Bombshell 10-Year JCOG0802 Results Show Lobectomy Is Superior to Segmentectomy for Lung Cancer

I am surprised that there is so little critical analysis of the "Asamura Bombshell". As a respected professor and one of the most senior thoracic surgeons in the world, Asamura has made these comments knowing he should be challenged, but no one has stepped up. Here are my 2 cents on this issue. 1. No RCT has found inferiority of segmentectomy against lobectomy for peripheral T1N0 NSCLC ≤ 2cm. 2. The median age of patients after 10 years in JCOG 0802 is 77, so many are over 80 and we have not yet achieved immortality, so of course the curves are coming together as part of this regression to the norm for very elderly people (Saji, the first author of the paper even said this in question time at AATS). The non-significance of the persistent 3.8% survival benefit of segmentectomy relates to the fact that loss-to follow-up and deaths have left just over half of the original population still at-risk for analysis. The study was not powered to detect a difference at 10 years in any case. No surprise to the power of 2! 3. In between the 5 and 10-year milestone, quite similar relapse and local recurrence occurred in the two groups, suggesting the durability of segmentectomy as an oncological therapy. 4. Local recurrence in segmentectomy group did not convert to death in all cases, as at least 7 cases received a completion lobectomy and 1 received a completion pneumonectomy based on Saji's comments at AATS. 5. The issue of "other cancer and new lung cancer deaths" deserves a mention. Without knowing the individual patient circumstances, it is possible that the "other cancer" diagnosis rate was similar between the two groups, but the intensity of therapy differed to the detriment of lobectomy patients. That makes a higher death rate from "other cancer and new lung cancer" certainly plausible. 6. the contention that there is no respiratory benefit of segmentectomy is just not plausible, and the method of reportage was misleading. There is no more insensitive method of comparing respiratory function than the difference of median FEV1. Blowing out birthday candles is not the same as running for the train. I have done waterfall plot simulations of the reported medians and 95% confidence intervals for both CALGB and JCOG trials and they show the relative risk of having a 10% loss of FEV1 is 3.3 (p<0.00001) and 1.68 (p<0.00001), respectively. Even for such a rubbish single indicator of wholistic lung function, the trials clearly showed more harm for lobectomy despite the statistically bizarre way of framing a highly significant and very clinically important difference. In summary, I don't think there is anywhere near enough evidence to say we should throw out the baby (healthy lung tissue) with the dirty bathwater (slightly higher local recurrence) or that there is no respiratory benefit of segmentectomy over lobectomy. Asamura should be proud of his legacy, not apologising for it.