4 Consideration of the evidence

4.1

The appraisal committee reviewed the data available on the clinical and cost effectiveness of rituximab for the first-line treatment of chronic lymphocytic leukaemia, having considered evidence on the nature of the condition and the value placed on the benefits of rituximab by people with chronic lymphocytic leukaemia, those who represent them, and clinical specialists. It also took into account the effective use of NHS resources.

Clinical effectiveness

4.2

The appraisal committee discussed current standard clinical management of people with chronic lymphocytic leukaemia. The committee heard from clinical specialists that approximately 90% of people with chronic lymphocytic leukaemia are asymptomatic and that diagnosis may be made as a chance finding on routine blood testing. People who are asymptomatic may not need immediate treatment, although some will need treatment later in life. The committee heard that fludarabine in combination with cyclophosphamide is frequently used for people who need immediate treatment. However, chlorambucil alone is normally used for people with poor performance status or comorbidities, especially impaired renal function. The committee heard from patient experts that people with chronic lymphocytic leukaemia who need treatment will often have a series of treatments following first-line treatment, with further lines of treatment used after each relapse. The committee specifically considered the clinical management of people whose chronic lymphocytic leukaemia has the p53 mutation. The committee heard from clinical specialists that these people have a poorer prognosis and are usually treated with alternative treatments (for example, alemtuzumab) rather than chemotherapy.

4.3

The committee noted that the evidence of clinical effectiveness was based mainly on a single unpublished randomised controlled trial (the CLL-8 trial), which compared rituximab in combination with fludarabine and cyclophosphamide with fludarabine and cyclophosphamide. The committee accepted that the CLL-8 trial demonstrated a benefit in progression-free survival, and increased overall and complete response rates for rituximab. The committee heard from clinical specialists that the CLL-8 trial population was younger and fitter than the population of people with chronic lymphocytic leukaemia seen in routine practice within the NHS in England and Wales. However, for the study population fludarabine in combination with cyclophosphamide was the appropriate comparator, and these people reflected the group who would receive fludarabine in combination with cyclophosphamide in clinical practice. The committee noted that an interim analysis of the clinical trial results had demonstrated a statistically significant gain in overall survival but this gain had not been maintained during longer follow-up. The committee accepted that crossover and subsequent lines of treatment in the trial made the overall survival benefit difficult to prove. The committee heard expert opinion that the degree of response to treatment and the duration of progression-free survival were generally accepted as surrogates for overall survival. In addition, the committee heard that cohort studies using historical controls had also shown survival benefits for people treated with rituximab-containing regimens, although results may have been influenced by changing clinical management, such as earlier identification of people with chronic lymphocytic leukaemia. On balance, the committee was persuaded that the benefits observed in progression-free survival and response rate were likely to lead to a gain in overall survival, although currently this would be difficult to quantify.

4.4

The committee recognised that the manufacturer had also provided evidence from uncontrolled phase 2 trials that reported the benefits of adding rituximab to other chemotherapy regimens for the first-line treatment of people with chronic lymphocytic leukaemia. The committee discussed the methodological limitations of obtaining an estimate of clinical effectiveness from uncontrolled and historical comparison studies. In addition, the committee discussed the further evidence from comparative studies of follicular lymphoma provided by the manufacturer in response to consultation on the appraisal consultation document. The committee was specifically mindful of comments from consultees about the addition of rituximab to chlorambucil for the treatment of people who are unable to tolerate fludarabine therapies. It accepted that it was reasonable to expect that rituximab would be of benefit when added to any chemotherapy regimen. However, the committee considered that the additional comparative studies provided by the manufacturer involved the use of more toxic regimens than chlorambucil and were likely to have enrolled people who were on average younger and had a better performance status than people treated with chlorambucil in clinical practice. The committee were not persuaded that the estimates of treatment effects from different studies were transferable. The committee heard from clinical specialists that they considered that there was no evidence to support adding rituximab to chlorambucil, but that there was an ongoing trial investigating this. The committee concluded that there was considerable uncertainty about the relative clinical benefit associated with adding rituximab to chemotherapy regimens other than fludarabine and cyclophosphamide in the treatment of chronic lymphocytic leukaemia.

4.5

The committee was aware that fludarabine and cyclophosphamide were administered intravenously in the CLL-8 trial. It heard from clinical specialists that these chemotherapy agents were routinely administered orally in the NHS. The committee accepted that the efficacy of both methods of administration was equivalent as long as doses were adjusted to ensure equivalent bioavailability.

4.6

The committee heard from patient experts that progression-free survival was associated with a marked improvement in quality of life compared with the symptomatic progressed state. Patient experts commented that the first treatment-induced remission was likely to be the longest and associated with the most substantial improvements in quality of life. For this reason, people with chronic lymphocytic leukaemia valued having a choice of first-line treatments.

Cost effectiveness

4.7

The committee discussed the economic model submitted by the manufacturer. It noted that the manufacturer had only presented estimates of cost effectiveness for rituximab in combination with fludarabine and cyclophosphamide, and that this was compared with fludarabine in combination with cyclophosphamide and chlorambucil monotherapy. The committee heard from clinical specialists that for those people for whom chlorambucil was the most appropriate treatment (that is, people with poor performance status or comorbidities), rituximab in combination with fludarabine and cyclophosphamide would not be considered an appropriate treatment option. Therefore, the committee was not persuaded that the comparison of rituximab in combination with fludarabine and cyclophosphamide with chlorambucil was valid as the chemotherapy regimens were used in different populations and such a choice between the 2 treatments was not expected to be clinically meaningful.

4.8

The committee reviewed the manufacturer's economic model and the critique of it by the ERG. The committee noted that the model was based on all people entering the model in the progression-free survival health state and moving to the progressed health state, and did not allow people to move from the progressed health state to the progression-free survival health state. The committee was mindful that the economic model allowed for costs of subsequent lines of therapy to be included but noted that this did not allow any benefit from further therapy to be taken into account. More importantly a single transition probability from the progressed health state to death was applied to people from both trial groups in the progressed health state. The committee recognised that although the manufacturer had not assumed any relative advantage in survival following progression, the use of a single transition probability from the progressed health state to death had the effect of associating improved progression-free survival with improved overall survival. The committee considered that the assumed association between progression-free survival and overall survival in the model could overestimate the benefits of the clinical effectiveness of rituximab as taken from the CLL-8 trial, because this had not demonstrated a statistically significant difference in overall survival between treatment groups.

4.9

The committee discussed the analysis by the ERG that suggested that two-thirds of the QALY gain (0.64 out of 0.88) in the model was because of the modelled improvement in overall survival, which is driven by gain in progression-free survival being reflected in gain in overall survival. The committee noted that when the overall survival benefit was removed from the model the ICER increased from £13,000 to £30,000 per QALY gained. The committee recognised therefore that the assumption about the amount of gain in overall survival from treatment with rituximab was an important assumption in the economic model and the use of different assumptions could have a large impact on the estimates of cost effectiveness.

4.10

The committee noted that condition-specific quality-of-life data collected in the rituximab trial had not been fully reported in the manufacturer's submission and that the utility values used in the model were not consistent with the NICE reference case because they were not preference-based. The committee considered that the manufacturer may have been able to map the health-related quality-of-life data from the rituximab trial to a preference-based measure to derive utilities and that this may have provided an alternative to the utility data used. The committee noted that, if there was no difference in overall survival between the treatment groups in the model, the results became very sensitive to the difference between the utility values used for the progression-free survival health state and the progressed health state. The committee considered the lack of appropriate utility data contributed to substantial uncertainty in the economic modelling.

4.11

The committee discussed the additional exploratory analysis done by the ERG using an assumption that the actual survival benefit from treatment with rituximab was somewhere between that presented in the base case and an assumption that there was no gain in survival (see section 3.23). The committee noted that the probability, using the base-case utilities (that is, 0.80 for the progression-free survival health state and 0.60 for the progressed health state), of rituximab in combination with fludarabine and cyclophosphamide being cost effective at £20,000 per QALY gained was 71% and at £30,000 per QALY gained was 87%. On balance, the committee was persuaded that even taking into account the additional uncertainty about the utility values, the economic analysis had demonstrated that rituximab in combination with fludarabine and cyclophosphamide for the first-line treatment of chronic lymphocytic leukaemia was a cost-effective use of NHS resources.

4.12

The committee considered the economic analysis provided by the manufacturer after consultation on the appraisal consultation document. It examined the assumptions used in the economic modelling for the comparison of rituximab and chlorambucil with chlorambucil alone. The committee noted that the economic analysis assumed that the hazard ratio observed in the CLL-8 trial was exactly transferable to other chemotherapy regimens. However, the committee noted that subgroup analyses of the CLL-8 trial demonstrated that there were differences in effect between different subgroups of patients based on age and staging. The committee recognised that the marketing authorisation for rituximab allowed its use with any chemotherapy regimen. However, the committee was not persuaded that the relative effects of the treatment or hazard ratio were transferable between various chemotherapies combined with rituximab and between different subgroups of people.

4.13

The committee was also mindful that the base-case ICER for rituximab in combination with chlorambucil from the manufacturer's additional analysis was £22,000 per QALY gained (section 3.25). The committee considered that this estimate would be higher if benefits in progression-free survival did not lead to benefits in overall survival (section 4.9). The committee noted that these uncertainties had increased the base-case ICER estimate for rituximab in combination with fludarabine and cyclophosphamide compared with fludarabine and cyclophosphamide in the main analysis from £13,000 to £30,000 per QALY gained. This estimate was also associated with uncertainty because no preference-based utility values were available (section 4.10). The committee was also aware that the analysis did not allow for the possibility of increased costs and disutilities because of adverse events that a population with poorer performance status or comorbidities may experience. The committee considered that these uncertainties could make the ICER for rituximab in combination with chlorambucil considerably less favourable.

4.14

The committee agreed that the uncertainty regarding the relative clinical effectiveness and the assumptions that had to be included in the additional economic analysis did not support the clinical and cost effectiveness of the use of rituximab in combination with chlorambucil. The committee was therefore not persuaded that it could recommend rituximab in combination with chlorambucil as a cost-effective use of NHS resources. The committee noted that the group of patients who are not suitable for a regimen of fludarabine in combination with cyclophosphamide, and who might therefore be treated with rituximab in combination with chlorambucil, would include a high proportion of people with poor performance status or comorbidities. The committee considered whether the equalities legislation and the requirement for fairness meant that it should make a positive recommendation for rituximab in combination with chlorambucil for this group. However, the committee noted that a negative recommendation for rituximab in combination with chlorambucil did not appear to have an impact on any group protected by the equalities legislation. It is not obvious that there is a clear correlation between the comorbidity factors which rendered this patient group unsuitable for certain chemotherapies and 'disability' as defined in the Disability Discrimination Act 1995. The committee could not be satisfied that a negative recommendation of rituximab in combination with chlorambucil represented less favourable treatment or loss of benefit, given the lack of clear evidence as to the relative clinical effectiveness of rituximab in combination with chlorambucil in this particular patient group. Given the lack of evidence for both the clinical and cost effectiveness of this combination, the committee could not justify a positive recommendation of rituximab in combination with chlorambucil.

4.15

The committee was aware that new data on the combination of rituximab with chlorambucil in chronic lymphocytic leukaemia would soon be available. It considered that even though this was not a comparative trial, the data could potentially provide more certain estimates of cost effectiveness of rituximab in combination with chlorambucil. The committee therefore agreed that the current guidance should be reviewed when all data from the ongoing trial become available.