The main results of this study are: i) the implantation of coils for ELVR does not result in a clinically relevant or sustained benefit for patients with severe COPD and ii) there is no long-term survival benefit after 48 months.
Given current literature, 29 patients can be considered as a relatively large cohort for a single center analysis, as a total of 60 patients were included in a European multicenter randomized controlled trial at the same time and the Swiss National Registry comprises 64 patients treated over a period of three years 11,12. Since all procedures were performed by the same interventionalist variance in procedure quality is unlikely to have an effect on patient’s outcomes. There was no statistically significant overall improvement in either lung function or exercise capacity 6 months after ELVR. Singling out distinct patients, some achieved greater improvements of FEV1, RV and 6-MWD than others. Donohue has previously defined “responders” from ELVR as patients achieving an increase in FEV1 by at least 100 ml 7. Similar propositions have been made for RV (at least -430 ml) 8 as well as 6-MWD9 (at least +23 m) and are commonly employed as ELVR response criteria. In this study, 48.3% of patients showed short-term increases in FEV1 of at least 100 mL or more thus potentially qualifying as short-term responders. Response in RV was less frequently observed (31.03%). Only 7 out of 29 patients benefited in both categories. Only 4 out of 25 patients had increases in 6-MWD of more than 26 m (4 patients lost in follow-up), but only two of them also qualified as responders in FEV1 or RV, suggesting that response in lung function
CT scans were submitted to two independent commercial providers of CT analysis software to exclude that failure to improve by ELVR was due to misjudgment of the investigator. When target lung lobes were determined by computed algorithms, the selected lobes differed from clinician’s choice in 18.75% and 30% of cases, respectively. Judging from both comparisons, determination of target lung lobes for endoscopic lung volume reduction coils by two commercially available software products was not superior to investigator’s choice. Admittedly, one software was not designed to guide coil implantation and the number of properly analyzed CT scans is low, so a definite conclusion if software guided coil deployment is superior to investigator guided isn’t possible from the data presented here.
Results from this single-center cohort study indicate that endoscopic lung volume reduction coils did not benefit most patients with severe emphysema. Since most published randomized controlled trials investigating ELVC come to more optimistic conclusions11,12,13, we first compared the results to the patients’ baseline characteristics. The patient populations were very comparable regarding age, sex distribution, body mass index and FEV111,12,13, while average forced vital capacity (FVC) and 6-MWD were slightly more severely impaired in our study population. Residual volume as well as RV/TLC on the other hand were quite congruent, indicating that our patient collective mostly represented that of larger RCTs investigating endoscopic lung volume reduction coils. In addition, the number of placed coils per sitting was also 10 on average per procedure. We therefore conclude that a possible beneficial effect was neither disguised by a more morbid cohort, nor by a difference in the procedure itself or the interventionalist’s experience.
In May 2019, Slebos et al proposed computed analysis of patients’ chest CT scans and RV > 200 % as predictors of response to coils 14. Interestingly, 60% of long-term responders in our study had baseline RV below 200 % of expected. Additionally, subgroup comparison revealed that response rate was not higher in those cases when investigator’s CT analysis overlapped with software-based analysis. We conclude that response criteria as suggested by Slebos et al. did not translate to our study cohort, as they did not identify those patients who ended up benefitting from ELVR.
Comparing the rate of major complications in our study, 6.9% ranked well below what was reported from most above-mentioned clinical trials. It should be noted though that 40% of our patients had hemoptysis to varying degrees post intervention, in one case requiring intensive care. Survival without transplant after 48 months was well comparable to what was predicted by BODE indices, indicating that ELVR coils had no relevant effect on patients’ long-term prognosis.
Our study has several important limitations that need to be addressed. The study is a single center retrospective real-life assessment of ELVR using coils. While the baseline parameters of treated patients and primary results of the intervention were comparable to those of other studies, our patients seemed to deteriorate earlier than those of other cohorts. However, we did not compare the results of the intervention to a cohort receiving optimal non-interventional therapy only. Additionally, we did not perform an analysis of the patients' quality of life. This would have been interesting additional information, however, the data had not been consequently collected in this real-world setting for most of the patients.
Another limitation is the fact that some patients did not receive treatment of the contralateral side and six patients had previously received valves, as they had complete lobar fissure without collateral ventilation in a target lobe.
Most importantly, some patients received predominantly coils of 100 mm length, which might have led to less lung tissue compression than what might have been achieved through larger coils. The strategy to implant bigger coils, eventually creating more elastic recoil, was adopted during the study period. However, the impact of coil-size on patient outcome after ELVR is currently unclear.