Dostarlimab with platinum-based chemotherapy for treating advanced or recurrent endometrial cancer with high microsatellite instability or mismatch repair deficiency

Endometrial cancer starts in the lining of the uterus. Symptoms can include vaginal bleeding, pelvic pain, unintended weight loss, nausea and fatigue. Around 23% of people with endometrial cancer have the subtype with high microsatellite instability (MSI‑H) or DNA mismatch repair (dMMR) deficiency biomarkers. Endometrial cancer has a significant effect on both life expectancy and quality of life. People with advanced or recurrent endometrial cancer (meaning that the cancer has spread beyond the uterus or come back after previous treatment) have a poor prognosis. Only 15% of women diagnosed at stage 4 live for 5 or more years. The impact is not just limited to physical health, but the mental health and wellbeing of people and their families also. The patient experts emphasised that effective treatment options at this stage are limited, leaving people feeling frustrated, hopeless and abandoned. They highlighted the lack of treatment options in endometrial cancer compared with other cancer types. The committee concluded that advanced or recurrent endometrial cancer has a devastating effect on life expectancy and quality of life, and that there is an unmet need for more effective treatments.

Standard care for recurrent or advanced endometrial cancer is platinum-containing chemotherapy (PCC), with the most common regimen being carboplatin plus paclitaxel. People whose cancer progresses after chemotherapy may be offered immunotherapy treatment. Pembrolizumab with lenvatinib is available for all people who have previously had treatment for endometrial cancer (see NICE's technology appraisal on pembrolizumab with lenvatinib for previously treated advanced or recurrent endometrial cancer). Dostarlimab is available through the Cancer Drugs Fund for people with MSI‑H or dMMR endometrial cancer (see NICE's technology appraisal on dostarlimab for previously treated advanced or recurrent endometrial cancer with high microsatellite instability or mismatch repair deficiency). This subtype of endometrial cancer is more likely to respond to immunotherapy treatment. The company submission moves dostarlimab to an earlier line of therapy, so instead of being given following disease progression on chemotherapy, dostarlimab could be given with chemotherapy. The company says that bringing an immunotherapy earlier into the treatment pathway will result in more people being offered and benefitting from the treatment. The clinical experts said that, once the cancer has progressed after chemotherapy, many people are unable to tolerate further treatment. This means that they are unable to access effective immunotherapy treatments in the second-line setting. The patient experts said the current treatment approach is geared towards expecting a recurrence and then adding a more effective second-line treatment. They thought that the most effective treatments should be offered earlier in the pathway to reduce the risk of recurrence. The committee concluded that earlier access to immunotherapy would be welcomed by patients and clinicians, and agreed that the company's positioning was appropriate.

The clinical evidence is based on an interim analysis of RUBY‑1, a phase 3, randomised, double-blind, multicentre, placebo-controlled study in people with advanced or recurrent endometrial cancer. It provides direct evidence to compare dostarlimab plus carboplatin with paclitaxel (from now, dostarlimab) with placebo plus carboplatin with paclitaxel (from now, placebo). People in the trial needed to have endometrial cancer with a low potential for cure by radiation therapy, surgery alone or in combination. In line with the marketing authorisation, the company gave efficacy data for people with dMMR or MSI‑H primary advanced or recurrent endometrial cancer (dMMR/MSI‑H population). Investigator-assessed progression-free survival (PFS) was the primary endpoint for this population. There was a statistically significant benefit for dostarlimab compared with placebo for PFS. At data cut-off (PFS data maturity of 56%), dostarlimab reduced the risk of progression or death by 72% compared with placebo (hazard ratio [HR] 0.28, 95% confidence interval [CI] 0.16 to 0.50). Overall survival (OS) was a prespecified subgroup analysis in the dMMR/MSI‑H population. At OS maturity of 26%, dostarlimab reduced the risk of death by 70% compared with placebo (HR 0.30, 95% CI 0.13 to 0.70). The company noted that, in addition to the hazard ratios, it was important to consider the percentage of people who were progression free at 24 months, and the longer duration of response with dostarlimab. The estimated probability of being progression free at 24 months was 61.4% (95% CI 46.3 to 73.4) in the dostarlimab arm and 15.7% (95% CI 7.2 to 27.0) in the placebo arm. The probability of the cancer remaining in response to treatment at 12 months was 62.1% (95% CI 44.4 to 75.5) in the dostarlimab arm, and 19.2% (95% CI 8.6 to 33.1) in the placebo arm. The clinical experts confirmed that the benefits seen in RUBY‑1 were clinically meaningful. They also said it was similar to trials for other immunotherapy drugs in similar populations. The committee concluded that RUBY‑1 is appropriate for decision making, and that dostarlimab shows clinical benefit in the relevant population.

As discussed in section 3.3, the clinical evidence for this appraisal is based on RUBY‑1. The EAG had concerns about whether the RUBY‑1 data reflected the true benefit of the treatment. It noted the small sample size of the dMMR/MSI‑H population (n=118) and the immaturity of the data. The EAG also had concerns about the randomisation between the 2 arms because there were more people in the placebo arm than the dostarlimab arm. It said that people in the placebo arm were generally older in age and had a higher body mass index but better performance status than the dostarlimab group. It said that these factors could have biased the treatment effect estimate. The company said that further analyses had been done to assess the potential impact of these differences, and the hazard ratio remained consistent in these. The clinical experts said that the small sample size was reasonable given the relatively small patient population in the NHS with this condition. The committee acknowledged the immaturity of the data, differences in baseline characteristics between the groups and small sample size. It concluded that adding dostarlimab to usual treatment is clinically effective, but the amount of benefit is uncertain.

The company gave a partitioned survival model with 3 health states to estimate the cost effectiveness of adding dostarlimab to PCC. The 3 health states were PFS, progressed disease and death. Health-state utilities and baseline characteristics were taken from the dMMR/MSI‑H population in RUBY‑1. The comparator arm of RUBY‑1, placebo with carboplatin plus paclitaxel, was used to inform the PCC arm in the model. Diagnostic testing for dMMR/MSI‑H is routine in the NHS, so the costs for this were not included in the model. In line with the marketing authorisation, a 3‑year stopping rule was applied. The EAG preferred to use a higher starting age in the model (67.1 years) to better reflect the NHS population. The committee concluded that the model structure is appropriate for decision making and preferred the EAG's starting age of 67.1 years.

Long-term PFS was estimated by extrapolating the data from RUBY‑1. The company and EAG agreed on the extrapolation approach for the placebo arm but disagreed on the preferred approach for the dostarlimab arm. The company selected a piecewise approach and the EAG selected Weibull. With the EAG's approach, the hazard rates for the dostarlimab arm and PCC arm crossed at around 5 years. From this point, the EAG set both arms to use the PCC arm hazard rate (so there were equal hazard rates between the 2 arms). In the company model, the PFS benefit for dostarlimab was sustained for the duration of the model, but the EAG said that this was implausible. The clinical experts said that it was reasonable to assume that dostarlimab had a sustained benefit, based on the durable responses that had been seen with dostarlimab and other immunotherapy treatments. So, they did not expect the hazard rates for dostarlimab to be equal with PCC from 5 years onwards. The clinical experts said that their own estimates for long-term PFS would be closer to the company's adviser estimates than the EAG estimates, which they thought were overly pessimistic. Once someone has been progression free for 5 years, their risk of progression is very low, so they thought the drops in PFS probability seen between 5 and 10 years were not realistic. But they noted that it is very difficult to give long-term estimates because it is not yet known how newer treatments will change the long-term outcomes for people. The NHS Cancer Drugs Fund lead said that the pattern from other dMMR/MSI‑H tumour types (for example, colorectal cancer) suggests that most relapses occur in 2 years and that there is a clear plateauing in PFS and OS from year 2 to year 3. They said that dMMR/MSI‑H cancers clearly respond better to immunotherapy treatments than non dMMR/MSI‑H cancers, regardless of the tumour location. In dMMR/MSI‑H cancer there tends to be a clear plateau at a relatively early stage. The committee concluded that it was reasonable to assume a sustained treatment effect benefit for dostarlimab. So, it preferred the company's model, in which the hazard rates did not cross.

The company's base case assumed that the dostarlimab treatment benefit was sustained for the full duration of the model. But the EAG assumed that the treatment effect began to wane from 80 weeks, over a period of 3 years. The EAG said this was applied to reflect the convergence of the dostarlimab and PCC hazard rates seen in the RUBY‑1 follow-up data. It explained that, because of limitations in the model structure, it had to converge the dostarlimab hazard rate to the PCC arm, rather than the other way round (which would have better reflected the RUBY‑1 data). It also allowed the extrapolated OS to sit within the range of survival predictions made by the EAG and company's clinical advisers. The company said that the EAG's approach was clinically implausible and not consistent with the RUBY‑1 data, in which the hazard rate for the dostarlimab arm declined over time. The clinical experts said that data from other immunotherapies had shown that this type of treatment has a durable treatment effect in people with dMMR/MSI‑H endometrial cancer. They also mentioned emerging data from the relapsed setting, in which the cancer has had durable responses to immunotherapy treatments beyond the 3‑year stopping rule. The NHS Cancer Drugs Fund lead noted that the treatment effect waning is typically applied after treatment has stopped, not while treatment is ongoing. They noted that, while the data for this indication is immature, immunotherapies have showed sustained benefit in other dMMR/MSI‑H tumour types (for example, in NICE's technology appraisal guidance on nivolumab with ipilimumab for previously treated metastatic colorectal cancer with high microsatellite instability or mismatch repair deficiency). The committee noted that there was not a clear waning effect for dostarlimab seen in the RUBY‑1 data. It also noted that immunotherapy treatments are particularly effective in dMMR/MSI‑H cancer. The committee acknowledged that the hazard rates did converge because of the hazard rate in the PCC arm declining over time. But it said that having longer-term data on both the PCC arm and the dostarlimab arm would be a better way to reflect converging hazard rates, rather than the EAG's approach of applying treatment waning to the dostarlimab arm while treatment is ongoing. The committee concluded that there was not enough evidence from RUBY‑1 to know if, and when, the dostarlimab treatment effect wanes. It thought that applying a treatment waning effect while treatment with dostarlimab was still ongoing was not justified. Once future data becomes available, the committee would like to see a range of treatment waning assumptions given as scenario analyses.

Long-term OS was estimated by extrapolating the data from RUBY‑1. But the estimates are uncertain because of the immaturity of the data and the small sample size of the dMMR/MSI‑H subgroup. There were also large differences between the survival estimates given by the EAG and the company's clinical advisers. These differences resulted in the company and EAG selecting different extrapolation approaches for their base cases. For the PCC arm, the company used the Kaplan–Meier (KM) data from RUBY‑1 for the full follow-up period, followed by the log-logistic curve. The EAG used the full log-logistic curve without the KM data. For the dostarlimab arm, the company used the KM data for the full follow-up period, followed by the hazard ratio applied to the PCC arm. The EAG preferred to use the exponential curve, with the application of treatment waning from 80 weeks. The company's model estimated that 72% of people in the dostarlimab arm would still be alive at 5 years, but the EAG clinical adviser suggested that only 20% would still be alive. The clinical experts said that long-term estimates were extremely difficult to make because there is no long-term experience of people who have benefited from new immunotherapy treatments. But they thought that the EAG's adviser estimates were pessimistic and did not reflect the full benefit of immunotherapies. The clinical experts said that people whose cancer responds to immunotherapy treatment have very durable responses. They said their own estimates for long-term OS would be somewhere in between the company's and the EAG's adviser estimates, but closer to the company's adviser estimates. The committee recalled the NHS Cancer Drugs Fund lead's comments that dMMR/MSI‑H cancers tend to show clear plateauing in PFS and OS at an early stage (see section 3.6). The company noted that there were only 7 OS events in the dostarlimab arm, so it thought that applying a hazard ratio to the dostarlimab arm (which was based on more events so was more robust) was the best approach to use, given the available data. The committee noted that the short duration of follow up also gave uncertainties in the PCC arm. It was specifically concerned that the benefit of subsequent immunotherapy treatment had not been reflected in the data because of the short follow-up time. So, the long-term OS for the PCC arm may have been underestimated, leading to an overestimate of the benefits of dostarlimab. The committee concluded that it was not possible to agree on preferred extrapolation curves because of the immaturity of the data. It preferred the EAG's approach of using parametric models for the full duration of the model, rather than using semi-parametric models. This was because the tail of the KM curves was informed by such small numbers. It also recalled that the clinical experts thought that the EAG's long-term estimates were too pessimistic. But the committee's view may change once more data becomes available.

People in RUBY‑1 were able to have subsequent treatments once they had stopped taking carboplatin plus paclitaxel with or without dostarlimab. The subsequent treatment use seen in RUBY‑1 included several treatments not available on the NHS. So, subsequent treatment use in the economic model was informed by clinical expert opinion, as well as RUBY‑1 data. The EAG agreed with this approach, but said that expert opinion varies significantly, so the use, costs and effects of subsequent treatment were a source of uncertainty in the model. The clinical experts said that the treatments they use in the NHS differ from those included in the model (for example, immunotherapy monotherapy). But the company noted that those drugs were either only available in the Cancer Drugs Fund or had only become available in routine commissioning very recently, which is why they were not included in the model. The committee acknowledged the sensitivity analysis that had been done to explore differences in subsequent treatment use in the model. It recalled that the NICE health technology evaluations manual (see section 2.2.12) specifies that comparators should be established clinical practice in the NHS. It also specifies that technologies which are recommended with managed access are not suitable comparators. The committee concluded that the company's approach was acceptable, given the differences between what is routinely available on the NHS and the data available from RUBY‑1. But it will reconsider this once further data has been collected during managed access.

The company's base case included the disutilities and treatment costs for adverse events (AEs) that were grade 3 or above. The EAG said that the limited follow up and small sample size may have resulted in the true impact of AEs being under-reported, and their impact underestimated. In the EAG's base case, the EAG preferred to include a broader range of AEs (those affecting at least 2% of people, rather than 5% used in the company's base case). It also preferred to increase the costs for ongoing monitoring of immune-related AEs in the dostarlimab arm (0.23 visits a week from cycle 19, rather than 0.13). The clinical experts said that most AEs can be managed in an outpatient setting. They noted that AEs can have a delayed onset and can be unpredictable, but clinicians are getting better at spotting and managing them. The patient experts said that, based on their research, most people are willing to accept additional AEs if it gives them the chance of living a meaningful life for longer. The patient experts said that, compared with chemotherapy, the side effects for dostarlimab were easy to manage. The committee concluded that, given the small sample size, it was appropriate to include the broader range of AEs (those affecting at least 2% of people). It also agreed that the company's approach for the ongoing monitoring of immune-related AEs in the dostarlimab arm was sufficient (0.13 visits a week from cycle 19) and did not need to be increased.

The committee also discussed the apparent lack of efficacy in people with stage 3 disease. The clinical experts said that this was unexpected, and that it may be a statistical quirk caused by the low number of people in this subgroup. The committee concluded that this issue did not need further consideration.

A small number of people may have previously been treated with platinum-based chemotherapy (before or after surgery) by this stage in the treatment pathway. These people would therefore be eligible for pembrolizumab with lenvatinib (see NICE's technology appraisal on pembrolizumab with lenvatinib for previously treated advanced or recurrent endometrial cancer), as an alternative to dostarlimab. The company did not include this comparator in its economic model because of a lack of data. The EAG agreed that there was not enough data to inform this comparison, and the clinical experts said they would be unlikely to use pembrolizumab with lenvatinib in this group. The committee concluded that the patient population for this comparison was very small and that the issue did not need further consideration.

NICE health technology evaluations manual notes that above a most plausible ICER of £20,000 per quality-adjusted life year (QALY) gained, decisions about the acceptability of a technology as an effective use of NHS resources will take into account the degree of certainty around the ICER. The committee will be more cautious about recommending a technology if it is less certain about the ICERs given. The committee noted several uncertainties, specifically:

• the robustness of the RUBY‑1 results, given the small sample size and imbalances in baseline characteristics (see section 3.3)

• if and when the treatment effect starts to wane and over what time period (see section 3.7)

• the long-term clinical benefit of dostarlimab because of data immaturity (see section 3.6 and section 3.8).
  
  The committee also considered the large unmet clinical need, the durable benefits that have been seen in dMMR/MSI‑H populations and the data seen in RUBY‑1 to date. Given the level of uncertainty, the committee concluded that the maximum acceptable ICER would be £25,000 per QALY gained. It noted that the maximum acceptable ICER may change once more data becomes available.

The committee considered the ICERs for dostarlimab plus PCC compared with PCC alone. Because of confidential commercial arrangements for dostarlimab and subsequent treatments in the pathway, the ICERs are confidential and cannot be reported here. The committee's preferred cost-effectiveness estimates included the following assumptions:

• AEs that affect at least 2% of people

• piecewise extrapolation (Odds K=1) for PFS

• baseline starting age in the model of 67.1 years

• 0.13 outpatient visits per week from cycle 19 for AE monitoring

• treatment waning should not apply in 3 years of starting treatment.
  
  The committee was unable to agree on a preferred assumption for the OS extrapolation, based on the immaturity of the data. It also agreed that there was too much uncertainty about the robustness of the ICERs produced by the economic model. Some of the most likely cost-effectiveness estimates are higher than what NICE usually considers an acceptable use of NHS resources. So, the committee concluded it was not possible to recommend dostarlimab with platinum-based chemotherapy for use in routine commissioning.

Having concluded that dostarlimab with platinum-based chemotherapy could not be recommended for routine use, the committee then considered whether it could be recommended in the Cancer Drugs Fund. The committee discussed the arrangements for the Cancer Drugs Fund agreed by NICE and NHS England in 2016, noting NICE's Cancer Drugs Fund methods guide (addendum; see NICE's managed access page):

• The company expressed an interest in dostarlimab being considered for funding through the Cancer Drugs Fund.

• The committee noted that RUBY‑1 is ongoing. Also, an updated data cut for RUBY‑1 became available shortly before the first committee meeting, but there was not enough time available for it to be included in the economic model. So, it was not considered by the committee in this appraisal.
  
  The committee considered that further data collection in the Cancer Drugs Fund could address some of the uncertainty in the company's estimates. Using the committee's preferred assumptions for PFS, starting age, treatment waning and AEs, dostarlimab has plausible potential to be cost effective. The committee concluded that dostarlimab met the criteria to be considered for inclusion in the Cancer Drugs Fund. It recommended dostarlimab with platinum-based chemotherapy use in the Cancer Drugs Fund as an option for people with advanced or recurrent endometrial cancer with MSI‑H or dMMR deficiency if the conditions in the managed access agreement are followed. When the guidance is next reviewed, the company should use the committee's preferred assumptions as set out in section 3.14, unless new evidence shows otherwise.

The committee noted that black ethnic groups have substantially higher mortality rates for endometrial cancer mortality than other ethnic groups in the UK. The company said that access to innovative treatment on the NHS for late-stage disease can help address severe inequalities in survival outcomes by ethnicity or socio-economic deprivation. The committee considered equality issues and concluded that its recommendations do not affect people protected by the equality legislation differently to the wider population.