Local/regional recurrence following surgery for early-stage lung cancer: A 10-year experience with 975 patients

Abstract

7542 Background: Surgery, with or without chemotherapy, is the preferred treatment for early-stage NSCLC. The risk of local/regional recurrence (LRR) after resection is not well defined, but generally considered to be small in comparison to the risk of distant recurrence. We herein evaluate the actuarial risk of LRR after surgery for stage I-II disease and assess surgical and pathologic factors affecting this risk. Methods: The medical records and pertinent radiographs of all patients who underwent surgery for pathological T1- 2N0–1 NSCLC at Duke between 1995–2005 were reviewed. Patients receiving preoperative therapy, or with synchronous/prior lung cancers were excluded. Disease recurrence at the surgical margin, ipsilateral hilum, and/or mediastinum was considered a LRR. Nodal failures were defined as a new or enlarging LN≥1 cm on short axis on CT or hypermetabolic on PET. The actuarial rates of LRR and distant recurrence (assessed independently) were estimated using the Kaplan-Meier method. A univariate and multivariate regression analysis assessed factors associated with LRR. Results: Of 975 patients, 45% were stage IA, 40% IB, and 15% stage II; 79% underwent ≥ lobectomy; 85% had sampling/dissection of mediastinal LNs; 4% had positive margins; 7% received adjuvant chemotherapy; 3% received adjuvant radiation therapy. The 5-year actuarial rate of LRR was 22% and distant recurrence was 31%. LRR was higher for stage IB, IIA, and IIB compared with IA (hazard ratio- 2, 2.6, and 1.9). Sublobar resections were associated with a higher risk of LRR on univariate, but not multivariate, analysis. On multivariate analysis, increasing size, squamous or large cell histology, visceral pleural invasion, stage IIA disease, and lack of nodal sampling were independently associated with a higher rate of LRR. Conclusions: The risk of LRR after surgery for stage I-II NSCLC is ≈22%, and is similar to the risk of distant recurrence. Thus, local control remains an important issue justifying continued clinical research in this area. Postoperative RT, utilizing modest doses and small conformal fields directed to sites most at risk, should be reevaluated. As systemic therapy improves, the relative importance of achieving local control will also increase. No significant financial relationships to disclose.