are worth the benefits when compared with observation. In this study, we evaluated and compared the economic impact of MR and RIT consolidation versus observation, respectively, following the first-line induction therapy for patients with advanced-stage FL.
MATERIALS AND METHODS
General approach
We developed three separate Markov models based on three phase 3 randomized clinical trials, respectively: one model compared RIT with observation following the first-line induction therapy based on the FIT trial (15,16), and two models compared two-year MR with observation based on the PRIMA (14) and the ECOG trials (13), respectively. We refer to each model using the corresponding trial name throughout the manuscript. There existed differences in patient characteristics and treatment regimens across the trials (Table 1). For example, the induction therapy was chemotherapy in ECOG and for most patients in FIT, but it was R-chemotherapy in PRIMA. In addition, the regimens of MR differed slightly between PRIMA and ECOG. The differences in induction therapy affected patients’ outcomes. Therefore, it was not appropriate to perform an indirect comparison between MR and RIT using the data from the trials with different induction therapies. Instead, we built three separate models, where each model inherited the specific clinical settings from the corresponding clinical trial.
Table 1.
Patients and treatment regimens in randomized trials.
| ECOG1496 | PRIMA | FIT | ||||
|---|---|---|---|---|---|---|
| MR* | OBS | MR† | OBS | RIT‡ | OBS | |
| N | 115 | 113 | 505 | 513 | 204 | 205 |
| Age | ||||||
| Median | 58 | 54 | 57 (26–79) | 55 (22–84) | 55 (29–78) | 53 (27–74) |
| ≥60 years-old | 47 (41%) | 38 (34%) | 176 (35%) | 180 (35%) | 58 (28%)§ | 28 (24%)§ |
| Advanced stage (III/IV) | 73 (64%)§ | 72 (64%)§ | 459 (91%) | 459 (89%) | 202 (99%) | 199 (97.1%) |
| Male sex | 59 (51%) | 62 (55%) | 270 (53%) | 263 (51%) | 97 (47.5%) | 103 (50.2%) |
| FLIPI score | ||||||
| Low | 23 (26%) | 24 (27%) | 106 (21%) | 110 (21%) | 56 (37.3%) | 62 (42.5%) |
| Intermediate | 32 (36%) | 32 (36%) | 183 (36%) | 187 (36%) | 38 (38.7%) | 54 (37.0%) |
| High | 33 (38%) | 33 (37%) | 215 (43%) | 216 (42%) | 36 (24.0%) | 30 (20.5%) |
| B symptoms | 22 (19%) | 34 (30%) | 160 (32%) | 156 (30%) | 46 (22.5%) | 42 (20.5%) |
| Induction therapy | CVP|| | CVP | RCHOP#: 7 6% RCVP: 22% RFCM : 3% | RCHOP : 75% RCVP: 22% RFCM: 3% | Chlorambucil: 9.8% CVP/COP: 25.5% CHOP: 31.4% CHOP-like: 14.7% FLU-comb**: 5.4% R-comb: 13.2% | Chlorambucil: 9.3% CVP/COP: 26.3% CHOP: 28.3% CHOP-like: 15.1% FLU-comb: 5.4% R-comb: 15.6% |
Rituximab 375 mg/m2 once per week for 4 weeks every 6 months for 2 years,
rituximab 375 mg/m2 every 8 weeks for 2 years,
rituximab 250 mg/m2 on day -7 and day 0 followed on day 0 by 90Y-ibritumomab tiuxetan 14.8 MBq/kg; maximum of 1,184 MBq,
Only stage IV reported,
CVP, cyclophosphamide, vincristine, and prednisone,
R, rituximab; CHOP, cyclophosphamide, doxorubicin, vincristine, prednisone, FCM, fludarabine with mitoxantrone and cyclophosphamide,
FLU, fludarabine.
In each Markov model, we evaluated the life-time cost and total quality-adjusted life years (QALYs) for MR, RIT, and observation following first-line induction therapy, respectively. Health states defined for the clinical course included: 1) before first progression; 2) first progression; 3) second progression; and 4) death (Fig. 1). The structure for the models was determined based on typical disease course of FL patient, in line with previously published models for FL treatment (21,22). The model simulated outcomes for patients upon completion of first-line induction therapy. During the long disease course of FL, patients’ age also in part affect their survival outcomes. As FL is commonly diagnosed in patients with a wide range of age, to better project the outcomes for the FL population beyond the narrow age distribution of the clinical trial, we sampled the initial age of each patient in the micro-simulation from the distribution of age at diagnosis of FL from the Surveillance, Epidemiology, and End Results (SEER) database between 1992–2009 (23). The Markov model simulated transitions between health states in each model cycle which represents one month. We considered up to three lines of treatment. Patients with relapsed FL were assumed to receive R-CHOP/CHOP therapy without MR in the second-line treatment (27) and bendamustine with rituximab (BR) in the third-line treatment (28). For the first- and second-line treatment, patients may remain in the same health state, progress to the next line of treatment, or die with estimated probabilities based on clinical data. Following the third-line treatment, only risks of death were considered. We applied half-cycle corrections (24) for model parameters to address the possibility that state transitions could occur at any time point within each cycle.
Figure 1. Markov model.
Description and data sources:
a: PFS for each arm (i.e., MR/RIT/observation), source: PRIMA/ECOG/FIT trials
b: OS for each arm (i.e., MR/RIT/observation), source: PRIMA/ECOG/FIT trials and the US life table
c: PFS after first progression, source: EROTC20981 trial (the second randomization, observation arm without rituximab maintenance)
d: OS after first progression, source: EROTC20981 trial
e: OS after second progression, source: Rummel et al. (2005).
Costs were associated with each health state. We adopted the US payer’s perspective, and therefore considered direct medical costs, which captured consumption of all resources directly attributable to the treatment strategy. Direct non-medical costs (e.g., transport costs) and indirect costs (e.g., loss of productivity) were not considered in the model. All cost estimates were converted to 2013 US dollars based on the Consumer Price Index in the medical care category (25). We followed standard recommendations for conducting health economic evaluation and discounted the clinical outcomes and costs at the annual rate of 3% (26). Each model was developed in Treeage Pro 2011 and statistical analyses were performed in R.
Progression risk and mortality
The probabilities of the first progression in each model were estimated based on PFS curves in the published randomized trials (13,14,16). Risks of the second progression were estimated based on the PFS curve of the observation group in the EORTC 20981 trial for relapsed FL patients (i.e., R-CHOP/CHOP therapy without MR) (22,27). A further analysis based on this trial has shown that the endpoints were not significantly different across different prior therapies, which enables us to safely assume that transition probabilities after the second-line treatment were identical for each arm. To estimate the monthly risk of progression, we used Engauge Digitizer to retrieve the data points from each PFS curve plot, and fitted parametric survival models using these data points. We considered Weibull, log-logistic, and Gompertz distributions for survival models. To maintain consistency between the arms for comparisons, we selected a common survival distribution that demonstrated good fit for all PFS and OS curves of treatment and observation arms in each trial. In the final analysis, we used log-logistic distributions for the first- and the second-line treatments, and Gompertz distributions for the third-line treatment. Other survival distributions were tested in the sensitivity analyses. We applied Bayes’ rule to derive the monthly risk of progression, as a conditional probability of progression in one month given that the patient has not progressed yet. Risk estimates beyond the follow-up time of clinical trials were extrapolated from the fitted survival model. As the differences in survival outcomes continued to separate beyond the median follow-up time, we assumed the duration of treatment effect to be 6 years and the risks of progression or death in MR and RIT arms to be the same as those in observation arms after 6 years, which is in line with a prior published model of FL treatment (22).
Mortality risk was defined as the maximum of cause-specific mortality and other-cause mortality at each cycle. The cause-specific mortality before the first progression was estimated based on the OS curves for each arm in the PRIMA, ECOG, and FIT trials, and cause-specific mortalities of the second and the third-line treatments were derived from the OS curves of CHOP/RCHOP treatment (27) and BR treatment (28) for relapsed FL, respectively. Other-cause mortality for each age group was estimated from the US life tables (29).
Utility and cost estimates
HRQL for MR and RIT consolidation have been assessed in clinical studies separately. We used the health utility estimates from the published literature. The utilities were estimated as 0.88 for MR and observation in the PRIMA and ECOG models (22,30,31) and 0.84 for RIT and 0.83 for observation in FIT model (17) before the first progression; 0.79 after the first progression, 0.62 after the second progression, and 0 for the death state (22,30–32).
Only direct costs were considered in this study, including drug costs, administration costs, monitoring costs, and adverse event costs. The drug costs for MR and RIT consolidation were calculated based on the dosing for the regimens in each trial, while no drug costs were incurred in the observational arm. The infusion dosage for FL treatments were computed based on a body surface area (BSA) with a mean value of 1.835 m2 (22). The wholesale acquisition cost (WAC) (33) was used for the unit cost of each drug. The cost estimate for the second-line treatment was based on an established practice pattern for relapsed FL previously described in a published FL modeling study (22), and the cost of third-line treatment was based on the bendamustine and rituximab regimen (28).
Unlike RIT which has only one-time drug infusion, MR and second- or third-line treatments extend for more than one model cycle, and require multiple drug administrations (e.g., two years for MR, 18 weeks for RCHOP/CHOP, and 16 weeks for BR). We divided the total costs of drugs and administration by the complete treatment duration, and allocated the same amount in each month over the treatment periods. This way, only resources used and corresponding costs occurred to that point were attributed to individual patients in the models, since patients in the model could die or progress to the next line of treatment during any model cycle as in clinical practice.
We estimated the cost of grade 3/4 adverse events based on our clinical co-authors’ expert opinion in the management strategies for each adverse event. We assumed that patients with grade 3/4 infection/febrile neutropenia were hospitalized, and others were managed as outpatients. The cost of hospitalization included inpatient physician service fees and hospital reimbursement based on the length of stay (LOS) diagnosis-related group (DRG) code corresponding to the adverse event (34). For other hematological adverse events, the cost included outpatient visit cost and the cost of the materials for the adverse event management (seeAppendix 1 for details inSupplemental Materials at: XXX). For each treatment strategy, the cost of each adverse event was obtained by multiplying the incidence and the unit cost for each type of adverse events.
Drug administration and adverse event costs were calculated using the Medicare physician fee schedule for 2013 (35). Each medical service performed by a physician is associated with specific the Health Care Procedure Coding System (HCPCS) code or/and the Current Procedure Terminology (CPT) code. For each procedure, we determined its relative value units (RVUs) based on the HCPCS/CPT code from the Centers of Medicare & Medicaid Services (CMS) database (35), and then computed the outpatient physician service cost based on RVUs (34). Monitoring costs included the costs of blood tests and physician visits every three months, and costs of CT scans every six months, which were independent of the treatment arm and the current line of treatment (Table 2).
Table 2.
Estimates, ranges, and distributions of model parameters.
| Value | Min | Max | Distribution for PSA* | |
|---|---|---|---|---|
| BSA† | 1.835 m† | |||
| Cost of rituximab maintenance‡ | ||||
| PRIMA regimen (12 doses) | $54,588 | |||
| PRIMA regimen: monthly cost | $2,275 | 2,047 | 2,502 | Lognormal(mu=7.725,sigma=0.1)§ |
| ECOG regimen (16 doses) | $72,784 | |||
| ECOG regimen: monthly cost | $3,033 | 2,729 | 3,336 | Lognormal(mu=8.012,sigma=0.1) |
| Cost of Radioimmunotherapy consolidation | $46,566 | 41,910 | 51,223 | Lognormal(mu=10.744,sigma=0.1) |
| Cost of second-line treatment (22) | ||||
| Average cost of second-line therapies | $46,504 | 41,854 | 51,155 | Lognormal(mu=10.742,sigma=0.1) |
| Cost of third-line treatment | ||||
| Bendamustine + rituximab (4 courses) | $45,433 | 40,891 | 49,977 | Lognormal(mu=10.719,sigma=0.1) |
| Administration cost|| | ||||
| Monthly administration cost of chemotherapy# | ||||
| PRIMA regimen | $88 | 62 | 117 | Lognormal(mu=4.431,sigma=0.304) |
| ECOG regimen | $117 | 82 | 155 | Lognormal(mu=4.718,sigma=0.303) |
| Radioimmunotherapy administration cost | ||||
| Radiopharmaceutical therapy, radiolabeled-monoclonal antibody, IV (CPT: 79403) | $190 | 151 | 234 | Lognormal(mu=5.223,sigma=0.215) |
| Monitoring cost | ||||
| CT Scans: Chest/Abdomen/Pelvis (CPT: 72129, 74160, 72193) | $828 | 598 | 1,083 | Lognormal(mu=6.678,sigma=0.287) |
| Laboratory tests (50) | $76 | 68 | 84 | Lognormal(mu=4.326,sigma=0.1) |
| Other procedures | ||||
| Outpatient physician visits (CPT: 99213) | $50 | 43 | 66 | Lognormal(mu=3.879,sigma=0.229) |
| Inpatient physician visit | ||||
| First visit (CPT: 99222) | $135 | 117 | 179 | |
| Subsequent visits (CPT: 99232) | $70 | 62 | 94 | |
| Discharge visit (CPT: 99238) | $71 | 61 | 93 | |
| Blood transfusion (CPT: 36430) | $35 | 23 | 47 | |
| Adverse event cost¶ | ||||
| Anemia | $1,881 | 1,826 | 1,910 | Lognormal(mu=7.54,sigma=0.022) |
| Neutropenia | $3,066 | 2,758 | 3,384 | Lognormal(mu=8.023,sigma=0.102) |
| Febrile Neutropenia/infection | $11,566 | 7,092 | 27,656 | Lognormal(mu=9.065,sigma=0.763) |
| Thrombocytopenia | $1,086 | 593 | 1,139 | Lognormal(mu=6.96,sigma=0.248) |
| Incidence of adverse events | ||||
| PRIMA study | ||||
| MR arm - neutropenia | 0.04 | 0.02 | 0.06 | Beta(a=3.84,b=92.16)** |
| MR arm - febrile neutropenia/infection | 0.04 | 0.02 | 0.06 | Beta(a=3.84,b=92.16) |
| OBS arm - neutropenia | 0.01 | 0 | 0.04 | Beta(a=0.99,b=98.01) |
| OBS arm - febrile neutropenia/infection | 0.01 | 0 | 0.04 | Beta(a=0.99,b=98.01) |
| MR and OBS arm - anemia, thrombocytopenia† | 0.005 | 0 | 0.04 | Beta(a=0.0.03,b=5.50) |
| ECOG study | ||||
| MR arm - neutropenia | 0.03 | 0.02 | 0.05 | Beta(a=3.88,b=125.45) |
| MR arm - febrile neutropenia/infection | 0.01 | 0 | 0.02 | Beta(a=3.96,b=392.04) |
| OBS arm - neutropenia | 0.01 | 0 | 0.04 | Beta(a=0.99,b=98.01) |
| OBS arm - febrile neutropenia/infection | 0.01 | 0 | 0.04 | Beta(a=0.99,b=98.01) |
| MR and OBS arm - anemia, thrombocytopenia† | 0.005 | 0 | 0.04 | Beta(a=0.0.03,b=5.50) |
| FIT study | ||||
| RIT arm - neutropenia | 0.667 | 0.4 | 0.7 | Beta(a=136.04,b=67.92) |
| RIT arm - febrile neutropenia/infection | 0.079 | 0.04 | 0.12 | Beta(a=3.68,b=42.95) |
| RIT arm - anemia | 0.034 | 0.018 | 0.228 | Beta(a=4.36,b=123.93) |
| RIT arm - thrombocytopenia | 0.608 | 0.4 | 0.65 | Beta(a=82.15,b=52.96) |
| OBS arm - neutropenia | 0.025 | 0 | 0.04 | Beta(a=2.71,b=105.63) |
| OBS arm - febrile neutropenia/ infection | 0.024 | 0 | 0.04 | Beta(a=2.20,b=89.30) |
| OBS arm - anemia† | 0 | 0 | 0.04 | Beta(a=0.0.03,b=5.50) |
| OBS arm - thrombocytopenia† | 0 | 0 | 0.04 | Beta(a=0.0.03,b=5.50) |
| Utility | ||||
| No progression – MR | 0.88 | 0.81 | 0.95 | Beta(a=18.96,b=2.59) |
| No progression – RIT | 0.84 | 0.77 | 0.91 | Beta(a=23.04,b=4.39) |
| No progression – observation | ||||
| PRIMA study | 0.88 | 0.81 | 0.95 | Beta(a=18.96,b=2.59) |
| FIT study | 0.83 | 0.76 | 0.9 | Beta(a=23.90,b=4.90) |
| Combined | 0.867 | 0.79 | 0.94 | Beta(a=18.55,b=2.86) |
| After the first progression | 0.79 | 0.72 | 0.86 | Beta(a=26.75,b=7.11) |
| After the second progression | 0.62 | 0.48 | 0.76 | Beta(a=7.45,b=4.57) |
| Discount factor | 0.03 | 0 | 0.05 | - |
| Effective horizon (months) | 72 | 36 | Until progression | - |
PSA, probabilistic sensitivity analysis,
BSA, body surface area,
WAC (wholesale acquisition cost) for the unit cost of the drug (33), values of drug costs varied by ±10% in one-way sensitivity analysis, and lognormal distributions were used in probabilistic sensitivity analysis,
parameters of lognormal distribution are determined based on the mean and standard deviation (stdev) of the variable (stdev is estimated based on the half of the range of parameter value),
for the cost estimates of procedures, we used the national payment amount (i.e., geographic pricing cost index = 1), conversion factor for 2013 was $34.02, all the cost estimates were based on the CY 2013 PFS final rule (35,51),
Chemotherapy administration takes 3 hours for the first course, and 2 hours for subsequent courses,
seeAppendix 1 for details,
parameters of beta distribution are determined based on the point estimate and the sample size in the original study.
Sensitivity analysis
We conducted sensitivity analyses to evaluate the robustness of the model and to address the uncertainty in parameter estimation. Ranges and distributions of the parameters used in sensitivity analysis are summarized inTable 2. Utilities were varied over their 95% confidence intervals. For each procedure, physician fees were adjusted by different geographic pricing cost index (GPCI) adjustment factors, and the ranges of physician fees were determined by the lowest and highest costs in the CMS fee schedule. Similarly, to define the ranges for DRG-based hospital reimbursement, we computed the DRG rate for each of 3,500 providers in the US (34), and found the lowest and the highest DRG rates. Drug costs were varied within ±10% of their baseline values. We acknowledge that the treatment duration effect may be conservative for ECOG and FIT models, since these trials have longer follow-up time. Therefore, we varied the duration from the minimum follow-up time of 3 years to life-time in the sensitivity analysis.
In univariate sensitivity analysis, we examined the effect of each parameter on ICERs separately. In probabilistic sensitivity analysis (PSA), parameters were sampled simultaneously according to their sampling distributions. We followed recommended distributions based on parameter types, and assumed lognormal distribution for cost and beta distribution for utility and incidence of adverse events (Table 2) (36). We ran 10,000 replications for the PSA.
In secondary sensitivity analyses examining the robustness of the model structure, the models were evaluated based on different fitted survival distributions. In addition, we compared MR and observation in a combined model, in which we aggregated the fitted PFS and OS curves of each arm using the weighted log-relative-risk method (37), and estimated risks based on aggregated survival curves. Other model parameters were combined by taking the average of the estimates from multiple studies weighted by the sample size of each study.
RESULTS
In primary analyses, effectiveness and costs were compared within each clinical trial. Based on the PRIMA study, MR therapy provided 7.64 QALYs at a cost of $112,780, compared with the observation arm that provided 6.55 QALYs at a cost of $68,855. Based on the ECOG study, MR therapy provided 6.51 QALYs at a cost of $124,405, compared with the observation arm that provided 5.11 QALYs at a cost of $72,066. Based on the FIT study, RIT therapy provided 6.60 QALYs at a cost of $115,011, compared with 5.57 QALYs at a cost of $73,098 for the observation arm. The ICERs for MR were $40,335 and $37,412/QALY-gained in PRIMA and ECOG study, respectively, and the ICER for RIT was $40,851/QALY-gained based on the FIT study. Our model also estimated total projected life years (LYs) without quality adjustment, and provided the effectiveness and cost estimates by treatment period as summarized in Table 4.
Results of univariate sensitivity analyses are presented in the tornado diagrams (Figure 2). The most influential common factors for all three models included: utility before first progression in maintenance/consolidation arm or observation arm in each trial, drug cost for MR and RIT, duration of treatment effect, and discount factor. In the PSA, the incremental cost and effectiveness from 10,000 samples are shown in the scatterplots for each trial (Appendix 2 in Supplemental Materials at: XXX). The results of the PSA also are presented in the cost-effectiveness acceptability curve (CEAC,Figure 3). The CEACs showed that MR and RIT consolidation is cost-effective compared with observation at $50,000/QALY WTP threshold with probability 58%, 74%, and 62%, in PRIMA, ECOG, and FIT model, respectively. At a WTP of $100,000/QALY, the probabilities become 79%, 92%, and 84%, respectively. The models also were shown to be robust to the input risk estimates. The results produced similar ICER estimates when different fitted survival models were used (Appendix 3 in Supplemental Materials at: XXX). In the combined model, MR provided additional 1.13 QALYs (1.15 LYs) at the incremental cost of $46,234 compared with observation, and the ICER was $40,956 per QALY gained.
Figure 2.
Tornado diagrams of most influential variables to ICERs.
Figure 3.
Cost-effectiveness acceptability curves.
DISCUSSION
Maintenance with rituximab and radio-immunotherapy consolidation is two commonly considered strategies following the first-line induction therapy for FL patients to improve patients’ response and survival without disease progression. In this study, we evaluated the long-term benefits and costs of these two treatments, MR and RIT, compared with observation. In our model, both MR and RIT showed that patients’ total QALYs were improved at reasonable costs with the estimated ICERs below the threshold of $50,000 per additional QALY. We assessed the uncertainty in the model results and demonstrated moderate confidence of cost-effectiveness for MR and RIT with about 60% likelihood of an ICER of ≤ $50,000/QALY and 80% of an ICER of ≤ $100,000/QALY.
Different induction therapies may affect the projected effectiveness of MR, RIT, and observation. Given that the majority of patients in FIT received frontline chemotherapy without rituximab while all patients in PRIMA received R-chemotherapy (Table 1), the two studies do not have similar PFS and OS in their observation arm. We observed that survival curves for the RIT arm of the FIT model were comparable with those for the observation arm of the PRIMA model, which led to similar model outputs: 6.60 QALYs (8.48 LYs) and 6.55 QALYs (8.28 LYs) for each arm, respectively. This was validated externally by a recent comparison between RCHOP followed by observation and CHOP followed by RIT (I131 tositumumab) in a phase 3 randomized clinical trial (S0016) (38). This study showed that both strategies had outstanding PFS and OS but there was no significant difference between the two groups (P=0.11 comparing 2-year PFS).
The cost before first progression was mainly driven by drug cost in each trial. According to the WAC, RIT costs approximately $46,000, and a full course of MR costs approximately $54,000 (12 courses) in PRIMA and $72,000 (16 courses) in ECOG. Adverse event costs were more significant for RIT compared with MR and observation, due to the high incidences of anemia and thrombocytopenia following RIT consolidation. This finding is in line with the high incidence of adverse events reported in a series of phase 2 studies (39–44). On the other hand, patients in the observation arms progressed earlier, leading to higher costs after first progression. These models help to describe and quantify to what extent the additional costs of treatment and adverse events following front line therapy for FL are counterbalanced by the benefit of maintaining more prolonged remissions.
Sensitivity analyses showed that the utility before first progression following MR or RIT was the most important factor in all three models, which suggests that improving patients’ HRQL following MR or RIT consolidation could make these approaches more cost-effective. Limited data, however, exist on the HRQL associated with MR or RIT based on small cohorts in trials: a UK study (31) and the FIT study. Additional assessments of HRQL for FL patients in the first remission would be warranted to better understand the robustness of these findings. Moreover, the duration of treatment benefit also played an important role in cost-effectiveness of MR and RIT. The models also showed that treatment can become more cost-effective if the treatment can prolong the duration of remission and reduce the risk of disease related mortality. The cost-effectiveness of MR in the PRIMA model in this study was in line with a published cost-effectiveness analysis (22). The estimates of total QALYs for both arms were slightly lower in our model, which could be attributed to the differences in the initial age distribution and fitted distribution of survival curves.
Our study has limitations. First, there are differences in induction therapies and patient characteristics across trials, which may influence the model parameters, such as risk and utility estimates. This also limited our ability to combine study results and to directly compare the ICER of MR and RIT. Second, the utilities for patients in MR and observation in PRIMA and ECOG models were estimated from a separate UK study (31), since health utility estimates are not available from the original trials. More comprehensive estimates of utilities for the general population would further improve the accuracy and robustness of the model. Third, we did not differentiate the health states based on response to the treatment, due to lack of data concerning response-stratified utility and survival estimates. In addition, in clinical practice, disease progression does not necessarily indicate an immediate treatment for FL patients. Therefore, time-to-next-treatment (TTNT) is preferable to PFS in representing the transition to the next line of treatment. TTNT data were not available, however, for the ECOG trial or relapsed FL studies.
Our analyses showed that the cost-effectiveness of MR and RIT was comparable to other treatments for advanced-stage FL that replaced an older standard of care. For example, the ICERs were $20,000–30,000 per QALY gained for first-line R-chemotherapy regimens compared with chemotherapy alone (30,45,46), for first-line bendamustine with rituximab compared with R-chemotherapy (47), and for MR compared with observation for relapsed FL (32,48). Despite differences in model settings, costs, and health systems between countries, these studies revealed a similar magnitude of ICERs for these treatment strategies for indolent lymphoma. Based on our model results, MR and RIT consolidation following first-line induction therapy for patients with advanced stage FL also appeared to be reasonably cost-effective approaches.
On the other hand, the favorable cost-effectiveness profile of MR and RIT does not imply a uniform approach of selecting treatments in the general population. For an individual patient, selection of either approach or observation should depend on individual patient characteristics, such as performance status. The risks of severe adverse events induced by RIT (such as cytopenias) also should be considered, although these have limited influence on the cost-effectiveness of RIT. Moreover, the length of treatment also may affect the decision. MR requires repeated treatment over two years, whereas RIT requires only one drug infusion, which may be more accessible to patients in certain circumstances (49).
In summary, we used the same modeling framework and consistent parameter estimates to evaluate the cost-effectiveness of MR and RIT compared to observation following frontline treatment for FL patients. All strategies showed favorable cost-effectiveness profile for these approaches to prevent FL disease progression when compared to observation following frontline therapy. While differences in induction therapies in these three trials should be noted when the ICERs of maintenance/consolidation therapies are compared, this work provides the most comprehensive assessment to date comparing the cost effectiveness of these strategies. Future analyses comparing these approaches would benefit from more direct assessment of HRQL during the period of MR or observation following frontline therapy and long term follow-up data from randomized trials directly comparing MR and RIT consolidation.
Supplementary Material
Table 3.
Base case result
| PRIMA model | ECOG model | FIT model | ||||
|---|---|---|---|---|---|---|
| MR | OBS | MR | OBS | RIT | OBS | |
| Life years | ||||||
| Before first progression | 6.239 | 4.520 | 4.751 | 2.451 | 5.042 | 3.204 |
| Total LYs | 9.376 | 8.277 | 8.136 | 6.744 | 8.481 | 7.447 |
| QALYs | ||||||
| Before first progression | 5.491 | 3.977 | 4.181 | 2.157 | 4.235 | 2.659 |
| Total QALYs | 7.643 | 6.553 | 6.505 | 5.107 | 6.597 | 5.572 |
| Costs ($) | ||||||
| Adverse Event | 600 | 161 | 222 | 161 | 3,683 | 354 |
| Before first progression | 63,015 | 9,620 | 71,515 | 5,160 | 61,028 | 6,979 |
| First progression to second progression | 26,612 | 31,463 | 28,216 | 35,630 | 28,913 | 35,184 |
| After second progression | 23,154 | 27,772 | 24,675 | 31,276 | 25,071 | 30,937 |
| Total | 112,781 | 68,856 | 124,406 | 72,066 | 115,012 | 73,099 |
| Incremental values | ||||||
| LYs | 1.099 | 1.391 | 1.034 | |||
| QALYs | 1.089 | 1.399 | 1.026 | |||
| Costs ($) | 43,925 | 52,339 | 41,913 | |||
| Incremental cost-effectiveness ratios (ICERs) | ||||||
| Incremental cost per LY gained | 39,968 | 37,627 | 40,535 | |||
| Incremental cost per QALY gained | 40,335 | 37,412 | 40,851 | |||
Footnotes
Dr. Flowers is an unpaid consultant for Genentech for service on the LymphoCare Scientific Advisory Board and for Celgene for service on the CONNECT CLL Scientific Advisory Board. He reports research support from AbbVie, Acerta, Celgene, Genentech, Gilead Sciences, Infinity Pharmaceuticals, Janssen, Millennium/Takeda, Onyx Pharmaceuticals, Pharmacyclics, and consulting fees from OptumRx and Seattle Genetics for work performed outside of the current study. Partial support for this study was provided by Spectrum
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Chan W, Armitage J, Gascoyne R, et al. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin’s lymphoma. Blood. 1997;89:3909–18. [PubMed] [Google Scholar]
- 2.Swerdlow SH. WHO classification of tumours of haematopoietic and lymphoid tissues. World Health Organization; 2008. [Google Scholar]
- 3.Howlader NNA, Krapcho M, Garshell J, Miller D, Altekruse SF, Kosary CL, Yu M, Ruhl J, Tatalovich Z, Mariotto A, Lewis DR, Chen HS, Feuer EJ, Cronin KA. SEER Cancer Statistics Review, 1975–2011. National Cancer Institute; 2014. [Google Scholar]
- 4.Hoppe BS, Hodgson DC, Advani R, et al. ACR Appropriateness Criteria® Localized Nodal Indolent Lymphoma. Oncology-New York. 2013;27:786–94. [PubMed] [Google Scholar]
- 5.Friedberg J, Huang J, Dillon H, et al. Initial therapeutic strategy in follicular lymphoma (FL): An analysis from the National LymphoCare Study (NLCS). ASCO Annual Meeting Proceedings; 2006. [Google Scholar]
- 6.Freedman A. Follicular lymphoma: 2012 update on diagnosis and management. Am J Hematol. 2012;87:988–95. doi: 10.1002/ajh.23313. [DOI] [PubMed] [Google Scholar]
- 7.Gleeson M, Carolina R, Danese M, et al. Cost of follicular lymphoma (FL) in Medicare patients. ASCO Annual Meeting Proceedings; 2008. [Google Scholar]
- 8.Tan D, Horning SJ. Follicular lymphoma: Clinical features and treatment. Hematology/oncology clinics of North America. 2008;22:863–82. doi: 10.1016/j.hoc.2008.07.013. [DOI] [PubMed] [Google Scholar]
- 9.van Oers MHJ, Kersten MJ. Treatment strategies in advanced stage follicular lymphoma. Best practice & research. Clinical haematology. 2011;24:187. doi: 10.1016/j.beha.2011.03.003. [DOI] [PubMed] [Google Scholar]
- 10.Flowers CR, Armitage JO. A decade of progress in lymphoma: advances and continuing challenges. Clinical Lymphoma Myeloma and Leukemia. 2010;10:414–23. doi: 10.3816/CLML.2010.n.086. [DOI] [PubMed] [Google Scholar]
- 11.Witzig TE, Gordon LI, Cabanillas F, et al. Randomized controlled trial of yttrium-90–labeled ibritumomab tiuxetan radioimmunotherapy versus rituximab immunotherapy for patients with relapsed or refractory low-grade, follicular, or transformed B-cell non-Hodgkin’s lymphoma. J Clin Oncol. 2002;20:2453–63. doi: 10.1200/JCO.2002.11.076. [DOI] [PubMed] [Google Scholar]
- 12.Cheson B. New response criteria for lymphomas in clinical trials. Ann Oncol. 2008;19:iv35–iv38. doi: 10.1093/annonc/mdn191. [DOI] [PubMed] [Google Scholar]
- 13.Hochster H, Weller E, Gascoyne RD, et al. Maintenance rituximab after cyclophosphamide, vincristine, and prednisone prolongs progression-free survival in advanced indolent lymphoma: results of the randomized phase III ECOG1496 Study. J Clin Oncol. 2009;27:1607–14. doi: 10.1200/JCO.2008.17.1561. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Salles G, Seymour JF, Offner F, et al. Rituximab maintenance for 2 years in patients with high tumour burden follicular lymphoma responding to rituximab plus chemotherapy (PRIMA): a phase 3, randomised controlled trial. Lancet. 2011;377:42–51. doi: 10.1016/S0140-6736(10)62175-7. [DOI] [PubMed] [Google Scholar]
- 15.Morschhauser F, Radford J, Van Hoof A, et al. 90Yttrium-ibritumomab tiuxetan consolidation of first remission in advanced-stage follicular non-Hodgkin lymphoma: Updated results after a median follow-up of 7.3 years from the international, randomized, phase III first-line indolent trial. J Clin Oncol. 2013 doi: 10.1200/JCO.2012.45.6400. [DOI] [PubMed] [Google Scholar]
- 16.Morschhauser F, Radford J, Van Hoof A, et al. Phase III trial of consolidation therapy with yttrium-90–ibritumomab tiuxetan compared with no additional therapy after first remission in advanced follicular lymphoma. J Clin Oncol. 2008;26:5156–64. doi: 10.1200/JCO.2008.17.2015. [DOI] [PubMed] [Google Scholar]
- 17.Gondek K, Shah S, Bischof-Delaloye A, et al. Health-related quality of life in patients with advanced-stage follicular lymphoma receiving consolidation with 90Y-Ibritumomab Tiuxetan (Zevalin(R)) of first remission: Results from the randomized phase 3 first-line indolent trial (FIT) ASH Annual Meeting Abstracts. 2007;110:3319. [Google Scholar]
- 18.Canales MA, Dlouhy I, Briones J, et al. A randomized Phase II study comparing consolidation with a single dose of 90Y ibritumomab tiuxetan (Zevalin®)(Z) vs. maintenance with rituximab (R) for two years in patients with newly diagnosed follicular lymphoma (FL) responding to R-CHOP. Preliminary results at 36 months from randomization. Blood. 2013;122:369–69. [Google Scholar]
- 19.Nastoupil LJ, Sinha R, Byrtek M, et al. The use and effectiveness of rituximab maintenance in patients with follicular lymphoma diagnosed between 2004 and 2007 in the United States. Cancer. 2014 doi: 10.1002/cncr.28659. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Illidge TM. Radioimmunotherapy of lymphoma: a treatment approach ahead of its time or past its sell-by date? J Clin Oncol. 2010;28:2944–46. doi: 10.1200/JCO.2009.26.8748. [DOI] [PubMed] [Google Scholar]
- 21.Soini EJ, Martikainen JA, Vihervaara V, et al. Economic evaluation of sequential treatments for follicular non-Hodgkin lymphoma. Clin Ther. 2012;34:915–25. e2. doi: 10.1016/j.clinthera.2012.02.019. [DOI] [PubMed] [Google Scholar]
- 22.Hornberger J, Chien R, Friedmann M, et al. Cost-effectiveness of rituximab as maintenance therapy in patients with follicular non-Hodgkin lymphoma after responding to first-line rituximab plus chemotherapy. Leukemia & Lymphoma. 2012;53:2371–7. doi: 10.3109/10428194.2012.694429. [DOI] [PubMed] [Google Scholar]
- 23.Surveillance, Epidemiology, and End Results (SEER) Program; Surveillance Research Program SSB, editor. Research Data (1973–2011) National Cancer Institute, DCCPS; Apr, 2014. (www.seer.cancer.gov) based on the November 2013 submission. [Google Scholar]
- 24.Sonnenberg FA, Beck JR. Markov models in medical decision making a practical guide. Medical Decis Making. 1993;13:322–38. doi: 10.1177/0272989X9301300409. [DOI] [PubMed] [Google Scholar]
- 25.Archived Consumer Price Index detailed report information. 2013 [Google Scholar]
- 26.Drummond MF, Sculpher MJ, Torrance GW, et al. Methods for the Economic Evaluation of Health Care Programmes. Oxford University Press; 2005. [Google Scholar]
- 27.van Oers MHJ, Klasa R, Marcus RE, et al. Rituximab maintenance improves clinical outcome of relapsed/resistant follicular non-Hodgkin lymphoma in patients both with and without rituximab during induction: results of a prospective randomized phase 3 intergroup trial. Blood. 2006;108:3295–301. doi: 10.1182/blood-2006-05-021113. [DOI] [PubMed] [Google Scholar]
- 28.Rummel MJ, Al-Batran SE, Kim S-Z, et al. Bendamustine plus rituximab is effective and has a favorable toxicity profile in the treatment of mantle cell and low-grade non-Hodgkin’s lymphoma. J Clin Oncol. 2005;23:3383–89. doi: 10.1200/JCO.2005.08.100. [DOI] [PubMed] [Google Scholar]
- 29.Arias E. United States life tables, 2007. National vital statistics reports: from the Centers for Disease Control and Prevention, National Center for Health Statistics, National Vital Statistics System. 2011;59:1. [PubMed] [Google Scholar]
- 30.Papaioannou D, Rafia R, Rathbone J, et al. Rituximab for the first-line treatment of stage III–IV follicular lymphoma (review of Technology Appraisal No. 110): a systematic review and economic evaluation. Health Technol Assess. 2012;16:1366–5278. doi: 10.3310/hta16370. [DOI] [PubMed] [Google Scholar]
- 31.Wild D, Walker M, Pettengell R, et al. Utility elicitation in patients with follicular lymphoma. Value Health. 2006;9:A294–A94. [Google Scholar]
- 32.Deconinck E, Miadi-Fargier H, Le Pen C, et al. Cost effectiveness of rituximab maintenance therapy in follicular lymphoma. Pharmacoeconomics. 2010;28:35–46. doi: 10.2165/11314070-000000000-00000. [DOI] [PubMed] [Google Scholar]
- 33.Curtiss FR, Lettrich P, Fairman KA. What is the price benchmark to replace average wholesale price (AWP)? J Manag Care Pharmacy. 2010;16:492–501. doi: 10.18553/jmcp.2010.16.7.492. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Tumeh JW, Moore SG, Shapiro R, et al. Practical approach for using Medicare data to estimate costs for cost-effectiveness analysis. 2005 doi: 10.1586/14737167.5.2.153. [DOI] [PubMed] [Google Scholar]
- 35.Medicare physician fee schedule. 2013.
- 36.Briggs AH, Weinstein MC, Fenwick EA, et al. Model parameter estimation and uncertainty analysis: A report of the ISPOR-SMDM modeling good research practices task force working group–6. Med Decis Making. 2012;32:722–32. doi: 10.1177/0272989X12458348. [DOI] [PubMed] [Google Scholar]
- 37.Earle CC, Wells GA. An assessment of methods to combine published survival curves. Medical Decision Making. 2000;20:104–11. doi: 10.1177/0272989X0002000113. [DOI] [PubMed] [Google Scholar]
- 38.Press OW, Unger JM, Rimsza LM, et al. Phase III randomized intergroup trial of CHOP plus rituximab compared with CHOP chemotherapy plus 131iodine-tositumomab for previously untreated follicular non-Hodgkin lymphoma: SWOG S0016. J Clin Oncol. 2013;31:314–20. doi: 10.1200/JCO.2012.42.4101. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Press OW, Unger JM, Braziel RM, et al. Phase II trial of CHOP chemotherapy followed by tositumomab/iodine I-131 tositumomab for previously untreated follicular non-Hodgkin’s lymphoma: five-year follow-up of Southwest Oncology Group Protocol S9911. J Clin Oncol. 2006;24:4143–49. doi: 10.1200/JCO.2006.05.8198. [DOI] [PubMed] [Google Scholar]
- 40.Zinzani PL, Tani M, Pulsoni A, et al. Fludarabine and mitoxantrone followed by yttrium-90 ibritumomab tiuxetan in previously untreated patients with follicular non-Hodgkin lymphoma trial: a phase II non-randomised trial (FLUMIZ) Lancet Oncol. 2008;9:352–58. doi: 10.1016/S1470-2045(08)70039-1. [DOI] [PubMed] [Google Scholar]
- 41.Link BK, Martin P, Kaminski MS, et al. Cyclophosphamide, vincristine, and prednisone followed by tositumomab and iodine-131-tositumomab in patients with untreated low-grade follicular lymphoma: Eight-year follow-up of a multicenter phase II study. J Clin Oncol. 2010;28:3035–41. doi: 10.1200/JCO.2009.27.8325. [DOI] [PubMed] [Google Scholar]
- 42.Jacobs SA, Swerdlow SH, Kant J, et al. Phase II trial of short-course CHOP-R followed by 90Y-ibritumomab tiuxetan and extended rituximab in previously untreated follicular lymphoma. Clin Cancer Res. 2008;14:7088–94. doi: 10.1158/1078-0432.CCR-08-0529. [DOI] [PubMed] [Google Scholar]
- 43.Leonard JP, Coleman M, Kostakoglu L, et al. Abbreviated chemotherapy with fludarabine followed by tositumomab and iodine I 131 tositumomab for untreated follicular lymphoma. J Clin Oncol. 2005;23:5696–704. doi: 10.1200/JCO.2005.14.803. [DOI] [PubMed] [Google Scholar]
- 44.Hainsworth JD, Spigel DR, Markus TM, et al. Rituximab plus short-duration chemotherapy followed by yttrium-90 ibritumomab tiuxetan as first-line treatment for patients with follicular non-Hodgkin lymphoma: a phase II trial of the Sarah Cannon Oncology Research Consortium. Clinical Lymphoma and Myeloma. 2009;9:223–28. doi: 10.3816/CLM.2009.n.044. [DOI] [PubMed] [Google Scholar]
- 45.Hornberger J, Reyes C, Lubeck D, et al. Economic evaluation of rituximab plus cyclophosphamide, vincristine and prednisolone for advanced follicular lymphoma. Leuk Lymphoma. 2008;49:227–36. doi: 10.1080/10428190701769665. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Ray JA, Carr E, Lewis G, et al. An Evaluation of the Cost-Effectiveness of Rituximab in Combination with Chemotherapy for the First-Line Treatment of Follicular Non-Hodgkin’s Lymphoma in the UK. Value Health. 2010;13:346–57. doi: 10.1111/j.1524-4733.2009.00676.x. [DOI] [PubMed] [Google Scholar]
- 47.Dewilde S, Woods B, Castaigne J-G, et al. Bendamustine-rituximab: a cost-utility analysis in first-line treatment of indolent non-Hodgkin’s lymphoma in England and Wales. J Med Econ. 2014;17:111–24. doi: 10.3111/13696998.2013.873044. [DOI] [PubMed] [Google Scholar]
- 48.Blommestein HM, Issa DE, Pompen M, et al. Cost-effectiveness of rituximab as maintenance treatment for relapsed follicular lymphoma: results of a population-based study. Euro J Haematol. 2014;92:398–406. doi: 10.1111/ejh.12264. [DOI] [PubMed] [Google Scholar]
- 49.Hagenbeek A. Rituximab maintenance versus radio-immunotherapy consolidation in first remission follicular non-Hodgkin lymphoma. Transfus Apher Sci. 2013;49:43. doi: 10.1016/j.transci.2013.05.024. [DOI] [PubMed] [Google Scholar]
- 50.Guadagnolo BA, Punglia RS, Kuntz KM, et al. Cost-effectiveness analysis of computerized tomography in the routine follow-up of patients after primary treatment for Hodgkin’s disease. J Clinical Oncol. 2006;24:4116–22. doi: 10.1200/JCO.2006.07.0409. [DOI] [PubMed] [Google Scholar]
- 51.Canales MA, Dlouhy I, Briones J, et al. A randomized phase II study comparing consolidation with a single dose of 90y ibritumomab tiuxetan (Zevalin®)(Z) Vs. maintenance with rituximab (R) For two years in patients with newly diagnosed follicular lymphoma (FL) responding to R-CHOP: Preliminary results at 36 months from randomization. Blood. 2013;122:369–69. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.