Comparative oncological features of centrally and peripherally located small-sized radiologically solid-dominant non-small-cell lung cancer

Abstract

OBJECTIVES

This study aimed to compare the oncological features of centrally and peripherally located small-sized (≤2 cm), radiologically solid-dominant, cN0 non-small-cell lung cancer (NSCLC).

METHODS

We retrospectively reviewed 1240 patients who underwent lobectomy or segmentectomy for radiologically solid-dominant cN0 NSCLC tumours ≤2 cm in size. Tumours were categorized as centrally (inner two-thirds of the pulmonary parenchyma) or peripherally (outer one-third) located. Clinicopathological characteristics and prognoses were compared between the 2 groups.

RESULTS

Among the 1240 patients, 299 had centrally located and 941 had peripherally located tumours. Centrally located tumours showed a significantly higher proportion of pure solid tumours and pathological lymph node upstaging than peripherally located tumours (P = 0.018 and P = 0.038, respectively). Multivariable logistic regression analysis identified central location as an independent predictor for pN1 (odds ratio, 1.91; 95% confidence interval, 1.09–3.36; P = 0.024), but not for pN2, upstaging. The cumulative incidence of loco-regional and distant recurrences did not significantly differ between the 2 groups (P = 0.455 and P = 0.383, respectively). Overall survival and recurrence-free survival rates were also similar among patients with central and peripheral tumours (P = 0.267 and P = 0.269, respectively).

CONCLUSIONS

Patient prognosis following complete anatomical resection was comparable between centrally and peripherally located radiologically solid-dominant cN0 NSCLC tumours ≤2 cm in size. However, centrally located tumours were associated with a higher risk of pN1 upstaging, highlighting the importance of thorough hilar lymph node dissection in these patients.

Open in new tabDownload slide

INTRODUCTION

Two large trials, JCOG0802/WJOG4607L and CALGB140503, recently revealed the efficacy of sublobar resection as compared to lobectomy in peripheral early-stage non-small-cell lung cancer (NSCLC) [1, 2]. These findings suggest that segmentectomy is likely to become more widespread in this capacity. Efforts to expand its indications have led to a growing interest in performing segmentectomy for centrally located tumours [3, 4]. While we previously reported the efficacy of segmentectomy for centrally located small-sized (≤2 cm), radiologically solid-dominant [0.5 < consolidation-to-tumour ratio (CTR) ≤ 1] NSCLC [5], lobectomy remains the standard treatment of these tumours. Additionally, the National Comprehensive Cancer Network (NCCN) version 11.2024 guidelines [6] recommend invasive mediastinal staging for centrally located tumours, even those diagnosed as stage IA. Therefore, in determining the optimal treatment strategy for small-sized early-stage NSCLC, distinguishing between peripherally and centrally located tumours is crucial.

The clinicopathological findings of centrally located solid-dominant NSCLC tumours, however, remain unclear, particularly in comparison to those of peripherally located tumours. While the central type is recognized as a risk factor for N1 upstaging [7, 8], differences in malignancy, recurrence patterns and prognosis between the 2 groups remain poorly understood. Understanding these differences is crucial for expanding the indications for segmentectomy, optimizing surgical strategies, and improving the outcomes of patients with NSCLC. Therefore, this study aimed to investigate the oncological features of centrally located small-sized (≤2 cm) and radiologically solid-dominant (0.5 < CTR ≤ 1) cN0 NSCLC compared to peripheral NSCLC.

PATIENTS AND METHODS

Ethics statement

This study was approved by the institutional review board of Hiroshima University Hospital (approval no. E2018-1216-02) on 30 November 2022. Given its retrospective nature, the requirement for informed consent was waived. Any collection and storage of data or biological material from research participants for multiple and indefinite use should be consistent with the requirements outlined in the WMA Declaration of Taipei. A research ethics committee must approve the establishment and monitor ongoing use of such databases.

Study cohort

Patient data were obtained from the HITOKA-3 database project, encompassing individuals who underwent either lobectomy or segmentectomy for small-sized (≤2 cm), radiologically solid-dominant (0.5 < CTR ≤ 1) cN0 NSCLC tumours at Hiroshima University, Tokyo Medical University Hospital, and Kanagawa Cancer Center between January 2010 and December 2022 (Supplementary Material, Fig. S1). All patients underwent preoperative high-resolution computed tomography and 18-fluoro-2-deoxyglucose positron emission tomography/computed tomography (FDG-PET/CT). Clinical nodal metastasis was defined as negative for lymph nodes <10 mm on high-resolution computed tomography with no FDG uptake detected on PET/CT. Preoperative endobronchial ultrasonography and mediastinoscopy were only performed when mediastinal lymph node metastasis was suspected based on CT or PET/CT findings. Patients who received neoadjuvant therapy were excluded. Staging was conducted according to the 8th edition of the TNM Classification of Lung and Pleural Tumours [9].

FDG-PET/CT

Standardized uptake values can vary across multicentre studies. We employed an anthropomorphic body phantom to minimize the inter-institutional variability of the maximum standardised uptake value (SUVmax). This method helped mitigate differences in preparation procedures, scan acquisition, image reconstruction, and data analysis among the 3 study centres.

Tumour location

Tumour locations were assessed using the distance ratio, as previously described [5]. This was calculated by dividing the distance from the pulmonary hilum to the centre of the tumour by the distance from the pulmonary hilum to the visceral pleural surface through the centre of the tumour. Based on this ratio, tumours were categorized as central-type if the distance ratio was ≤0.67, indicating the inner two-thirds of the lung field, and peripheral-type if the distance ratio was >0.67, indicating the outer third of the lung field.

Surgical procedure and follow-up examination

The surgical approach and procedures were determined by a multidisciplinary cancer board comprised surgeons, physicians, pathologists, and radiologists, based on the tumour and patient characteristics. Segmentectomy was chosen when preoperative CT scans demonstrated a sufficient surgical margin larger than the tumour size. For both lobectomy and segmentectomy, the pulmonary artery, vein, and bronchus were dissected. The intersegmental plane was identified using either the inflation-deflation line or indocyanine green injection. The lung parenchyma was dissected using staplers or electrocautery to ensure adequate surgical margins, defined as the distance from the tumour to the nearest incision line, which could be the intersegmental plane, interlobar cutting line, or bronchial stump. If lymph node metastasis was detected, or the margins were closed during surgery, conversion to lobectomy or additional resection was considered. Surgical margin assesment was performed exclusively for patients who underwent segmentectomy based on pathological findings. Adjuvant chemotherapy was administered with oral tegafur uracil for pathological stage I tumours >20 mm in size, and platinum-based chemotherapy was administered for pathological stage II or III tumours.

All patients underwent follow-up evaluations every 3–6 months during the first 5 years after surgery, followed by annual assessments for the next 5 years. These evaluations included physical examinations, tumour marker measurements, chest radiography, and chest CT. If recurrence was suspected, further investigations such as FDG-PET/CT and brain magnetic resonance imaging were performed. Recurrence was classified as follows: local recurrence, tumour relapse at or within the resection margins of the lungs or bronchus; regional recurrence, tumour relapse in the hilar, mediastinal, or ipsilateral lung; and distant recurrence, metastasis to other organs or diffuse pleural disease.

Statistical analysis

Continuous variables are reported as medians with interquartile ranges and were compared using the Mann–Whitney U-test. Categorical variables are reported as numbers and percentages, with comparisons performed using Fisher’s exact test. Recurrence-free survival (RFS) was defined as the duration from surgery to recurrence, death from any cause, or final follow-up. Overall survival (OS) was defined as the duration from surgery to death from any cause or last follow-up. Survival data were calculated using the Kaplan–Meier method and compared using the log-rank test. Multivariable analysis was conducted using a backward stepwise method, incorporating clinical variables relevant to treatment decisions. Although multicollinearity was not assessed in this study, the selection of variables was based on their clinical relevance and prior evidence in the literature. Time-to-event end-points were analysed using a competing risk analysis, and the risk of recurrence was estimated using a cumulative incidence function that accounted for deaths without recurrence as competing events. The patients were censored if they were alive and had experienced no recurrence upon their final follow-up. Differences in the cumulative incidence of recurrence (CIR) between groups were assessed using Gray’s test. All statistical analyses were performed using JMP (version 17.0; SAS Institute, Cary, NC, USA) and EZR version 1.51 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was set at P < 0.05.

RESULTS

Patient characteristics

Among 1240 patients, 299 had central-type tumours, and 941 had peripheral-type tumours. Figure 1 shows the relationship between the distance ratio and the number of cases, revealing a decrease in cases as the distance ratio approaches the hilar. The clinical and pathological characteristics of the 2 groups are shown in Table 1. Pure solid tumours were more frequently observed in the central than in the peripheral type (72% vs 64%, respectively; P = 0.018). The median surgical margin for segmentectomies was significantly shorter among the central type than the peripheral type (10 vs 15 mm; P < 0.001). Although the number of resected segments was higher in the central type than in the peripheral type, the difference was not statistically significant (P = 0.102). The proportion of patients with pathological lymph node upstaging was higher in the central type than in the peripheral type (pN1: 7% vs 4%; pN2: 2% vs 3%, respectively; P = 0.038), while pleural invasion was significantly more common in the peripheral type than in the central type (16% vs 10%, respectively; P = 0.024). No significant differences in epidermal growth factor receptor mutation-positive and anaplastic lymphoma kinase-positive statuses were observed between the central type and the peripheral type (P = 0.219 and P = 0.898, respectively).

Risk of occult N1 and N2 lymph node metastasis

Multivariable logistic regression analysis (Table 2) revealed that central-type [odds ratio (OR), 1.91; 95% confidence interval [CI], 1.09–3.36; P = 0.024) was an independent risk factor for occult N1 lymph node metastasis as well as solid tumour size (OR 1.22; 95% CI 1.12–1.35; P < 0.001) and SUVmax (OR 1.10; 95% CI 1.05–1.16; P < 0.001). While the central type was not an independent risk factor for occult N2 lymph node metastasis, older age (OR 0.96; 95% CI 0.94–0.99; P = 0.019), pure solid tumour (OR 2.50; 95% CI 1.00–6.22; P = 0.049), and SUVmax (OR 1.07; 95% CI 1.01–1.14; P = 0.030) were significantly associated with occult N2 lymph node metastasis.

Survival outcomes

The median follow-up period after surgery was 4.3 years (interquartile range, 2.0–5.7). OS and RFS were not significantly different between central and peripheral types: 5-year RFS (95% CI), 81.9% (72.0–88.8) vs 84.8% (80.0–88.7); P = 0.269 and 5-year OS (95% CI), 88.5% (79.3–93.9) vs 91.0% (86.7–94.0); P = 0.267, respectively; Fig. 2.

Recurrence occurred in 26 (9%) and 81 (9%) patients with central- and peripheral-type tumours, respectively (Table 3), while only 1 (0.3%) and 4 (0.4%) patients developed local recurrence, respectively. Although the loco-regional and distant CIRs were not significantly different between the 2 groups (Fig. 3), central type had a lower loco-regional CIR and a higher distant CIR than the peripheral type. Multivariable Cox analysis revealed that older age [OS: hazard ratio (HR) 1.09; 95% CI 1.07–1.12; P < 0.001; RFS: HR 1.04; 95% CI 1.02–1.06; P < 0.001], male sex (OS: HR 2.51; 95% CI 1.60–3.95; P < 0.001; RFS: HR 1.72; 95% CI 1.25–2.39; P = 0.001), solid tumour size (OS: HR 1.16; 95% CI 1.10–1.21; P < 0.001; RFS: HR 1.11; 95% CI 1.06–1.17; P < 0.001), and increasing SUVmax (OS: HR 1.06; 95% CI 1.02–1.10; P = 0.004; RFS: HR 1.06; 95% CI 1.03–1.09; P < 0.001) were independent prognostic factors for both OS and RFS; however, central-type was not (Table 4).

DISCUSSION

This study demonstrated that the proportion of occult hilar lymph node metastases was significantly higher in the central- than in the peripheral-type tumours. Although the prognoses were comparable between the groups, the central type showed a trend towards a slightly worse prognosis, likely due to the higher prevalence of the prognostically unfavourable pN1 population in the central type. Notably, despite the higher proportion of patients with pN1, the loco-regional recurrence was comparable between the 2 groups. Securing surgical margins and performing hilar lymph node dissection may have contributed to reducing loco-regional recurrence. These findings are important for the treatment of centrally located small NSCLC.

Central-type tumours are recognized as a risk factor for occult lymph node metastasis [7, 8, 10–12]. The underlying mechanism remains unclear; however, the differences in the development and distribution of lymphatic vessels among the tumour locations may be partially involved [13]. The central regions of the lung, particularly around the bronchial and peribronchial areas, contain a denser lymphatic network, potentially facilitating the rapid transport of metastatic cells to the hilar lymph nodes. In contrast, the peripheral regions, characterized by fewer and smaller lymphatic vessels [13, 14], may lead to slower lymphatic spread and lower rates of lymph node metastasis. Interestingly, the proportion of occult N2 metastases was not higher in the central type than in the peripheral type, which is consistent with previous studies [7]. This result may be attributed to the widespread use of PET/CT. In this study, all patients underwent PET/CT, which facilitated accurate preoperative staging and allowed for the exclusion of cases with evident nodal metastases preoperatively. As a result, many cases in this study were likely limited to pN1.

According to the NCCN guidelines version 11. 2024, invasive mediastinal staging is recommended for the central type of even clinical stage I NSCLC [6]. However, since the central type was not a risk of pN2 compared to the peripheral type, routine invasive mediastinal staging may not always be necessary given its highly invasive nature. This is especially relevant as the diagnostic rate of small-sized tumours using bronchoscopy remains poor (30.8% for tumours 15–25 mm in size) [15]. These findings suggest that the necessity of invasive mediastinal staging should be reconsidered in future guidelines for preoperative node staging, particularly for small-sized central-type NSCLC.

Despite the higher rate of lymph node metastasis in central-type tumours, the loco-regional recurrence remained comparable between the 2 groups, while distant recurrence was more frequent in the central type than in the peripheral type. These findings underscore the clinical relevance of hilar lymph node metastasis, which is more associated with a higher risk of distant recurrence than loco-regional recurrence [16]. Adjuvant therapy plays a critical role in mitigating the risk of distant recurrence, particularly with recent advancements in adjuvant therapy, including tyrosine kinase inhibitors and immune checkpoint inhibitors [17–19]. Therefore, for centrally located tumours, thorough lymph node dissection is essential for achieving local disease control and ensuring accurate pathological staging.

Additionally, negative margins were consistently secured in both central and peripheral types. When the surgical margins were closer, additional resection, including conversion to lobectomy, was performed. While the definition of an adequate surgical margin remains controversial [20–22], the NCCN guidelines recommend that the surgical margin should be at least equal to the tumour size or ≥2 cm [6]. Since our study showed that the number of resected segments was higher in the central than in the peripheral type, it is necessary to remove not just a single segment but also adjacent segments or even perform basal segmentectomy for central-type tumours. This approach may help lower the risk of local recurrence, ultimately enabling comparable local control regardless of tumour location. Multivariable Cox analysis also indicated that surgical procedure was not an independent prognostic factor for either OS or RFS. Previous reports [23, 24] have also demonstrated that the prognosis after segmentectomy for the central type is not significantly different from that for the peripheral type. Unlike wedge resection, segmentectomy involves the dissection and transection of the pulmonary artery, vein and bronchus. These segmentectomy processes facilitate more extensive hilar lymph node dissection and larger surgical margins [25].

Our study has a few limitations. First, it was a retrospective study that lacked preoperative data such as performance status, pulmonary function and comorbidities. Second, the selection criteria of surgical procedures, preoperative surgical margins, and the number of patients who had segmentectomy converted to lobectomy were unclear. Regardless of the selection criteria or the conversion strategy, they were applied consistently for both central and peripheral types. The prognostic results reflect this consistent strategy. Third, more than half of the patients were not examined using driver mutation tests. Although there were no significant differences in patients with epidermal growth factor receptor mutations or anaplastic lymphoma kinase positivity between central- and peripheral-type tumours, additional analyses are needed to clarify the molecular differences. Fourth, the definition of central and peripheral location used in this study represents one of several approaches described in the literature [5, 7, 11, 12]. While we followed the method used in JCOG0802/WJOG4607L [1] and CALGB140503 [2], which defined peripheral tumours as those located in the outer third of the lung, different classification methods could lead to varying results. This variability highlights the need for a standardized definition to improve comparability across studies. Fifth, excluding tumours with clinical lymph node metastasis ensured a more homogeneous study population; however, this may have introduced selection bias. Future analyses including tumours with lymph node metastasis are needed. Lastly, this study is retrospective in nature, which inherently limits the ability to control for all potential confounders. Selection bias may have influenced the choice of surgical procedures and patient outcomes. Additionally, we did not formally test the proportional hazards assumption in the log-rank tests and Cox regression analysis. Violation of this assumption could potentially impact the validity of the results. Prospective studies are necessary to confirm the definitive conclusions.

In conclusion, central-type tumours had a higher prevalence of occult N1 lymph node metastasis than peripheral-type among radiologically solid-dominant cN0 NSCLC tumours ≤ 2 cm in size, indicating the importance of thorough hilar lymph node dissection. However, despite the higher rate of pN1 upstaging, loco-regional recurrence and prognosis are comparable between central- and peripheral-type tumours if complete anatomical resection with thorough hilar lymph node dissection and adequate margins is performed.

SUPPLEMENTARY MATERIAL

Supplementary material is available at EJCTS online.

FUNDING

This work was supported by grants from the Japan Society for the Promotion of Science KAKENHI (grant number 23K15557).

Conflict of interest: none declared.

ACKNOWLEDGMENTS

We thank Editage (www.editage.com) for English language editing.

DATA AVAILABILITY

The data underlying this study are available in this article and its online supplementary material.

Author contributions

Norifumi Tsubokawa: Conceptualization; Data curation; Formal analysis; Writing—original draft; Writing—review & editing. Takahiro Mimae: Validation; Writing—review & editing. Yoshihiro Miyata: Supervision; Writing—review & editing. Chiaki Kanno: Data curation. Yujin Kudo: Data curation; Writing—review & editing. Takuya Nagashima: Data curation; Writing—review & editing. Hiroyuki Ito: Supervision. Norihiko Ikeda: Supervision. Morihito Okada, PhD, MD: Supervision; Writing—review & editing

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Alex Fourdrain, Frank A. Baciewicz Jr, Jose Belda-Sanchis and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• CI

  Confidence interval

 
• CIR

  Cumulative incidence of recurrence

 
• CT

  Computed tomography

 
• CTR

  Consolidation-to-tumour ratio

 
• FDG-PET/CT

  18-Fluoro-2-deoxyglucose positron emission tomography/computed tomography

 
• HR

  Hazard ratio

 
• NSCLC

  Non-small-cell lung cancer

 
• OR

  Odds ratio

 
• OS

  Overall survival

 
• RFS

  Recurrence-free survival

 
• SUVmax

  Maximum standardized uptake value