Dynamic fluorine-18 fluorodeoxyglucose PET for evaluating different-sized metastatic lymph nodes in patients with non-small cell lung cancers

Introduction

Lung cancer is the leading cause of cancer-related mortality worldwide, with over 85% of cases being non-small cell lung cancer (NSCLC) (1). For patients with locally advanced NSCLC, regional lymph node (LN) staging is the most important factor influencing therapeutic decisions, as it enables clinicians to distinguish between two critical groups of patients: one group of patients with stage I–II resectable tumors or with stage IIIA potentially resectable tumors and another group with unresectable stage III tumors. For patients with stage N0 disease, the 5-year survival rate can reach 56%, while for those with stage N3, this rate is only 6% (2). Therefore, accurate and detailed LN staging of NSCLC is essential to selecting the most suitable treatment and determining patient prognosis.

Computed tomography (CT) has long been used as the primary imaging method for LN staging. According to clinical guidelines, LNs with a short-axis diameter ≥1 cm are commonly considered metastatic (3). However, relying solely on this criterion provides limited sensitivity (55%) and specificity (81%) (4). The National Comprehensive Cancer Network (NCCN) recommends the use of fluorine-18 fluorodeoxyglucose (F-18 FDG) positron emission tomography-CT (PET/CT) for clinical staging (5). Using a cutoff of maximum standardized uptake value (SUVmax) ≥2.5, a meta-analysis found that FDG PET/CT had higher sensitivity (0.81; range, 0.70–0.89) and specificity (0.79; range, 0.70–0.87) for N staging in NSCLC compared to CT (6). However, it should be noted that false positives and false negatives can occur in certain conditions, such as in the setting of infectious diseases or with small LNs (7,8). In one study (9), the sensitivity of F-18 FDG PET/CT was significantly higher among enlarged (>1 cm) LNs than among non-enlarged (≤1 cm) LNs (74% vs. 40%), but the specificity and accuracy were significantly lower (81% vs. 98% and 78% vs. 90%, respectively; P<0.05). Furthermore, as a semiquantitative metabolic parameter, the SUV is influenced by various factors, including scan time and blood glucose levels. Therefore, there is a critical need to develop methods that can identify metastatic LNs in NSCLC, especially those of different sizes.

Dynamic chest F-18 FDG PET/CT captures the spatial and temporal distribution of F-18 FDG from injection, providing real-time insights into blood flow, perfusion, and metabolic activity, such as the net influx rate (Ki), tumor blood flow, tumor blood efflux rate, and phosphorylation rate, offering a more accurate assessment of changes in tumor metabolism (10,11). Our previous studies examined the diagnostic value of dynamic chest F-18 FDG PET/CT in LN staging (12,13). The results indicated that dynamic metabolic parameters, especially Ki at a cutoff value of 0.022 mL/g/min, were more effective than SUVmax in distinguishing metastatic from nonmetastatic LNs, with a specificity of 0.918 and an area under the curve (AUC) of 0.672 (12). Additionally, the combination of SUVmax and Ki yielded a higher sensitivity (0.843), specificity (0.946), and AUC (0.907) (13). Nevertheless, there were certain limitations to this work, including the relatively modest sample size and the absence of size stratification of the LNs.

Thus far, no studies have specifically examined the performance of dynamic F-18 FDG PET in distinguishing metastatic from nonmetastatic LNs of different sizes. It is particularly unclear whether Ki outperforms other parameters in diagnosing non-enlarged LNs. Given the critical role of accurate LN staging in lung cancer treatment and prognosis, this study aimed to evaluate the ability of dynamic F-18 FDG PET metabolic parameters to diagnose different-sized LNs in patients with NSCLC. We present this article in accordance with the STARD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1887/rc).

Methods

Patients

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments and was approved by the Institutional Ethics Committee of The Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences (No. KYKT2024-19-1). The requirement for informed consent was waived, as this study retrospectively analyzed pre-existing de-identified data. The data from a total of 366 consecutive patients who underwent a dynamic chest F-18 FDG PET/CT scan (65 minutes) and a static whole-body F-18 FDG PET/CT scan (10–20 minutes) between May 2021 and March 2025 were retrospectively collected. All patients had lung nodules or masses identified on chest CT scans and had undergone endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) after F-18 FDG PET/CT scans. Meanwhile, none of the patients had received anti-inflammatory or antitumor treatments. LNs with SUVmax ≥2.5 on PET/CT scan and suspicious malignant EBUS sonographic features (i.e., size >1 cm, round shape, distinct margin, absence of central hilar structure, and heterogeneous echogenicity) were punctured. The inclusion criteria were as follows: (I) pathologic diagnosis of NSCLC; (II) LNs confirmed by biopsy/surgical pathology; and (III) a time interval of 2 weeks or less between the PET/CT scan and pathology. Meanwhile, the exclusion criteria were as follows: (I) dynamic chest F-18 FDG PET/CT scan failure; (II) LNs with an SUVmax <2.5 or no F-18 FDG uptake; and (III) inability to delineate LNs with volume of interest (VOI) software via semiautomatic and manual segmentation. Ultimately, the study included 98 patients comprising 369 LNs, of which 172 were metastatic and 197 were nonmetastatic. The flowchart of enrollment for patients with NSCLC and pathologically confirmed LNs is provided in Figure 1.

Figure 1 Flowchart for the enrollment of patients with NSCLC and pathologically confirmed LNs. F-18 FDG, fluorine-18 fluorodeoxyglucose; LN, lymph node; NSCLC, non-small cell lung cancer; PET/CT, positron emission tomography-computed tomography; SCLC, small cell lung cancer; SUVmax, maximum standardized uptake value; VOI, volume of interest.

PET/CT scanning and image reconstruction

All patients fasted for at least 6 hours before being PET/CT scanned on a Discovery MI PET/CT device (GE HealthCare, Chicago, IL, USA) and avoided strenuous exercise within 24 hours prior to imaging. Blood glucose levels were maintained below 8.0 mmol/L. First, breath-holding chest CT and whole-body CT scans (from the head to mid-thigh in the supine position with arms raised) were performed under the following parameters: tube voltage, 120 kV; tube current, 10–220 mA; pitch, 1.375:1; and noise index, 20. After intravenous injection of F-18 FDG (281.3±42.1 MBq), the dynamic chest PET scan was immediately initiated. A total dynamic scan duration of 65 minutes was divided into 28 frames: 6 frames × 10 seconds, 4 frames × 30 seconds, 4 frames × 60 seconds, 4 frames × 120 seconds, and 10 frames × 300 seconds. Following the dynamic scan, a static whole-body PET scan (from the head to mid-thigh in the supine position with arms raised) was performed at 1.5 min/bed. Attenuation correction was performed based on CT data, and reconstruction was completed via the block-sequential regularized expectation maximization (BSREM) algorithm with 25 iterations, 2 subsets, and a matrix size of 256×256.

PET/CT data analysis

The dynamic metabolic parameter, Ki, was obtained with a two-tissue irreversible compartment model. The image-derived input function (IDIF) was extracted from the ascending aorta in the early time frames (0–60 seconds), with a 10-mm diameter region of interest (ROI) placed on six consecutive slices. Due to motion and partial volume effects, the effect from the left ventricle was less prominent compared to the ascending aorta. Blood and plasma uptake differences were not considered in this study. The model assumed unidirectional uptake of F-18 FDG (i.e., k4=0) and its irreversible conversion to F-18 fluorodeoxyglucose-6-phosphate. Voxel-based analysis generated parameter images for each dynamic scan. To address linearization issues, the Lawson-Hanson nonnegative least squares algorithm was applied instead of conventional nonlinear methods. The VOI was delineated semiautomatically with a 40% SUVmax threshold in ITK-SNAP software version 4.9, as shown in Figure 2A. Two nuclear medicine physicians, with over 10 years of experience, manually delineated the three-dimensional VOIs slice by slice for lesions with failed semiautomated segmentation. The segmented VOIs were then applied to the Ki parameter images to extract quantitative Ki values.

Figure 2 The flowchart for the automatic drawing of the VOI for Ki images with ITK-SNAP software (version 4.9) (A) and an example image of VOI (red circle) drawn to determine the SUVmax of the mediastinal blood pool (B). 3D, three dimensional; Ki, net influx rate; PET, positron emission tomography; PET/CT, positron emission tomography-computed tomography; SUV, standardized uptake value; SUVmax, maximum standardized uptake value; VOI, volume of interest.

Static images were independently reviewed by two experienced nuclear medicine physicians with over 10 years of experience. LN locations were classified according to the International Association for the Study of Lung Cancer (IASLC) LN map and divided into mediastinal (stations 1–9) and hilar (stations 10–12) regions. N stage was assigned according to the eighth edition of the TNM classification for lung cancer. The SUVmax of LNs was calculated through placement of a spheroid-shaped VOI encompassing the entire LN, and the mediastinal blood pool SUVmax was measured through placement of a VOI in the lumen of the aortic arch, as shown in Figure 2B. The longest diameter, shortest diameter, and density of LNs were measured on the maximum cross-sectional slice of LNs displayed on axial breath-holding chest CT imaging. Muscle density was also measured at the same level. The ratio of LN density to muscle density (DR), ratio of LN SUVmax to mediastinal blood pool SUVmax (SUVR), and ratio of LN long diameter to short diameter (L/S) were recorded.

Statistical analysis

Quantitative parameters of metastatic and nonmetastatic LNs within different groups were compared via the Wilcoxon rank-sum test or independent samples t-test. Receiver operating characteristic (ROC) curves for each parameter were constructed, and the AUC was calculated, with histopathologic diagnosis being used as the gold standard. Differences in AUC, sensitivity and specificity were assessed via the DeLong test and χ² test, respectively. A P value of <0.05 was considered statistically significant. All statistical analyses were performed with MedCalc version 22.018 (MedCalc Software Ltd., Ostend, Belgium).

Results

Patient and LN characteristics

The characteristics of the patients and LNs are summarized in Table 1. Among the 98 patients who underwent dynamic chest PET/CT imaging, the mean age was 61.55±10.76 years (range, 45–85 years), and there were 66 males and 32 females. Regarding the histological type of the primary tumors, 71 (72.45%) were adenocarcinomas, 19 (19.39%) were squamous cell carcinomas, and 8 (8.16%) were other primary lung malignant tumors. Moreover, 58 patients underwent EBUS-TBNA alone, and 203 LNs (160 metastatic and 43 nonmetastatic) were included. Each patient had 2–4 biopsied LNs, with a mean of 3.5 LNs (3 patients with 2 LNs, 23 patients with 3 LNs, and 32 patients with 4 LNs). Additionally, 40 patients underwent surgery (39 patients underwent lobectomy, 1 patient underwent wedge resection, 3 patients underwent mediastinal LN sampling, 1 patient underwent lobectomy-specific LN dissection, and 36 patients underwent systematic mediastinal LN dissection), and a total of 166 LNs with an SUVmax ≥2.5 were included (12 metastatic and 154 nonmetastatic).

Of the 369 pathologically confirmed LNs, 197 were nonmetastatic (160 from adenocarcinoma, 34 from squamous cell carcinoma, and 3 from other primary lung malignant tumors) and 172 were metastatic (131 from adenocarcinoma, 22 from squamous cell carcinoma, and 19 from other primary lung malignant tumors). The LNs were grouped according to size: 158 had a maximum short diameter (D_max) ≥1 cm (including 127 metastatic and 31 nonmetastatic), and 211 had a D_max <1 cm (including 45 metastatic and 166 nonmetastatic).

Analysis of LN features

Regarding the CT features, the mean D_max of metastatic LNs was 1.30±0.59 cm, while that of nonmetastatic LNs was 0.70±0.22 cm. The L/S of metastatic LNs was 0.67±0.18, while that of nonmetastatic LNs was 1.54±0.31. The D_max of metastatic LNs was significantly larger than that of nonmetastatic LNs (P<0.05; Table 2), while the L/S was significantly smaller (P<0.05; Table 2). For metastatic LNs and nonmetastatic LNs the density was 37.39±9.78 and 43.21±11.87 HU, respectively, while the DR was 0.67±0.18 and 0.71±0.19, respectively, with both parameters being significantly different (P<0.05; Table 2).

Among the PET features, static metabolic parameters, including SUVmax (13.12±6.74 vs. 5.23±2.42) and SUVR (7.02±3.65 vs. 2.84±1.34), were significantly different between the metastatic and nonmetastatic LNs (P<0.05; Table 2). The dynamic metabolic parameter Ki differed significantly between metastatic and nonmetastatic LNs (0.039±0.025 vs. 0.013±0.006 mL/g/min; P<0.05; Table 2).

ROC curve analysis was further performed, the results of which are shown in Table 3. The optimal cutoff value for D_max was 1.00 cm, with an AUC of 0.852, a sensitivity of 74.40%, and a specificity of 84.30%; for SUVmax, the optimal cutoff value was 7.65, with an AUC of 0.894, a sensitivity of 79.70%, and a specificity of 87.80%; for SUVR, the optimal cutoff value was 4.54, with an AUC of 0.888, a sensitivity of 76.70%, and a specificity of 92.30%; for Ki, the optimal cutoff value was 0.018 mL/g/min, with an AUC of 0.936, a sensitivity of 86.00%, and a specificity of 84.70%. The AUC of the other CT features, including L/S, density and DR, were <0.50. The features with an AUC >0.8 were subjected to the DeLong test, which showed that Ki had a higher diagnostic efficacy than did SUVmax, SUVR, and D_max (P<0.05; Table 3 and Figure 3). No statistical differences were found between SUVmax, SUVR, and D_max (P>0.05; Table 3 and Figure 3).

Figure 3 The ROC curves of each metabolic parameter for the differential diagnosis of nonmetastatic and metastatic LNs in the overall LN group (A), the D_max ≥1 cm group (B), and the D_max <1 cm group (C). D_max, maximum short diameter; Ki, net influx rate; LN, lymph node; ROC, receiver operating characteristic; SUVmax, maximum standardized uptake value; SUVR, ratio of lymph node maximum standardized uptake value to mediastinal blood pool maximum standardized uptake value.

Comparison of characteristics between subgroups of LNs with different Dmax values

Considering that the D_max of LNs has often been used as a criterion for determining malignancy on CT imaging in clinical work and that the cutoff value (D_max ≥1 cm) was consistent with the results obtained in our study, we further investigated the differences between metastatic and nonmetastatic LNs in the groups with D_max ≥1 and <1 cm.

The D_max ≥1 cm group comprised 158 LNs (127 metastatic and 31 nonmetastatic). In the subgroup of metastatic LNs, the SUVmax, SUVR, and Ki were 14.59±7.15, 7.79±3.86, and 0.044±0.027 mL/g/min, respectively, which were all significantly higher than those in the nonmetastatic LNs group (6.24±2.66, 3.41±1.69, and 0.017±0.008 mL/g/min, respectively) (all P values <0.05; Table 2). The ROC curves were plotted in 158 LNs to determine the diagnostic accuracy of SUVmax, SUVR, and Ki in differentiating between metastatic and nonmetastatic LNs. The optimal cutoff value of SUVmax was 10.45, with an AUC of 0.884, a sensitivity of 69.30%, and a specificity of 96.80%. The optimal cutoff value of SUVR was 4.54, with an AUC of 0.878, a sensitivity of 81.90%, and a specificity of 90.30%. The optimal cutoff value of Ki was 0.026 mL/g/min, with an AUC of 0.910, a sensitivity of 81.10%, and a specificity of 87.10%. According to the results of the Delong test, no significant difference in the diagnostic accuracy was found between Ki and SUVmax, Ki and SUVR, or between SUVmax and SUVR (P>0.05; Table 3).

The D_max <1 cm group comprised 211 LNs (45 metastatic and 166 nonmetastatic). In the subgroup of metastatic LNs, the SUVmax, SUVR, and Ki were 9.14±3.87, 4.96±2.05, and 0.026±0.011 mL/g/min, respectively, which were all significantly higher than those in the nonmetastatic LNs group (5.11±2.41, 2.77±1.28, and 0.012±0.006 mL/g/min, respectively) (all P values <0.05; Table 2). The ROC curve analysis revealed that the optimal cutoff value for SUVmax was 7.65, with an AUC of 0.836, a sensitivity of 66.70%, and a specificity of 89.90%. The optimal cutoff value of SUVR was 3.69, with an AUC of 0.813, a sensitivity of 73.30%, and a specificity of 83.10%. The optimal cutoff value of Ki was 0.015 mL/g/min, with an AUC of 0.894, a sensitivity of 91.11%, and a specificity of 74.70%. The Delong test indicated that Ki had a higher diagnostic efficacy than did SUVmax and SUVR (P<0.05; Table 3), while no significant difference was found between SUVmax and SUVR (P>0.05; Table 3).

Furthermore, among metastatic LNs and nonmetastatic LNs, the SUVmax, SUVR, and Ki were higher in LNs with D_max ≥1 cm than in those with D_max <1 cm (P<0.05).

Comparison of diagnostic sensitivity and specificity of Ki between LNs with different Dmax values at varying cutoff values

As shown in Table 4, in the D_max ≥1 cm group, the corresponding diagnostic sensitivity and specificity were 91.34% and 70.97%, respectively, when the Ki cutoff value was 0.018 mL/g/min. Compared with a Ki cutoff value of 0.026 mL/g/min, the cutoff of 0.018 yielded a significantly lower specificity (P<0.05; Table 4 and Figure 4). However, while diagnostic sensitivity remained relatively high, the difference was not statistically significant (P>0.05; Table 4). In the D_max <1 cm group, when the Ki cutoff value was 0.018 mL/g/min, the corresponding diagnostic sensitivity and specificity were 71.10% and 86.75%, respectively. Compared with a Ki cutoff value of 0.015 mL/g/min, the cutoff of 0.018 mL/g/min yielded a significantly higher specificity and a significantly lower sensitivity (P<0.05; Table 4 and Figure 5).

Figure 4 PET/CT images of nonmetastatic F-18 FDG-avid LNs. A 69-year-old male patient who had surgical pathology confirmed lung adenocarcinoma in the upper lobe of the right lung (A; PET/CT fusion image, black arrow). The PET/CT scan showed F-18 FDG-avid LNs in zone 4R of the mediastinal region, with a maximum short diameter of 1.10 cm (C; CT image, white arrow), an SUVmax of 11.10 (B; PET/CT fusion image, white arrow), and a Ki of 0.017 mL/g/min (D; Ki image, white arrow), which surgical pathology confirmed to be nonmetastatic. F-18 FDG, fluorine-18 fluorodeoxyglucose; Ki, net influx rate; LN, lymph node; PET/CT, positron emission tomography-computed tomography; R, right; SUVmax, maximum standardized uptake value.

Figure 5 PET/CT images of metastatic F-18 FDG-avid LNs. A 41-year-old female patient who had biopsy pathology-confirmed lung adenocarcinoma in the upper lobe of the right lung (A; PET/CT fusion image, black arrow). PET/CT scanning showed a F-18 FDG-avid LN in zone 4R of the mediastinal region, with a maximum short diameter of 0.80 cm (C; CT image, white arrow), an SUVmax of 4.40 (B; PET/CT fusion image, white arrow), and a Ki of 0.017 mL/g/min (D; Ki image, white arrow), which EBUS-TBNA pathology confirmed to be metastatic. EBUS-TBNA, endobronchial ultrasound-guided transbronchial needle aspiration; F-18 FDG, fluorine-18 fluorodeoxyglucose; Ki, net influx rate; LN, lymph node; PET/CT, positron emission tomography-computed tomography; R, right; SUVmax, maximum standardized uptake value.

Discussion

In this study on patients with NSCLC, we found that CT features (including D_max, L/S, density, and DR) and PET features (SUVmax, SUVR, and Ki) were significantly different between metastatic and nonmetastatic LNs. Among these variables, D_max, SUVmax, SUVR, and Ki demonstrated good diagnostic performance, while Ki had superior diagnostic performance compared to SUVmax and SUVR among all LNs and LNs with D_max <1 cm. In addition, by grouping different-sized LNs, the respective optimal cutoff values were determined.

In chest CT, the traditional diagnosis criterion of D_max ≥1 cm was widely used in imaging evaluation (14). However, this criterion yields variable diagnostic results, with the sensitivity ranging from 41% to 67% and a specificity ranging from 79% and 86% (15,16). Moreover, a meta-analysis confirmed that its diagnostic performance is suboptimal, with sensitivity and specificity values of 0.590 and 0.780, respectively (3). In our study, the D_max ≥1 cm yielded an AUC of 0.852, a sensitivity of 0.744, and a specificity of 0.843, indicating inadequate diagnostic performance, particularly in terms of sensitivity. It has been reported that LN metastases from adenocarcinoma often occur in smaller LNs (17,18). In our study, adenocarcinoma accounted for 72.45% of cases, which is higher than that indicated in other work (18-21), which may explain the poor sensitivity. Furthermore, with the growing prevalence of lung cancer screening, early-stage lung cancer is becoming more common, which could result in microscopic rather than bulky metastases. Compared to a previous study (22), our study included a higher sensitivity than that of the traditional criterion of D_max ≥1 cm. It is possible that the pathological method for diagnosing LNs contributed to the inconsistent results, as in our study, 55.01% LNs were confirmed by EBUS-TBNA. EBUS-TBNA may not only obtain the ultrasound image features of LNs for differential diagnosis but may also allow for selection of larger LNs for puncture and is needed for any patient with LNs greater than 1 cm in the short axis (22). In addition to the size of the LNs, the shape and density of the LNs are indicators that can be referred to. However, for small metastatic LNs, the shape and density of the LN observed by CT may appear normal, while these structural changes may be observed in benign lesions such as inflammation, granulomas, and reactive hyperplasia of the LNs, making them less sensitive and specific. The results of our study may also be explained by the fact that density, DR, and L/S were lower in metastatic LNs than in nonmetastatic LNs, but the AUCs were all lower than 0.5.

SUVmax is the most commonly used PET/CT parameter and is widely applied for diagnosis and follow-up. According to previous studies, it has a higher diagnostic accuracy compared to the traditional short-axis diameter criterion (3), but false-positive or false-negative results are common (7,8). In our study, we found that the SUVmax and SUVR of metastatic LNs were significantly higher than those of nonmetastatic LNs, including in the D_max ≥1 and <1 cm groups; moreover, SUVmax had higher sensitivity and specificity in distinguishing LNs in the overall group, which is in line with the work by Yang et al. (23). For D_max ≥1 and <1 cm, the AUC and optimal cutoff value of SUVmax in LNs differed, but both had high diagnostic specificity (96.80% vs. 89.80%) and relatively low diagnostic sensitivity (69.30% vs. 66.70%). This is inconsistent with the findings of Al-Sarraf et al. (9), who found that SUVmax on PET/CT at a cutoff value of >2.5 had a significantly higher sensitivity for enlarged (>1 cm) than for non-enlarged (≤1 cm) LNs (74% vs. 40%) but had a significantly lower specificity (81% vs. 98%) and a lower accuracy (78% vs. 90%). This may be attributed to the different SUVmax diagnostic criteria and the number and proportion of enlarged and non-enlarged LNs. However, a meta-analysis reported that SUVmax has high specificity and relatively poor sensitivity for distinguishing mediastinal LNs metastases in NSCLC (pooled sensitivity: 0.610, 95% confidence interval: 0.582–0.636; pooled specificity: 0.924, 95% confidence interval: 0.918–0.930) (24), which is highly similar with our findings for LNs with D_max ≥1 and <1 cm. Similar to SUVmax, SUVR also demonstrated a high diagnostic performance in our study, yielding superior results to those reported by Lee et al. (25).

Compared with static parameters, including SUVmax and SUVR, dynamic PET/CT metabolic parameters (such as Ki) have demonstrated superior performance in differential diagnosis and treatment evaluation (26-29). Our team’s previous studies have also shown that Ki, at a cutoff value of 0.022 mL/g/min, can more effectively differentiate metastatic from nonmetastatic F-18 FDG-avid LNs compared to static parameters (12). Additionally, combining SUVmax with Ki showed promising diagnostic performance (13). As mentioned above, the diagnostic efficacy and optimal cutoff value of SUVmax varied in LNs of different D_max values, but whether the results of Ki vary remains to be confirmed further. In our study, we increased the sample size to three times that of the previous study and grouped according to different D_max. Similar to previous work (12,13), the Ki of metastatic LNs was significantly higher than that of nonmetastatic LNs. The ROC analysis indicated that Ki had excellent diagnostic performance, with an AUC of 0.936 at an optimal cutoff value of 0.018 mL/g/min, which differs from the findings reported by Wumener et al. (with an AUC of 0.895 at an optimal cutoff value of 0.022 mL/g/min) (13).

In the LNs with D_max ≥1 and <1 cm, the Ki of metastatic LNs were both significantly higher than those of nonmetastatic LNs. For the D_max ≥1 cm group, the diagnostic performance of the Ki was excellent (AUC =0.910) and comparable to that of SUVmax (AUC =0.884). Compared to the cutoff value of 0.018 mL/g/min, the cutoff value of 0.026 mL/g/min improved the specificity in the diagnosis of enlarged LNs (87.10% vs. 70.97%) but yielded low sensitivity (81.10% vs. 91.34%). In general, EBUS-TBNA staging is needed for any patient with LNs with D_max ≥1 cm on CT or PET-positive mediastinal LNs. It is particularly important to improve the diagnostic specificity for metastatic LNs in order to avoid unnecessary invasive EBUS-TBNA. Therefore, for the LNs with D_max ≥1 cm, the cutoff value of Ki should be 0.026, which may help to prevent overdiagnosis and better meet current clinical needs.

In the D_max <1 cm group, Ki had the same excellent diagnostic performance (AUC =0.894) and outperformed SUVmax (AUC =0.836). Furthermore, at the optimal cutoff value of 0.015, Ki yielded considerably good sensitivity and specificity (91.11% and 74.70%, respectively). This suggests that Ki reduces the false-negative rate in diagnosing small LNs. The relatively poor diagnostic performance of SUVmax in small LNs may be due to the lower metabolic activity or shorter F-18 FDG uptake time of small LNs, which results in greater fluctuations in SUVmax values, or due to a greater susceptibility to partial-volume effects, which affects its accuracy (30). Compared to the cutoff value of 0.018 mL/g/min, the cutoff value of 0.015 mL/g/min improved the sensitivity in the diagnosis of small LNs (91.10% vs. 71.10%); however, it had low specificity (74.70% vs. 86.75%). Early detection and diagnosis of small metastatic LNs could better meet clinical needs. Therefore, for the LNs with D_max <1 cm, the cutoff value of Ki was set to 0.015, which is more clinically valuable.

Several limitations to this study should be noted. First, we employed a single-center design with a limited sample size, which could have affected the generalizability of the results. Future studies should consider the integration of data from multiple centers to verify the stability and reliability of the findings. Second, the effect of motion correction on dynamic parameters was not considered, but chest motion may impact the quality of F-18 FDG PET/CT images and quantitative results (31-33). Future research could explore how to combine motion correction techniques to further improve the accuracy of dynamic metabolic parameters. Finally, we used SUVmax instead of mean SUV due to the lesser influence of partial volume effects on SUVmax; however, it should be noted that SUVmax may be affected by boundary effects. Despite these limitations, this study, with its expanded sample size, further validates the clinical value of dynamic chest F-18 FDG PET/CT in diagnosing LN metastasis, particularly in small LNs. Our findings provide valuable reference information for clinicians in the early diagnosis of LN metastasis in patients with NSCLC.

Conclusions

Compared with static PET features and CT features, the dynamic PET parameter Ki demonstrated superior diagnostic performance in differentiating metastatic from nonmetastatic LNs, particularly for small LNs. Through the optimization of cutoff values, the clinical applicability of the dynamic parameter, Ki, could be improved.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1887/rc

Data Sharing Statement: Available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1887/dss

Funding: This work was supported by Shenzhen High-level Hospital Construction Fund, and National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Shenzhen (grant No. E010324031), Shenzhen Municipal Basic Research Program (Natural Science Foundation) (No. 20220525171717003), and the Shenzhen Science and Technology Program of China (No. JCYJ20220818101804009).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-2025-1887/coif). The authors have no conflicts of interest to declare.

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. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Institutional Ethics Committee of Cancer Hospital & Shenzhen Hospital, Chinese Academy of Medical Sciences (No. KYKT2024-19-1) and the informed consent was waived due to it was a retrospective analysis of pre-existing de-identified data.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer statistics, 2022. CA Cancer J Clin 2022;72:7-33. [Crossref] [PubMed]
2. Woodard GA, Jones KD, Jablons DM. Lung Cancer Staging and Prognosis. Cancer Treat Res 2016;170:47-75. [Crossref] [PubMed]
3. Birim O, Kappetein AP, Stijnen T, Bogers AJ. Meta-analysis of positron emission tomographic and computed tomographic imaging in detecting mediastinal lymph node metastases in nonsmall cell lung cancer. Ann Thorac Surg 2005;79:375-82. [Crossref] [PubMed]
4. Silvestri GA, Gonzalez AV, Jantz MA, Margolis ML, Gould MK, Tanoue LT, Harris LJ, Detterbeck FC. Methods for staging non-small cell lung cancer: Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest 2013;143:e211S-50S.
5. Ettinger DS, Wood DE, Aisner DL, Akerley W, Bauman JR, Bharat A, et al. NCCN Guidelines Insights: Non-Small Cell Lung Cancer, Version 2.2021. J Natl Compr Canc Netw 2021;19:254-66. [Crossref] [PubMed]
6. Schmidt-Hansen M, Baldwin DR, Zamora J. FDG-PET/CT imaging for mediastinal staging in patients with potentially resectable non-small cell lung cancer. JAMA 2015;313:1465-6. [Crossref] [PubMed]
7. Deppen S, Putnam JB Jr, Andrade G, Speroff T, Nesbitt JC, Lambright ES, Massion PP, Walker R, Grogan EL. Accuracy of FDG-PET to diagnose lung cancer in a region of endemic granulomatous disease. Ann Thorac Surg 2011;92:428-32; discussion 433. [Crossref] [PubMed]
8. Metser U, Even-Sapir E. Increased (18)F-fluorodeoxyglucose uptake in benign, nonphysiologic lesions found on whole-body positron emission tomography/computed tomography (PET/CT): accumulated data from four years of experience with PET/CT. Semin Nucl Med 2007;37:206-22. [Crossref] [PubMed]
9. Al-Sarraf N, Gately K, Lucey J, Wilson L, McGovern E, Young V. Lymph node staging by means of positron emission tomography is less accurate in non-small cell lung cancer patients with enlarged lymph nodes: analysis of 1,145 lymph nodes. Lung Cancer 2008;60:62-8. [Crossref] [PubMed]
10. Rahmim A, Lodge MA, Karakatsanis NA, Panin VY, Zhou Y, McMillan A, Cho S, Zaidi H, Casey ME, Wahl RL. Dynamic whole-body PET imaging: principles, potentials and applications. Eur J Nucl Med Mol Imaging 2019;46:501-18. [Crossref] [PubMed]
11. Meijer TWH, de Geus-Oei LF, Visser EP, Oyen WJG, Looijen-Salamon MG, Visvikis D, Verhagen AFTM, Bussink J, Vriens D. Tumor Delineation and Quantitative Assessment of Glucose Metabolic Rate within Histologic Subtypes of Non-Small Cell Lung Cancer by Using Dynamic 18F Fluorodeoxyglucose PET. Radiology 2017;283:547-59. [Crossref] [PubMed]
12. Wumener X, Zhang Y, Wang Z, Zhang M, Zang Z, Huang B, Liu M, Huang S, Huang Y, Wang P, Liang Y, Sun T. Dynamic FDGPET imaging for differentiating metastatic from non-metastatic LNs of lung cancer. Front Oncol 2022;10:1005924. [Crossref] [PubMed]
13. Wumener X, Zhang Y, Zang Z, Ye X, Zhao J, Zhao J, Liang Y. The value of net influx constant based on FDG PET/CT dynamic imaging in the differential diagnosis of metastatic from non-metastatic lymph nodes in lung cancer. Ann Nucl Med 2024;38:904-12. [Crossref] [PubMed]
14. Gross BH, Glazer GM, Orringer MB, Spizarny DL, Flint A. Bronchogenic carcinoma metastatic to normal-sized lymph nodes: frequency and significance. Radiology 1988;166:71-4. [Crossref] [PubMed]
15. McLoud TC, Bourgouin PM, Greenberg RW, Kosiuk JP, Templeton PA, Shepard JA, Moore EH, Wain JC, Mathisen DJ, Grillo HC. Bronchogenic carcinoma: analysis of staging in the mediastinum with CT by correlative lymph node mapping and sampling. Radiology 1992;182:319-23. [Crossref] [PubMed]
16. Beadsmoore CJ, Screaton NJ. Classification, staging and prognosis of lung cancer. Eur J Radiol 2003;45:8-17. [Crossref] [PubMed]
17. Takahashi Y, Takashima S, Hakucho T, Miyake C, Morimoto D, Jiang BH, Numasaki H, Tomita Y, Nakanishi K, Higashiyama M. Diagnosis of regional node metastases in lung cancer with computer-aided 3D measurement of the volume and CT-attenuation values of lymph nodes. Acad Radiol 2013;20:740-5. [Crossref] [PubMed]
18. Funakoshi Y, Maeda H, Takeda S, Nojiri T, Kawamura T. Tumor histology affects the accuracy of clinical evaluative staging in primary lung cancer. Lung Cancer 2010;70:195-9. [Crossref] [PubMed]
19. Arita T, Kuramitsu T, Kawamura M, Matsumoto T, Matsunaga N, Sugi K, Esato K. Bronchogenic carcinoma: incidence of metastases to normal sized lymph nodes. Thorax 1995;50:1267-9. [Crossref] [PubMed]
20. Suzuki K, Nagai K, Yoshida J, Nishimura M, Takahashi K, Nishiwaki Y. Clinical predictors of N2 disease in the setting of a negative computed tomographic scan in patients with lung cancer. J Thorac Cardiovasc Surg 1999;117:593-8. [Crossref] [PubMed]
21. Kerr KM, Lamb D, Wathen CG, Walker WS, Douglas NJ. Pathological assessment of mediastinal LNs in Lung cancer: implications for non-invasive mediastinal staging. Thorax 1992;47:337-41. [Crossref] [PubMed]
22. Fielding DI, Kurimoto N. EBUS-TBNA/staging of lung cancer. Clin Chest Med 2013;34:385-94. [Crossref] [PubMed]
23. Yang W, Fu Z, Yu J, Yuan S, Zhang B, Li D, Xing L, Zhao D, Mu D, Sun X, Fang Y, Huang Y, Li W. Value of PET/CT versus enhanced CT for locoregional lymph nodes in non-small cell lung cancer. Lung Cancer 2008;61:35-43. [Crossref] [PubMed]
24. Zhao L, He ZY, Zhong XN, Cui ML. (18)FDG-PET/CT for detection of mediastinal nodal metastasis in non-small cell lung cancer: a meta-analysis. Surg Oncol 2012;21:230-6. [Crossref] [PubMed]
25. Lee AY, Choi SJ, Jung KP, Park JS, Lee SM, Bae SK. Characteristics of Metastatic Mediastinal Lymph Nodes of Non-Small Cell Lung Cancer on Preoperative F-18 FDG PET/CT. Nucl Med Mol Imaging 2014;48:41-6. [Crossref] [PubMed]
26. Dunnwald LK, Doot RK, Specht JM, Gralow JR, Ellis GK, Livingston RB, Linden HM, Gadi VK, Kurland BF, Schubert EK, Muzi M, Mankoff DA. PET tumor metabolism in locally advanced breast cancer patients undergoing neoadjuvant chemotherapy: value of static versus kinetic measures of fluorodeoxyglucose uptake. Clin Cancer Res 2011;17:2400-9. [Crossref] [PubMed]
27. Nishimura M, Tamaki N, Matsushima S, Kiba M, Kotani T, Bamba C, Nakamura Y, Yamada K. Dynamic whole-body 18F-FDG PET for differentiating abnormal lesions from physiological uptake. Eur J Nucl Med Mol Imaging 2020;47:2293-300. [Crossref] [PubMed]
28. de Prost N, Feng Y, Wellman T, Tucci MR, Costa EL, Musch G, Winkler T, Harris RS, Venegas JG, Chao W, Vidal Melo MF. 18F-FDG kinetics parameters depend on the mechanism of injury in early experimental acute respiratory distress syndrome. J Nucl Med 2014;55:1871-7. [Crossref] [PubMed]
29. Sachpekidis C, Hassel JC, Kopp-Schneider A, Haberkorn U, Dimitrakopoulou-Strauss A. Quantitative Dynamic 18F-FDG PET/CT in Survival Prediction of Metastatic Melanoma under PD-1 Inhibitors. Cancers (Basel) 2021;13:1019. [Crossref] [PubMed]
30. Cuaron J, Dunphy M, Rimner A. Role of FDG-PET scans in staging, response assessment, and follow-up care for non-small cell lung cancer. Front Oncol 2012;2:208. [Crossref] [PubMed]
31. Hoffman EJ, Huang SC, Phelps ME. Quantitation in positron emission computed tomography: 1. Effect of object size. J Comput Assist Tomogr 1979;3:299-308. [Crossref] [PubMed]
32. Rousset OG, Ma Y, Evans AC. Correction for partial volume effects in PET: principle and validation. J Nucl Med 1998;39:904-11.
33. Chang G, Chang T, Pan T, Clark JW Jr, Mawlawi OR. Joint correction of respiratory motion artifact and partial volume effect in lung/thoracic PET/CT imaging. Med Phys 2010;37:6221-32. [Crossref] [PubMed]