Does the positron emission tomography-computed tomography decrease the risk of thoracoscopic lobectomy for benign disease?—a propensity score matched cohort study

Abstract

Background

Preoperative characterization of solitary pulmonary nodules is critical yet challenging, as distinguishing early-stage neoplasms from benign lesions is essential to prevent unwarranted anatomical resections. The study aimed to analyze the impact of preoperative positron emission tomography-computed tomography (PET-CT) on the incidence of video-assisted thoracoscopic surgery (VATS) lobectomy for unsuspected benign disease, in patients without preoperative diagnosis.

Methods

The study included 864 patients, after VATS lobectomy, without the preoperative diagnosis, between January 1, 2014 and December 31, 2020, with 654 receiving preoperative PET-CT and 210 without it. The propensity score matching analysis was done, calculating the predicted probability using: age, sex, cardiovascular comorbidities, previous malignancy, chronic obstructive pulmonary disease (COPD), diabetes, and T category, obtaining equal groups of 210 patients with preoperative PET-CT (PET group) and without preoperative PET-CT (non-PET group). Univariate and multivariate analyses assessed the impact of the various factors on the incidence of benign postoperative diagnosis.

Results

No significant difference was found in benign postoperative diagnosis between the PET and non-PET groups [n=6 (2.9%) vs. n=10 (4.8%), P=0.31; risk difference (RD): −0.019, 95% confidence interval (CI): −0.060 to 0.022] In the univariate analysis, younger age [60.75 vs. 65.5 years, P=0.02] and absence of COPD (4.6% vs. 0%, P=0.046), significantly influenced the risk of the postoperative unsuspected benign diagnosis. The younger age emerged as the sole significant risk factor for VATS lobectomy for benign lesions [P=0.042, odds ratio (OR) =1.062; 95% CI: 1.002–1.127].

Conclusions

The use of preoperative PET-CT did not decrease the risk of performing VATS lobectomy on unsuspected benign tumors. That risk should be considered when qualifying for VATS lobectomy, especially in younger patients.

Highlight box.

Key findings

• Preoperative positron emission tomography-computed tomography (PET-CT) scans did not significantly reduce the rate of video-assisted thoracoscopic surgery (VATS) lobectomies performed on unsuspected benign lesions (2.9% in the PET group vs. 4.8% in the non-PET group).

• Younger age was identified as the only significant independent risk factor for undergoing surgery that resulted in a benign postoperative diagnosis.

What is known and what is new?

• PET-CT is a standard diagnostic tool for the metabolic characterization of pulmonary nodules; however, its specificity is often limited by inflammatory or infectious processes that mimic malignancy.

• Through a propensity score matching analysis of 864 patients, this study demonstrates that PET-CT does not provide a superior diagnostic safeguard against “benign” lobectomies. It highlights that the risk of unnecessary resection remains present regardless of metabolic imaging, particularly in younger cohorts and those without chronic obstructive pulmonary disease.

What is the implication, and what should change now?

• The utilization of preoperative PET-CT does not inherently reduce the risk of VATS lobectomy for unsuspected benign lesions. This diagnostic limitation must be explicitly accounted for during the surgical qualification process, as metabolic imaging cannot replace definitive histological or clinical confirmation.

• Emphasis should be placed on a comprehensive risk assessment by a multidisciplinary team when qualifying patients for anatomical resection. A more rigorous diagnostic pathway—potentially involving preoperative biopsy or short-term surveillance—should be prioritized, particularly in younger patients, to mitigate the risk of performing VATS lobectomy for benign disease.

Introduction

The diagnostics of solitary pulmonary nodules presents an actual challenge, particularly in distinguishing neoplastic from benign lesions without subjecting patients to unnecessary interventions and associated complications. Despite continuous advancements in invasive diagnostic techniques for lung cancer, obtaining a preoperative diagnosis remains challenging (1). This is especially true for small, centrally located nodules within the lung parenchyma and those near the hilum, which are inaccessible via bronchoscopy. In these cases, the likelihood of successful transthoracic biopsy or wedge resection with frozen section analysis during surgery to confirm malignancy is low. Consequently, pathological results are typically obtained only after anatomical resection (2).

Minimally invasive video-assisted thoracoscopic surgery (VATS) has proven to carry a lower risk of complications and mortality compared to thoracotomy and is now recommended for surgical resection of pulmonary nodules and early-stage lung cancer (3-6). The widespread adoption and perceived safety of VATS can sometimes lead to proceeding directly to surgery for a suspicious lung nodule without prior confirmation of malignancy. This approach carries the risk of performing a lobectomy on a patient with a benign lesion, with its associated complications (7-9).

Positron emission tomography-computed tomography (PET-CT) with 18F-fluorodeoxyglucose (FDG), which assesses the metabolic activity of lesions, plays a crucial role in staging and evaluating the malignant potential of lung tumors (10,11). This imaging modality has demonstrated high sensitivity and intermediate specificity for predicting malignancy and is considered a standard of care in the evaluation of suspicious lung nodules (12-14). While some studies have analyzed the spectrum, safety, and cost-effectiveness of lung resections resulting in a benign postoperative diagnosis (15-18), no study has specifically examined the impact of PET-CT on the incidence of benign disease in patients undergoing exclusively VATS lobectomies.

Most previously published analyses on benign findings after pulmonary resection combine both open and minimally invasive procedures, which makes it difficult to assess the diagnostic and decision-making accuracy in the era of minimally invasive thoracic surgery. Our study focuses exclusively on patients undergoing VATS lobectomy, reflecting contemporary surgical standards and allowing for a more precise evaluation of preoperative diagnostic pathways and the clinical utility of PET-CT in this specific population.

The primary aim of this study was to determine whether preoperative PET-CT significantly reduces the risk of “benign” VATS lobectomies performed in patients with lung nodules suspicious for cancer but without a preoperative diagnosis. The secondary aim was to identify potential risk factors associated with VATS lobectomy for benign lesions. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1393/rc).

Methods

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This retrospective, non-interventional study did not require approval by the Bioethics Committee of Poznan University of Medical Sciences, as under its regulations retrospective analyses based on anonymized medical records are exempt from ethical review. Given the fully retrospective design and use of anonymized data, the requirement for obtaining individual informed consent from patients was formally waived in accordance with the institutional and Bioethics Committee regulations.

This single-center, retrospective cohort study included patients who underwent VATS lobectomy for suspected primary lung cancer without a preoperative pathological diagnosis between January 1, 2014, and December 31, 2020. All patients underwent a standard diagnostic workup, including CT of the chest, bronchoscopy, and, when clinically indicated, invasive mediastinal staging and transthoracic needle biopsy. PET-CT was indicated for solitary pulmonary nodules >1 cm in diameter to differentiate benign from malignant etiology when other diagnostic methods were inconclusive. However, due to limited PET-CT availability and the insufficient capacity of diagnostic centers resulting in prolonged waiting times, some patients did not undergo this test preoperatively. Pulmonary function tests and diffusing capacity of the lungs for carbon monoxide (DLCO) were performed to assess performance status and perioperative risk. Patients deemed high-risk underwent exercise testing, echocardiography, and cardiology consultation. Treatment decisions were made by a multidisciplinary team.

Patients diagnosed with middle lobe syndrome or lung sequestration (n=7) and those with clinical stage T3 or T4 disease (n=131) were excluded from the study (Figure 1).

Figure 1.

Flow chart of study design including inclusion and exclusion criteria. CT, computed tomography; PET, positron emission tomography; TNM, tumor-node-metastasis; VATS, video-assisted thoracoscopic surgery.

The percentage of VATS lobectomies performed with preoperative PET-CT was calculated for each year of the study period. The patient cohort was divided into two groups: those who underwent preoperative PET-CT (PET group, n=654) and those who did not (non-PET group, n=210).

Statistical analysis

Categorical variables were analyzed using the Chi-squared test or Fisher-Freeman-Halton test. Continuous variables with a normal distribution and homogeneous variances were analyzed using the unpaired t-test. Normality was assessed using the Shapiro-Wilk test, and homogeneity of variance was assessed using Levene’s test. Non-normally distributed data were analyzed using the Mann-Whitney U test. All tests were two-sided, and a P value <0.05 was considered statistically significant. Propensity score matching was used to create comparable groups. The propensity score, representing the predicted probability of receiving preoperative PET-CT, was calculated using the following variables: age, sex, cardiovascular comorbidities, prior malignancy, chronic obstructive pulmonary disease (COPD), diabetes mellitus, and T stage. Nearest neighbor matching without replacement and a caliper width of 0.01 were employed.

The primary endpoint was the pathological diagnosis (benign or malignant). Secondary endpoints included FDG maximum standardized uptake value (SUVmax) uptake value and predictors of VATS lobectomy for benign lesions.

Logistic regression analysis was performed to identify factors associated with a benign diagnosis after surgery.

Data manipulation and statistical analysis were performed using IBM^® SPSS^® Statistics version 27 (PS Imago Pro 8).

Results

In the total cohort of patients without a preoperative diagnosis (n=864), 28 (3.2%) had a benign pathological diagnosis postoperatively: 18 (2.8%) in the PET group and 10 (4.8%) in the non-PET group. The percentage of VATS lobectomies performed with preoperative PET-CT during the study period is shown in Figure 2.

Figure 2.

VATS lobectomies with preoperative PET-CT in years 2014–2020. CT, computed tomography; PET, positron emission tomography; VATS, video-assisted thoracoscopic surgery.

Propensity score matching yielded two groups of 210 patients each: one with and one without preoperative PET-CT. The groups were similar with respect to age, sex, T stage, and comorbidities. Patient characteristics and results of the univariate analysis are presented in Table 1. There was no statistically significant difference in the frequency of benign postoperative diagnoses between the PET group (2.9%, n=6) and the non-PET group (4.8%, n=10). Univariate analysis identified younger age at surgery {60.75 years [interquartile range (IQR), 55.5–67.5 years] vs. 65.5 years (IQR, 61.0–70.0 years), P=0.02} and absence of COPD (4.6% vs. 0%, P=0.046) as significantly associated with a postoperative benign diagnosis. The pathological diagnoses of patients with benign disease are detailed in Table 2.

Table 1. Characteristics and basic comparison of the study groups with or without preoperative PET-CT examination after propensity score matching.

Feature	PET group (n=210)	Non-PET group (n=210)	P value	RD (95% CI)
Age (years)	65 [61–71]	65 [61–70]	0.8	–
Sex, male	118 (56.2)	116 (55.2)	0.84	0.010 (−0.109 to 0.090)
T stage			0.09	–
1a	10 (4.7)	7 (3.3)
1b	68 (32.4)	40 (19.0)
1c	52 (24.8)	40 (19.0)
2a	72 (34.3)	97 (46.2)
2b	8 (3.8)	26 (12.4)
Comorbidities
COPD	36 (17.1)	36 (17.1)	0.89	0 (−0.072 to 0.072)
CVD	24 (11.4)	24 (11.4)	0.87	0 (−0.061 to 0.061)
Previous history of cancer	16 (7.6)	15 (7.1)	0.85	0.005 (−0.060 to 0.050)
Diabetes mellitus	19 (9.0)	20 (9.5)	0.87	−0.005 (−0.056 to 0.065)
Final diagnosis			0.85	−0.019 (−0.060 to 0.022)
Malignant disease	204 (97.1)	200 (95.2)
Benign disease	6 (2.9)	10 (4.8)

Data are presented as median [interquartile range] or number (%). CI, confidence interval; COPD, chronic obstructive pulmonary disease; CT, computed tomography; CVD, cardiovascular disease; PET, positron emission tomography; RD, risk difference; T, tumor.

Table 2. Results of postoperative pathological examination of a group with benign lesions.

Diagnosis	PET-group (n=6)	No-PET group (n=10)
Inflammation	3 (50.0)	3 (30.0)
Granulomatous inflammation	1 (16.6)	3 (30.0)
Tuberculosis	1 (16.6)	1 (10.0)
Hamartoma	0	2 (20.0)
Aspergilloma	1 (16.6)	0
Abscess	0	1 (10.0)

Data are presented as n (%). PET, positron emission tomography.

Logistic regression analysis, including age, sex, PET-CT, cardiovascular disease, COPD, prior malignancy, diabetes, and T stage, identified younger age at surgery as the only statistically significant predictor of lobectomy for a benign lesion [P=0.042, odds ratio (OR) =1.062; 95% confidence interval (CI): 1.002–1.127]. Results of the multivariate analysis are shown in Table 3.

Table 3. Logistic regression results assess a given parameter’s impact on the risk of a malignant or benign lesion diagnosis.

Feature	Odds ratio (95% CI)	P value	P value adjusted to the T stage
Age	1.063 (1.002–1.127)	0.042*	0.044*
Sex	1.057 (0.377–2.962)	0.92	0.92
COPD	2.483 (0.210–15.084)	0.99	0.99
CVD	1.657 (0.382–14.876)	0.99	0.99
Previous history of cancer	1.483 (0.182–12.084)	0.71	0.99
Diabetes mellitus	1.937 (0.242–15.474)	0.53	0.71
PET-CT	1.784 (0.624–5.102)	0.28	0.28

*, statistically significant (P<0.05). CI, confidence interval; COPD, chronic obstructive pulmonary disease; CT, computed tomography; CVD, cardiovascular disease; PET, positron emission tomography; T, tumor.

In the PET group, all suspected lung nodules were metabolically active. The median FDG [SUVmax was 6.1 (IQR, 4.5–7.9) in the benign disease group and 7.95 (IQR, 4.7–12.0] in the lung cancer group. This difference was not statistically significant (P=0.26).

Discussion

The primary finding of this study is that preoperative PET-CT does not decrease the risk of lobectomy for benign lesions.

The increasing use of imaging modalities like CT scans and lung cancer screening programs has led to increased detection of pulmonary nodules and ground-glass opacities. Consequently, a greater number of patients require further diagnostic evaluation to differentiate benign from malignant lesions. This distinction remains a significant clinical challenge. Balancing the need for timely intervention for potentially malignant nodules with the avoidance of unnecessary procedures for benign lesions is a key concern. Adherence to established guidelines incorporating clinical factors, imaging characteristics, and validated risk prediction models is crucial (19). However, the feasibility of obtaining a preoperative tissue diagnosis and patient preferences also play a role (6,13,20). In cases of suspected lung nodules, the goal is to obtain a preoperative pathological diagnosis and accurate staging to guide treatment decisions.

While advancements in invasive diagnostic procedures, such as ultrasound or CT-guided transthoracic needle biopsy, endobronchial ultrasound-guided transbronchial needle aspiration, radial endobronchial ultrasound, bronchoscopic lung cryobiopsy and electromagnetic navigation bronchoscopy have improved diagnostic accuracy (1), each technique has limitations, varying false-negative rates, associated complications, and the potential to prolong the time to definitive treatment (16).

When a preoperative diagnosis cannot be obtained and malignancy cannot be confidently excluded, some authors suggest proceeding with surgical exploration to establish a diagnosis (2,16,17). Intraoperative frozen section analysis of a wedge resection specimen is often attempted. However, for centrally located tumors that are not amenable to wedge resection, a definitive diagnosis may only be possible after lobectomy (17).

In our study, all patients with a benign diagnosis had centrally located nodules within the lobar parenchyma or near the hilum, precluding wedge resection and frozen section analysis. Mori et al. demonstrated that standard lobectomy is an acceptable treatment option in carefully selected patients with nodules clinically diagnosed as malignant, as the accuracy of clinical diagnosis in these cases was high and comparable to frozen section biopsy (2). Similarly, several studies have shown the safety of proceeding directly to surgery without a prior tissue diagnosis, thereby reducing the time from initial presentation to definitive treatment (15-17). This is a critical consideration in lung nodule management, particularly in the context of high caseloads and wait times for lung cancer surgery. However, proceeding with surgery without a preoperative diagnosis or intraoperative frozen section carries the risk of resecting benign lesions, exposing patients to the risks of surgery and potential deterioration in pulmonary function. Therefore, in cases where histopathological confirmation through an open approach (thoracotomy) is technically feasible and considered essential for subsequent clinical decision-making—particularly in patients with substantial perioperative risk or reduced cardiopulmonary function—the risk–benefit balance should be thoroughly evaluated.

Over recent decades, VATS has become the preferred surgical approach for lung resections due to its lower complication rate, reduced mortality, and improved postoperative quality of life compared to open thoracotomy (3,6,13). In our study period [2014–2020], we observed a steady increase in the utilization of VATS lobectomy, reaching nearly 70% by 2020. The minimally invasive nature of VATS may lower the threshold for proceeding with surgery for undiagnosed lung nodules, potentially increasing the risk of resecting benign lesions (2,16,17).

Concurrently, the availability of PET-CT has increased substantially, exceeding 90% in recent years. PET-CT is now considered essential for assessing the malignant potential and staging of lung nodules (6,21). The increasing use of preoperative PET-CT prompted us to investigate whether this imaging modality reduces the rate of unintended VATS lobectomies for benign disease.

Prior studies examining the frequency of benign diagnoses after lung resection have often included heterogeneous groups of patients undergoing both thoracotomy and VATS, as well as wedge resections (2,7-9,15-17). To our knowledge, no prior study has specifically evaluated the impact of PET-CT on the incidence of benign diagnoses in patients undergoing exclusively VATS lobectomy.

In our study, 3.2% of patients undergoing VATS lobectomy had a benign pathological diagnosis, a rate consistent with previously reported data. After propensity score matching, the rate of benign diagnoses was 2.9% in the PET group and 4.8% in the non-PET group, a difference that was not statistically significant. Previous studies have reported benign pathology rates ranging from 2.5% to 15% in resected lung nodules suspicious for lung cancer (9,18). In published research, Sihoe et al. reported benign disease after surgery in 7.8%, similarly to Smith et al. (9%) (8,16). Similar to our findings, Kaaki et al. found no significant difference in benign pathology rates between patients undergoing attempted biopsy preoperatively (3.2%) and those proceeding directly to surgery without biopsy (7.0%), concluding that forgoing preoperative biopsy does not reduce the risk of benign resection (15,16).

PET-CT demonstrates excellent performance in differentiating benign from malignant solitary pulmonary nodules (14). Reported sensitivity ranges from 75% to 95%, and specificity from 70% to 90% (10-12,14). Feng et al. reported a diagnostic accuracy of 93.5% and a false-positive rate of 6.5% for PET-CT in the evaluation of lung nodules, with age, diabetes, interleukin-6 levels, and T-spot test results identified as risk factors for false-positive results (14). In our study, all benign lesions in the PET group demonstrated FDG uptake (3.2%), a rate consistent with, and not exceeding, the reported false-positive rate of PET-CT for lung cancer.

PET and PET-CT can detect 18F-FDG accumulation not only in malignant lesions but also in areas of infection, inflammation, and granulomatous diseases (11). In our study, all nodules in the PET group were metabolically active, with a median SUVmax of 6.1 (IQR, 4.5–7.9) in the benign disease group and 7.95 (IQR, 4.7–12.0) in the lung cancer group. Although this difference was not statistically significant (P=0.26), the substantial disparity in group sizes should be considered. While extensive literature exists regarding optimal SUVmax cutoff values for malignancy, this remains a subject of ongoing debate.

In our study, most patients with benign lesions had a diagnosis of granulomatous or inflammatory disease (tuberculosis, aspergilloma) (Table 2). Only two cases of hamartoma were observed, both in the non-PET group. These findings are consistent with prior reports documenting granulomatous inflammation, histoplasmosis, chronic abscess, lipoid pneumonia, hamartoma, fibrosis, and blastomycosis as causes of benign lung nodules (11,12). Previous studies have sought to identify risk factors for both malignant and benign lung nodules (6,13,20). Reported risk factors for benign disease include small size, homogeneous solid appearance, and smooth perifissural or subpleural location (6,13,20). In our study, logistic regression analysis, including age, sex, PET-CT status, cardiovascular disease, COPD, prior malignancy, diabetes, and T stage, identified younger age at surgery as the only statistically significant predictor of “benign lobectomy”. This finding is consistent with the established association between age and lung cancer risk, with age being a component of many lung nodule risk prediction models (6,13,17,20).

In univariate analysis, younger age and absence of COPD were associated with “benign lobectomy”. This aligns with existing literature identifying age and smoking (and consequently COPD) as independent risk factors for lung cancer (6,13,17,20). These findings suggest that greater caution is warranted when considering surgery for younger patients without COPD, and that efforts should be focused on obtaining a pathological diagnosis through invasive techniques or continued surveillance in this subgroup.

Advanced diagnostic techniques, such as virtual bronchoscopy, electromagnetic navigation systems, and robotic bronchoscopy, may further improve diagnostic accuracy. Currently, establishing a histopathological diagnosis is particularly important due to the introduction of preoperative and perioperative immunochemotherapy. While lobectomy was the standard of care for suspected lung cancer during our study period, segmentectomy is increasingly recognized as a safe alternative for small nodules when pathological confirmation is unavailable, providing a less invasive diagnostic approach (6,22,23). This may be particularly relevant in the management of undiagnosed, centrally located lung nodules.

This study has several limitations. The retrospective design, relying on a multi-year database, may reflect variations in data quality and completeness, including potential deficiencies in chart documentation and missing data. However, to create the most homogenous study group possible, we included only VATS lobectomies, performed propensity score matching, and excluded patients with T3 and T4 disease. While this selection criterion ensured a uniform cohort of individuals eligible for minimally invasive surgery, it may also limit the external validity of our findings for the entire surgical population. Therefore, our results should be interpreted as representative primarily for patients with early-stage (T1–T2) disease who are candidates for minimally invasive resection. Interestingly, no unintentional lobectomies for benign disease were observed in the excluded group of patients with T3 and T4 tumors (n=131). This may be due to several factors. Larger tumors are more likely to be diagnosed preoperatively via transthoracic biopsy or endobronchial techniques. Furthermore, tumor size itself is correlated with increased malignancy risk (6,20). It is therefore plausible that tumors >5 cm carry a low risk of benign pathology after lobectomy, justifying proceeding directly to resection in these cases.

It should also be emphasized that our analysis was limited to patients who underwent surgical resection; therefore, individuals with PET-negative nodules managed conservatively or under pulmonology follow-up were not included, as these data were not available in our institutional database.

The absence of complete SUV data for all lung cancer patients and information regarding intraoperative frozen sections limits our ability to make more robust comparisons. It is possible that some patients with low or negative PET-CT results underwent wedge resection with frozen section analysis, which revealed malignancy and prompted lobectomy. This could represent a false-negative PET-CT result, particularly in low cellular density or low-avidity lesions such as carcinoid tumors and adenocarcinoma in situ. Additionally, information on wedge resections performed for benign pathology would undoubtedly offer valuable additional context, but those verified intraoperatively as benign do not represent a diagnostic or decision-making challenge. Access to this data would have allowed for better study group selection and more comprehensive evaluation.

Conclusions

In conclusion, this study demonstrates that preoperative PET-CT does not decrease the risk of unintended VATS lobectomy for benign disease in patients with lung nodules suspicious for lung cancer. The low number of benign lesions in both groups reflects the expertise of the multidisciplinary team of thoracic surgeons and radiologists at this high-volume center. When considering lung resection without pathological confirmation, younger age should be considered a relative risk factor for benign disease. Given that positive PET-CT results can occur in inflammatory and granulomatous conditions, careful patient selection by a multidisciplinary team is essential.

Supplementary

The article’s supplementary files as

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This retrospective, non-interventional study did not require approval by the Bioethics Committee of Poznan University of Medical Sciences, as under its regulations retrospective analyses based on anonymized medical records are exempt from ethical review. Given the fully retrospective design and use of anonymized data, the requirement for obtaining individual informed consent from patients was formally waived in accordance with the institutional and Bioethics Committee regulations.

Data Sharing Statement

Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1393/dss

jtd-18-02-113-dss.pdf ^{(66.3KB, pdf)}

DOI: 10.21037/jtd-2025-1393