The surgical management of lung cancer has undergone tremendous evolution since the first case series of pulmonary resection for lung cancer was reported by Oschner and DeBakey in 1939.¹ Since that time, the extent of surgical resection for lung cancer has been a point of heated debate among thoracic surgeons. In the early 20th century, most surgeons considered pneumonectomy the treatment of choice for lung cancer despite its associated high operative mortality.¹ Later, noncomparative retrospective case series showed that lobar resection can result in equivalent survival to that achieved by pneumonectomy with lower morbidity and mortality.² Anatomic segmentectomy was first described by Churchill and Belsey³ in 1939 for resection of bronchiectasis and tuberculosis. Anatomic segmentectomy for resection of lung cancer was performed by some thoracic surgeons.⁴,⁵,⁶,⁷ However, the relative complexity of the operation and the perceived higher risk of local recurrence dampened the enthusiasm of most surgeons.⁸ In 1995, the Lung Cancer Study Group (LCSG) reported a randomized trial of 247 patients with clinical T1N0 (cT1N0) peripheral non-small cell lung cancer (NSCLC) who were randomly assigned to either limited resection (anatomical segmentectomy or wedge resection) or lobectomy. ⁹ Absence of nodal metastases to the draining segmental, lobar, hilar, and mediastinal lymph nodes was confirmed by frozen section examination prior to randomization. The study showed that locoregional recurrence was significantly higher in patients treated by sublobar resections. Locoregional recurrence rates per person-year were 0.022 following lobectomy, 0.044 following segmental resection, and 0.086 after wedge resection. Although patients treated by lobectomy had a higher 5-year survival compared with those treated by limited resection, the difference in survival did not achieve statistical significance (73% vs. 56%, p = 0.06).⁹ This trial established lobectomy as the surgical standard of care for early stage disease. Limited resection (segmentectomy or wedge) has since been largely reserved for patients with compromised cardiopulmonary function who could not tolerate a lobectomy.

Recently, there has been renewed enthusiasm for the use of limited resection as a result of increased detection of smaller tumors by more widespread use of computed tomography (CT). Evidence derived mostly from retrospective case series and a few prospective studies suggested that sublobar resection was feasible and possibly associated with 5-year survival comparable to that attained by lobectomy in patients with stage I peripheral NSCLC measuring 2 cm or less in size. Ongoing prospective randomized clinical trials are currently underway to determine the relative merits of sublobar resection compared with lobectomy in this subset of patients. In this chapter, we will review the results reported after limited resection done either as a compromise procedure in patients with suboptimal cardiopulmonary reserve or intentionally in patients considered suitable candidates for lobar resection. We will also discuss some of the factors one needs to consider in selecting patients for limited resection including tumor size, location, intralobar satellites, cell type, nodal metastases, and surgical margins. Methods such as brachytherapy to extend the effective surgical margin will also be discussed.

Several studies have examined the outcome of limited resection as a compromise procedure in patients with limited cardiopulmonary reserve. For example, Landreneau et al.¹⁰ analyzed the results of 219 consecutive patients with pathologic T1N0 NSCLC who underwent wedge resection or lobectomy. The wedge resection group of patients was significantly older and had reduced pulmonary function. At 5 years, overall survival (OS) was 58% following treatment by open wedge resection, 65% for patients treated by video-assisted wedge resection, and 70% for those treated by lobectomy. The difference in survival between the lobectomy group and the entire wedge resection group approached but did not attain statistical significance (p = 0.56). The difference in survival at 5 years was a result of
a significantly greater frequency of non-cancer-related deaths among patients treated by open wedge resection (38% vs. 18% for the lobectomy group; p = 0.014). Furthermore, there was no significant difference in the local/systemic recurrence rates between patients treated by wedge resection and those treated by lobectomy. The authors concluded that wedge resection was a viable surgical option for patients with stage I NSCLC who have impaired cardiopulmonary function.

Keenan et al.¹¹ retrospectively reviewed patients with stage I NSCLC who had either lobectomy (n = 147) or anatomical segmentectomy (n = 54). The segmentectomy group of patients had a significantly greater degree of preoperative pulmonary impairment compared with the lobectomy group. The authors reported no statistically significant difference in overall and disease-free survival (DFS) between the two groups of patients. The 4-year OS was 67% for lobectomy and 62% for segmentectomy (p = 0.86). The authors also compared pulmonary function between the two groups 1 year postoperatively. Patients treated by segmentectomy had preservation of forced vital capacity, forced expiratory volume in 1 second, and maximum voluntary ventilation postoperatively. In contrast, patients treated by lobectomy experienced a significant decline in all pulmonary function parameters.

In 2006, El-Sherif et al.¹² reported the results of sublobar resection for stage I NSCLC in 207 patients with cardiopulmonary impairment and compared them with the results following lobectomy in 577 patients with similar stage disease. For stage IA patients, both sublobar resection and lobectomy resulted in identical DFS of 65% at 7 years (p = 0.308). For stage IB patients, there was a significantly worse DFS after sublobar resection compared with lobectomy, 50% versus 58% (p = 0.009). The authors concluded that sublobar resection seemed appropriate for stage IA patients.

Most, if not all, of the evidence supporting the use of limited resection in early stage NSCLC is derived from case series reported by Japanese surgeons (Table 32.1). For example, Kodama et al.¹³ reported a case-control series of 46 patients with cT1N0 NSCLC treated by intentional limited resection with curative intent. All patients had peripheral tumors less than 2 cm in size, and all were treated by segmentectomy with regional lymph node dissection. The comparator or control group comprised 77 patients with stage I treated by lobectomy. The segmentectomy group had a 5-year survival of 93%, which was similar to survival in the control group. Locoregional recurrence was 2.2% in the segmentectomy group and 1.3% in the lobectomy group.

Similarly, Okada et al.¹⁴ reviewed their prospective experience with 70 patients who had T1N0 NSCLC 2 cm or less treated by intentional limited resection. The authors performed what they termed an extended segmentectomy, which is essentially a segmental resection where parenchymal division extends slightly beyond the anatomical segmental boundary. In this study, segmentectomy was performed only after frozen section examination of the segmental, hilar, and mediastinal nodes confirmed the absence of metastatic disease. Patients with nodal disease were treated by lobar resection. Five-year survival of pathologic T1N0 patients was 87.1% in the extended segmentectomy group compared with 87.7% in the lobectomy group (p = 0.8). There were no local recurrences after limited resection. The local recurrence rate
for the lobectomy group was not reported in that study. The authors concluded that extended segmentectomy is an acceptable alternative to lobectomy for T1N0 NSCLC 2 cm or less.

Yoshikawa et al.¹⁵ published their multi-institutional prospective Japanese trial of limited resection for peripheral NSCLC less than 2 cm. A total of 55 patients were enrolled. Intraoperatively, an extended segmentectomy with frozen section of hilar and mediastinal lymph node was performed. The authors reported a 5-year OS of 81.8% and lung cancer-specific survival of 91.8%.

Koike et al.¹⁶ reported their results of intentional limited resection for peripheral T1N0 NSCLC 2 cm or less. The limited resection group consisted of 60 segmentectomies and 14 wedge resections, which was compared with 159 patients treated by lobectomy. Intraoperative frozen analysis of hilar and mediastinal lymph nodes was performed, and a standard lobectomy was done when lymph node metastasis were detected. The 3- and 5-year OS was 94.0% and 89.1% in the limited resection group compared with 97.0% and 90.1% in the lobectomy group. Tumor recurrence was noted in five patients after limited resection and in nine patients after lobectomy. Both the OS and the local recurrence rate were not significantly different between the two groups.

Watanabe et al.¹⁷ reported their results of intentional limited resection for T1N0 peripheral NSCLC 2 cm or less in size. Limited resection was done in 34 patients, wedge resection in 14 patients, and extended segmentectomy in 20 patients. Again, intraoperative examination of hilar and mediastinal lymph nodes showed no evidence of nodal metastases. Impressively, the 5-year survival after extended segmentectomy was 93% with no local recurrences. This compared favorably with 5-year survival after lobectomy patients, which was 84%.

Collectively, these studies suggest equivalence between anatomical segmentectomy (the data are less robust for wedge resection) and lobectomy in a select group of patients with peripheral small (≤2 cm) NSCLC who are meticulously staged to rule out the presence of segmental, hilar, and mediastinal nodal metastases. Nonetheless, the data are largely derived from retrospective or prospective noncomparative studies that are inherently limited by unavoidable selection biases. Furthermore, one cannot reasonably exclude the possibility of a significant impact of ethnicity on biological tumor characteristics and survival.

Tumor Size Tumor size is a critical factor in determining the feasibility and safety of limited resection. It is well established that tumor size is an important prognostic factor for survival in NSCLC.¹⁸,¹⁹,²⁰,²¹,²²,²³ The current staging system recognizes a difference in survival between tumors >3 cm and those ≤3 cm. This is supported by several retrospective studies that showed a survival advantage for T1 tumors compared with T2.²⁴,²⁵,²⁶,²⁷,²⁸ Even within T1 tumors, several retrospective studies have shown that survival is better with smaller tumor size. Gajra et al.²⁹ reported their experience with 246 patients surgically resected for stage IA NSCLC. Patients with tumors ≤1.5 cm had a significantly improved DFS and OS compared with those patients with tumors 1.6 to 3.0 cm. Tumor size was an independent prognostic factor for survival in their multivariate analysis. Similarly, Port et al.¹⁸ reviewed 244 patients with resected stage IA NSCLC. The 5-year OS for patients with tumors ≤2 cm was 77.2% compared with 60.3% in patients with tumor >2 cm in size, again supporting the relationship between tumor size and survival. For subcentimeter stage IA tumors, Lee et al.¹⁹ reported 5- and 10-year OS of 94% and 75%, respectively, with the disease-specific survival was 100% at both time points without any recurrences. Other investigators have also reported excellent survival following resection of subcentimeter tumors.³⁰,³¹ The aforementioned studies along with many others have led to the new proposed staging system, which divides T1 tumors into T1a and T2b at the 2-cm cutoff.³²,³³,³⁴,³⁵,³⁶ Given the preceding evidence of tumor sizes, sublobar resection may be more appropriate for the T1a tumors ≤2 cm in size.