Mifamurtide for the treatment of osteosarcoma

Management of high-grade resectable non-metastatic osteosarcoma in UK clinical practice

4.1 The Committee discussed the clinical needs of patients with high-grade resectable non-metastatic osteosarcoma. The patient experts stated that diagnosing and treating osteosarcoma has a significant impact on patients and their families and friends. This includes disruption of family life, strain on family relationships, additional stress at work and financial pressures, and a negative effect on the health of families, friends and carers. The Committee heard from the clinical specialists and patient experts that the main aim of treatment is to cure the patient. The patient experts and clinical specialists stated that there had been few developments that had improved treatment outcomes for osteosarcoma over the past 20 years, and that any improvement in overall survival from adding mifamurtide to standard chemotherapy was clinically significant and important. The clinical specialists stated that the only development in the past 10 years had been to add high-dose methotrexate to the treatment regimen for osteosarcoma, but that there is currently limited evidence available about whether overall survival rates in the UK have improved in the last decade. The Committee heard from the clinical specialists that current UK clinical practice is neoadjuvant multi-agent chemotherapy and surgical resection, followed by postoperative adjuvant multi-agent chemotherapy. The clinical specialists stated that the standard adjuvant multi-agent chemotherapy regimen in the UK is doxorubicin, methotrexate and cisplatin and the 5-year overall survival rate is approximately 55%. The Committee noted that this survival rate is for all patients, including some with more advanced disease for whom mifamurtide would not be indicated. The Committee heard from the clinical specialists that ifosfamide is currently being investigated in an ongoing European and US osteosarcoma study (EURAMOS 1) as part of an adjuvant regimen (with etoposide, cisplatin, doxorubicin and methotrexate). Only patients with tumours showing a poor histological response to pre-operative chemotherapy can receive ifosfamide. The clinical specialists stated that study recruitment should be complete in 2011, with results anticipated in 2015–20, and the study may establish a role for ifosfamide in UK clinical practice. They explained that meaningful research in osteosarcoma is difficult to carry out because of the small number of patients and the long follow-up required. The Committee heard that a significant number of patients in the UK are taking part in EURAMOS 1 and some may be taking ifosfamide in that context. Patients would not be eligible for mifamurtide while they are receiving the study drug regimens. It also heard that a follow-up trial to EURAMOS 1 is in development. It is expected that the EURAMOS 2 trial will commence in 2012/2013. The Committee welcomed continued research into the best regimen for this condition. The Committee concluded that the current established chemotherapy regimen in England and Wales is doxorubicin, methotrexate and cisplatin, and that the extent of ifosfamide use in UK clinical practice outside the EURAMOS 1 study had not been quantified.

4.2 The Committee noted evidence from the clinical specialists and patient experts that treatment with mifamurtide is safe and well tolerated. The Committee noted that standard neoadjuvant and adjuvant chemotherapy in the UK (regimen A) is completed in approximately 30 weeks, and that an additional 18 weeks of treatment with mifamurtide would be required to be consistent with the administration schedule in INT-0133. The Committee heard from the clinical specialists that in INT-0133 a significant proportion of patients (22–30%) did not complete treatment with mifamurtide, and, based on evidence from the ongoing EURAMOS 1 study, this may have been because patients did not want to extend treatment beyond the duration of standard multi-agent chemotherapy in a trial setting. Patient experts stated that increased overall survival is so important that patients would accept the option of prolonged treatment with mifamurtide if it was shown to improve overall survival. The Committee agreed that patients would be more willing to extend treatment in clinical practice if mifamurtide provided them with a higher chance of cure.

Clinical effectiveness

4.3 The Committee considered the evidence on the clinical effectiveness of mifamurtide as presented in the manufacturer's submission and the ERG's critique. It considered the evidence from the only relevant randomised clinical trial (INT-0133). The Committee noted that the study was generally well conducted, but it agreed that there were substantial methodological issues identified by the ERG that led to uncertainty about the estimates of disease-free survival and overall survival. These included delayed or non-administration of mifamurtide and an imbalance in histological response to neoadjuvant therapy between treatment groups. The imbalance was particularly pronounced for patients assigned to regimen A+, who had a greater proportion of tumours showing a poor (greater than 5% remaining viable tumour) histological response, which may have disadvantaged mifamurtide. The Committee concluded that these aspects of the study made interpretation of the survival data more difficult, and that the effect of these factors on the results could not be reliably predicted.

4.4 The Committee noted that the manufacturer had presented an analysis of all the INT-0133 data from the three- and four-agent chemotherapy regimens (that is, regimens A+ and B+ combined and regimens A and B combined) for overall survival and a number of post hoc efficacy analyses. The Committee was aware that the combined analysis was the primary prespecified analysis of the trial and noted that this suggested a statistically significant improvement in overall survival, from 71% in the control arm to 78% in the mifamurtide arm over a median follow-up period of 7.9 years (hazard ratio 0.72, 95% CI 0.53 to 0.97). Although the study was powered for the intermediate endpoint of disease-free survival, it did not show a statistically significant increase in disease-free survival for regimens of chemotherapy plus mifamurtide (regimens A+ and B+ combined) compared with chemotherapy alone (regimens A and B combined) (hazard ratio 0.78, 95% CI 0.61 to 1.01). The Committee noted that a greater proportion of patients assigned to regimen A+ had tumours showing a poor (greater than 5% remaining viable tumour) histological response to neoadjuvant pre-operative therapy. It accepted the view of the clinical specialists that there was evidence of a link between poor histological response to neoadjuvant therapy and prognosis, but concluded that it was not possible to establish whether this variation in histological response before adjuvant therapy in the different treatment groups might have affected the results, or by how much. The Committee also noted the ERG's concerns that there may have been interaction between treatments (that is, there may be synergy between ifosfamide and mifamurtide), given that the test for statistical interaction for disease-free survival was very close to the prespecified threshold for interaction of 0.10 (p = 0.102). The Committee heard from the clinical specialists that factorial trials are designed on the assumption that there is no interaction between the study drugs, and that power to detect plausible interactions requires greatly increased sample sizes. The Committee accepted that the statistical test for interaction did not suggest a strong interaction between the drugs in the analysis of overall survival. It also accepted the clinical specialists' views that there was no biologically plausible reason for such an effect.

4.5 The Committee then discussed the post hoc analyses requested by the ERG that compared regimen A+ with A, and regimen B+ with B. It was aware that this was an alternative approach to the analysis and that INT-0133 was not designed for these comparisons or powered for this analysis. The Committee noted that for regimen A+ compared with A, there was a non-statistically significant improvement in overall survival (hazard ratio 0.75; 95% CI 0.49 to 1.16). For regimen B+ compared with B, there was also a non-statistically significant improvement in overall survival (hazard ratio 0.68; 95% CI 0.44 to 1.05). Both comparisons were consistent with the overall estimate but the confidence intervals were wider, possibly because of smaller patient numbers in the subgroups. The Committee understood that in trials for the treatment of rare diseases such as this, recruiting the numbers of patients needed to adequately power the trial is difficult, and even more so to allow subgroup analyses of this nature. The Committee then discussed the analyses in the context of UK clinical practice. It noted that currently ifosfamide is usually only administered in a clinical trial setting in the UK. The comparison most relevant to UK clinical practice was therefore agreed to be the mifamurtide regimen (A+) compared with the regimen reflecting current UK clinical practice (A). However, for the reasons above, the Committee accepted that the combined analysis of all the INT-0133 data was more appropriate in determining the effect of adding mifamurtide to the standard UK regimen than the post hoc analysis directly comparing regimen A+ with A.

4.6 The Committee discussed the adverse effects of mifamurtide plus multi-agent chemotherapy and noted that the combined analysis of all the INT-0133 data from the three- and four-agent chemotherapy regimens showed a statistically significant increase in subjective and objective hearing loss in patients receiving mifamurtide regimens. The Committee was aware that in the post hoc subgroup analyses an increased incidence of hearing loss occurred only in patients treated with regimen A+. It noted that there was uncertainty about which agent in the regimen could be associated with hearing loss. The Committee accepted the clinical specialists' views that cisplatin was used in all arms of the study and there is a known risk of hearing loss associated with its use (usually in the range 5–15%). Accordingly, the rate of hearing loss seen in INT-0133 was not unusual and could be an effect of cisplatin rather than mifamurtide. The Committee also accepted the clinical specialists' views that objective hearing loss after treatment may not be clinically important or necessarily require the use of hearing aids, and that this risk should be considered in the context of a possible higher cure rate for osteosarcoma.

Cost effectiveness

4.7 The Committee considered the manufacturer's economic analyses and the respective critiques and exploratory sensitivity analyses performed by the ERG. The Committee noted that the efficacy data for the manufacturer's analyses were taken from INT-0133 for regimens A+ and B+ combined compared with regimens A and B combined, but that on the request of the ERG the manufacturer had also presented cost-effectiveness estimates for the post hoc analyses for regimen A+ compared with regimen A and regimen B+ compared with regimen B. The Committee noted that the manufacturer's most recent additional analyses incorporated a revised patient access scheme agreed by the Department of Health (see section 2.5). The Committee discussed the following assumptions in the analyses:

4.8 The Committee considered including the costs associated with amputation and limb salvage. It noted from the scenario analyses carried out by the manufacturer (see section 3.19) that there was a marginal increase in the ICER when these costs were included. The Committee agreed with the ERG that it was appropriate to include amputation and limb salvage costs in the model.

4.9 The Committee noted that hearing loss adverse events were not included in the manufacturer's economic analyses. However, the Committee accepted the views of the clinical specialists that although hearing loss was the main adverse event, occurring more frequently with mifamurtide treatment in the clinical study, the rate of hearing loss seen in INT-0133 was not unusual in cisplatin-containing regimens and its exclusion from the model could be justified.

4.10 The Committee considered the mortality rates used by the manufacturer in its economic analyses. The Committee heard from the clinical specialists that 25% of patients with recurrent disease may be cured and that the prognosis after recurrence depends on time to recurrence (that is, patients with a longer time to recurrence have a better prognosis). The Committee agreed that after 5 years free of disease, it was reasonable to use the mortality rates of the general population.

4.11 The Committee considered the utility values used in the model. It noted that the manufacturer's model contained utility values from two different sources: a review of NICE technology appraisals for cancer treatments and a small study using the EQ-5D in patients from INT-0133. The Committee noted that the technology appraisals included in the review were from very different populations and did not generally use NICE's preferred method to derive the utility values. The Committee also noted that although the sample size for the study using the EQ-5D was small, it included the population of interest (that is, only patients with osteosarcoma) and used a method to derive the utility values that met NICE's reference case. The Committee was aware that a utility value of 0.85, derived from the manufacturer's review of NICE technology appraisals, had been applied to the disease-free state in the model. The Committee discussed whether this utility value should be maintained throughout the length of the model. It noted that the ICER increased when age-adjusted utility values were used. The Committee heard from the patient experts that young people with osteosarcoma are able to live full lives and they have a similar quality of life to their peers, with many adapting well to any remaining disability, and in some cases being empowered by their experience. The Committee agreed that in the general population utility declines with age, and therefore age-adjusted utility values should be used in the model.

4.12 The Committee considered the most recent additional analyses carried out by the manufacturer (see section 3.24), including the Committee's preferred assumptions (see sections 4.8–4.11), the results for regimens A+ and B+ combined compared with regimens A and B combined (that is, independent of ifosfamide) over a 60-year time horizon and the revised patient access scheme. The Committee noted that the ERG had carried out exploratory sensitivity analyses (see section 3.25) on the manufacturer's most recent additional analyses using data from regimen A+ compared with regimen A rather than all the INT-0133 data from the three- and four-agent chemotherapy regimen. The Committee agreed that it was appropriate for the discussion to focus on the manufacturer's most recent additional analyses rather than the ERG's exploratory sensitivity analyses. The most recent analyses by the manufacturer (regimens A+ and B+ combined compared with regimens A and B combined) reported 'best-case' ICERs of £60,200 per QALY gained (deterministic analysis) and £56,700 per QALY gained (probabilistic analysis).

4.13 The Committee discussed the sensitivity of the manufacturer's 'best-case' ICER to the discount rate applied (see section 3.26). The Committee noted the exploratory sensitivity analysis carried out by the manufacturer which showed that applying a discount rate of 0% on QALYs gained (but keeping the discount rate on costs at 3.5%) decreased the manufacturer's deterministic ICER from £60,200 to £25,100 per QALY gained and a discount rate of 6% increased the manufacturer's deterministic ICER to £95,100 per QALY gained. It also noted the wide range in these figures. The Committee noted the clarification to the 'Guide to the methods of technology appraisal' issued by the Board of NICE, which states that 'where the Appraisal Committee has considered it appropriate to undertake sensitivity analysis on the effects of discounting because treatment effects are both substantial in restoring health and sustained over a very long period (normally at least 30 years), the Committee should apply a rate of 1.5% for health effects and 3.5% for costs'. The Committee discussed whether these criteria were met in the case of mifamurtide. It noted that mifamurtide is a treatment with curative intent that increased the overall survival from 71% to 78% compared with chemotherapy alone in the whole trial (regimens A+ and B+ combined versus regimens A and B combined). It also noted that patients who are cured are expected to have a long and sustained benefit and regain normal life expectancy. The Committee concluded that both criteria were met and a discount rate of 1.5% should be used for health effects. This resulted in a manufacturer's best-case probabilistic ICER of £36,000 per QALY gained (see section 3.26).

4.14 The Committee noted that the ICER of £36,000 per QALY gained is higher than what is normally considered an effective use of NHS resources and that the NICE 'Guide to the methods of technology appraisal' states that a strong case should be identified for an ICER that is higher than £30,000 per QALY gained. The Committee noted that, in these circumstances, the NICE 'Guide to methods of technology appraisal' states that judgements about the acceptability of the technology, as an effective use of NHS resources, will specifically take account of any strong reasons to indicate that the assessment of the change in health-related quality of life has been inadequately captured or whether the innovative nature of the technology may not have been adequately captured in the QALY measure. The Committee discussed whether the assessment of the change in health-related quality of life had been inadequately captured in the economic analysis. It heard from patient experts that successfully treated patients could lead an active and fulfilling life and were able to contribute to society. The Committee also heard from the patient experts that supporting a young person with osteosarcoma has a profound impact on the health-related quality of life of the family and friends of the person affected, particularly when treatment is not successful. For example, parents and siblings may develop mental health problems and family relationships may be strained. The Committee concluded that these are very important issues affecting the health-related quality of life of those close to the person with osteosarcoma which should be taken into account but on this occasion had not been adequately captured in the economic analysis.

4.15 Furthermore the Committee accepted that mifamurtide plus adjuvant chemotherapy may represent a potentially valuable new therapy and that the mechanism of action was novel. It acknowledged that few advances had been made in the treatment of osteosarcoma in recent years and mifamurtide could be considered a significant innovation for a rare disease. The Committee concluded that the combined value of these factors, in addition to the potential uncaptured QALY benefits, meant that mifamurtide could be considered a cost-effective use of NHS resources.

4.16 The Committee considered whether there were issues relating to equality to be taken into account in light of its duties under the equalities legislation. The Committee discussed comments made at the scoping stage. These included the observation that osteosarcoma mainly affects children, teenagers and young adults, and that osteosarcoma is a rare disease. The Committee considered that no different recommendations were made for the patient population within the licensed indication, that is, the recommendations are not based on age and do not vary according to the age of the patient. The Committee was therefore satisfied that there were no equalities issues relating to age in this appraisal and that the recommendations were consistent with NICE's obligations under the equalities legislation and the requirement for fairness.

Summary of the Appraisal Committee's key conclusions