Date of Award
Open Access Thesis
Medical Doctor (MD)
Anthony W. Kim
Very large primary non-small cell lung cancers (NSCLC), defined as those >7 cm, remain a therapeutic challenge due to known survival disadvantage compared to smaller tumors and lack of specific studies in this population. This study compares the effect of various treatment modalities on survival of patients with large NSCLC with none or positive hilar lymph node involvement (T3>7cmN0 and T3>7cmN1, respectively).
The Surveillance, Epidemiology, and End Results (SEER) database was used to identify patients undergoing a lobectomy or pneumonectomy for T3>7cmN0 NSCLC from 1999 to 2008. Patients were categorized into groups based on type of surgery performed and whether neoadjuvant radiation therapy (NRT) was used. The National Cancer Data Base (NCDB) was used to identify adult patients who were diagnosed with T3>7cmN1 NSCLC from 1999-2005. Nonsurgical treatments included chemoradiation, chemotherapy, radiation therapy, or no treatment whereas primary surgical treatments included surgery only, chemoradiation or chemotherapy prior to surgery (CxR-S or C-S, respectively), chemoradiation or chemotherapy after surgery (S-CxR or S-C, respectively), or postoperative radiation therapy (S-PORT). Five-year overall (OS) and lung cancer specific survival (LCSS) were estimated by the Kaplan-Meier method and comparisons made using log-rank tests and Cox regression models.
A total of 1,301 surgical patients with T3>7cmN0 NSCLC were evaluated using the SEER database, including 1,232 undergoing primary surgical therapy (PST) and 69 receiving NRT. NRT was not associated with improvements in 5-year OS (48% vs. 41%, P=.06) or LCSS (59% vs. 52%, P=.12) compared to PST. Lobectomies were associated with better 5-year OS (43% vs. 33%; P=.006) and LCSS (54% vs. 43%, P=.005) compared to pneumonectomies. On multivariate analysis, NRT did not produce any significant advantage in OS (P=.24) and LCSS (P=.21). Using the NCDB, a total of 642 patients with T3>7cmN1 NSCLC were evaluated: 425 nonsurgical and 217 primary surgical treatments. Primary surgical treatments were associated with an improved 5-year OS of 28% compared to 8% and 4% for primary nonsurgical treatments and no treatments, respectively (P<.001). Specific nonsurgical treatment 5-year OS were 11%, 5%, 2%, 4% for chemoradiation, chemotherapy, radiation therapy, and no treatment, respectively (P<.001). Primary surgical treatment 5-year OS were 16%, 44%, 40%, 40%, 38%, and 18% for surgery only, CxR-S, C-S, S-CxR, S-C, and S-PORT, respectively (P<.001). On multivariate analysis, surgery and chemotherapy in most combinations were associated with significantly improved OS compared to chemoradiation only (C-S hazard ratio (HR), 0.4 [95% confidence interval, 0.18-0.88], P=.02; CxR-S HR, 0.41 [0.19-0.9], P=.03; S-C HR, 0.4 [0.19-0.85], P=.02).
Our results demonstrate that neoadjuvant radiation therapy, which most likely was a combination of chemotherapy and radiation, was not associated with improvements in OS or LCSS compared to primary surgical therapy for patients with T3>7cmN0 NSCLC. When feasible, lobectomy appears more beneficial than pneumonectomy in terms of long-term survival. For patients with T3>7cmN1 NSCLC, surgery with systemic therapy delivered in a neoadjuvant or adjuvant fashion is associated with improvements in long-term overall survival. Finally, when surgical resection is not feasible, definitive chemoradiation therapy should be considered as an equal alternative to surgical resection alone.
Moreno, Amy, "Treating Very Large Non-Small Cell Lung Cancers: A Survival Analysis Using National Cancer Databases" (2014). Yale Medicine Thesis Digital Library. 1908.