Key Ideas
The surgery-versus-radiation debate in stage III NSCLC is increasingly the wrong framing; the better approach is starting with the best available systemic therapy, then reassessing with a multidisciplinary team to determine which local therapy, surgery or radiation, is right for that specific patient. Defining "resectability" itself remains genuinely difficult and inconsistent, and induction chemo-immunotherapy is converting a meaningful number of previously unresectable patients into surgical candidates.
A Reasonable Bias to Disclose
As a surgeon, I'll acknowledge upfront that I likely carry some bias toward surgery, and I'd like to avoid the "radiation versus surgeon" framing that shows up too often on social media. My honest view is that as perioperative chemo-immunotherapy has advanced, there's growing consideration in my own practice for surgery even in patients who might be considered borderline or previously unresectable, and I think the field may be heading toward a model where treatment starts with neoadjuvant systemic therapy, followed by multidisciplinary reassessment to select the optimal local therapy, surgery or radiation, for that specific patient, rather than deciding upfront.
Unresectable Stage 3 Disease: A Real, if Incremental, Evolution
The PACIFIC trial (published in 2017) was genuinely practice-changing for unresectable stage III disease: adding durvalumab after chemoradiation improved PFS to 17 months compared with under 6 months for chemoradiation alone. Since then, progress has been more incremental. The COAST trial (testing novel agents added to the PACIFIC backbone, with updated 2025 data) showed modest PFS improvement, roughly 20-21 months, though the PACIFIC-arm comparator in that specific trial was only 6.3 months, meaningfully lower than in the original PACIFIC trial itself, likely reflecting differences in patient selection, worth keeping in mind before over-interpreting the magnitude of benefit. Separately, PACIFIC-2 and INCMGA0012-181 (EA5181) both tested giving immunotherapy concurrently with chemoradiation rather than sequentially after, and neither trial met its primary endpoint; no significant PFS improvement was found. It's worth noting these two trials also measured PFS from randomization rather than from the end of chemoradiation (as PACIFIC did), a six-to-eight-week difference in start point that likely makes their reported PFS numbers actually shorter by comparison.
Perioperative Chemo-Immunotherapy Is Genuinely Changing the Surgical Conversation
Since CheckMate 816 (2021), a series of trials has shown genuine benefit from neoadjuvant and perioperative chemo-immunotherapy: impressive two-year event-free survival, high rates of R0 resection, and, notably, a five-year overall survival of 95% in the CheckMate 816 update specifically for patients with a complete pathologic response, a genuinely remarkable number for locally advanced disease that's hard to compete with using any other current approach.
Resectability Is Genuinely Hard to Define, and That's Not New
A 1966 paper on resectability criteria noted that defining resectability is particularly difficult given the rapid pace of progress in thoracic surgery, and if anything, today's expanding treatment options have made this harder, not easier, to pin down. It's worth distinguishing resectability from operability: operability refers to patient-specific factors (performance status, comorbidities, lung and heart function) that determine whether someone should go to the operating room at all, while resectability refers to whether the tumor itself can technically be removed with a clear (R0) margin, generally not possible with aortic or esophageal invasion, or bulky N2/N3 nodal disease. Recent consensus efforts (an EORTC document published earlier this year, and separate expert consensus document from the Society of Thoracic Surgeons) both show a large "potentially resectable" gray zone, underscoring that even expert consensus doesn't fully resolve this question.
Converting "Unresectable" Into "Resectable"
A retrospective series from several high-volume centers (including Dana-Farber and Memorial Sloan Kettering), looking at patients with clinical T4 or bulky N2/N3 disease, found a 75% rate of proceeding to surgical resection after induction treatment, with meaningful downstaging and favorable event-free survival in those who went on to surgery, despite the retrospective design. A separate phase 2 trial from China similarly restaged patients initially deemed unresectable after 2-4 cycles of neoadjuvant chemo-immunotherapy and found a 77% surgical conversion rate, driven by high pathologic response rates.
LungMate-013, presented at ESMO, went a step further: 100 patients with unresectable stage IIIB/IIIC NSCLC received induction chemo-immunotherapy, were restaged, and those deemed resectable were then randomized to surgery or radiation. Of the 57 patients who were resectable following induction, 50 patients were then randomized to receive local therapy. All patients randomized to surgery achieved R0 resection, with event-free survival favoring surgery over radiation in this converted, resectable population. The ongoing MDT-bridge trial is testing a similar concept prospectively: patients with stage IIB-IIIB disease (without EGFR or ALK mutations) receive induction chemo-durvalumab, are restaged, and either continue to surgery (with additional chemo-durvalumab first) or move to chemoradiation followed by durvalumab if deemed unresectable. Preliminary results show an 86% overall resection rate in this cohort, with most achieving R0 resection, suggesting that starting with chemo-durvalumab and reassessing via multidisciplinary review can move more patients toward curative-intent local therapy overall, whichever specific local therapy that ends up being.
For Patients
For stage III lung cancer, deciding between surgery and radiation is a genuinely complex decision that depends on multiple factors, including whether the tumor can technically be safely and completely removed, which isn't always straightforward to determine even for experienced surgeons. Chemotherapy combined with immunotherapy given before any local treatment (surgery or radiation) is increasingly able to shrink tumors enough that some patients initially thought to be inoperable may become candidates for surgery after all. Given how complex and evolving this decision-making process is, getting an opinion from a multidisciplinary team (involving medical oncology, radiation oncology, and thoracic surgery together, rather than any one specialist alone) is genuinely valuable for this stage of disease. Ask your care team whether your case has been reviewed by a full multidisciplinary team, and whether starting with chemotherapy and immunotherapy before deciding on surgery or radiation might be appropriate for your specific situation.
Key Takeaways
-
The PACIFIC trial remains foundational for unresectable stage 3 NSCLC, but subsequent trials adding novel agents or giving immunotherapy concurrently with chemoradiation have not clearly improved on it.
-
Perioperative chemo-immunotherapy (since CheckMate 816) has produced genuinely striking outcomes in resectable disease, including 95% five-year overall survival in patients with a complete pathologic response.
-
Resectability and operability are distinct concepts; resectability (whether a tumor can be technically removed with a clear margin) remains genuinely difficult to define consistently, even among expert consensus groups.
-
Induction chemo-immunotherapy is converting a meaningful proportion (roughly 75-86% in several studies) of previously unresectable or borderline patients into successful surgical candidates.
-
Multidisciplinary reassessment after induction systemic therapy, rather than deciding upfront between surgery and radiation, may be an increasingly supported approach for selecting the optimal local therapy for stage 3 disease.
References
-
Antonia SJ, et al. Durvalumab after chemoradiotherapy in stage 3 NSCLC (PACIFIC). N Engl J Med. 2017.
-
Forde PM, et al. Neoadjuvant nivolumab plus chemotherapy in resectable NSCLC (CheckMate 816), 5-year update.
-
Taylor MD, et al. Surgical conversion after induction therapy in initially unresectable stage 3 NSCLC: a multi-institutional retrospective series.
-
LungMate-013: induction chemo-immunotherapy, restaging, and randomization to surgery versus radiation in converted-resectable stage 3B/3C NSCLC. Presented at ESMO.