Segmentectomy versus wedge resection for radiological solid predominant and low metabolic non-small cell lung cancer

Abstract OBJECTIVES The prognosis of segmentectomy and wedge resection for solid predominant early-stage non-small cell lung cancer with low metabolic activity is unclear. METHODS This study aimed to assess patients who underwent segmentectomy or wedge resection with curative intent for clinically node-negative non-small cell lung cancer presenting as a solid predominant tumour (consolidation tumour ratio >50%) with a whole size ≤3 cm and [18F]-fluoro-2-deoxy-D-glucose accumulation weaker than that of the mediastinum tissue (Deauville score, 1 or 2) on positron emission tomography/computed tomography. The cumulative incidence of recurrence (CIR) was compared using the Gray method, and the predictive factor of CIR was analysed using the Fine and Gray method. RESULTS Of 140 patients included in this study, 93 (66.4%) underwent segmentectomy and 47 (33.6%) underwent wedge resection. No significant difference in the clinical stage was found between the 2 groups. The CIR was higher with wedge resection than with segmentectomy (P = 0.004). Recurrence after wedge resection was noted in 4 (8.5%) patients, 2 of whom had a recurrent site containing lung parenchyma of the preserved lobe and hilum lymph node, which would have been resected if segmentectomy had been performed. In the multivariable analysis for CIR using inverse probability of treatment weighting and the procedure, wedge resection was a significantly worse predictive factor (hazard ratio, 12.280; P = 0.025). CONCLUSIONS Segmentectomy rather than wedge resection should be considered for solid predominant, small-size non-small cell lung cancer even if [18F]-fluoro-2-deoxy-D-glucose accumulation is low.


INTRODUCTION
Lobectomy is the standard procedure in the treatment of nonsmall cell lung cancer (NSCLC), as supported by the results of a randomized trial [1]. However, some studies present a favourable prognosis with sublobar resection for small size (< _30 mm), clinically node-negative NSCLC [2][3][4][5]. Currently, a prospective trial to investigate the feasibility of segmentectomy for early-stage NSCLC is ongoing [6,7].
In contrast, NSCLC with high [18F]-fluoro-2-deoxy-D-glucose (FDG) accumulation on positron emission tomography (PET)/ computed tomography (CT) has invasive characteristics [8], even if its size is small [9]. Based on visual evaluation by FDG-PET/CT, we previously showed that early-stage adenocarcinoma with low FDG accumulation rarely has lymph node metastasis [10] and its prognosis is favourable, irrespective of the procedure [11]. In that study, even in patients with high consolidation tumour ratio (CTR >50%), which is a worse prognostic factor of early-stage NSCLC [12], the incidences of recurrence and lymph node metastasis were low in those with a low FDG accumulation [11]. The frequency of invasive characteristics, such as lymphatic invasion, vascular invasion and pleural invasion, were also low in patients with low FDG accumulation [11]. Therefore, this study was conducted based on the hypothesis that wedge resection may have a prognosis similar to that of segmentectomy for NSCLC with low FDG accumulation, even if its CTR is high.

Ethics statement
This retrospective study was approved by the Institutional Review Board of Hiroshima University Hospital, Hiroshima, Japan (approval number, E-1216). The requirement for informed consent from individual patients was waived because it was a retrospective study.

Patients
We analysed the clinicopathological data and prognosis of patients who underwent segmentectomy or wedge resection with curative intent for NSCLC that presented as a nodenegative, solid component predominant tumour (CTR >50%) with a whole tumour size < _3 cm on preoperative CT and had a low FDG accumulation on preoperative FDG-PET/CT between April 2007 and March 2019 at Hiroshima University Hospital. FDG accumulation on PET/CT was evaluated using the Deauville criteria, a visual 5-point scale evaluating FDG-PET/CT characteristics [12]. Based on the results of our previous study, a Deauville score of 1-2 was considered low accumulation, whereas a score of 3-5 was considered high accumulation [10,11].
Supplementary Material, Table S1 shows additional details of the Deauville criteria. The exclusion criteria were as follows: previous induction therapy; small cell lung cancer or carcinoid; whole tumour size >3 cm or high FDG accumulation on PET/CT (Deauville score 3-5); and suspicious lymph node metastasis (enlargement >1 cm in preoperative CT or significant FDG accumulation on PET/CT) preoperatively. Patients without complete resection and those with obvious tumour progression after FDG-PET/CT were also excluded from this study. Figure 1 presents a flowchart of patient selection. The primary endpoint of this study was cumulative incidence of recurrence (CIR).

Preoperative examination
Preoperative evaluations, including chest CT, whole-body FDG-PET/CT, brain magnetic resonance imaging and pulmonary function tests, were performed to determine the clinical stage and treatment strategies. Staging was based on the TNM Classification for Lung Cancer, 8th edition, for malignant tumours [13].

Surgical procedures
Surgery was performed with hybrid video-assisted thoracic surgery [14]. In our institution, sublobar resection is performed for patients with tumour <2 cm, those with tumour with low metabolic activity, and patients considered intolerant to lobectomy. The decision regarding which sublobar procedure to perform (i.e. segmentectomy or wedge resection) is usually based on a combination of patient performance status, tumour characteristics and surgeon preference. Wedge resection is preferred for small, pleural-based tumours, particularly for patients with poor performance status and comorbidities, whereas segmentectomy is usually required for large tumours that settle within a resectable anatomical segment.

Histologic and pathologic evaluations
The determination of pathologic stages was based on the TNM Classification for Lung Cancer, 8th edition, for malignant tumours [13]. Determination of the histologic subtype was based on the World Health Organization classification [15]. The diagnosis of lymphatic invasion was based on pathologic examination using immunostaining for D2-40 to clarify the location of the lymphatic ducts. The presence of vascular invasion and pleural invasion was evaluated using elastic van Gieson staining to determine tumour invasion above the elastic layer of the vessels and visceral pleura.

Follow-up evaluation
Postoperative follow-up procedures, including physical examination and chest CT every 6 months, were performed for 5-10 years after surgical resection. Recurrence was determined based on radiographic features or histological evidence.

Statistical analysis
The results were presented as medians and interquartile ranges for continuous variables and numbers and percentages for categorical variables. Continuous variables that were normally distributed were analysed using Student's t-test. Continuous variables that were non-normally distributed were analysed using Wilcoxon's rank-sum test. Pairwise analysis was used for analysis that contains missing data. McNemar's test for categorical variables and paired t-tests for continuous variables were used to analyse propensity-matched patient pairs. Prognosis was analysed using competing risk analysis. The risk of recurrence (defined as CIR), which is the period from surgery to recurrence, was the main outcome of this study and estimated using a cumulative incidence function that accounted for mortality without recurrence as a competing event. The risk of death without recurrence [defined as cumulative incidence of death without recurrence (CIDWR), which is the period from surgery to death without recurrence] was also estimated using a cumulative incidence function that accounted for recurrence as a competing event. Patients were censored if they were alive and without recurrence at the time of the last follow-up. Differences in CIR and CIDWR between groups were assessed using the Gray method.
Propensity scores were estimated using a logistic regression model that included solid component size (continuous value), CTR (continuous value) and histology (adenocarcinoma or nonadenocarcinoma). Matching cohorts were formed using this propensity score; segmentectomy and lobectomy group pairs with an equivalent propensity score were selected by a 1-to-1 match with a calliper width of 0.2 of standard deviation.
Inverse probability of treatment weighting (IPTW) based on propensity scores was also used to adjust for differences in covariates between both groups, and multivariable analysis for CIR was performed using the IPTW and surgical method to investigate whether the surgical procedure affected prognosis using the Fine and Gray methods.  All statistical analyses were performed using EZR version 1.51 (Saitama Medical Center, Jichi Medical University, Saitama, Japan) [16], which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria).

RESULTS
A total of 140 patients were included in this study. The distribution of patients is shown in Supplementary Material, Fig. S1. Patient characteristics are presented in Table 1. Of the 140 study patients, 93 (66.4%) underwent segmentectomy and 47 (33.6%) underwent wedge resection. Although the whole tumour size was larger in patients who underwent segmentectomy than in those who underwent wedge resection (P = 0.030), no difference was noted in the solid component size (P = 0.533) and clinical stage (P = 0.536). The incidence of adenocarcinoma was higher in patients who underwent segmentectomy (P = 0.035). There was no difference in the pathological stage between patients who underwent segmentectomy and those who underwent wedge resection (P = 0.826). Eight patients had pathological stage IB. The diagnosis of stage IB was based on pleural invasion, not tumour diameter.
The characteristics and preoperative image of patients with relapse are shown in Table 2 and Fig. 3, respectively. No recurrences were noted in patients with Deauville score 1; however, relapse occurred in 4 patients (8.5%) after wedge resection and 1 (1.1%) patient after segmentectomy. Among patients who underwent wedge resection, 2 patients had recurrence only at the lesion site where the tissue would not have been resected even if segmentectomy were performed (Case 1, pleural dissemination; Case 4, bone metastasis). In contrast, relapse occurred in 2 other patients, in whom the tissue may have been resected if segmentectomy were performed (Case 2, hilar and mediastinum lymph node; Case 3, lung at preserved lobe).
In the multivariable analysis for CIR using IPTW and procedure (characteristics of IPTW cohort are shown in Supplementary Material, Table S3), wedge resection was a significantly worse predictive factor of CIR (hazard ratio 12.980, 95% CI 1.601-105.300, P = 0.016, Table 3).

DISCUSSION
In this study, CIR after wedge resection was worse than that after segmentectomy, and no significant difference was noted between procedures in CIDWR. In the multivariable analysis, wedge resection was a significantly worse predictive factor of recurrence. These results indicate that, even for a tumour with low metabolic activity, segmentectomy should be considered first as a sublobar resection for NSCLC that presents as a solid predominant (CTR > 50%) tumour.
In this study, CIR was the primary endpoint. The background characteristics differed between patients who underwent segmentectomy and those who underwent wedge resection. In deciding the surgical procedure, selection bias may have been present. Recurrence-free survival and overall survival are often used as endpoints of this type of retrospective study. However, recurrence and death from causes other than lung cancer are treated as events in the analysis using recurrence-free survival. In the analysis using overall survival, all deaths are equally treated   Table 2.
as events, regardless of cause of death. In the analysis of earlystage NSCLC, the incidence of death from causes other than lung cancer is high; therefore, CIR was used as a primary endpoint to assess the oncologic effects between segmentectomy and wedge resection, considering the competing risk such as non-cancer events.
It is already known that wedge resection can provide sufficient prognosis for NSCLC with low malignancy, such as a tumour with low CTR (< _50%) [4,17]. Regarding early-stage NSCLC with high CTR, a randomized trial is ongoing, and segmentectomy can be a standard treatment strategy for NSCLC with whole tumour size < _2 cm. To the best of our knowledge, no randomized study has been conducted to compare segmentectomy and lobectomy for NSCLC with whole tumour size < _3 cm. However, the retrospective study shows that prognosis after segmentectomy is equivalent to that after lobectomy, including NSCLC with whole tumour size < _3 cm [18][19][20]. Therefore, segmentectomy may become the standard treatment for NSCLC with whole tumour size < _3 cm. Generally, as a sublobar resection, segmentectomy is superior to wedge resection in terms of postoperative prognosis [21]. In contrast, it is also known that FDG-PET/CT is useful in predicting invasive characteristics of NSCLC. Therefore, we hypothesized that wedge resection can provide sufficient prognosis in patients with a tumour that has low metabolic activity, even if CTR is high.
We reported that the Deauville score is a significant predictive factor of lymph node metastasis [10] and prognosis [11] for earlystage lung adenocarcinoma; in our study, no difference was found in the prognosis after lobectomy, segmentectomy and wedge resection in patients with low Deauville score [11]. Conversely, CTR is also a worse prognostic factor of early-stage NSCLC [22]; therefore, we analysed the incidence of lymph node metastasis and recurrence by classifying patients as having CTR < _50% or CTR >50%. In that analysis, most patients with CTR < _50% had a Deauville score of 1 or 2, and no recurrence and lymph node metastasis were noted. This result is consistent with the results of previous studies on CTR, and it was important to investigate the relationship between prognosis, procedure and FDG accumulation in patients with a high CTR. Because it was already known that NSCLC with high FDG accumulation develop lymph node metastasis even if it is clinically node-negative, patients with a high Deauville score were excluded from this study that compared segmentectomy with lymph node dissection and wedge resection because usually lymph node dissection is not performed. Initially, we expected that the prognosis would be the same regardless of whether we performed segmentectomy or wedge resection. However, the results were unexpected. In our study, 4 patients had recurrence after wedge resection. Of these patients, 2 had relapse with a site outside the area to be resected if segmentectomy were performed. Recurrence in these patients may have been inevitable, even if they had undergone segmentectomy. Recurrence for 2 other patients could have been prevented because their recurrent site contained lung parenchyma of preserved lobe and hilum lymph node, which would have been resected if segmentectomy were performed. The merit of segmentectomy compared with wedge resection is that an adequate surgical margin can be obtained for a centrally located tumour that is difficult to treat by wedge resection. Moreover, hilar lymph nodes may be dissected by segmentectomy but not by wedge resection. These factors probably contribute to better prognosis after segmentectomy.

Limitations
This study has several limitations. First, this study is from a single institution, and the number of study patients was restricted. Particularly, wedge resection is not a standard procedure for NSCLC with high CTR; thus, the number of patients who underwent wedge resection was small. Hazard ratios in multivariable analysis are large and 95% CIs in prognostic analysis and multivariable analysis are wide. It is possible that collinearity, small sample size or low rate of recurrence affected results of our study. Second, because this study design is retrospective and the decision on the procedure is finally influenced by the preference of the attending surgeon and patients, some bias may have affected the results. Although the reason was unclear, the proportion of patients who underwent segmentectomy and wedge resection was different for each year. To confirm the results of this study, a prospective study is needed. However, wedge resection was found to be a worse prognostic factor of CIR in the multivariable analysis. This finding indicates that segmentectomy should be considered first. Therefore, the significance of this study is profound because no other study has compared wedge resection and segmentectomy for a solid predominant tumour with low metabolic activity.

CONCLUSION
In patients with NSCLC presenting with a solid predominant tumour on preoperative CT with a whole tumour size < _3 cm, prognosis after wedge resection was worse than that after segmentectomy, even if FDG accumulation was low. Regarding the treatment strategy for early-stage NSCLC with high CTR, segmentectomy should be considered first even if FDG accumulation is low.

SUPPLEMENTARY MATERIAL
Supplementary material is available at ICVTS online.

Funding
This work was supported by Japan Society for the Promotion of Science KAKENHI (Grant Number JP20K17749).

Data availability statement
The data underlying this article are available in the article and in its online supplementary material.