Abstract

Following curative surgery for non-small cell lung cancer, around half of patients ultimately succumb to recurrent/metastatic disease. This may occur in many ways, such as recurrence at the surgical site, within the pleura, or by metastatic spread via lymphatics or blood vessels. This study aims to identify tumour-intrinsic pathological risk factors which predict particular modes of recurrence of lung adenocarcinoma, to improve our understanding of both the clinical course of disease and the biology underlying tumour recurrence. A single-centre cohort of 1025 patients who had surgically resected primary lung adenocarcinoma between the years of 1998 and 2015 was used. Full clinicopathological data were collected from patient records and slide review. All available radiology image data were examined to identify first recurrence site(s) for each case; these were grouped as surgical site, ipsilateral lung, ipsilateral pleura, nodal recurrence and distant recurrence. 824 cases with complete recurrence data were included in the final analysis. Univariate and multivariate Cox models of time to recurrence after surgery for each recurrence site were constructed. Depending on recurrence site, models included growth pattern (predominance or presence), tumour size, nodal involvement, vascular invasion, pleural involvement, surgical procedure and R stage. 408 patients (49.5%) had recurrent disease, with 200 (24.2%) patients showing simultaneous recurrence at multiple sites. Of the patients who developed recurrences, 99 (24.3%) recurred at the surgical site. 118 (28.9%) patients recurred within the ipsilateral lung and 75 (18.4%) within the ipsilateral pleura. There were 168 instances of nodal recurrence, most often within N2 lymph nodes (n=95, 56.5%). 316 cases of distant recurrence were reported, with the most common sites being the contralateral lung (n=110, 34.8%), bone (n=71, 22.5%) and brain (n=67, 21.2%). Distant recurrence is associated with significantly shorter post-recurrence survival (PRS) than locoregional recurrence (50% PRS 7.9 vs. 11.1 months, P=0.008). Surgical site recurrence is related to subanatomical wedge resection (HR=4.98, 95% CI: 2.45-10.12, P<0.001), tumour size (HR=1.01, 95% CI: 1.00-1.03, P=0.042) and positive resection margins (HR=4.05, 95% CI: 2.01-8.15, P<0.001). Ipsilateral pleural recurrence is strongly predicted by advancing PL stage (PL2 vs. PL0, HR=2.61, 95% CI: 1.18-5.77, P=0.018; PL3 vs. PL0, HR=3.74, 95% CI: 1.64-8.51, P=0.002). Notably, recurrence in the ipsilateral lung (HR=2.08, 95% CI: 1.29-3.35, P=0.003) and nodal recurrence (HR=2.11, 95% CI: 1.29-3.46, P=0.003) are both independently predicted by the presence of any micropapillary component of tumour growth. In contrast, distant recurrence is related to solid pattern predominance (HR=2.50, 95% CI: 1.43-4.35, P=0.001) univariately, and with near independence in multivariate analysis (HR=1.79, 95% CI: 0.99-3.24, P=0.054). In multivariate models, only solid-predominant tumours are significantly associated with recurrence in the brain (HR=8.52, 95% CI: 1.13-64.01, P=0.037). We present evidence that the two high-risk patterns of tumour growth, solid and micropapillary, have predilections for recurrence/metastasis via haematogenous and lymphatic vascular systems respectively. This distinction raises important biological questions about mechanisms of tumour spread, and may help to inform future treatment and prevention strategies.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call