4 Consideration of the evidence

4.1 The Appraisal Committee reviewed the data available on the clinical and cost effectiveness of fludarabine monotherapy for the treatment of chronic CLL, having considered evidence on the nature of the condition and the value placed on the benefits of fludarabine by people with CLL, those who represent them, and clinical specialists. It was also mindful of the need to take account of the effective use of NHS resources.

4.2 The Committee noted that the decision problem submitted by the manufacturer included both fludarabine monotherapy and combination therapy with cyclophosphamide. However, the Committee noted that the current marketing authorisation does not specifically provide a recommendation that fludarabine should be used concurrently with other drugs for the treatment of CLL. Therefore the Committee was unable to issue guidance on fludarabine plus cyclophosphamide.

Clinical effectiveness

4.3 The Committee considered the ERG's review of the manufacturer's submission on the clinical effectiveness of fludarabine monotherapy compared with chlorambucil, and noted that the manufacturer did not make use of evidence from all relevant RCTs to provide a more precise estimate of the clinical effectiveness of the treatments. The Committee discussed the appropriateness of basing the clinical effectiveness of fludarabine monotherapy on preliminary results from the CLL4 study (median follow-up of 45 months) while the trial follow-up period was incomplete. The Committee noted the ERG's view that using all available sources of direct and indirect clinical data comparing fludarabine monotherapy with chlorambucil would have been an appropriate approach to reducing the uncertainty over the clinical effectiveness of fludarabine. However, the Committee heard from clinical specialists that patient inclusion criteria in all the relevant RCTs (including the CLL4 study) were substantially the same and therefore they would expect treatment outcomes to be similar across the trials.

4.4 The Committee accepted, on the basis of the preliminary results of the CLL4 trial, that fludarabine monotherapy showed improved response rates (overall and complete response) and 3-year progression-free survival compared with chlorambucil. The Committee was aware that improvements in progression-free survival with fludarabine monotherapy may not translate directly into overall survival benefits. However, the Committee heard from the clinical specialists that an international workshop in CLL had agreed that it was appropriate to use progression-free survival as a surrogate endpoint for overall survival in CLL. This was principally because the prolonged nature of the CLL disease pathway and the use of sequential therapies at different times in the treatment pathways make estimation of differences in overall survival problematic and unreliable. The Committee was persuaded that progression-free survival is a meaningful clinical endpoint for CLL patients.

4.5 The Committee heard from clinical specialists that fludarabine monotherapy is more toxic than chlorambucil, and is associated with a higher incidence of adverse events (particularly neutropenia and infections) and potentially higher mortality rates. The clinical specialists also advised that although the incidence of autoimmune haemolytic anaemia (AIHA) was higher in the chlorambucil arm of the CLL4 trial, the severity of AIHA triggered by fludarabine monotherapy is considered to be greater. Additionally the Committee considered the issue of possible malignant transformations associated with the use of fludarabine monotherapy including the incidence of acute myeloid leukaemia. The Committee understood that there were differences in opinion about this but was persuaded that such malignant transformations were rare.

4.6 The Committee considered the evidence available on quality of life during treatment for CLL. It noted that results from the CLL4 study to date, which showed improvements in response rates and progression-free survival with fludarabine monotherapy, did not lead to improvements in overall quality of life. The clinical specialists advised that quality of life in the early stages of treatment may have been adversely affected by the toxicity of fludarabine monotherapy. However, they stated that they would expect an overall improvement in quality of life for patients who have completed and responded to fludarabine monotherapy because it confers an extended period of progression-free survival.

4.7 The Committee further considered evidence from clinical specialists that the choice of treatment for CLL and the sequence in which treatments are used is made on an individual patient basis, taking into account their general health and fitness and clinical measures of disease activity, in particular the rate of disease progression. The Committee was persuaded that because of the indolent and long-term nature of CLL, watchful waiting is appropriate for some patients who are asymptomatic. Chemotherapy is reserved for those patients who are symptomatic or who are showing signs of progressive disease. The Committee heard from clinical specialists that when a decision to start chemotherapy has been made, first-line treatment and retreatment with chlorambucil may be effective and more appropriate for patients with a less aggressive form of CLL and for those with comorbidities and lower levels of general fitness. On the other hand, first-line treatment with fludarabine monotherapy may be more appropriate for patients with more aggressive forms of CLL disease and those who are considered fit enough to withstand more challenging treatments. The Committee noted statements from patient experts that said that people with CLL were cautious about fludarabine monotherapy because they were aware of its potential toxicity. The patient experts stated that fludarabine monotherapy might not, therefore, be suitable for use in all people with CLL.

Cost effectiveness

4.8 The Committee discussed the differences between the ICERs for the 5-year, 10-year, 15-year and 20-year time horizons in the manufacturer's economic analysis. It considered that the assumptions underlying the extrapolation of data beyond the current CLL4 trial evidence were overly optimistic given the ageing of the population cohort over the lifetime of the model. The Committee noted the assumption in the manufacturer's economic model of a constant risk of disease progression and death over time and considered that this was not a true reflection of the course of the disease. The Committee considered the results of the ERG's survival analysis using patient-level data from the CLL4 trial, which showed that an increasing risk of disease progression and death over time was more appropriate for both responders and non-responders to fludarabine monotherapy and chlorambucil. The Committee noted that the ERG's survival analysis indicated that extrapolation of the model data as presented in the manufacturer's submission was likely to give an overly optimistic estimate of the ICER for fludarabine monotherapy compared with chlorambucil.

4.9 The Committee discussed the manufacturer's assumption that overall survival is the same for fludarabine monotherapy and chlorambucil. The Committee noted that the way in which the manufacturer's economic model equalised overall survival meant that people in the fludarabine arms spent less time in the salvage treatment state, which is associated with lower utilities and additional costs. The Committee further considered that the manufacturer's survival equalisation approach may be inconsistent with the higher mortality rates observed with fludarabine monotherapy compared with chlorambucil in the CLL4 trial (although this was not statistically significant). The Committee concluded that the manufacturer's assumption of equal overall survival in the economic model may have resulted in an underestimation of the ICERs for fludarabine monotherapy compared with chlorambucil.

4.10 The Committee considered the clinical and economic significance of the adverse events associated with fludarabine and noted that these adverse events were important for the cost-effectiveness modelling. The Committee noted however that the manufacturer's model did not incorporate these costs, specifically those related to increased infections, (for example, hospitalisations and prophylaxis for Pneumocystis carinii pneumonia and herpes virus infections) or for treatment of AIHA associated with fludarabine monotherapy. The Committee agreed that omitting the number, type, severity and treatment costs of adverse events from the manufacturer's economic model would have contributed to an underestimate of the ICERs for fludarabine monotherapy compared with chlorambucil.

4.11 The Committee considered the manufacturer's one-way sensitivity analysis on retreatment response rates for fludarabine monotherapy and chlorambucil. It noted that the ICER increased from £26,105 to £86,770 per QALY when the base-case assumptions for the retreatment response rates were changed (see section 3.6). The Committee heard evidence from the clinical specialists that there is limited evidence on retreatment response rates, and complete follow-up of the CLL4 trial may provide more evidence. The Committee concluded that even a small improvement in the retreatment response rates over the ones used in the base-case, particularly for chlorambucil, is likely to increase the ICER for fludarabine monotherapy compared with chlorambucil.

Summary of the considerations

4.12 In summary, the Committee noted that the ICERs for fludarabine monotherapy in comparison with chlorambucil were associated with substantial uncertainties related to the extrapolation of the model data, the exclusion of costs of adverse events and consideration of retreatment response rates; all of which would have resulted in an underestimation of the ICERs for fludarabine monotherapy compared with chlorambucil. The Committee noted that although additional evidence could help to clarify these uncertainties, it was unlikely that such evidence would result in ICERs for fludarabine monotherapy within the range of cost effectiveness that is usually considered to be appropriate for the NHS. The Committee was therefore unable to recommend fludarabine monotherapy for the first-line treatment of CLL as a cost-effective use of NHS resources.