Abstract
Objective:
To investigate disease free survival (DFS) of sublobar resection vs. lobectomy for stage IA non-small cell lung cancer (NSCLC) with preoperative high-risk features.
Methods:
Data were abstracted from a prospective database to identify patients with clinical T1a-T1bN0M0 NSCLC (≤2 cm) who underwent lobectomy or sublobar resection (wedge resection or segmentectomy). 1:1 propensity matching was used to balance the dataset for FEV1 ≥60% and high-risk features: cT1b vs. cT1a, SUV of the primary tumor on PET, solid vs. subsolid tumor texture on CT, and micropapillary (MIP)/solid histology. The primary outcome was DFS.
Results:
825 patients met inclusion criteria: 52% (N=426) patients underwent lobectomy and 48% (N=399) of patients underwent sublobar resection (45% segmentectomy, 55% wedge resection). Lobectomy was associated with more preoperative high-risk features: cT1b (p<0.001), higher SUV (p<0.001), solid tumor texture on CT (p<0.001), and MIP/solid histology (p < 0.001). 660 patients were included in the matched analysis with all high-risk features balanced. Nodal upstaging (N1) was higher in lobectomy patients (9.1% vs. 3.4%, p=0.004). 5-year DFS (85% vs. 74%, p=0.12) was equivalent in the matched cohort. Lobectomy was protective for recurrence in the presence of two or greater high-risk features: sublobar resection patients with 2 high-risk features (HR 1.77, 95% CI 1.13–2.76, p=0.012) or 3–4 high-risk features (HR 1.97, 95% CI 1.25–3.10, p=0.004) had worse DFS.
Conclusion:
Lobectomy should be considered over sublobar resection for stage IA NSCLC ≤2 cm in the presence of multiple high-risk features.
Keywords: Sublobar resection, non-small cell lung cancer, high-risk
Central Picture:
Compared to lobectomy, sublobar resection with 2 or more high-risk features have worse DFS.
Introduction
The recent randomized CALGB1405031 and JCOG08022 studies demonstrated that for peripheral ≤ 2 cm non-small cell lung cancers (NSCLC) without nodal metastases, sublobar resection is non-inferior to lobectomy with regards to oncologic outcomes. Despite the finding of equivalent overall and disease free survival in both trials, locoregional recurrence was significantly higher in the segmentectomy group in JCOG0802 (11% vs 5%, p = 0.0018); a similar but nonsignificant trend was seen in the North American CALGB140503 study2. This is in keeping with historical data from the Lung Cancer Study Group in which local recurrence was significantly higher in patients undergoing less than lobectomy3. Although overall survival is favorable for early NSCLC, the main driver of disease-specific mortality in these patients is secondary to relapse. Most recurrences occur at distant sites and are associated with 30% 5-year survival4,5. Curative intent resection aims to minimize the risk of long-term recurrence, with negative margins and complete oncologic resection of the primary tumor and draining lymph nodes.
In early-stage NSCLC, there are well-described high-risk features that are associated with higher recurrence rate and worse overall survival (OS), thought to be related to inherent more aggressive biology of these tumors. Many of these features can be identified preoperatively and potentially guide appropriate surgical resection. Specifically, patients with cT1a tumors have significantly improved OS compared to cT1b tumors, and radiographically solid lesions have shorter recurrence-free survival than pure ground glass opacity (GGO) nodules6–8. Additionally, high standardized uptake value (SUV) on positron emission tomography (PET) of the primary tumor, specifically ≥ 2.6, is associated with an increased risk of recurrence among stage I lung adenocarcinoma patients9. Finally, micropapillary (MIP) and solid histologic subtypes are associated with worse disease-specific survival10 and, specifically among patients with stage I NSCLC, MIP/solid histologic subtypes are independently associated with increased risk of recurrence11,12.
It is not currently known whether extent of resection impacts recurrence in the subset of patients with these high-risk features in early-stage NSCLC. We hypothesize that there is a subset of patients with stage IA NSCLC and high-risk features that, although otherwise eligible for sublobar resection, may have improved disease-free survival (DFS) with lobectomy.
Methods
Study Design and Participants
We retrospectively reviewed our prospectively maintained institutional database to identify patients with clinical stage IA1–2 (cT1a-b N0 M0, ≤2 cm) NSCLC (American Joint Committee on Cancer, AJCC, eighth edition) who underwent lobectomy or sublobar resection (wedge resection or segmentectomy) between 2010 to 2022 (IRB#18–391, 09/07/2018). Medical records were systematically reviewed to identify patient demographic and clinicopathologic characteristics as well as perioperative outcomes. The primary endpoint was DFS, defined as time to recurrence after complete resection. Recurrence was defined by radiologic evidence of recurrent disease, with tissue confirmation performed when feasible.
Routine staging workup was performed for all patients, comprising CT of the chest (with high resolution contrast-enhanced imaging when available) as well as PET/CT imaging to evaluate extent of disease.13 Additional imaging including magnetic resonance imaging (MRI) of the brain was performed when clinically appropriate. Preoperative biopsy of the primary tumor was not required for inclusion but was performed when clinically indicated. If indicated, invasive mediastinal staging was performed using endobronchial ultrasound or mediastinoscopy to evaluate for lymph node metastases14. Patients who were known to have nodal or distant disease preoperatively were excluded, as were patients without CT or PET images available for review. No patients received induction therapy.
All patients underwent curative intent resection via either lobectomy, segmentectomy, or non-anatomic wedge resection. Prior to resection, routine cardiopulmonary assessment with pulmonary function testing (PFT) was performed to determine operability. Patients without PFT data were excluded. Any sublobar anatomic resection with division of a segmental pulmonary artery, vein, and bronchus was considered a segmentectomy for the purposes of this study. This included either an individual segment (e.g. S6, or superior segmentectomy) or composite segmentectomy of 2 or more adjacent segments (e.g. S1–2+S3, or trisegmentectomy). Choice of operation as well as surgical approach was at the discretion of the operating surgeon. A formal mediastinal and hilar lymph node dissection or sampling was performed for all patients per routine protocol. Patients with occult nodal disease discovered on final pathology were included in the analysis.
Surgical specimens underwent routine pathologic evaluation with reporting of TNM pathologic stage by trained thoracic pathologists as part of routine clinical care. For patients with adenocarcinoma, histologic subtype and relative percentage was reported15.
For patients who were found to have nodal disease at the time of surgery, adjuvant therapy was administered on the basis of multidisciplinary discussion. Surveillance was performed with CT of the chest every 6 months for the first two years postoperatively and yearly thereafter13.
High-Risk Features
Four previously validated high-risk features which have been shown to predict recurrence in early-stage NSCLC were selected for matching: clinical tumor size (cT1b vs cT1a)6, radiographic tumor density7, SUV of the primary tumor per PET9, and presence of MIP or solid histologic subtype on final pathology10. Tumor size was determined by highest quality, most recent CT scan of the chest and categorized according to AJCC 8th edition (i.e., cT1a ≤1cm and cT1b if 1–2cm). Tumor density was determined on review of the highest quality, most recent CT chest by members of the study team. Tumors were categorized as subsolid in the presence of any ground glass opacity (GGO) component or as pure solid if no GGO component. Maximum SUV of the primary tumor on preoperative PET/CT imaging was utilized as a continuous variable in the matching procedure. Patients with tumors consisting of at least 5% MIP or solid histologic subtypes were categorized as having MIP/solid histology.
Statistical Analysis
Patient characteristics were summarized using frequency (percentages) for categoric variables and median and interquartile range (IQR) for continuous variables. DFS was defined as the time from date from surgery to date of recurrence, death, or last follow-up. OS was defined as the time from date of surgery to date of death or last follow-up. Patients were censored at date of last follow-up. DFS and OS were estimated using the Kaplan-Meier approach and log-rank tests were utilized to compare survival curves. Univariable cox regression models were utilized to determine if there are associations between number of high-risk features and DFS.
Propensity score matching was used to balance the cohorts who received a sublobar resection and those who received a lobectomy for FEV1 ≥60% as well as the following high-risk preoperative features: cT1b vs. cT1a, solid vs. subsolid tumor texture on CT, SUVmax, and presence of MIP/solid histology. The caliper was defined as 20% of the standard deviation of the propensity score. The standardized mean difference (SMD) was utilized to assess balance between surgery groups. A SMD ≤ 0.1 is conventionally considered to indicate a good balance between the two groups16. Comparison of OS and DFS between the two groups was conducted using the log-rank test and quantified using a Cox proportional hazards model. All statistical tests were 2-sided and p < 0.05 was considered to indicate statistical significance. Statistical analyses were conducted using R (version 4.4.1, R Core Development Team, Vienna, Austria).
Data were reviewed after approval from the Memorial Sloan Kettering Institutional Review Board (IRB#18–391 (09/07/2018)), which waived the need for patient consent.
Results
Unmatched Cohort
A total of 825 patients met inclusion criteria. Table 1 summarizes the demographic and clinical characteristics by surgical procedure. The majority of patients were female (65%) of white non-Hispanic race (80%) and most were former smokers (67%). A total of 494 patients (60%) underwent a preoperative biopsy. The majority of tumors were adenocarcinoma (85%). A minimally invasive approach was utilized in 90% of patients (56% video-assisted thoracoscopy, VATS; 34% robot-assisted). Lobectomy was performed in 52% of patients (N=426) and 48% (N=399) of patients underwent sublobar resection (45% segmentectomy, 55% wedge resection). Supplementary Table 1 provides a detailed description of the resected segments.
Table 1.
Demographic and clinical characteristics according to extent of resection in the unmatched and matched cohorts.
| Unmatched Cohort | Matched Cohort | |||||||
|---|---|---|---|---|---|---|---|---|
| Characteristic | Overall (n = 825) |
Lobectomy (n = 426) |
Sublobar Resection (n = 399) |
P value | Lobectomy (n = 330) |
Sublobar Resection (n = 330) |
SMD* | 95% CI |
| Female Sex, N (%) | 536 (65) | 279 (65) | 257 (64) | 0.7 | 227 (69) | 212 (64) | 0.10 | −0.06, 0.25 |
| White Non-Hispanic Race, N (%) | 631 (80) | 335 (82) | 296 (78) | 0.6 | 258 (82) | 249 (79) | 0.21 | 0.05, 0.36 |
| ASA Score, N (%) | 0.8 | 0.10 | −0.06, 0.25 | |||||
| II | 144 (17) | 74 (17) | 70 (18) | 58 (18) | 52 (16) | |||
| III | 644 (78) | 335 (79) | 309 (77) | 260 (79) | 260 (79) | |||
| IV | 37 (4.5) | 17 (4.0) | 20 (5.0) | 12 (3.6) | 18 (5.5) | |||
| Former Smoker, N (%) | 556 (67) | 291 (68) | 265 (66) | 0.8 | 226 (68) | 225 (68) | 0.06 | −0.09, 0.22 |
| FEV1 (%), Median (Q1, Q3) | 94 (80, 106) | 93 (79, 106) | 95 (81, 107) | 0.4 | 93 (79, 107) | 94 (81, 108) | 0.06 | −0.21, 0.10 |
| DLCO (%), Median (Q1, Q3) (N=813) | 85 (71, 98) | 85 (72, 98) | 84 (71, 98) | 0.7 | 84 (70, 98) | 83 (71, 97) | 0.08 | −0.07, 0.23 |
| CT Tumor Texture, N (%) | <0.001 | 0.04 | −0.11, 0.19 | |||||
| Subsolid | 502 (61) | 221 (52) | 281 (70) | 207 (63) | 213 (65) | |||
| Solid | 323 (39) | 205 (48) | 118 (30) | 123 (37) | 117 (35) | |||
| PET SUVmax, Median (Q1, Q3) | 2.2 (1.2, 4.2) | 2.7 (1.5, 5.1) | 1.8 (1.0, 3.4) | <0.001 | 2.5 (1.3, 4.5) | 2.1 (1.3, 3.7) | 0.09 | −0.07, 0.24 |
| Adenocarcinoma Histology, N (%) | 423 (85) | 234 (84) | 189 (88) | 0.4 | 180 (85) | 152 (86) | 0.03 | −0.17, 0.23 |
| MIP/Solid Histology, N (%) | 369 (45) | 220 (52) | 149 (37) | <0.001 | 150 (45) | 145 (44) | 0.03 | −0.12, 0.18 |
| cT Stage, N (%) | <0.001 | 0.01 | −0.14, 0.16 | |||||
| 1a | 120 (15) | 44 (10) | 76 (19) | 43 (13) | 42 (13) | |||
| 1b | 705 (85) | 382 (90) | 323 (81) | 287 (87) | 288 (87) | |||
SMD ≤0.1 is considered balanced between the 2 groups.
(Abbreviations: SMD, standardized mean difference; CI, confidence interval; CT, computed tomography; PET, positron emission tomography; SUV, standard uptake value; MIP, micropapillary)
Patients that underwent lobectomy, compared to a sublobar resection, were more likely to have solid tumor texture on preoperative CT imaging (48% vs. 30%, p < 0.001), higher rate of cT1b (90% vs. 81%, p < 0.001), higher median SUVmax (2.7 vs. 1.8, p < 0.001). Table 2 summarizes the surgical and pathologic characteristics by surgical procedure. Lobectomy patients, compared to sublobar resection patients, had a higher rate of MIP/solid histology (52% vs. 37%, p < 0.001). Similarly, upon pathologic review of the resected specimens, the lobectomy cohort, compared to the sublobar resection group, had more aggressive features including higher median tumor size (1.5 cm vs. 1.2 cm, p < 0.001), higher rate of lymphovascular invasion (LVI; 30% vs. 18%, p < 0.001), higher rate of visceral pleural invasion (VPI; 3.1% vs. 1.3%, p = 0.045), and a higher rate of spread through air spaces (STAS; 54% vs. 38%, p < 0.001).
Table 2.
Surgical and pathologic characteristics according to extent of resection in the unmatched and matched cohorts.
| Unmatched Cohort | Matched Cohort | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Characteristic | Overall (n = 825) |
Lobectomy (n = 426) |
Sublobar Resection (n = 399) |
P value | Lobectomy (n = 330) |
Sublobar Resection (n = 330) |
SMD* | 95% CI | P value |
| Minimally Invasive Approach, N (%) | 742 (89) | 391 (92) | 351 (88) | 0.2 | 306 (93) | 290 (88) | 0.17 | 0.01, 0.32 | 0.11 |
| Procedure, N (%) | - | - | - | - | |||||
| Lobectomy | 426 (52) | 426 (100) | 0 | 330 (100) | 0 | ||||
| Segmentectomy | 178 (22) | 0 | 178 (45) | 0 | 153 (46) | ||||
| Wedge | 221 (27) | 0 | 221 (55) | 0 | 177 (54) | ||||
| Segmentectomy Type, N (%) | - | - | - | - | |||||
| Composite | - | 0 | 103 (26) | - | 88 (27) | ||||
| Individual | - | 0 | 75 (19) | - | 65 (20) | ||||
| Closest Margin (cm), Median (Q1, Q3) (N=582) | 1.1 (0.5, 2.2) | 1.5 (0.4, 2.7) | 1.0 (0.5, 2) | 0.006 | 1.4 (0.4, 2.7) | 1.0 (0.5, 1.9) | 0.30 | 0.12, 0.49 | 0.045 |
| LVI, N (%) (N=817) | 199 (24) | 127 (30) | 72 (18) | <0.001 | 90 (28) | 68 (21) | 0.16 | 0.00, 0.31 | 0.044 |
| VPI, N (%) (N=819) | 18 (2.2) | 13 (3.1) | 5 (1.3) | 0.045 | 9 (2.8) | 5 (1.5) | 0.11 | −0.04, 0.27 | 0.4 |
| STAS Present, N (%) (N=573) | 265 (46) | 152 (54) | 113 (38) | <0.001 | 112 (51) | 103 (43) | 0.18 | −0.01, 0.36 | 0.5 |
| Tumor Size (cm), Median (Q1, Q3) (N=824) | 1.3 (0.96, 1.7) | 1.5 (1.1, 1.8) | 1.2 (0.8, 1.5) | <0.001 | 1.4 (1.1, 1.8) | 1.2 (0.8, 1.6) | 0.39 | 0.23, 0.54 | <0.001 |
| Occult Nodal Disease, N (%) | 67 (8.1) | 50 (12) | 17 (4.3) | <0.001 | 34 (10) | 17 (5.2) | 0.19 | 0.04, 0.35 | 0.013 |
| N1 Upstaging, N (%) | 53 (6.7) | 43 (10) | 10 (2.8) | <0.001 | 30 (9.1) | 10 (3.4) | 0.24 | 0.08, 0.39 | 0.004 |
| N2 Upstaging, N (%) | 26 (3.3) | 17 (4) | 9 (2.5) | 0.2 | 11 (3.3) | 9 (3.1) | 0.02 | −0.14, 0.17 | 0.8 |
| pStage, N (%) | <0.001 | 0.46 | 0.30, 0.61 | 0.023 | |||||
| I | 756 (91.6) | 377 (88.1) | 379 (95.5) | 295 (89.4) | 312 (94.5) | ||||
| II | 43 (5.2) | 34 (7.9) | 9 (2.3) | 24 (7.3) | 9 (2.7) | ||||
| III | 25 (3) | 16 (3.7) | 9 (2.3) | 10 (3) | 9 (2.7) | ||||
| IV | 1 (0.1) | 1 (0.2) | 0 | 1 (0.3) | 0 | ||||
SMD ≤0.1 is considered balanced between the 2 groups.
(Abbreviations: SMD, standardized mean difference; CI, confidence interval; LVI, lymphovascular invasion; VPI, visceral pleural invasion; STAS, spread through airspaces)
Postoperative complication rate was higher in the lobectomy group, compared to the sublobar resection group (30.3% vs. 17.2%, p = 0.001), however there were no differences in grade 3 or higher complications (2.6% vs. 1.1%, p = 0.08). There was one in-hospital death in the lobectomy group due to venous thromboembolism.
Median follow-up was 3.76 years (95% CI 3.39–4.06 years). Five-year DFS (lobectomy 83%, 95% CI 78%−87% vs. sublobar 79%, 95% CI 73%−84%, log-rank p = 0.5) and OS (lobectomy 89%, 95% CI 85%−93% vs. sublobar 85%, 95% CI 80%−90%, log-rank p = 0.3) did not differ between the unmatched cohorts. A total of 15 patients (3.8%) in the sublobar resection group and 49 patients (11.4%) in the lobectomy group underwent adjuvant therapy (p < 0.001).
Matched Cohort
In order to balance the groups based on high-risk characteristics and pulmonary function, patients underwent propensity score matching. 660 patients were included in the 1:1 matched analysis: 330 patients in the lobectomy group and 330 patients in the sublobar resection group. After matching, all preoperative high-risk features were balanced between groups (Supplementary Table 2). FEV1 was also matched in this cohort. Similarly, there was no difference in the rates of aggressive pathologic features including LVI, VPI, and STAS between matched cohorts (Table 2).
On final pathology, the lobectomy group had larger median closest margin (1.4 cm vs. 1.0 cm; p = 0.045). Occult nodal disease was higher in the lobectomy group compared to the sublobar resection group (10% vs. 5.2%; p = 0.013), with a higher rate of N1 upstaging (9.1% vs. 3.4%; p = 0.004). There was no difference in the incidence of N2 disease between groups (Table 2).
Overall, 9.4% of patients recurred in the matched cohort: 9.6% in the lobectomy and 9.1% in the sublobar resection groups. Distant recurrence was more common in the lobectomy group (13/330, 4%) vs. sublobar resection group (9/330, 2.7%), whereas locoregional recurrence was more common in sublobar resection (13/330, 4%) vs. lobectomy patients (10/330, 3%), although not statistically significant (p = 0.6). After propensity matching, 5-year DFS was numerically higher in the lobectomy group but not statistically significant (lobectomy 85%, 95% CI 80%−90% vs. sublobar resection 74%, 95% CI 68%−81%, log-rank p = 0.12). Similarly, five-year OS was numerically higher in the lobectomy group (lobectomy 91%, 95% CI 87%−95% vs. sublobar resection 82%, 95% CI 76%−88%, log-rank p = 0.053) but also did not differ statistically between matched groups.
Analysis of Disease-Free Survival Stratified by High-Risk Features
Next, to evaluate the impact of surgical resection and high-risk features on DFS, we stratified the matched lobectomy and sublobar resection cohorts by number of high-risk features. The majority of patients had 1 or more high-risk features (Supplementary Table 3). In the sublobar group, DFS was significantly decreased for each additional high-risk feature present (log-rank p < 0.001) (Figure 1A). In contrast, DFS in the lobectomy group was not affected based on the number of high-risk features present (log-rank p = 0.4) (Figure 1B). There were too few events to perform a similar analysis with OS.
Figure 1.
Disease-free survival stratified by number of high-risk features (HRF) in the matched sublobar resection cohort alone (A) and the matched lobectomy cohort alone (B; Confidence Intervals equal 95%).
DFS was compared between sublobar resection and lobectomy cohorts. In univariable analysis, for patients with zero or 1 high-risk feature undergoing a sublobar resection, there was no statistical difference in DFS between sublobar resection or lobectomy (Table 3). However, patients with two high-risk features (HR 1.77, 95% CI 1.13–2.76, p = 0.012) or 3–4 high-risk features (HR 1.97, 95% CI 1.25–3.10, p = 0.004) had improved DFS with lobectomy when compared to sublobar resection (Figure 2).
Table 3.
Disease-free survival hazard ratios stratified by number of high-risk features in the matched sublobar resection cohort.
| Event N | HR | 95% CI | P value | |
|---|---|---|---|---|
| Lobectomy | 61 | — | — | |
| 0 high-risk features | 1 | 0.56 | 0.08, 4.04 | 0.6 |
| 1 high-risk features | 6 | 0.39 | 0.17, 0.89 | 0.026 |
| 2 high-risk features | 29 | 1.77 | 1.13, 2.76 | 0.012 |
| 3–4 high-risk features | 27 | 1.97 | 1.25, 3.10 | 0.004 |
(Abbreviations: HR, hazard ratio; CI, confidence interval)
Figure 2.
Disease-free survival stratified by number of high-risk features (HRF) in the matched sublobar resection cohort compared with the matched lobectomy cohort (Confidence Intervals equal 95%).
Finally, we performed a univariable analysis of association of high-risk features and resection extent with DFS (Table 4). The following variables were statistically associated with worse DFS: sublobar resection vs. lobectomy (HR 1.43, 95% CI 1–2.05, p = 0.048), solid vs. subsolid tumor texture on CT (HR 1.80, 95% CI 1.24–2.61, p = 0.002), and SUV of the primary tumor (HR 1.05, 95% CI 1–1.10, p = 0.0034).
Table 4.
Univariable analysis of association of high-risk features and resection extent with disease-free survival.
| Characteristic | Event N | HR | 95% CI | P value |
|---|---|---|---|---|
| Procedure | ||||
| Lobectomy | 61 | — | — | |
| Sublobar | 63 | 1.43 | 1.00, 2.05 | 0.048 |
| CT Tumor Texture | ||||
| Subsolid | 62 | — | — | |
| Solid | 62 | 1.80 | 1.24, 2.61 | 0.002 |
| PET SUVmax | 124 | 1.05 | 1.00, 1.10 | 0.034 |
| cT | ||||
| IA | 12 | — | — | |
| IB | 112 | 1.18 | 0.65, 2.16 | 0.6 |
| MIP/Solid Histology | ||||
| No | 59 | — | — | |
| Yes | 65 | 1.24 | 0.87, 1.77 | 0.2 |
(Abbreviations: HR, hazard ratio; CI, confidence interval; CT, computed tomography; PET, positron emission tomography; SUV, standard uptake value; MIP, micropapillary)
Discussion
Sublobar resection has recently been shown to be oncologically noninferior to lobectomy for resection of small, peripheral NSCLC1,2. However, optimal management of patients with high-risk features indicative of an increased risk of recurrence at the time of surgery has not been studied in a randomized fashion. While our group has established the prognostic role of high-risk radiological17 and histological11,12,18 features as isolated criteria, in the current study we validated previous observations in a recent patient cohort and developed a score that is clinically useful. We limited our selection of high-risk variables to factors that can be known preoperatively so they can be incorporated into surgical planning prior to having complete pathologic data. Tumor size, tumor texture on CT, PET avidity of the primary tumor, and histologic subtype, if a biopsy was performed, are accessible parameters that can be applied in clinical practice. The results of the present study suggest that for clinical stage IA patients with 2 or more preoperative high-risk features, there may be a decreased risk of recurrence for patients undergoing lobectomy when compared to those undergoing sublobar resection. We also demonstrate that there is a significantly higher rate of nodal upstaging in the lobectomy group.
MIP/solid histology is one high-risk factor associated with worse outcomes and detection of these histologic subtypes prior to resection may help guide treatment planning11,12. Studies of preoperative biopsy techniques have shown that core needle biopsy and cryobiopsy have high specificity for identifying MIP/solid histology19,20 as well as moderate concordance rates of 56.5–64% at our institution. When obtaining a tissue diagnosis, we advocate for modalities that would allow for histologic subtype analysis such as core needle biopsy or cryobiopsy.
Recurrence rates in CALGB140503 and JCOG0802 ranged from 8% to 30%1,2; clinical and pathologic high-risk features were not comprehensively studied in these trials, although both studies excluded pure GGO lesions. Interestingly, in JCOG0802, the subgroup of patients with solid lesions had improved overall survival with segmentectomy over lobectomy2. No difference in disease free survival was seen in CALGB140503 by tumor size1. Although OS and DFS did not differ statistically in the overall matched cohorts in our study by resection type, we did identify a subset of patients with multiple high-risk features that may benefit from lobectomy over sublobar resection.
In addition to preoperative factors identified above, our data suggests that patients with high-risk clinical features also have more aggressive pathologic features, consistent with aggressive tumor biology. This was seen with higher rates of VPI, LVI, STAS, pathologic nodal upstaging in the unmatched lobectomy group that resolved in the matched group. These characteristics have also been correlated with a more aggressive molecular profile21.
Regardless of type of surgical resection, adequacy of resection margin and systematic lymph node dissection or sampling is correlated with oncologic outcomes. The National Comprehensive Cancer Network guidelines recommend a margin ≥ 2 cm or greater than the size of the nodule13. In our cohort, the median closest pathologic margin was 1.4 cm in the lobectomy group and 1.0 cm in the sublobar resection group with four patients undergoing R1 resection. Pathologic tumor size was ≤ 2 cm in 87% of the patients undergoing sublobar resection and 76% of patients undergoing lobectomy. Although close pathologic margin may contribute to rate of locoregional recurrence in cohort, there was no statistical difference between groups and overall locoregional recurrence rate with or without distant disease was approximately 6% for the overall matched cohort (not stratified based on high-risk characteristics), which is commiserate with published data.
Patients routinely underwent standardized lymph node dissection or sampling in our study. Due to the retrospective nature of this study, frozen section analysis of lymph nodes was not required during sublobar resection, and conversion to a larger anatomic procedure was at the discretion of the operating surgeon. Similar to previously published reports22, we identified a significantly higher rate of N1 upstaging in the lobectomy group, likely due to the intraparenchymal nodal dissection that is required in anatomic resections. Patients with diagnosed nodal disease received standard of care adjuvant systemic therapy. Undiagnosed nodal disease in the sublobar group may have contributed to worse DFS in patients with high-risk features.
Our study is primarily limited by its retrospective nature. As shown in this study, there is a wide range of tumor biology in even early-stage NSCLC. Although we attempted to control for both patient (pulmonary) and tumor high risk factors to create balanced groups with propensity score matching, it is not possible to eliminate confounding factors or selection bias. As seen in the unmatched cohorts, patients who underwent lobectomy were likely selected for more aggressive characteristics whereas the sublobar group was used for less aggressive lesions. Although matching controls for a number of these factors and shows that there may indeed be a difference in outcomes to this selective approach, it is likely that there are other factors that were considered at the time of operation that are not accounted for. Similarly, a wide range of operations were compared, with wedge and individual segmentectomy grouped with composite segmentectomy due to sample size and institutional practice. Finally, FEV1 was used as a surrogate for baseline pulmonary reserve. All patients had relatively low burden of systemic disease and using a cutoff of FEV1 60% was chosen to select for a cohort of patients that could have undergone either operation. However, it is not possible to retrospectively evaluate individual patient fitness, and it is possible that some patients in the sublobar group would not have tolerated lobectomy. To evaluate this question with limited bias will require a randomized trial.
In conclusion, we show here that high-risk features including tumor size, solid vs. subsolid tumor texture on CT, primary tumor SUV, and presence of MIP/solid histology should be considered in preoperative planning. Patients with stage IA NSCLC ≤2 cm and two or more high-risk preoperative features may have improved DFS with a lobectomy. Lobectomy should be considered over sublobar resection in the presence of multiple high-risk features.
Supplementary Material
Figure 3.
Summary of study findings.
Audio.
You can listen to the audio recording of the presentation and discussion associated with this paper in supplementary materials.
Central Message:
Lobectomy should be considered over sublobar resection for stage IA NSCLC ≤2 cm in the presence of multiple high-risk features.
Perspective Statement:
Sublobar resection has been validated as non-inferior for small, peripheral NSCLC. Data is lacking regarding the oncologic outcomes of stage IA NSCLC patients with high-risk features undergoing a sublobar resection. We show that patients with stage IA NSCLC ≤2 cm and two or more high-risk preoperative features may have improved disease-free survival with lobectomy when compared to sublobar resection.
Funding:
National Institutes of Health/National Cancer Institute (P30 CA008748 to Memorial Sloan Kettering Cancer Center).
Disclosures:
P.S.A. declares research funding from ATARA Biotherapeutics; Scientific Advisory Board Member and Consultant for ATARA Biotherapeutics, Bayer, Bio4T2, Carisma Therapeutics, Imugene, ImmPactBio, Johnson & Johnson, Orion Pharma, Outpace Bio; Patents, royalties, and intellectual property on mesothelin-targeted CAR and other T-cell therapies, which have been licensed to ATARA Biotherapeutics, issued patent method for detection of cancer cells using virus, and pending patent applications on PD-1 dominant negative receptor, wireless pulse-oximetry device, and on an ex vivo malignant pleural effusion culture system. Memorial Sloan Kettering Cancer Center has licensed intellectual property related to mesothelin targeted CARs and T-cell therapies to ATARA Biotherapeutics and has associated financial interests.
M.J.B. is a consultant for AstraZeneca Pharmaceuticals, Iovance Biotherapeutics, and Intuitive Surgical and receives research support from Obsidian Therapeutics. J.M.I. has served as an advisory board member for AstraZeneca and Merck, as an uncompensated steering board member for Genentech, has received institutional research support from ArcherDx/Invitae, Guardant Health, GRAIL, and Intuitive Surgical and travel support from Intuitive Surgical, and has equity/ownership interest in LumaCyte. D.M. serves on a steering committee for AstraZeneca, as a consultant for Johnson & Johnson, Bristol-Myers Squibb, AstraZeneca, and Boston Scientific, and has been an invited speaker for Merck and Genentech. B.J.P. serves as a consultant for Intuitive Surgical, CEEVRA, Medtronic, and Becton Dickinson. G.R. has a financial relationship with Scanlan, Merck, and Medtronic. V.W.R. reports unreimbursed participation in Data Safety and Monitoring Committees for Cancer Research UK MARS II and RAMONA trials. S.S. is a member of the AstraZeneca Advisory Board. D.R.J. is a member of the Advisory Council for AstraZeneca and receives research grant support from Merck.
Abbreviations
- NSCLC
non-small cell lung cancer
- GGO
ground glass opacity
- CT
computed tomography
- SUV
standard uptake value
- PET
positron emission tomography
- DFS
disease-free survival
- MIP
micropapillary
- LVI
lymphovascular invasion
- VPI
visceral pleural invasion
- STAS
spread through airspaces
Footnotes
IRB/Consent: IRB#18-391 (09/07/2018)
References
- 1.Altorki N, Wang X, Kozono D, et al. Lobar or Sublobar Resection for Peripheral Stage IA Non-Small-Cell Lung Cancer. N Engl J Med. Feb 9 2023;388(6):489–498. doi: 10.1056/NEJMoa2212083 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Saji H, Okada M, Tsuboi M, et al. Segmentectomy versus lobectomy in small-sized peripheral non-small-cell lung cancer (JCOG0802/WJOG4607L): a multicentre, open-label, phase 3, randomised, controlled, non-inferiority trial. Lancet. Apr 23 2022;399(10335):1607–1617. doi: 10.1016/S0140-6736(21)02333-3 [DOI] [PubMed] [Google Scholar]
- 3.Ginsberg RJ, Rubinstein LV. Randomized trial of lobectomy versus limited resection for T1 N0 non-small cell lung cancer. Lung Cancer Study Group. Ann Thorac Surg. Sep 1995;60(3):615–622; discussion 622–3. doi: 10.1016/0003-4975(95)00537-u [DOI] [PubMed] [Google Scholar]
- 4.Brandt WS, Bouabdallah I, Tan KS, et al. Factors associated with distant recurrence following R0 lobectomy for pN0 lung adenocarcinoma. J Thorac Cardiovasc Surg. Mar 2018;155(3):1212–1224 e3. doi: 10.1016/j.jtcvs.2017.09.151 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lou F, Huang J, Sima CS, Dycoco J, Rusch V, Bach PB. Patterns of recurrence and second primary lung cancer in early-stage lung cancer survivors followed with routine computed tomography surveillance. J Thorac Cardiovasc Surg. Jan 2013;145(1):75–81; discussion 81–2. doi: 10.1016/j.jtcvs.2012.09.030 [DOI] [PubMed] [Google Scholar]
- 6.Van Schil PE, Asamura H, Nishimura KK, et al. The International Association for the Study of Lung Cancer Lung Cancer Staging Project: Proposals for the Revisions of the T-Descriptors in the Forthcoming Ninth Edition of the TNM Classification for Lung Cancer. J Thorac Oncol. May 2024;19(5):749–765. doi: 10.1016/j.jtho.2023.12.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Fu F, Zhang Y, Wen Z, et al. Distinct Prognostic Factors in Patients with Stage I Non-Small Cell Lung Cancer with Radiologic Part-Solid or Solid Lesions. J Thorac Oncol. Dec 2019;14(12):2133–2142. doi: 10.1016/j.jtho.2019.08.002 [DOI] [PubMed] [Google Scholar]
- 8.Ito H, Suzuki K, Mizutani T, et al. Long-term survival outcome after lobectomy in patients with clinical T1 N0 lung cancer. J Thorac Cardiovasc Surg. Jan 11 2020;doi: 10.1016/j.jtcvs.2019.12.072 [DOI] [PubMed] [Google Scholar]
- 9.Fick CN, Dunne EG, Vanstraelen S, et al. High-risk features associated with recurrence in stage I lung adenocarcinoma. J Thorac Cardiovasc Surg. Feb 2025;169(2):436–444 e6. doi: 10.1016/j.jtcvs.2024.05.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hung JJ, Yeh YC, Jeng WJ, et al. Predictive value of the international association for the study of lung cancer/American Thoracic Society/European Respiratory Society classification of lung adenocarcinoma in tumor recurrence and patient survival. J Clin Oncol. Aug 1 2014;32(22):2357–2364. doi: 10.1200/JCO.2013.50.1049 [DOI] [PubMed] [Google Scholar]
- 11.Nitadori J, Bograd AJ, Kadota K, et al. Impact of micropapillary histologic subtype in selecting limited resection vs lobectomy for lung adenocarcinoma of 2cm or smaller. J Natl Cancer Inst. Aug 21 2013;105(16):1212–1220. doi: 10.1093/jnci/djt166 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Ujiie H, Kadota K, Chaft JE, et al. Solid Predominant Histologic Subtype in Resected Stage I Lung Adenocarcinoma Is an Independent Predictor of Early, Extrathoracic, Multisite Recurrence and of Poor Postrecurrence Survival. J Clin Oncol. Sep 10 2015;33(26):2877–2884. doi: 10.1200/JCO.2015.60.9818 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Network NCC. NCCN Clinical Practice Guidelines in Oncology. Non-small cell lung cancer. Accessed March 15, 2025, https://www.nccn.org/professionals/physician_gls/pdf/nscl.pdf
- 14.Dunne EG, Fick CN, Jones DR. Mediastinal Staging in Non-Small-Cell Lung Cancer: Saying Goodbye to Mediastinoscopy. J Clin Oncol. Aug 1 2023;41(22):3785–3790. doi: 10.1200/JCO.23.00867 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Travis WD, Brambilla E, Noguchi M, et al. International association for the study of lung cancer/american thoracic society/european respiratory society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol. Feb 2011;6(2):244–285. doi: 10.1097/JTO.0b013e318206a221 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. Nov 10 2009;28(25):3083–3107. doi: 10.1002/sim.3697 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Kadota K, Colovos C, Suzuki K, et al. FDG-PET SUVmax combined with IASLC/ATS/ERS histologic classification improves the prognostic stratification of patients with stage I lung adenocarcinoma. Ann Surg Oncol. Oct 2012;19(11):3598–3605. doi: 10.1245/s10434-012-2414-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Eguchi T, Kameda K, Lu S, et al. Lobectomy Is Associated with Better Outcomes than Sublobar Resection in Spread through Air Spaces (STAS)-Positive T1 Lung Adenocarcinoma: A Propensity Score-Matched Analysis. J Thorac Oncol. Jan 2019;14(1):87–98. doi: 10.1016/j.jtho.2018.09.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Suzuki M, Matsumoto Y, Imabayashi T, et al. Cryobiopsy as a reliable technique for the preoperative identification of micropapillary/solid components in early-stage lung adenocarcinoma. Lung Cancer. Dec 2021;162:147–153. doi: 10.1016/j.lungcan.2021.11.004 [DOI] [PubMed] [Google Scholar]
- 20.Kim TH, Buonocore D, Petre EN, et al. Utility of Core Biopsy Specimen to Identify Histologic Subtype and Predict Outcome for Lung Adenocarcinoma. Ann Thorac Surg. Aug 2019;108(2):392–398. doi: 10.1016/j.athoracsur.2019.03.043 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Caso R, Sanchez-Vega F, Tan KS, et al. The Underlying Tumor Genomics of Predominant Histologic Subtypes in Lung Adenocarcinoma. J Thorac Oncol. Dec 2020;15(12):1844–1856. doi: 10.1016/j.jtho.2020.08.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Jacobs RC, Rabin EE, Logan CD, et al. Pathologic upstaging and survival outcomes for patients undergoing segmentectomy versus lobectomy in clinical stage T1cN0M0 non–small cell lung cancer. JTCVS Open. 2025/April/01/ 2025;24:394–408. doi: 10.1016/j.xjon.2025.01.014 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.