The Appraisal Committee reviewed the data available on the clinical and cost effectiveness of fulvestrant, having considered evidence on the nature of locally advanced or metastatic breast cancer and the value placed on the benefits of fulvestrant by women with the condition, those who represent them, and clinical specialists. It also took into account the effective use of NHS resources.
The Committee considered the views of the patient experts on their experience of fulvestrant as a treatment for locally advanced or metastatic breast cancer. It heard from 1 patient expert who is currently receiving fulvestrant and understood that patients value the availability of a further treatment option after aromatase inhibitors and anti-oestrogen therapies, both as a treatment and because it delays the need for chemotherapy. The Committee also heard from this patient expert that she found the disadvantages of having 2 injections and the associated side effects of fulvestrant were outweighed by the benefits of remaining fit and well on this therapy. The Committee recognised the importance of additional treatment options for post-menopausal women with locally advanced or metastatic breast cancer.
The Committee considered the licensed indication for fulvestrant. It noted that fulvestrant has a marketing authorisation 'for the treatment of postmenopausal women with oestrogen receptor positive, locally advanced or metastatic breast cancer for disease relapse on or after adjuvant anti-oestrogen therapy, or disease progression on therapy with an anti-oestrogen'. The Committee noted comments from the ERG that it is not clear from the wording of the marketing authorisation that eligibility for treatment depends on the last therapy received and may include women who have received more than 1 previous line of treatment for metastatic breast cancer. However, the manufacturer confirmed that fulvestrant has a marketing authorisation as a second-line treatment for metastatic breast cancer in postmenopausal women after adjuvant or first-line treatment of advanced disease with an anti-oestrogen therapy (for most patients this is usually tamoxifen). The Committee was aware that 42% of the patients in the CONFIRM trial had received an aromatase inhibitor as their last treatment before fulvestrant. It also heard from the manufacturer that the European Medicines Agency had regarded the results for the group that had received an aromatase inhibitor as being inconclusive and had rejected the manufacturer's request for an extension of the marketing authorisation for fulvestrant to include patients who have experienced treatment failure with an aromatase inhibitor. Therefore, the Committee was aware of the restriction of the marketing authorisation to patients who had been treated previously with an anti-oestrogen, and the manufacturer's confirmation about the marketing authorisation, which places fulvestrant, within its licensed indication, as an alternative to aromatase inhibitors after anti-oestrogen treatment.
The Committee considered the likely position of fulvestrant in the treatment pathway for women with oestrogen-receptor-positive advanced breast cancer in the UK. The Committee also examined the recommendations on the use of hormone therapy in NICE's guidelines on early and locally advanced breast cancer and advanced breast cancer. It observed that aromatase inhibitors were recommended either as the sole, or as a significant part of, adjuvant treatment for most postmenopausal women with oestrogen-receptor-positive early breast cancer. The Committee also understood that aromatase inhibitors were recommended for postmenopausal women with oestrogen-receptor-positive advanced breast cancer who had no history of hormone therapy or who had been treated previously with tamoxifen. It heard from the clinical specialist that clinical practice follows these guidelines, in that most postmenopausal women receive an aromatase inhibitor as adjuvant hormone therapy for early breast cancer or as first-line treatment if presenting with advanced breast cancer. The Committee understood that the use of tamoxifen in clinical practice in postmenopausal women as a sole adjuvant treatment or as a first-line treatment for new locally advanced or metastatic breast cancer is diminishing, apart from in a small group of women with early breast cancer who have a very poor prognosis and in the small proportion of women who are unable to tolerate any aromatase inhibitor. The manufacturer stated in its submission that the split is approximately 20:80 between treatment with aromatase inhibitors and treatment with anti-oestrogens for patients whose disease progresses on or after adjuvant therapy. However, the Committee heard from the clinical specialist that the proportion of patients receiving aromatase inhibitors is increasing because of changes in clinical practice and therefore aromatase inhibitors are now increasingly favoured over anti-oestrogens. The clinical specialist also indicated that there was not thought to be any significant clinical difference between the effectiveness of anastrozole and letrozole for treating advanced disease. The clinical specialist informed the Committee that exemestane or tamoxifen may be offered as a second-line treatment to women whose disease has failed to respond to an aromatase inhibitor given as either adjuvant therapy or first-line treatment for advanced disease, with the choice depending on a variety of factors, including an assessment of how well previous treatment had worked. The Committee heard from the clinical specialist that fulvestrant is currently considered to be a third-line or fourth-line treatment for postmenopausal women with metastatic breast cancer in UK clinical practice. It further heard that there is little or no clinical evidence about the optimal treatment sequence for advanced breast cancer beyond first-line treatment. The Committee considered that the most likely position of fulvestrant in UK clinical practice would remain as a third-line or fourth-line treatment after therapy with aromatase inhibitors and/or an anti-oestrogen therapy. The Committee again noted the difference between the manufacturer's submission and clinical practice, and that fulvestrant was restricted by its marketing authorisation to use after treatment with an anti-oestrogen. However, based on the manufacturer's confirmation about the marketing authorisation for fulvestrant (see section 4.3), the Committee considered that third-line or fourth-line use was not within the remit of this technology appraisal.
The Committee considered the relevant comparator treatments for fulvestrant within its licensed indication. It understood that the scope listed low-dose (250 mg) fulvestrant and aromatase inhibitors (anastrozole, exemestane and letrozole) as the relevant treatment comparators. It heard from the manufacturer that fulvestrant 250 mg has been replaced by fulvestrant 500 mg as the licensed dose. The Committee considered the remaining comparators. It noted that for the only positions in the treatment pathway for which evidence for fulvestrant was available, non-steroidal aromatase inhibitors such as anastrozole and letrozole are the most likely treatments to be used in clinical practice. It heard from the clinical specialist that, for women who are unable to tolerate non-steroidal aromatase inhibitors, exemestane would be the appropriate comparator if they have been treated previously with an anti-oestrogen. The Committee concluded that the aromatase inhibitors anastrozole, letrozole and exemestane are the most appropriate comparators for the appraisal of fulvestrant.
The Committee considered the clinical-effectiveness data from the CONFIRM trial. It noted that the only comparator in CONFIRM was low-dose (250 mg) fulvestrant. Relative to this comparator, the Committee noted that fulvestrant 500 mg offered benefits in increasing the TTP, but that the difference between groups was statistically significant only for those patients whose last therapy was an anti-oestrogen, and not for patients whose last therapy was an aromatase inhibitor. However, the Committee was also aware that the CONFIRM trial was not powered to detect a statistically significant difference in TTP between fulvestrant 500 mg and fulvestrant 250 mg in the 2 patient subgroups. The Committee concluded that fulvestrant 500 mg offered some clinical benefit compared with fulvestrant 250 mg.
The Committee noted that the results of the network meta-analyses by the manufacturer showed no statistically significant differences in overall survival between fulvestrant, anastrozole and letrozole, although fulvestrant resulted in statistically significantly longer TTP compared with anastrozole (but not letrozole). In addition, the Committee observed that parametric survival models had been used to estimate TTP (log-normal distribution) and overall survival (Weibull distribution) for fulvestrant and other comparators included in the network meta-analysis. The Committee agreed with the issues identified by the ERG about the fit of the log-normal survival model used by the manufacturer to estimate TTP for fulvestrant and other comparators included in the network. It agreed that this resulted in a small number of patients with very long TTPs, which was likely to significantly influence the mean TTP. The Committee also noted that, although the parametric model the manufacturer used to estimate overall survival in the network meta-analysis appeared to be a reasonable match with the CONFIRM trial data, the projections of overall survival beyond the period of observation may have been substantially over- or under-estimated because of the complex changes in risk that were likely to apply at later times. Therefore, the Committee concluded that there was high uncertainty about the validity of the results of the network meta-analyses used to estimate TTP and overall survival.
The Committee considered the populations of the trials included in the network meta-analysis, which included aromatase inhibitors as comparators. It was aware that, although the marketing authorisation for fulvestrant 500 mg is for patients who have received previous anti-oestrogen treatment, the CONFIRM, FINDER-1 and FINDER-2 trial populations included some patients who had last received an aromatase inhibitor, whereas all other trials in the network included only patients who had previously received an anti-oestrogen. The Committee further noted differences in the previous anti-oestrogen and previous aromatase inhibitor groups relating to the position of fulvestrant as a first-line or second-line therapy. Data from CONFIRM showed that most (65.5%) patients receiving fulvestrant after an anti-oestrogen therapy received fulvestrant as a first-line treatment for metastatic breast cancer and the remainder (34.5%) received fulvestrant as a second-line treatment for metastatic breast cancer. Conversely, of the patients who received an aromatase inhibitor as their last treatment before fulvestrant, most (66.8%) received fulvestrant as a second-line treatment for advanced breast cancer. The Committee also noted the differences in demography in the anti-oestrogen and aromatase inhibitor populations and agreed with the ERG that these 2 groups were heterogeneous. Therefore, the Committee agreed that only data from the subgroup in the CONFIRM trial who had received an anti-oestrogen as their last treatment before fulvestrant should be included in the network meta-analyses, in line with the marketing authorisation for fulvestrant.
The Committee considered the eligibility criteria for trials included in the network meta-analysis. It was aware that CONFIRM, FINDER-1 and FINDER-2 were the only trials with an entire patient population documented as having oestrogen-receptor-positive breast cancer. The Committee noted that the manufacturer had sought advice from key opinion leaders about setting firm criteria for the selection of trials for inclusion in the meta-analysis (for example, including only recent trials, or agreeing a stipulated percentage of patients with cancer of unknown oestrogen receptor status), but that no such criteria could be agreed. The main inclusion criterion was relaxed by the manufacturer to include trials for comparators with at least 70% of patients with documented oestrogen-receptor-positive status. The Committee was aware that, based on this criterion, exemestane was excluded as a comparator because of the lack of any relevant trials with at least 70% of patients with oestrogen-receptor-positive cancer. The Committee noted that the percentage of patients with oestrogen-receptor-negative cancer in the trials included in the network meta-analysis ranged from 0% to 33%. The Committee also highlighted sources of heterogeneity between the trials included in the network meta-analysis, including inclusion criteria, median duration of follow-up, amount of previous chemotherapy given, types of recurrent and metastatic disease and the wide timespan of the included trials, which were published between 1996 and 2010. The Committee noted that fulvestrant 500 mg was linked to other treatments in the network only through fulvestrant 250 mg, which was used as the baseline comparator in the manufacturer's network meta-analysis. The Committee further noted that the baseline characteristics of the patients enrolled in the CONFIRM, FINDER-1 and FINDER-2 trials may not be directly comparable with those of patients enrolled in earlier studies that compared fulvestrant 250 mg with anastrozole. The Committee also observed that the results of the network meta-analyses suggested better outcomes in terms of overall survival and TTP for letrozole 0.5 mg (which does not have a marketing authorisation for this indication) than for letrozole 2.5 mg (which does have a marketing authorisation for this indication) when compared with fulvestrant 500 mg. The Committee noted the results of 2 other trials (Dombernowsky et al. 1998; Gershanovich et al. 1998) that were excluded from the network meta-analyses (because they did not meet the oestrogen-receptor-positive status inclusion criterion) in which there was a trend suggesting clinical superiority of letrozole 2.5 mg over letrozole 0.5 mg. The Committee concluded that the results of the manufacturer's network meta-analysis were subject to bias from the selection of studies included in the network and this therefore increased the uncertainty about the outputs of this analysis.
The Committee again discussed the parametric survival models used to project TTP and overall survival by the manufacturer. It accepted the ERG's exploratory analyses that derived the best estimate of overall survival from modelling post-progression on the basis of trial data and combining pre-progression and post-progression estimates. Overall, the Committee concluded that the manufacturer did not provide sufficient commentary or analysis of uncertainty about the fit of alternative parametric survival models and concluded that the estimates of TTP and overall survival based on the ERG's exploratory analysis were more appropriate and were therefore preferred.
The Committee considered the manufacturer's economic model and the ERG's critique of the model. The Committee agreed with the ERG that because 2 different probability distributions were fitted to the 2 sets of data, it is possible for projected estimates of TTP to exceed the corresponding estimates of overall survival, which can lead to negative values for the number of patients alive in the post-progression state in the economic model. The Committee noted that, although the model corrected any negative post-progression survival estimates to zero, it did not compensate for any resulting overestimation of survival. The Committee concluded that the manufacturer's economic model is unlikely to provide a robust basis for projecting survival data beyond the observed data from the CONFIRM trial.
The Committee discussed the utility values applied to the pre-progression and post-progression health states by the manufacturer. The Committee agreed that, although the utility values (taken from Lloyd et al. 2006) were not generated in line with the NICE reference case, they probably represent the best published estimates available, although methodological uncertainty remains. However, it agreed that the utility values based on the age of the participants in the study by Lloyd et al. (2006) should have been adjusted to the mean age of patients used to estimate UK EQ-5D tariff scores. The Committee also agreed with the ERG that the 'responder status' of patients should have been incorporated in the estimation of utility values for the pre-progression and post-progression health states. The Committee concluded that the ERG's adjusted utility values used in its exploratory analysis were preferable to those used by the manufacturer.
The Committee discussed the validity of the cost inputs used in the manufacturer's economic model. The Committee agreed with the ERG in relation to uncertainty about some of the cost inputs used and that modifications to these parameters resulted in small increases in the ICERs. The Committee also noted that the list price of anastrozole reported in the manufacturer's submission may not be what the NHS usually pays. Therefore, the Committee concluded that it is likely that the ICERs for fulvestrant compared with anastrozole would be underestimated.
The Committee noted that the key drivers of the cost-effectiveness results in the manufacturer's model were the estimates of TTP and overall survival for all treatments and the utility values assigned to the pre-progression and post-progression health states. The Committee highlighted the results of the manufacturer's probabilistic sensitivity analysis, which indicated that fulvestrant 500 mg had a low probability of being cost effective (2%) at a threshold of £20,000 per QALY gained. The Committee also considered that the ICERs generated using the manufacturer's model were not reliable because of problems with the design of the model and the inclusion of the mixed patient population from the CONFIRM trial. The Committee agreed that the ICERs generated by the ERG's exploratory analysis – which used different estimates of TTP and overall survival, included only the population from the CONFIRM trial whose last treatment had been an anti-oestrogen, revised cost inputs and adjusted the utility estimates – would be more reliable. However, the Committee noted that the ERG's exploratory analysis was based on the same trials in the network meta-analysis as those used in the manufacturer's network meta-analysis. The Committee considered that the network meta-analysis contained considerable uncertainty, which was unaccounted for in the ICERs. Overall, the Committee concluded that the ERG's ICER of £35,000 per QALY gained for fulvestrant 500 mg compared with anastrozole was more plausible than the manufacturer's base-case estimate but there remained considerable uncertainty about this estimate.
The Committee noted that no comparison with exemestane could be made in the manufacturer's base-case cost-effectiveness analysis because of a lack of any relevant trials in which 70% or more of patients had oestrogen-receptor-positive advanced breast cancer in a population who had received an anti-oestrogen. The Committee noted that, as a result, the cost effectiveness of fulvestrant compared with exemestane in this patient population remains unknown. The Committee further concluded that a cost-effectiveness analysis for a subgroup of patients with contraindications to non-steroidal aromatase inhibitors would be desirable, but because the comparator treatment in such an analysis would be exemestane, this could not be undertaken. The Committee concluded that it was unable to recommend fulvestrant as an alternative to exemestane in postmenopausal women with oestrogen-receptor-positive, locally advanced or metastatic breast cancer, or disease progression on therapy with an anti-oestrogen who have contraindications to non-steroidal aromatase inhibitors.
The Committee considered the small subgroup of women who are unable to tolerate treatment with any aromatase inhibitor. The Committee noted that there was no available evidence on the clinical and cost effectiveness of fulvestrant for this small subgroup, and concluded that it was unable to recommend fulvestrant for women unable to tolerate both non-steroidal and steroidal aromatase inhibitors. However, the Committee was aware of alternative funding arrangements available for providing treatment for women who are unable to tolerate an aromatase inhibitor, such as individual funding requests based on exceptionality.
The Committee considered supplementary advice from NICE that should be taken into account when appraising treatments that may extend the life of patients with a short life expectancy and that are licensed for indications that affect small numbers of people with incurable illnesses. For this advice to be applied, all of the following criteria must be met:
The treatment is indicated for patients with a short life expectancy, normally less than 24 months.
There is sufficient evidence to indicate that the treatment offers an extension to life, normally of at least an additional 3 months, compared with current NHS treatment.
The treatment is licensed or otherwise indicated for small patient populations.
In addition, when taking these criteria into account, the Committee must be persuaded that the estimates of the extension to life are robust and that the assumptions used in the reference case of the economic modelling are plausible, objective and robust.
The Committee discussed whether fulvestrant fulfilled the criteria for a life-extending, end-of-life treatment. The Committee agreed that, based on the results of the CONFIRM trial, fulvestrant is indicated for patients with a life expectancy of more than 24 months (the ERG's estimate of mean overall survival for patients taking fulvestrant 500 mg was 36.33 months compared with 32.31 months for those taking anastrozole and 30.90 months for those taking letrozole) and so the criterion of patients with a short life expectancy was not met. Therefore, the Committee agreed that it was not necessary for the criteria of extension to life of at least an additional 3 months and a small patient population to be established. The Committee concluded that fulvestrant did not fulfil the end-of-life criteria.
The Committee considered the most plausible ICER for fulvestrant compared with anastrozole, which it had agreed was likely to be at least £35,000 per QALY gained (see section 4.14). The Committee also noted the considerable uncertainty associated with this estimate because of the network meta-analysis. The Committee concluded that fulvestrant could not be considered a cost-effective use of NHS resources as an alternative to aromatase inhibitors for the treatment of oestrogen-receptor-positive, locally advanced or metastatic breast cancer in post-menopausal women whose cancer has relapsed on or after adjuvant anti-oestrogen therapy, or who have disease progression on anti-oestrogen therapy.
The Committee considered a potential equalities issue highlighted during consultation about the use of fulvestrant for patients unable to swallow oral aromatase inhibitor medication. The Committee was aware that women who are unable to swallow (for example, following a stroke) would be fed using an enteral tube, and that oral medication can also be given by this route. In addition, given that the recommendation did not differentiate between any groups of people, the Committee concluded that its recommendations did not limit access to the technology for any specific group compared with other groups.