Bendamustine for the first-line treatment of chronic lymphocytic leukaemia

Clinical effectiveness

4.2 The Committee discussed the place of bendamustine in the clinical pathway for chronic lymphocytic leukaemia. The Committee heard from the clinical specialists that bendamustine is used as a first-line treatment in UK clinical practice when fludarabine combination chemotherapy is not considered an appropriate treatment, that a second round of bendamustine may be offered, and that bendamustine is sometimes used as a second-line treatment. The Committee noted that second-line treatment is currently outside of bendamustine's UK marketing authorisation.

4.3 The Committee heard from the clinical specialists that bendamustine is less toxic than fludarabine combination therapy and is a useful addition to the available treatments for patients with chronic lymphocytic leukaemia in whom fludarabine combination therapy is unsuitable. The only available treatment for these patients is chlorambucil. The Committee heard that although bendamustine is slightly more toxic than chlorambucil and is associated with more adverse events, the clinical specialists considered bendamustine to be the more effective treatment. The Committee also noted the views of the patient groups in their submissions to NICE that because of its improved efficacy compared with chlorambucil, people with the condition would be willing to accept the side effects associated with bendamustine. The Committee was satisfied from the testimonies of the clinical specialists and patient experts that bendamustine represents an important treatment for patients with chronic lymphocytic leukaemia for whom fludarabine combination therapy is not appropriate.

4.4 The Committee heard from the clinical specialists that there are no definitively agreed criteria for deciding when fludarabine combination therapy is unsuitable as a first-line treatment for patients with chronic lymphocytic leukaemia. They commented that there is a growing consensus that patients should be offered the most effective treatment that they can tolerate first. Therefore fludarabine combination therapy (that is, fludarabine, cyclophosphamide and rituximab) is first choice unless there are important factors related to age, physical fitness and the presence of comorbidities to suggest that fludarabine combination therapy should not be used. The Committee accepted, therefore, that any future NICE-recommended use of bendamustine would be determined by clinical judgement based on the factors listed above.

4.5 The Committee discussed the clinical trial data from study 02CLLIII and agreed with the ERG's comments that it was a well-conducted RCT. It noted the higher response rates and longer progression-free survival (21.6 months versus 8.3 months) in patients treated with bendamustine compared with patients treated with chlorambucil. The Committee was concerned, however, about two issues related to the clinical trial evidence. The Committee's first concern was that the trial population may not have been representative of the population that would be treated with bendamustine in clinical practice. The Committee noted the exclusion from trial 02CLLIII of patients with comorbidities including abnormal liver, renal or cardiac function. It was also aware of the high performance status of the majority of participants and the relatively low mean age (63–64 years). However the Committee was reassured by the subgroup analysis conducted by the manufacturer, which demonstrated the clinical effectiveness of bendamustine relative to chlorambucil in the trial participants who had a lower performance status and in patients aged 65 years and over. It also accepted that the exclusion criteria were standard and that there was no reason to suppose that the results would not hold in people with a lower performance status, or people with comorbidities, in particular renal impairment (which is a contraindication of fludarabine). The Committee agreed that inferences could be made about the clinical effectiveness of bendamustine for the population specified in the marketing authorisation, using the available trial data.

4.6 The Committee's second concern about the evidence from study 02CLLIII was that it may have underestimated the clinical effectiveness of chlorambucil. The Committee heard from the clinical specialists how patients treated with chlorambucil in another trial, CLL4 (which compared chlorambucil with fludarabine and fludarabine plus cyclophosphamide), experienced higher response rates and longer progression-free survival compared with the patients treated with chlorambucil in trial 02CLLIII. The Committee discussed the possible reasons for the differences in the results between the two trials. It noted the views of the clinical specialists that the variation in the results may have been because of differences in the patient populations and differences in the doses of chlorambucil used.

4.7 The Committee explored the differences in the patient populations between the two trials. It heard from the manufacturer that one way in which the patient populations of the two trials differed was that the 02CLLIII study did not include people with Binet stage A chronic lymphocytic leukaemia. It also heard from the clinical specialists that the patient population in the CLL4 trial may have been healthier than the patient population in trial 02CLLIII. The Committee was satisfied that the differences between the patient populations in the two trials may have contributed to the differences in the results for chlorambucil.

4.8 The Committee discussed the different doses of chlorambucil used in the 02CLLIII study compared with the CLL4 trial. The clinical specialists explained that the dose used in trial 02CLLIII was consistent with the dose used for other chronic lymphocytic leukaemia studies, and that the dose used in the CLL4 study was unique at the time the study was set up. The clinical specialists also explained that the cumulative dose for chlorambucil in the 02CLLIII study was approximately 85% of the dose used in the CLL4 trial, which might explain the difference in progression-free survival in the chlorambucil arms of the two trials. The Committee heard from the clinical specialists that many different doses of chlorambucil were used in UK clinical practice, but that there was an increasing shift towards the dose used in the CLL4 study, following on from the results of that trial. Furthermore, new chronic lymphocytic leukaemia trials with chlorambucil as a comparator were increasingly using the same dose as was used in the CLL4 study. The Committee accepted that the doses of chlorambucil used in the two trials may have contributed to the differences in the results but that the precise impact of this was unknown, and that this represented an important area for future research. The Committee concluded that there was sufficient evidence to demonstrate that bendamustine was more clinically effective than chlorambucil, leading to higher response rates and longer progression-free survival.

Cost effectiveness

4.9 The Committee discussed the manufacturer's economic model. It agreed with the ERG that the manufacturer's model was of high quality and was more sophisticated than other models in the disease area. The Committee discussed that because of the added complexity of the model, the data to inform some of the model parameters, such as transition probabilities (particularly for second-line treatment) were sparse. On balance, the Committee considered that the model was appropriate and fit for purpose.

4.10 The Committee noted that the cost per QALY gained of treatment with bendamustine compared with chlorambucil was £12,000 in the manufacturer's base-case analysis. The Committee discussed the robustness of the ICER to the subgroup and sensitivity analyses conducted by the manufacturer. The Committee was satisfied with the effect of the subgroup and sensitivity analyses, noting that the ICER remained lower than £15,000 in all of the analyses.

4.11 The Committee was aware that the ERG had made some adjustments to the assumptions used in the manufacturer's economic model about the frequency of blood transfusions, the hazard ratio for overall survival, the dose intensity of bendamustine and chlorambucil and the frequency of visits to a haematologist. The Committee heard from the clinical specialists that the ERG's changes to the assumptions reflected clinical practice. The Committee noted that the effect of these adjustments caused the ICER to fall from £12,000 to £9400, thus becoming more favourable to bendamustine than the base-case ICER presented by the manufacturer. The Committee agreed that the adjustments made by the ERG were reasonable and was satisfied with the resulting effect on the ICER.

4.12 The Committee discussed the potential effect on the ICER of comparing bendamustine with the higher dose of chlorambucil given in the CLL4 study. The Committee considered that an increased dose of chlorambucil would push up the cost of chlorambucil as well as increase the number of QALYs gained. The Committee accepted that since the influence of a higher chlorambucil dose on the clinical effectiveness of the treatment had not been determined, any resulting effect on the ICER could not be quantified with any precision. The Committee agreed, however, that the ICER of bendamustine would be unlikely to increase above the level generally considered to be a cost-effective use of NHS resources. It therefore concluded that bendamustine should be recommended as a first-line treatment option for patients with chronic lymphocytic leukaemia for whom fludarabine combination chemotherapy is not appropriate.

Summary of Appraisal Committee's key conclusions