GreenLight XPS for treating benign prostatic hyperplasia

3.1 The company submissions included 3 publications of a single trial (the GOLIATH study: Bachmann et al. 2014, Bachmann et al. 2015, Thomas et al. 2015), which compared GreenLight XPS with transurethral resection of the prostate (TURP). GOLIATH was a European multicentre randomised controlled trial in 281 people aged between 40 years and 80 years with a prostate volume less than 100 ml who were not on active anticoagulation therapy. Results showed no statistically significant difference in benign prostatic hyperplasia (BPH) symptom improvement (measured on the International Prostate Symptom Score [IPSS] or as maximum urinary flow [Qmax]) between GreenLight XPS and TURP up to 2 years. Using GreenLight XPS resulted in a significantly shorter duration of catheterisation and shorter lengths of stay. The committee at the time of the original guidance concluded that GreenLight XPS was as effective as TURP in treating BPH in non-high-risk groups.

3.2 The EAG identified 10 studies in total, including 2 from the 3 that the company submitted, that were relevant to the high-risk groups. High risk was defined in the previous guidance as people with a higher risk of bleeding (such as those on anticoagulants), larger prostates (over 100 ml) and urinary retention. Five of these studies included comparative clinical data. There were significant improvements from baseline in all clinical outcomes (p<0.001). The clinical experts said that GreenLight XPS may be a safe alternative to TURP in this population. However, the committee agreed that there was not enough evidence to show any notable differences in effectiveness or adverse events using GreenLight XPS in the high-risk population compared with TURP and holmium laser enucleation of the prostate (HoLEP). The committee therefore concluded in the original guidance that multicentre prospective studies with GreenLight XPS were needed in this population.

3.3 Following a review of the evidence base in 2019, NICE decided to update the guidance to consider the new evidence on its use in high-risk groups since the original guidance.

3.4 For the guidance update, the EAG considered a total of 58 new studies relevant to the decision problem. Because of the size of the evidence base, the EAG prioritised 37 of these studies:

3.5 No new RCTs comparing GreenLight XPS with TURP or HoLEP were identified at guidance update. The GOLIATH trial remained the only randomised controlled evidence comparing GreenLight XPS with TURP. There was no randomised evidence comparing GreenLight XPS with HoLEP. At guidance update, 6 observational studies compared GreenLight XPS with TURP. The EAG said that further randomised comparative studies in people at high risk exclusively may not be ethical. This is because of the increased risk of bleeding, complications and longer hospital stays associated with TURP.

3.6 The new evidence suggested that, when compared with TURP, GreenLight XPS was associated with a significantly shorter hospital stay, significantly shorter postoperative catheterisation period and significantly higher preservation of ejaculatory function at 12 months (Reimann et al. 2019, Cimino et al. 2017, Mattevi et al. 2017, Gondran-Tellier et al. 2021, Mathieu et al. 2017). While most of this new evidence included people at high risk (50 of 58 studies), only 4 reported on high-risk populations exclusively. Two were comparative studies (Gondran-Tellier et al. 2021, Mesnard et al. 2021) and 2 were retrospective cohorts (Meskawi et al. 2017, Eken and Soyupak 2018). An additional 4 cohort studies stratified by anticoagulation status (Lee et al. 2016, Knapp et al. 2017, Meskawi et al. 2019, Eken and Soyupak et al. 2018). Details of these studies are in section 4.2 of the assessment report update in the supporting documentation (NICE 2022).

3.7 The original guidance included 2 published cost-effectiveness studies, both of which compared GreenLight XPS with TURP (Thomas 2015, Bunejam-Gual 2014). Both studies suggested that reduced length of stay, or an increased proportion of procedures done as day cases would be associated with a cost saving when using GreenLight XPS.

3.8 The EAG identified 6 economic studies published since the original guidance. None was done in the UK. Two studies reported GreenLight XPS to be cost saving against TURP (Masucci et al. 2018, Ulchaker and Martinson 2018), one reported GreenLight XPS to be more costly but more effective than TURP (Caicedo et al. 2019). Two studies reported TURP to be more cost effective (Erman et al. 2018, Ulchaker and Martinson 2018) and 1 reported cost savings when compared with HoLEP or ThuLEP in people with a prostate volume greater than 80 ml (Mathieu et al. 2017). For full details of the cost evidence, see section 9 of the assessment report update in the supporting documentation (Newcastle EAG 2022).

3.9 In the original guidance the company developed a decision tree model, which was used to inform the committee's recommendation. These compared the cost consequences of using GreenLight XPS with:

3.10 With the updated clinical and cost parameters, the EAG's base case results suggested that GreenLight XPS remains cost saving by £70 per person compared with TURP, but is cost incurring when compared with HoLEP (an additional cost of £114 per person). The latter was because of reduced capital costs associated with increased use per year of HoLEP in the updated model. Base case estimates assume 4% of TURP procedures and 36% of GreenLight XPS procedures were done as a day case procedure. The key driver of the cost saving was the proportion of procedures that could be carried out as day cases. The EAG's threshold analysis suggested that GreenLight XPS would be cost incurring if the proportion of day case procedures for TURP or HoLEP was above 43.6% and 56% respectively. This is assuming the proportion of GreenLight XPS procedures done as day cases stayed at 68%. The EAG and clinical experts agreed that these thresholds were clinically unlikely in the NHS.

3.11 The company submitted a new cost model during the guidance update. It had a Markov model structure, which allowed for retreatment, and had a 4‑year time horizon. The model included everyone who needed treatment for BPH and had a high-risk group scenario, which was informed by the results of an unpublished systematic review. The EAG considered the unpublished systematic review to be low quality and the results of the review not robust because of methodological concerns. Details of the EAG's critique are in the economic model parameters of section 9.4 of the assessment report update in the supporting documentation (Newcastle EAG 2022).

3.12 During the consultation, the systematic review was published (Burtt et al. 2022). The EAG reviewed and critiqued the published review. It considered that the publication provided no additional new evidence and the main methodological concerns remained. The EAG concluded that the published review was not sufficiently robust to inform a cost model for the high-risk population.

3.13 Given these limitations the EAG judged that the GOLIATH trial remained the most robust comparative evidence and that there was no new prospective comparative evidence specifically on using GreenLight XPS in high-risk populations, since the original guidance. The EAG considered that modelling all people whose BPH was treated, including those in high-risk groups, was more appropriate and more generalisable to the NHS. The EAG made some changes in the model costs and clinical parameters, including: extending the time horizon to 5 years, reducing the capital cost of HoLEP, and increasing the length of stay. Full details are in the assessment report update in the supporting documentation (Newcastle EAG 2022).

3.14 The EAG's revised base case analysis for the Markov model showed that GreenLight XPS is cost saving, per person, by £304.83 compared with TURP and £269.52 compared with HoLEP. The EAG did a limited probabilistic sensitivity analysis, varying just 2 parameters, because of the lack of data. Base case cost savings were driven by the duration of procedures and the length of stay after procedures. Threshold analyses suggested that GreenLight XPS would become cost incurring if TURP and HoLEP were done in less than 43.7 minutes and 60.0 minutes respectively (relative to 49.6 minutes for GreenLight XPS). GreenLight XPS would also become cost incurring if the length of hospital stay after TURP or HoLEP was less than 1.5 days and 0.9 days respectively (relative to 1.6 days for GreenLight XPS). The clinical experts agreed that the scenarios of length of stay or proportion of day cases that would make GreenLight XPS cost incurring are unlikely in the NHS.

3.15 There was no comparative data on length of stay. The company estimated a length of stay of 0.7 days for GreenLight XPS based on a single arm, single-centre study in Canada (Ajib et al. 2018). The EAG applied a 1.6‑day length of stay derived from NHS activity data (hospital episode data). Both data sources had limitations (see details in table 22 of the assessment report update in the supporting documentation, Newcastle EAG 2022). After the public consultation, the EAG did additional analyses to consider the possible size of cost savings with GreenLight XPS by applying different values for length of stay. One scenario was informed by a clinical expert's opinion that length of stay with GreenLight XPS was 1 day. Length of stay with HoLEP and TURP was kept at 1.6 days and 2.3 days respectively. Two additional scenarios were informed by the British Association of Urological Surgeons Bladder Outflow Obstruction Audit data (2019), which reported mean lengths of stay (for people with and without a catheter pre-operatively, respectively) of: