• cost-effectiveness analysis;
  • economics;
  • lymphoma;
  • rituximab


  1. Top of page
  3. Introduction
  4. Methods
  5. Data Sources
  6. Analyses
  7. Results
  8. Discussion
  9. References

Purpose:  To estimate the cost-effectiveness from a French payer perspective of CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) alone compared with CHOP plus rituximab (R-CHOP) for treatment of patients with diffuse large B-cell lymphoma.

Methods:  Mean patient survival, days of hospitalization, and chemotherapy costs during treatment were estimated from a Phase III trial in France, Belgium, and Switzerland. Survival during the trial was estimated using the Kaplan–Meier method; survival beyond the trial period was projected based on mortality rates from the Scottish and Newcastle Lymphoma Group database. French diagnosis-related group (DRG) payment schedules were applied to trial data to estimate cost of adverse events and drug administration. We estimated survival and cost-effectiveness [the incremental cost per quality-adjusted life-year (QALY) gained] from 4 years (median clinical trial follow-up period) to 15 years, discounted at a fixed annual rate of 3%. We used published patient preferences. We converted currency to euros, based on 2003 exchange rates.

Results:  R-CHOP resulted in a 20.6% relative increase in complete response rate (absolute increase from 63% to 76%), and a 31% decrease in risk of death at 4 years (95% CI 8–49%). Over a 15-year time horizon, mean overall survival (OS) duration was estimated to be 6.90 years for R-CHOP and 5.74 years for CHOP, a mean increase in OS of 1.16 years (or 1.07 QALYs). Total direct medical costs were €13,170 higher with R-CHOP, with an incremental cost-effectiveness ratio of €12,259 per QALY gained.

Conclusion:  R-CHOP significantly increases mean OS up to 4 years compared with CHOP, and its cost-effectiveness ratio compares favorably with other oncology treatments in widespread use.


  1. Top of page
  3. Introduction
  4. Methods
  5. Data Sources
  6. Analyses
  7. Results
  8. Discussion
  9. References

Diffuse large B-cell lymphoma (DLBCL) is the most common non-Hodgkin's lymphoma (NHL). Before the advent of CHOP (cyclophosphamide, doxorubicin, vincristine, and prednisone) chemotherapy in the 1970s, DLBCL resulted in a patient's death within weeks or months of diagnosis [1]. Nevertheless, with CHOP chemotherapy, a complete response (CR) is achievable in approximately 50% of patients and disease-free survival (DFS) occurs in 30% to 40% of patients [2]. The success of these regimens has encouraged research to increase response rates and prolong survival [3–5].

Patients aged 60 and older constitute a substantial proportion of patients with DLBCL, and they experience worse outcomes than patients aged 60 and younger [6]. Attempts to modify regimens to reduce toxicity while maintaining the survival benefits of CHOP in elderly patients have met with limited success [7–9].

In the late 1990s, rituximab (MabThera®, Roche, Switzerland), a chimeric anti-CD20 IgG1 monoclonal antibody, emerged as a novel treatment for the management of patients with relapsing indolent or aggressive lymphoma [10–12]. Phase II studies of rituximab in combination with CHOP in patients with intermediate grade NHL showed a comparable safety profile to CHOP and response rates in excess of 90%[13]. A phase III trial of 399 elderly patients (older than 60 years) with DLBCL, conducted by the Groupe d’Etude des Lymphomes de l’Adulte (GELA LNH 98-5), showed that the addition of rituximab to an 8-week CHOP regimen increases the CR rate from 63% to 76%[14], and reduced the risk of death by 33% at 3-year median follow-up [15].

Because of rising health-care utilization and expenditures relative to other public policy goods and services, there is increased policy interest in forecasting outcomes and economic consequences of novel therapeutic options in oncology [16,17]. In this study, we forecast the effects of rituximab plus CHOP compared with CHOP on years of life gained and health-care costs, and estimate cost-effectiveness.


  1. Top of page
  3. Introduction
  4. Methods
  5. Data Sources
  6. Analyses
  7. Results
  8. Discussion
  9. References

We developed a model that applies to patients aged 60 to 80 that have untreated DLBCL with stage II, III, or IV disease and a performance status of 0 to 2 according to the criteria of the Eastern Clinical Oncology Group. A representative, or “reference-case” patient [18–20] is assumed to have the initial characteristics (age, performance status, stage of illness, B symptoms, number of extranodal sites, bone marrow involvement and age-adjusted IPI score) as the average patient in the LNH 98-5 trial [14]. Characteristics of patients enrolled in LNH 98-5 are summarized in Table 1.

Table 1.  Characteristics of patients in trial GELA LNH 98-5 of treatment of diffuse large B-cell lymphoma and patients treated with CHOP for high-grade lymphoma/diffuse large B-cell lymphoma in the Scottish and Newcastle Lymphoma Group (SNLG) database
  • Performance status was defined according to the criteria of the Eastern Cooperative Oncology Group (with an increasing score indicating declining performance).

  • *

    Percentages do not add to 100% because of rounding.

Number of patients197202816
Age (year)*
Performance status*
 I 1%025%
B symptoms36%39%NA
Number of extranodal sites
Age-adjusted International
Prognostic Index score

The analyses compare two treatments: 1) CHOP, or 2) rituximab plus CHOP (R-CHOP). CHOP consisted of a combination of cyclophosphamide, 750 mg/m2 on day 1; doxorubicin, 50 mg/m2 on day 1; vincristine, 1.4 mg/m2, up to a maximal dose of 2 mg, on day 1; and prednisone 40 mg/m2/day for 5 days. CHOP was administered every 3 weeks for eight cycles. R-CHOP consisted of CHOP plus rituximab, 375 mg/m2, on day 1 of each of the 8 cycles of CHOP. Doses of cyclophosphamide and doxorubicin were adjusted for toxicity [14]. The doses of rituximab were not modified, but rituximab was discontinued when CHOP was stopped. Treatment also was stopped if progression of the lymphoma or severe toxicity occurred.

Data Sources

  1. Top of page
  3. Introduction
  4. Methods
  5. Data Sources
  6. Analyses
  7. Results
  8. Discussion
  9. References

Three data sources were used for the model. Our first data source was the LNH 98-5 trial data; details about the trial design and primary clinical findings are published elsewhere [14]. We used the LNH 98-5 data to estimate 1) mean overall and DFS; 2) incidence of hospitalizations for adverse events and drug administration; and 3) the cumulative doses of chemotherapeutic agents.

Our second data source was the Scottish and Newcastle Lymphoma Group (SNLG) database, used to estimate the tail of the survival distribution beyond the 48 months of median follow-up of the LNH 98-5 data. The SNLG database was established in 1979. Since 1994, it has captured comprehensive data on treatment and outcomes on more than 95% of the lymphomas presenting in a population of 8.5 million [21]. Approximately 1000 new patients are added to the database each year, with participating clinicians providing a yearly update on existing patients. The SNLG team provided data on 1359 patients with DLBCL/high-grade lymphoma who were treated with CHOP, with a median age of 63 years (53.6% of the patients were younger than 65 years of age; 16.7% were older than 75 years of age). The SNLG database used the Kiel classification until 1995, under which DLBCL patients would have made up the majority of high-grade lymphoma patients. The CR rates, DFS, and overall survival (OS) for the 816 patients aged 60 and above who received CHOP are shown in Table 2. To be consistent with the age distribution in the GELA trial we used the data for the persons aged 60 years and above to estimate OS and DFS beyond the trial period. Characteristics of the 816 patients are shown in Table 1.

Table 2.  Outcomes after treatment with CHOP for high-grade lymphoma/DLBCL among patients aged 60 or more in the Scottish and Newcastle Lymphoma Group database
Complete response rateDisease-free survival for patients achieving complete response (months)*Overall survival (months)*
Median (95% CI)Mean (95% CI)Median (95% CI)Mean (95% CI)
  • *

    Kaplan–Meier estimates; restricted means were calculated.

  • CI, confidence interval; DLBCL, diffuse large B-cell lymphoma.

62.1%Not reached at 140 months87 (79.5–95.6)33 (26.1–39.9)55 (50.1–60.8)

Third and last, we used published data to estimate the costs associated with cancer surveillance care (routine monitoring), salvage therapy (intensive chemotherapy only), bone marrow transplantation, end-of-life care (palliative care), hospitalizations for drug administration and adverse events, and chemotherapy acquisition costs, and to determine adjustments for quality of life.


  1. Top of page
  3. Introduction
  4. Methods
  5. Data Sources
  6. Analyses
  7. Results
  8. Discussion
  9. References


Survival data were collected for as many patients as possible. Where no additional data were available, patients were to be censored at the date last seen. All patients who were alive after the last database closure on January 31, 2002 were to be contacted by the GELA LNH 98-5 study team and information was obtained about their survival status.

All randomized patients (N = 399) were included in the survival update analysis and the data cut-off was July 1, 2003. The median follow-up time, defined for all patients alive as the difference between date of randomization and date of last contact, was 1475 days (approximately 48 months). Median observation time was similar in the two treatment groups: 1485 days in the CHOP group and 1464 days in the R-CHOP group. All patients, except one patient, had a follow-up of more than 36 months. About one third of the population was followed for 48 months or longer.

The relative risk of death among CHOP-treated patients in SNLG was 9% higher than for patients treated with CHOP in LNH 98-5. To forecast survival with CHOP, we recalibrated the mortality risks from SNLG to match those observed in LNH 98-5. Outcomes in nontrial settings may not reflect those found in trials, so we tested in further analyses whether the forecasts were sensitive to this calibration. The effect of R-CHOP mortality risk was based on the 31% further reduction estimated from the 4-year follow-up analyses. Response rates were based on the LNH 98-5 data. Because of the limited number of patients observed in SNLG after 15 years, we used 15 years for the forecast period, or time horizon.

Mortality rates in both CHOP and R-CHOP after 48 months were based on rates derived from CHOP-treated patients in the SNLG database. This implies that R-CHOP has no benefit compared with CHOP on risk of death beyond the 4-year update of the LNH 98-5 trial. The survival curve is shown in Figure 1. We estimated DFS for patients with CR, from the date of randomization to the date of the first relapse.


Figure 1. Survival (trial-based up to 4 years and projected using Scottish and Newcastle Lymphoma Group data). CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisone.

Download figure to PowerPoint

For some patients aged 60 and older who fail to achieve a response, an option is “salvage” treatment with high-dose chemotherapy with or without radiation therapy. Some patients also may be eligible for transplantation [22]. In a selected population of patients who relapse, mortality rates of transplantation compared with conventional therapy were reduced by 30%[23]. Other studies have confirmed the benefits of various transplantation techniques in patients who fail to respond to primary therapy [24–26]. Before the advent of transplantation, mean survival of elderly patients who failed to respond to primary treatment was 1 to 1.5 years [6]. Based on the 4-year LNH 98-5 data, we assumed that salvage therapy (intensive chemotherapy only) would be offered to 25% of R-CHOP patients and 43% of CHOP patients who failed to respond to primary therapy. Salvage therapy was assumed to increase OS an additional 0.5 years [27]. One patient in the LNH 98-5 trial received transplantation; for the base case, we assumed 0% probability of receiving transplantation in the French setting.

Quality-Adjusted Survival

Survival was partitioned into two periods: DFS and postprogression survival. Mean DFS was computed as the product of the CR rate and mean time to progression among patients with CR. Postprogression survival was computed as the difference of mean OS and mean DFS. We adjusted outcomes for quality of life using published utility estimates (scaled between 0 for death and 1 for normal health) [28–30, J. K. Doorduijn, et al., pers. comm., 2003]. For our base case analysis, we calculated utility scores for DFS and progression based on a study by Doorduijn et al. of 132 of 411 patients aged 65 to 90 with aggressive NHL who were entered into a randomized multicenter study comparing CHOP with CHOP plus G-CSF [28, J. K. Doorduijn, et al., pers. comm., 2003]. We weighted the 3-month utility scores from Doorduijn et al.'s study [28] by the proportion of patients in the GELA study who had a CR and no CR/progression to obtain a utility score of 0.83 for DFS and 0.39 for progression. We accounted for a wide range in utility scores in our sensitivity analyses.


The cost of cancer surveillance was estimated to equal €308 per year [31]. Chemotherapy administration costs were calculated by estimating drug acquisition costs and costs of administering the drug. All chemotherapy was administered at the hospital—as inpatient (i.e., spent at least one night in the hospital) or outpatient (patient did not stay in hospital overnight). LNH 98-5 files were analyzed to assess whether a patient was hospitalized, the reason for hospitalization, and total days of hospitalization per cycle. Each adverse event requiring hospitalization was classified by the principle diagnosis using the International Statistical Classification of Diseases (Version 10, WHO, 2003) [32], and as outpatient or inpatient, based on length of stay. Hospital admission data collection in LNH 98-5 occurred only until the treatment period or the patient chose to drop out of the trial.

The complete costs in euro in 2003 of the French diagnosis-related groups (DRGs) were used to estimate hospitalization costs for drug administration and treatment of adverse events. Patient admissions for chemotherapy administration and adverse events were classified in a DRG based on his/her principle diagnosis and length of stay, inpatient or outpatient DRG. When several adverse events in the same cycle leading to a hospitalization were documented for a patient a Delphi panel consisting of 5 GELA and Groupe Ouest-Est d’étude des leucémies et autres maladies du sang (GOELAMS) members selected the hospitalization principle diagnosis. The complete cost per DRG was estimated, including a daily cost component (fixed costs) and a variable cost component (direct medical costs) [33]. The theoretical cost of a stay was developed based on the DRG where a patient was classified, its length of stay, and specific information related to the drug acquisition costs.

Two studies have reported costs in Europe of transplantation for lymphoma. In France, Perrier et al. reported an average cost per patient of €34,630 for autologous blood progenitor cell transplantation for 95 patients with lymphoma, including costs for mobilization, treatment period, and secondary hospitalization [34]. In The Netherlands, van Agthoven et al. reported costs of €15,000 to €19,000 for patients undergoing autologous peripheral blood stem cell transplantation or autologous bone marrow transplantation, respectively, for refractory or relapsed non-Hodgkin's lymphoma or Hodgkin's disease [35]. In the base case, no patients received transplantation. We used €15,000 for the cost of high-dose chemotherapy after relapse. In sensitivity analyses, we applied a cost of €34,630 and probability of 3% to 8% for transplantation in the event of nonresponse. We also assumed that end-of-life and palliation care was €15,000. All costs were converted to euro based on exchange rate of €1 equals 6.56 French francs in 2003 [36].

Perspective, Cost-Effectiveness, and Other Analyses

The  analysis  was  performed  from  the  perspective  of the French Social Security system (i.e., French payer). The French Social Security funds public hospitals through DRGs and reimburses certain innovative drugs to the hospital. Patients pay a small fee per day for a hospitalization, which is generally fully funded by a third party payer (private insurance). In many instances, priority is given to interventions that improve survival, with economic considerations focusing on direct medical costs. Cost-effectiveness thus was estimated as the difference in direct medical costs between R-CHOP and CHOP divided by the difference in mean survival between the two regimens. All costs and benefits were discounted at a fixed annual rate of 3%, a standard method used in economic evaluation to account for time preferences [18–20].

We determined distributions for each of the key variables using either a uniform or normal distribution. Probabilistic simulation analyses were conducted to estimate the impact of varying all assumptions simultaneously. In one-way sensitivity analysis, we looked at the effect of each variable independently when assigned extreme values of their distributions. For uniform distribution, we used the highest and lowest values in the distribution. For normal distributions, we used the 5th and 95th percentiles of the distributions.


  1. Top of page
  3. Introduction
  4. Methods
  5. Data Sources
  6. Analyses
  7. Results
  8. Discussion
  9. References

Overall and DFS

The relative risk of death over 4 years was reduced by 31% in the R-CHOP group (univariate Cox regression, 95% CI 8–49%). Kaplan–Meier estimates of the survival rates at 12 months were 69.5% for CHOP and 82.7% for R-CHOP. At 24 months, the difference was less pronounced but still present: 57% in CHOP compared with 70% in R-CHOP [14]. Thereafter, there was little change in the survival rates. At 4 years, 47.2% of patients receiving CHOP were alive, compared with 59.6% receiving R-CHOP. In 48 months of follow-up in the LNH 98-5 study, the mean difference in OS between R-CHOP and CHOP was 0.33 years. Projected to 15 years, R-CHOP resulted in mean OS of 6.90 years compared with 5.74 years in mean OS for CHOP, mean difference in OS equal to 1.16 undiscounted years (or 1.00 discounted years, as shown in Table 3).

Table 3.  Discounted survival results over 15-year time horizon (years)
Disease-free survival3.805.40
Non–disease-free survival1.040.44
Total years of survival4.845.83
Quality-adjusted life-years3.594.66

R-CHOP increased CR rate by an absolute of 13%, from 63% for CHOP to 76% for R-CHOP. R-CHOP also reduced the relative risk of relapse among patients with CR by 43.9% (univariate Cox regression, 95% CI 19–61%) after 4 years. Considering the higher response rate and longer duration until relapse associated with R-CHOP, the mean DFS increased from 4.39 years to 6.27 years, for a mean difference of 1.88 undiscounted years over 15 years. Non-DFS was projected to be 1.27 years for CHOP compared with 0.59 for R-CHOP, for a mean difference of 0.68 years (undiscounted) over 15 years. R-CHOP resulted in 5.45 quality-adjusted life-years (QALYs) compared with 4.17 QALYs for CHOP, for a difference of 1.28 QALYs, undiscounted (or 1.07 discounted QALYs).

Hospitalizations, Costs, and Cost-Effectiveness

The mean total cost of chemotherapy agents for the R-CHOP group was €14,364 higher than for CHOP (Table 4). Patients receiving R-CHOP completed more courses of chemotherapy, on average (7.5 courses), compared with CHOP patients (7.1 courses) because patients receiving CHOP progressed more often on therapy than R-CHOP-treated patients. Thus, the cost of hospitalization related to chemotherapy was higher for R-CHOP patients.

Table 4.  Mean costs of CHOP versus R-CHOP (2003 euro) for treatment of diffuse large B-cell lymphoma over 15 years
CategoryCHOP (€)R-CHOP (€)
  1. CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisone; R-CHOP, rituximab plus CHOP.

Drug acquisition
 Rituximab 14,095
 For adverse events3,0973,014
 For chemotherapy5,8026,764
Salvage therapy (high dose chemotherapy)2,358890

Table 5 shows the adverse events that occurred in the trial during the treatment period and were used in the analysis; only events that occurred in at least five patients are shown. There was no difference in costs for hospitalizations for adverse events based on DRG payment schedules for specific diagnoses. Patients receiving R-CHOP had slightly fewer days in hospital during the 6-month treatment period. The average number of hospitalizations without an overnight stay was 15.1 for CHOP, compared with 14.7 for R-CHOP patients. The average number of inpatient hospitalization days was 5.4 for CHOP compared  with  4.9  for  R-CHOP  patients.  Based  on DRG methods, overall, mean hospital costs increased by €879. Post-treatment surveillance increased costs by €342. Mean salvage therapy costs decreased by €1468 and end-of-life care decreased by €948.

Table 5.  Diagnoses and incidence*, by diagnosis-related group (DRG)
CHOP (N = 197)R-CHOP (N = 202)
  • *

    Diagnoses included if at least 5 occurrences.

574Febrile neutropenia1013
671Health status alteration—asthenia76
256Abdominal and gastric pains55
604Septic shock09

Total direct medical costs increased with R-CHOP by €13,170 over 15 years. The incremental cost-effectiveness for 4 years (the median follow-up currently available for the trial) was €37,168 per life-year gained and €29,976 per QALY gained. The incremental cost-effectiveness at 15 years was estimated to be €11,392 per life-year gained and €12,259 per QALY gained.

Sensitivity Analyses

We conducted further analyses to determine if R-CHOP may be more or less cost-effective under less likely scenarios using one-way and probabilistic sensitivity analyses (Table 6). The most influential variable  on  cost-effectiveness  in  the  analysis  was the time horizon. When using a 4-year horizon, the cost-effectiveness ratio equaled €29,976 per QALY gained, compared with €17,120 and €12,259 when using an 8-year or 15-year horizon, respectively. Other variables had little influence on the cost-effectiveness ratio, all being less than €17,000. In probabilistic analyses, which simulate the parameters’ probabilistic distributions, the 5th and 95th percentiles of the cost-utility distribution were €6,818 and €22,789, respectively (Fig. 2).

Table 6.  Results of one-way sensitivity analyses (euro)
VariableBase case estimateRangeCost per quality adjusted life-year gained (€)
  1. CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisone; R-CHOP, rituximab plus CHOP.

Time horizon (years)15  41529,97612,259
Rituximab cost per cycle1,9141,665 2,20110,55214,427
Utility for disease-free0.830.720.9514,66310,399
Utility for progression0.390.340.8311,89816,736
Discount rate (%)3  0510,51913,497
Relative risk reduction in disease-free survival (%)43.9 385112,83911,616
Relative risk reduction in overall survival (%)31 273613,05111,350
Relative increase in complete response rates (%)20.6 182412,92211,481
Salvage therapy cost (high-dose chemotherapy)15,0005,00025,00013,17011,348
Probability of receiving salvage
 R-CHOP (%)254312,756
 CHOP (%)43 2513,021
Life extension with salvage therapy0.5  0111,91512,624
Surveillance costs per year308  052511,94012,484
CHOP cost per cycle672 58577312,22712,297
Probability that patients who receive salvage therapy also get transplantation (%); transplantation costs = 34,6300312,25912,165
Probability that patients who receive salvage therapy also get transplantation (%); transplantation costs = 34,6300812,25912,007
Probability that patients who receive salvage therapy also get transplantation (%); transplantation costs = 40,0000812,25911,968
End-of-life care15,0005,00025,00012,84711,671
Future unrelated medical costs0  0 5,00012,25917,438

Figure 2. Probabilistic analysis of CHOP versus R-CHOP for diffuse large B-cell lymphoma. CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisone; R-CHOP, rituximab plus CHOP.

Download figure to PowerPoint


  1. Top of page
  3. Introduction
  4. Methods
  5. Data Sources
  6. Analyses
  7. Results
  8. Discussion
  9. References

The addition of rituximab to CHOP was shown in a phase III trial to significantly improve the rate of CR and duration of event-free survival, DFS, and OS for elderly patients with DLBCL. R-CHOP also was well tolerated, had a similar incidence of serious adverse events compared with CHOP, and more patients were able to complete all eight courses of chemotherapy [14]. The economic analyses presented here show that R-CHOP has a projected cost-effectiveness ratio less than €20,000 per QALY gained.

Numerous economic analyses have been conducted in oncology [15]. No cost-effectiveness studies have been published for DLBCL, but there exist studies for other hematological malignancies. The cost-effectiveness of interferon-alpha therapy versus hydoxyurea therapy for chronic myelogenous leukemia (CML) was estimated to be $37,000 per QALY gained for a representative 50-year-old patient with median duration survival of 69 and 58 months, respectively [15,37]. Interferon alpha for CML increased quality-adjusted life expectancy by 15.5 months compared with conventional chemotherapy, for a cost per QALY of $96,000 [15–38]. Messori estimated that maintenance treatment for CML with interferon improved lifetime survival from 56 to 93 years (discounted) per 100 patients compared with busulfan or hydroxyurea, at $226,000 per life-year gained [39].

A key issue of this analysis was how long to model survival beyond the end of the trial. International economic guidelines recommend assessing implications of lifetime forecasts [40], allowing all potential benefits and costs to be considered. This is of particular relevance for R-CHOP because most of the costs are incurred within the first 6 months, but the benefit in survival persists at the median follow-up of 4 years. The longest time horizon in our analyses was 15 years, where at least 20% of patients in low-risk subgroups are likely to be alive at the end of 15 years [6].

A survival advantage persisted more than 4 years after last dose of R-CHOP was administered in the LNH 98-5  trial,  thereby  suggesting  the  possibility  of increasing the rate of cures and, thus, long-term DFS. There exists no published consensus about how long a patient should be followed before confidently claiming that a patient is cured. Fisher et al. considered a patient cured if he had not experienced a relapse or death by 2 years [41]. Armitage et al. reported that a continuous remission for 2 years has approximately 70% to 90% likelihood of maintaining a durable cure [42]. By contrast, Frei et al. stated that the DFS needed to be maintained up to 4 years [43]. In Shipp et al. the mortality rates declined to baseline at 3 to 4 years after initiating treatment [6]. Lee et al. found that patients with advanced-stage diffuse large-cell lymphoma treated with MACOP-B had a 78% chance of being in remission at 10 years if they were in remission after 2 years [44]. Additional follow-up is warranted to learn whether the persistent survival advantage of R-CHOP found at 4 years translates into long-term cures of DLBCL.

Costs, costing methodologies, and treatment patterns vary by different settings and patient populations (e.g., by center, country, and patient subgroups) [45]. The cost-effectiveness results may not be generalizable to other countries or to other patient populations. Economic analyses from additional studies will help to further refine the cost-effectiveness of R-CHOP.

Adding rituximab to CHOP significantly improves survival for elderly patients with DLBCL. R-CHOP is even cost-effective [46,47] during the period observed in the trial and compares favorably with other oncology treatments [16]. If the survival advantage persists on further follow-up, then rituximab will be among the more cost-effective therapies available in oncology.

The authors would like to acknowledge Jo White for her assistance in analyzing the SNLG data.

Funding: Research was funded in part by F. Hoffmann-La Roche, Ltd.


  1. Top of page
  3. Introduction
  4. Methods
  5. Data Sources
  6. Analyses
  7. Results
  8. Discussion
  9. References
  • 1
    DeVita VJ, Canellos G, Chabner B, et al. Advanced diffuse histiocytic lymphoma, a potentially curable disease. Lancet 1975;1: 248.
  • 2
    Coltman C, Dahlberg S, Jones S. CHOP is curative in thirty percent of patients with large cell lymphoma: a twelve-year Southwest oncology group follow-up. In: SkarinA, ed., Advances in Cancer Chemotherapy. New York: Park Row, 1986.
  • 3
    Yi P, Coleman M, Saltz L, et al. Chemotherapy for large cell lymphoma: a status update. Semin Oncol 1990;17: 6073.
  • 4
    Fisher R, Gaynor E, Dahlberg S, et al. Comparison of a standard regimen (CHOP) with three intensive chemotherapy regimens for advanced non-Hodgkin's lymphoma. N Engl J Med 1993;328: 10026.
  • 5
    Messori A, Vaiani M, Trippoli S, et al. Survival in patients with intermediate or high grade non-Hodgkin's lymphoma: meta-analysis of randomized studies comparing third generation regimens with CHOP. Br J Cancer 2001;84: 3037.
  • 6
    Shipp M, Harrington D, Anderson J, et al. A predictive model for aggressive non-Hodgkin's lymphoma. N Engl J Med 1993;329: 98794.
  • 7
    Zinzani P, Storti S, Zaccaria A, et al. Elderly aggressive-histology non-Hodgkin's lymphoma: First-line VNCOP-B regimen experience on 350 patients. Blood 1999;94: 338.
  • 8
    Balducci L, Extermann M. Management of cancer in the older person: a practical approach Oncologist. 2000;5:22437.
  • 9
    Tirelli U, Errante D, Van Glabbeke M, et al. CHOP is the standard regimen in patients > or = 70 years of age with intermediate-grade and high-grade non-Hodgkin's lymphoma: results of a randomized study of the European Organization for Research and Treatment of Cancer Lymphoma Cooperative Study Group. J Clin Oncol 1998;16: 2734.
  • 10
    Maloney D, Grillo-Lopez A, White C, et al. IDEC-C2B8 (rituximab) anti-CD20 monoclonal antibody therapy in patients with relapsed low-grade non-Hodgkin's lymphoma. Blood 1997;90: 218895.
  • 11
    McLaughlin P, Grillo-Lopez A, Link B, et al. Rituximab chimeric anti-CD20 monoclonal antibody therapy for relapsed indolent lymphoma: half of patients respond to a four-dose treatment program. J Clin Oncol 1998;16: 282533.
  • 12
    Coiffier B, Haioun C, Ketterer N, et al. Rituximab (anti-CD20 monoclonal antibody) for the treatment of patients with relapsing or refractory aggressive lymphoma: a multicenter phase II study. Blood 1998;92: 192732.
  • 13
    Vose J, Link B, Grossbard M, et al. Phase II study of rituximab in combination with CHOP chemotherapy in patients with previously untreated, aggressive non-Hodgkin's lymphoma. J Clin Oncol 2001;19: 38997.
  • 14
    Coiffier B, Lepage E, Briere J, et al. CHOP chemotherapy plus rituximab compared with CHOP  alone  in  elderly  patients  with  diffuse  large-B-cell lymphoma. N Engl J Med 2002;346: 23542.
  • 15
    Coiffier B, Herbrecht H, Tilly C. LNH 98-5 study comparing CHOP and R-CHOP in elderly patients with DLCL. 3-year median follow-up with an analysis according to co-morbidity factors. In: 2003 Meeting of the American Society of Clinical Oncology (Abstract 2395). Chicago, IL: American Society of Clinical Oncology, 2003.
  • 16
    Earle C, Chapman R, Baker C, et al. Systematic overview of cost-utility assessments in oncology. J Clin Oncol 2000;18: 330217.
  • 17
    Brown M, Glick H, Harrell F, et al. Integrating economic analysis into cancer clinical trials: the National Cancer Institute—American Society of Clinical Oncology Economics Workbook. J Natl Cancer Inst Monogr 1998;24: 128.
  • 18
    Russell L, Gold M, Siegel J, et al. The role of cost-effectiveness analysis in health and medicine. Panel on Cost-Effectiveness in Health and Medicine. JAMA 1996;276: 11727.
  • 19
    Siegel J, Weinstein M, Russell L, et al. Recommendations for reporting cost-effectiveness analyses. Panel on Cost-Effectiveness in Health and Medicine. JAMA 1996;276: 133941.
  • 20
    Weinstein M, Siegel J, Gold M, et al. Recommendations of the Panel on Cost-effectiveness in Health and Medicine. JAMA 1996;276: 12538.
  • 21
    Proctor SJ, Mackie M, Dawson A, et al. A population-based study of intensive multi-agent chemotherapy with or without autotransplant for the highest risk Hodgkin's disease patients identified by the Scotland and Newcastle Lymphoma Group (SNLG) prognostic index. A Scotland and Newcastle Lymphoma Group study (SNLG HD III). Eur J Cancer 2002;38: 795806.
  • 22
    Stockerl-Goldstein KE, Reddy SA, Horning SF, et al. Favorable treatment outcome in non-Hodgkin's lymphoma patients with “poor” mobilization of peripheral blood progenitor cells. Biol Blood Marrow Transpl 2000;6: 50612.
  • 23
    Philip T, Guglielmi C, Hagenbeek A, et al. Autologous bone marrow transplantation as compared with salvage chemotherapy in relapses of chemotherapy-sensitive non-Hodgkin's lymphoma (Comment). N Engl J Med 1995;333: 15405.
  • 24
    Hamlin PA, Zelenetz AD, Kewalramani T, et al. Age-adjusted international prognostic index predicts autologous stem cell transplantation outcome for patients with relapsed or primary refractory diffuse large B-cell lymphoma. Blood 2003;102: 198996.
  • 25
    Kewalramani T, Zelenetz AD, Hedrick EE, et al. High-dose chemoradiotherapy and autologous stem cell transplantation for patients with primary refractory aggressive non-Hodgkin lymphoma: an intention-to-treat analysis. Blood 2000;96: 2399404.
  • 26
    Vose JM, Zhang MJ, Rowlings PA, et al. Autologous transplantation for diffuse aggressive non-Hodgkin's lymphoma in patients never achieving remission: a report from the autologous blood and marrow transplant registry. J Clin Oncol 2001; 19: 40613.
  • 27
    Philip T, Guglielmi C, Hagenbeek A, et al. Autologous bone marrow transplantation as compared with salvage chemotherapy in relapses of chemotherapy-sensitive non-Hodgkin's lymphoma. N Engl J Med 1995;333: 15405.
  • 28
    Doorduijn J, Buijt I, Uyl-de Groot C, et al. Quality of life (QOL) in elderly patients with aggressive non-Hodgkin's lymphoma (NHL) treated with CHOP (Abstract). Blood 2001;98: 1803.
  • 29
    Djulbeovic B, Hollenberg J, Woodcock T, Herzig R. Comparison of different treatment strategies for diffuse large-cell lymphomas: a decision analysis. Med Decis Making 1991;11: 18.
  • 30
    Uyl-de Groot CA, Okhuijsen SY, Hagenbeek A. Costs of introducing autologous BMT in the treatment of lymphoma and acute leukaemia in The Netherlands. Bone Marrow Transplant 1995; 16: 46370.
  • 31
    Edelman M, Meyers F, Siegel D. The utility of follow-up testing after curative cancer therapy. A critical review and economic analysis. J Gen Intern Med 1997;12: 31831.
  • 32
    ICD-10. The international statistical classification of diseases and related health problems, tenth revision. Available from:[Accessed June 10, 2004].
  • 33
    Ministère de l’Emploi et de la Solidarité. Journaux Officiels. Dans: L’Echelle Nationale de Coûts par Groupe Homogène de Malades 1999: Bulletin Officiel. 1999;99/2 bis: 3955.
  • 34
    Perrier L, Sebban C, Philip I, et al. Necessary harmonization of health cost assessment. Autologous Peripheral Blood Progenitor Cell Transplantation in France Revue D Epidemiologie Et De Sante Publique 2002;50: 393403.
  • 35
    Van Agthoven M, Vellenga E, Fibbe WE, et al. Cost analysis and quality of life assessment comparing patients undergoing autologous peripheral blood stem cell transplantation or autologous bone marrow transplantation for refractory or relapsed non-Hodgkin's lymphoma or Hodgkin's disease. A prospective randomised trial. Eur J Cancer 2001;37: 17819.
  • 36
    FXHistory: historical currency exchange rates. Available from:[Accessed June 10, 2004].
  • 37
    Kattan M, Inoue Y, Giles F, et al. Cost-effectiveness of interferon-alpha and conventional chemotherapy in chronic myelogenous leukemia. Ann Intern Med 1996;125: 5418.
  • 38
    Liberato N, Quaglini S, Barosi G. Cost-effectiveness of interferon alfa in chronic myelogenous leukemia. J Clin Oncol 1997;15: 267382.
  • 39
    Messori A. Cost-effectiveness of interferon in interferon in chronic myeloid leukaemia: analysis of four clinical studies. Ann Oncol 1998;9: 38996.
  • 40
    Weinstein MC, O'Brien B, Hornberger J, et al. Principles of good practice for decision analytic modeling in health-care evaluation: report of the ISPOR Task Force on Good Research Practices—Modeling Studies. Value Health 2003;6: 917.
  • 41
    Fisher RI. Treatment of aggressive non-Hodgkin’ lymphomas. Lessons from the past 10 years. Cancer 1994;74: 265761.
  • 42
    Armitage J, Cheson B. Interpretation of clinical trials in diffuse large-cell lymphoma. J Clin Oncol 1988;6: 133547.
  • 43
    Frei E. Curative cancer chemotherapy. Cancer Res 1985;45: 652337.
  • 44
    Lee A, Connors J, Klimo P, et al. Late relapse in patients with diffuse large-cell lymphoma treated with MACOP-B. J Clin Oncol 1997;15: 174553.
  • 45
    Frenk J, Gómez-Dantés O. Globalization and the challenges to health systems. Health Affairs 2002; 21: 1605.
  • 46
    Laupacis A, Feeny D, Detsky AS, Tugwell PX. Tentative guidelines for using clinical and economic evaluations revisited. CMAJ 1993;148: 9279.
  • 47
    Ubel PA, Hirth RA, Chernew ME, Fendrick AM. What is the price of life and why doesn’t it increase at the rate of inflation? Arch Intern Med 2003; 163: 163741.