Authors' objectives: The manufacturer, PULMONX submitted a single-technology assessment (STA) of Zephyr® valve system to the Norwegian Institute of Public Health for evaluation of effect, safety, and health economics. The present report is an appraisal of this STA.

Authors' results and conclusions: The manufacturer’s literature search identified four randomized trials (IMPACT, STELVIO, LIBERATE, TRANSFORM) comparing Zephyr® valves with standard medical care and one trial (BeLieVeR-HIFI) that compared Zephyr® valves with a sham procedure. The five studies included a total of 498 patients: 295 got Zephyr®, 178 received standard treatment, and 25 received sham valve procedure. Effect and safety There was no clear difference in mortality between the Zephyr®-group and the control group (risk ratio: 1.61, 95% CI from 0.44 to 5.93; ⨁◯◯◯). These results depend on a risk ratio estimated using only seven deaths (7/270) in the Zephyr®-group and two (2/178) in the standard treatment group, implying that no clear conclusions could be reached regarding Zephyr® valve treatment and the risk of all-cause mortality. Moreover, there was no clear association between the Zephyr® procedure and the risk of COPD exacerbations. None of the studies found a statistically significant difference, and the pooled effect is estimated to OR 1.15 (95% CI 0.70 to 1.88; ⨁⨁◯◯), but the meta-analysis is based on a limited number of events (n=91) and the certainty of evidence is low. All studies showed an important improvement in FEV1 for the Zephyr® group compared to standard care. Studies included in the meta-analyses had different follow-up periods ranging from three to twelve months, but there was no serious heterogeneity in the results. The mean difference between the groups were 0.14 litres (95% CI 0.13 to 0.16; (⨁⨁⨁◯), which is higher than the suggested minimal clinically important difference (MCID) at 0.12 litres. The use of Zephyr® valves probably result in more favorable scores on the St. Georges’s Respiratory Questionnaire (SGRQ) after three to twelve months than standard care. A suggested MCID for this outcome is four points, and the estimated difference was almost eight points (95% CI 5 to 11; ⨁⨁⨁◯) in favour of Zephyr®. Zephyr® may improve 6-minute walking distance (6MWD) as compared to standard care three to six months after the procedure. The available studies showed heterogenous results, but random-effect meta-analyses still showed results in favour of Zephyr® (MD 57 metres, 95% CI 36 to 78; ⨁⨁◯◯). MCID for 6MWD in severe COPD is estimated to approximately 30 metres. Measurements of the BODE index were in favour of the Zephyr®-valve in all studies. Pooled estimate across the four studies showed that the BODE index was 1.3 points lower (95% CI -1.6 to -1.0; ⨁⨁⨁◯) after Zephyr® than after standard care (MCID = -1 point). The Zephyr®-valve procedure may increase the risk of pneumothorax (OR 34, 95% CI 8 to 142; ⨁⨁◯◯). Health economy Based on the submitter’s economic model, the ICER of Zephyr® valve treatment, when compared to standard care, is NOK XXXXX per QALY in a three-year perspective and NOK XXXXX per QALY in a ten-year perspective. However, 10- year estimates are considerably more uncertain than those in a 3-year perspective due to a lack of longer term efficacy and safety data. Clinical experts also consider a ten-year time horizon to be too long compared to the life expectancy og the relevant patient group. Hence, Zephyr® valve treatment may be cost-effective in the Norwegian setting, but there remain important uncertainties. The manufacturer anticipates a gradual increase in the number of patients undergoing a Zephyr® valve procedure with five percent each year, and the clinical expert anticipate a target population around 25 patients per year in Norway. In this scenario the cumulative budget consequences during a five-year period are estimated to NOK XXXXX.

Authors' methods: We used the documentation provided by the manufacturer. The literature search was checked by two librarians at NIPH. The manufacturer did not report whether they used independent screening and data extraction. We used the risk of bias assessments provided by the manufacturer. The manufacturer did not use GRADE, but we have graded the evidence. One researcher did the GRADE assessments, and another researcher checked the assessments. We categorized the certainty of evidence as high (⨁⨁⨁⨁), moderate (⨁⨁⨁◯), low (⨁⨁◯◯) or very low (⨁◯◯◯) according to GRADE. According to information in the submission file , the following databases were searched in April 2020: MEDLINE, EMBASE, Cochrane (CENTRAL) and ClinicalTrials.gov. The search was meant to update a search from a NICE guideline from 2017 using the following PICO: Population: People with emphysema (heterogeneous or homogeneous) Intervention: Endobronchial valves (EBV), Zephyr® valves, PulmonX Inc after assessment of collateral ventilation with Chartis® flow sensor and catheter Comparison: Sham procedure or standard care Outcome: FEV1, SGRQ, 6MWD test, BODE Index, pneumothorax episodes, COPD exacerbations episodes and death (for all causes, for respiratory complications, for IHD)

Project Status: Completed

Year Published: 2022

URL for published report: https://www.fhi.no/en/publ/2022/Zephyr-valves-system-in-the-treatment-of-emphysema/

English language abstract: An English language summary is available

Publication Type: Rapid Review

Country: Norway

Organisation Name: Norwegian Institute of Public Health

Contact Address: P.O. Box 222 Skoyen, N-0123, Oslo