Naloxegol for treating opioid‑induced constipation

Clinical effectiveness

4.5 The Committee considered the clinical effectiveness of naloxegol as seen in the KODIAC 4 and 5 trials, which compared naloxegol with placebo in people with OIC. It was aware that the marketing authorisation for naloxegol covers only people whose constipation had not adequately responded to laxatives and that an inadequate response to laxatives was defined in the trials and in the summary of product characteristics (see section 4.3). It noted the Evidence Review Group's (ERG's) concerns that it may take more than 4 days to assess the effectiveness of some laxatives. The company stated that most patients in the trial had used the laxatives for more than 4 days before being classed as inadequate responders, and that 4 days was the minimum duration of use. The clinical experts stated that although some laxatives will take longer than 4 days for their effectiveness to be established, the majority will work within this time. The Committee questioned whether the trial definition of laxative‑inadequate response was relevant in clinical practice in England. It heard from the clinical experts that the definition of laxative‑inadequate response is subjective in clinical practice, and that the company's definition was reasonable. The clinical experts also stated that the decision to consider alternative treatments in practice would depend on a person's quality of life after using laxatives. The Committee accepted the definition of inadequate response to laxatives used in the clinical trials and in the summary of product characteristics for naloxegol.

4.6 The Committee discussed the generalisability of the KODIAC 4 and 5 studies to the population in England with OIC. It was aware that the trials were mainly done in the USA and that some of the baseline characteristics of the trial population differed from the population of England with OIC. It noted that the proportion of people who were obese in the studies was higher than the proportion of people who are obese in England. In addition, the average age of the patients included in the studies was 52.2 years, whereas the population in England who take opioids is likely to be older, particularly in people with cancer pain. The Committee heard from the clinical experts that although these baseline characteristics differed between the trial population and the population in England with OIC, the efficacy of naloxegol was not expected to be affected by age or weight and that the results of the trials would be generalisable to people with OIC in England. The Committee also heard from the company that its post‑hoc subgroup analyses reported in the European Public Assessment report for naloxegol, showed naloxegol to be effective compared with placebo regardless of age or weight. The Committee concluded that the KODIAC 4 and 5 trials could be generalised to the population seen in clinical practice in England and that it could use those results in its decision making.

4.7 The Committee also considered the generalisability of the results of the KODIAC 4 and 5 studies to people with cancer pain who have OIC. The Committee noted that both KODIAC 4 and 5 specifically excluded people with cancer. The Committee heard from the company that a separate study which included people with cancer was started, but was stopped due to extreme difficulties of enrolment for this group of people. The Committee heard from the company and clinical experts that it was difficult to enrol patients with cancer mainly because of short life‑expectancy. The Committee heard from the clinical experts that naloxegol was likely to be effective in people with cancer because naloxegol targets the OIC, rather than the underlying condition causing the pain. The Committee noted that the marketing authorisation for naloxegol did not exclude people taking opioids for cancer pain. However, it was aware that the summary of product characteristics includes some special warnings about the lack of clinical trial evidence in people with cancer and the contraindications for certain patients who are at heightened risk of gastrointestinal perforation (including people with cancer pain). Having heard from the clinical experts and considering the marketing authorisation, the Committee was persuaded that naloxegol would be equally effective in people with cancer pain who have OIC taking into account the special warnings highlighted in the summary of product characteristics. It therefore concluded that its recommendations regarding the use of naloxegol in clinical practice also applies to people with cancer pain who have OIC.

4.8 The Committee considered the results of KODIAC 4 and 5 presented by the company. It noted that the results suggested a statistically significant improvement over weeks 1 to 12 in the proportion of laxative‑inadequate responders who had spontaneous bowel movements (SBMs) in the naloxegol arm compared with the placebo arm (see section 3.8). The Committee heard that although the number of SBMs during the trial was important, the most clinically relevant outcome measure was the mean difference from baseline in SBMs. This is because there are large differences in the number of SBMs normally experienced by people without OIC. The Committee also noted that naloxegol was associated with statistically significant improvements in secondary outcomes, including health‑related quality of life compared with placebo in the KODIAC 4 and 5 studies (see section 3.8). Therefore, the Committee concluded that naloxegol was effective compared with placebo as shown by the data on SBM frequency in the KODIAC 4 and 5 studies in people with OIC that has not responded adequately to laxatives.

4.9 The Committee considered evidence on the company's mixed treatment comparison of naloxegol, methylnaltrexone and naloxone‑oxycodone. The Committee was aware that in the mixed treatment comparison, the company had used response rates from the laxative‑inadequate responder population in the naloxegol KODIAC 4 and 5 studies, and conversely, rates for the overall population from the methylnaltrexone and naloxone‑oxycodone studies. The Committee noted that the results from the ERG's exploratory mixed treatment comparison, in which it used data for the overall population from the KODIAC trials, did not differ substantially from the company's estimate. Nevertheless, the Committee considered that using 2 different populations introduced uncertainty in the company's analyses. The Committee also noted that none of the results from the mixed treatment comparison for naloxegol compared with methylnaltrexone and naloxone‑oxycodone were statistically significant, indicating that there was also uncertainty regarding its relative efficacy in the mixed treatment comparison. The Committee concluded that there was insufficient evidence that naloxegol's clinical effectiveness differed from that of methylnaltrexone and naloxone‑oxycodone, and that it was not unreasonable to use the mixed treatment comparison analysis in its decision‑making.

Cost effectiveness

4.10 The Committee discussed the company's cost‑effectiveness analysis. It considered all the comparisons provided by the company, including naloxegol compared with placebo (with and without bisacodyl), methylnaltrexone and naloxone‑oxycodone, noting that methylnaltrexone was the most appropriate comparator for this appraisal (see section 4.4), The Committee noted that the company's comparison of naloxegol with placebo included analyses with and without bisacodyl. However, it was aware that most patients in KODIAC 4 and 5 had bisacodyl as a rescue medication. It also noted the ERG's comments that the use of the rescue medication may have positively affected the SBM rates in the trial. The Committee concluded that the analysis without bisacodyl was neither clinically relevant nor consistent with the KODIAC 4 and 5 trials. Therefore, for the comparison with placebo, it did not consider the analysis without bisacodyl for both groups any further.

4.11 The Committee considered the company's economic model and the ERG's critique of the model. The Committee noted that the company's base case ICER for naloxegol plus bisacodyl compared with placebo plus bisacodyl was £11,200 per QALY gained. The Committee noted that for the analyses comparing naloxegol with placebo, the company designed its model to include time‑specific and treatment‑specific utilities. The Committee heard from the ERG that it would have been more appropriate to use health state‑dependent utility values only, rather than assuming different utilities for the treatment arms. The Committee noted that the company's base case ICER increased to £38,900 per QALY gained when the company used health state‑dependent utility values for naloxegol compared with placebo (without bisacodyl in both treatment groups). Similar analysis was not presented for the comparison that included bisacodyl in both treatment groups. The Committee understood from the ERG that the large increase in the ICER was a result of the model structure, in that the non‑OIC (on treatment) state in the model was too broad, that is, it included a heterogeneous group of patients with different number of SBMs during the same period (see section 3.34), and that applying a single utility value to that health state would not accurately reflect patient experience in that state. The ERG stated that the model should have included more discrete health states which were more reflective of patient experience and this would have allowed the company to apply health‑state specific utilities. However, the Committee understood from the ERG that taking this approach may not necessarily have changed the model results.

4.12 The Committee noted that there was less uncertainty when comparing naloxegol with methylnaltrexone and naloxone‑oxycodone, because health state‑specific utilities were used for these comparisons rather than treatment‑specific utilities. It noted that when compared with methylnaltrexone and naloxone‑oxycodone, naloxegol dominated (that is, naloxegol was both more effective and cheaper than) these treatments in almost every scenario (see sections 3.28 and 3.29), except when naloxegol was given with oxycodone compared with naloxone‑oxycodone (which produced an ICER of £34,100 per QALY gained). It also noted that the comparison of naloxegol with naloxone‑oxycodone was for the subgroup of people taking a step 3 opioid and not the full population covered by the marketing authorisation. In addition, the Committee had previously decided that naloxone‑oxycodone was not the most relevant comparator for this appraisal because it is not used frequently in UK clinical practice (see section 4.4). The Committee concluded that although the company's model had some limitations, overall it was acceptable for modelling treatment in this population.

4.13 The Committee considered the assumptions used by the company to model the duration of treatment response – that is, transition from the non‑OIC (on treatment) health state to the OIC health state. The Committee noted that for its base case, the company had chosen the exponential function for both naloxegol and placebo because it was the most conservative of the available functions. However, the Committee observed that after only 2 years nearly all patients having naloxegol transitioned from the non‑OIC (treated) health state to the OIC health state, which suggested that naloxegol lost its treatment effect over time. The Committee queried whether this was clinically plausible, and heard from clinical experts that if a patient were having opioids and naloxegol, there was no reason to expect that the treatment effect of naloxegol would lessen over time. The Committee also heard from the company that the discontinuation rates in the KODIAC 8 trial were very small, even though the trial lasted for 1 year. The Committee noted that in the ERG's exploratory analyses, the ICERs for naloxegol plus bisacodyl compared with placebo plus bisacodyl were mostly less than £13,000 per quality‑adjusted life year (QALY) gained (see sections 3.35 and 3.36). The Committee concluded that the ERG's exploratory analyses had little impact on the cost‑effectiveness results presented by the company.

4.14 The Committee considered the effect of the company's not conducting a fully incremental analysis (that is, calculating the incremental QALY gains and costs for all treatment options and ordered by increasing costs). The company stated that an incremental analysis was not possible, because the definition of response from the KODIAC trials as used in the placebo comparisons differed from the definition used for the mixed treatment comparison for the active comparators. The mixed treatment comparison analyses did not include all the comparators in the economic model (as requested by NICE during clarification) and therefore no comparable ICERs or incremental analysis was available from the mixed treatment comparison either. The Committee stated that it would have preferred to see a fully incremental analysis as described in the guide to methods of technology appraisals. However, the Committee considered methylnaltrexone to be the most relevant comparator, and because naloxegol dominated methylnaltrexone in the pairwise analysis (that is, naloxegol was both more effective and cheaper); it would also dominate it in an incremental analysis. The Committee concluded that the ICERs presented in the pairwise analyses were sufficient evidence on which to base its decisions.

4.15 The Committee considered whether naloxegol was a cost‑effective use of NHS resources compared with the comparators. The Committee noted that for all comparisons, the company's base‑case results and most of the company's sensitivity analyses resulted in ICERs below £20,000 per QALY gained (see section 3.27). It was aware that all sensitivity analyses by both the company and the ERG showed that the model was stable and the ICERs robust to most model changes (see section 3.37). The Committee noted that the ICERs were above £20,000 per QALY gained only when health state‑specific utilities were used and when the outcomes were not extrapolated beyond the trial period for naloxegol compared with placebo, and also when naloxegol was given with oxycodone compared with naloxone‑oxycodone (see sections 4.11 and 4.12). The Committee concluded that in light of the robustness of the company's model, the ICERs being mostly below £20,000 per QALY gained for the comparison of naloxegol plus bisacodyl with placebo plus bisacodyl, and naloxegol mostly dominating methylnaltrexone and naloxone‑oxycodone, naloxegol was considered a cost‑effective use of NHS resources. The Committee therefore recommended naloxegol as an option within its marketing authorisation for people with OIC that has not responded adequately to laxatives.

4.16 The Committee considered whether naloxegol could be considered innovative in its potential to make a substantial effect on health‑related benefits for people with opioid‑induced constipation and whether it could be considered a step‑change in the management of opioid‑induced constipation. The Committee heard from the company that the innovativeness of naloxegol was in the combination of its mode of action and formulation, because it offers more flexibility in dosing than naloxone‑oxycodone and can be used independently of the opioid being prescribed. The Committee heard from the clinical specialists that naloxegol would be a useful option for treating opioid‑induced constipation. It noted that naloxone has been in use for many years and that the only innovation it could discern was the attachment of a polyethylene glycol molecule to naloxone in order to prevent it from crossing the blood‑brain barrier. The Committee considered that although this addition was novel, there were no additional gains in health‑related quality of life over those already included in the QALY calculations.

4.17 The Appraisal Committee considered whether it should take into account the consequences of PPRS 2014, and in particular the PPRS Payment Mechanism, when appraising naloxegol. The Appraisal Committee noted NICE's position statement in this regard, and accepted the conclusion 'that the 2014 PPRS Payment Mechanism should not, as a matter of course, be regarded as a relevant consideration in its assessment of the cost effectiveness of branded medicines'. The Committee heard nothing to suggest that there is any basis for taking a different view with regard to the relevance of the PPRS to this appraisal of naloxegol. It therefore concluded that the PPRS Payment Mechanism was irrelevant for the consideration of cost effectiveness of naloxegol.

Summary of Appraisal Committee's key conclusions