Lymph nodes metastatic burden as a prognosticator for advanced non-small cell lung cancer: a real-world study

Introduction

Non-small cell lung cancer (NSCLC) accounts for 85% of all lung cancer cases and represents a major global health burden due to its high incidence and mortality rates (1-4). Recent advances in immunotherapy have introduced new treatment options for patients with advanced tumors (5). However, the lack of reliable biomarkers continues to hinder progress, contributing to persistently high mortality. Thus, there is a critical need for improved assessment methods to predict patient prognosis and inform treatment decisions, particularly in selecting immunotherapies. The current method for assessing lymph node involvement is the Tumor-Node-Metastasis (TNM) staging system, which has recently undergone revisions in its ninth edition. Notably, changes to the N staging focus on patients with N2 involvement, further distinguishing between N2a and N2b based on the number of mediastinal lymph nodes affected (6).

While these updates refine the classification of lymph node metastasis, they also underscore the limitations of an anatomical-based staging approach. The heterogeneity of metastatic lymph nodes in the same anatomical region requires greater attention. In many other cancers, factors like the number and size of metastatic lymph nodes are crucial for staging and treatment decisions, as seen in breast, kidney, and gastrointestinal cancers (7-10). However, in lung cancer, the anatomical location remains the primary determinant for the nodal classification.

Given these limitations, we propose that the ninth edition of the TNM staging system would benefit from incorporating additional morphological characteristics, such as the number, size, and fusion of metastatic lymph nodes. These factors could serve as valuable biomarkers to better predict prognosis and guide treatment.

This study aims to explore the performance of lymph nodes metastatic burden (size, number, and fusion of metastatic lymph nodes) in predicting the prognosis of advanced NSCLC, to provide new potential biomarkers for the selection of immunotherapy treatment options, and to offer suggestions for the improvement of the next edition of the TNM staging system. We present this article in accordance with the TRIPOD reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2201/rc).

Methods

Study population

This study retrospectively analyzed 3,426 patients with advanced lung cancer treated at Jinling Hospital (Including Qinhuai Medical District), Affiliated Hospital of Medical School, Nanjing University, from January 2018 to December 2023. The inclusion criteria were as follows: (I) ages ranging from 18 to 90 years old; (II) staged as stage III and IV according to the International Association for the Study of Lung Cancer (IASLC) TNM staging system, with no indication for surgical treatment; (III) received first-line immunotherapy, including monotherapy with immunotherapy or in combination with platinum-based chemotherapy; (IV) had undergone high-resolution computed tomography (HRCT) within 2 months prior to first-line treatment; The exclusion criteria were as follows: (I) pathologically confirmed as small cell lung cancer; (II) had undergone radical surgery for lung cancer or other procedures involving mediastinal lymph node dissection; (III) underwent radical surgery for lung cancer after neoadjuvant immunotherapy; (IV) mutations in EGFR, ALK, ROS1 genes; (V) incomplete or missing basic information, treatment plans, or poor imaging quality in imaging examinations. Patients’ treatment plans were assessed by a multidisciplinary team. Ultimately, 339 patients were included in the study.

Radiologic examination

All patients underwent HRCT scans of the chest before receiving immunotherapy as first-line treatment. Thin-section chest computed tomography (CT) was acquired with the following parameters: slice thickness =2 mm, pitch =0.75, tube voltage =120 kVp, tube current-time product =200 mAs, gantry rotation time =0.8 s/rotation; matrix =512×512. Convolution kernel B31f and an iterative algorithm were used for image reconstructions.

Radiologic evaluation

Currently, measuring the largest short diameter of lymph nodes is considered the simplest and most reproducible method to determine the benignancy or malignancy of lymph nodes, with a largest short diameter greater than 1.00 cm being indicative of pathological lymph nodes. In this study, we measured the largest short diameter of lymph nodes from cross-sectional imaging before treatment, defined lymph nodes with a largest short diameter greater than 1.00 cm as pathological lymph nodes and counted them (regardless of whether they were N1, N2, or N3), and took the largest short diameter of the largest lymph node (regardless of whether it was N1, N2, or N3) as the size of the lymph node. Lymph node fusion was defined as two or more lymph nodes adhering to each other and not being separable on thin-section CT scans (when two lymph nodes are fused and both have a largest short diameter greater than 1.00 cm, they were counted as two lymph nodes rather than one). This study involved two radiologists with extensive experience in reviewing imaging data and statistical analysis. When discrepancies occurred in the measurement of the largest short diameter of lymph nodes, the average value was taken as the final result.

Statistical analysis

Categorical and continuous variables were compared using Pearson’s χ² test and Student’s t-test, respectively. Optimal thresholds for lymph node size and number were determined using X-tile software (Version 3.6.1, Yale University). Kaplan-Meier survival curves and multivariable Cox regression analysis were used to assess overall survival (OS) and progression-free survival (PFS). To explore the potential benefits of combining lymph node size and number, we constructed a composite score using Z-score standardized values of lymph node size and number, weighted by their respective contributions derived from multivariate Cox regression analysis. The composite score was calculated as follows:

CompositeScore=w1×Zsize+w2×ZNumber[1]

Where Z_size and Z_Number are the standardized values of lymph node size and number, respectively, and w₁ and w₂ are the weights assigned based on their regression coefficients. The performance of different N staging systems in predicting prognosis was evaluated using time-dependent receiver operating characteristic (ROC) curves, calculated with the DeLong method. Model improvement was assessed using decision curve analysis (DCA). All statistical analyses were conducted using SPSS 26.0 (IBM Corporation) and R software (Version 4.4.1, http://R-project.org). A two-sided P value of less than 0.05 was considered statistically significant.

Ethical considerations

The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. This retrospective study was approved by the Ethics Committee of Jinling Hospital, Affiliated Hospital of Medical School, Nanjing University (No. 2024DZKY-049-01). Patient data were de-identified prior to analysis, and the need for informed consent was waived by the ethics committee.

Results

Patients characteristics

Table 1 shows a total of 339 patients were included in this study, consisting of 291 males (86%) and 48 females (14%). The majority of the patients (252, 74%) had a history of smoking, and 178 (53%) patients had squamous cell carcinoma. Lymph node fusion was observed in 54 (16%) patients on imaging studies. A total of 318 patients were diagnosed with lymph node metastasis through examinations such as positron emission tomography-computed tomography (PET-CT) or endoscopic ultrasound-guided fine-needle aspiration (EUS/EBUS). Among the 339 patients, 301 received immunotherapy in combination with chemotherapy or anti-angiogenic therapy, while 38 patients received monotherapy (Figure S1). For the 208 patients with available PFS data, the average number of treatment cycles was calculated. Patients in the combined therapy group received an average of 12.2 cycles, while those in the monotherapy group received an average of 16.7 cycles. These treatment patterns were analyzed in relation to patient outcomes. Patients were divided into a training set (Main Medical District, 238 patients, 75%) and a validation set (Qinhuai Medical District, 80 patients, 25%) based on the hospital where they received treatment (Figure 1).

Figure 1 Flowchart of patient selection. From an initial 3,426 advanced lung cancer patients (Jinling Hospital, 2018–2023), 339 met the inclusion criteria (stage III/IV NSCLC, first-line immunotherapy + HRCT) after excluding ineligible cases (e.g., EGFR/ALK mutations, prior surgery, or incomplete data). HRCT, high-resolution computed tomography; NSCLC, non-small cell lung cancer.

Determination of optimal cutoff value of size and number of metastatic lymph nodes

Imaging studies revealed 254 patients with metastatic lymph nodes exhibiting a maximum short axis greater than 1.00 cm. There were significant differences in the size and number distribution of metastatic lymph nodes among patients with different N stages, with the average values being N1: (<1.00 cm, <1), N2: (1.27 cm, 2), and N3: (1.76 cm, 5.8) (Figure S2). Using the X-tile software in training set, the optimal cutoff value for the size of metastatic lymph nodes was found to be 1.60 cm, and for the number was 3 (Figure 2).

Figure 2 The optimal cut-off values of SLNs (A) and NLNs (B) were determined using the X-tile program (N=238, training cohort). Grey shows inverse association, while blue indicates direct association for OS. NLNs, the number of metastatic lymph nodes >1.00 cm; OS, overall survival; SLNs, size of lymph nodes.

The predictive performance of metastatic lymph node size, number, and fusion

Figure 3 displays the Kaplan-Meier survival curves from the validation set, indicating that in univariate survival analysis, the size, number, and fusion of metastatic lymph nodes are all factors affecting OS {hazard ratio (HR) [95% confidence interval (CI): 2.179 (1.432–3.316), 1.859 (1.226–2.821), and 3.635 (1.796–7.358)}. Figure 4 presents the forest plot of univariate and multivariate Cox regression analysis for the validation set, demonstrating that only the size and fusion of lymph nodes are independent prognostic factors [HR (95% CI): 6.21 (1.19–32.25), 3.20 (1.32–7.75)]. Figure S3 demonstrates that the size, number, and fusion of metastatic lymph nodes are not prognostic factors for PFS [HR (95% CI): 1.523 (0.845–2.747), 1.324 (0.738–2.378), and 0.979 (0.469–2.041)].

Figure 3 Kaplan-Meier analysis of OS (N=80, validation cohort). (A) OS according to the SLNs; (B) OS according to the NLNs; (C) OS according to LNs fusion status. P values and HR (95% CI) were obtained from the log-rank test using R software. CI, confidence interval; HR, hazard ratio; LNs, lymph nodes; LNs fusion, multiple swollen lymph nodes adhere to each other; NLNs, the number of metastatic lymph nodes >1.00 cm; OS, overall survival; SLNs, size of lymph nodes.

Figure 4 Univariate and multivariate Cox regression analysis for OS (N=80, validation cohort). Group 1: univariate Cox regression analysis; Group 2: multivariate Cox regression analysis. CI, confidence interval; HR, hazard ratio; LNs, lymph nodes; LNs fusion, multiple swollen lymph nodes adhere to each other; NLNs, the number of metastatic lymph nodes >1.00 cm; OS, overall survival; SLNs, size of lymph nodes.

Prognostic performance of a composite score incorporating lymph node size and number

To further explore the potential benefits of combining lymph node size and number, we constructed a composite score weighted by their respective contributions from multivariate analysis. The prognostic performance of the composite score was evaluated using time-dependent ROC curves. The area under the curve (AUC) values for the composite score were 0.718 at 12 months, 0.629 at 24 months, and 0.734 at 36 months, compared to 0.708, 0.689, and 0.727 for lymph node size alone. These results suggest that the composite score may have some advantages at certain time points, but the overall improvement is limited.

Predictive performance of the new N staging based on lymph node size and fusion

This study incorporated the size and fusion of lymph nodes into the N staging criteria based on anatomical location.

N staging based on anatomical location and mediastinal metastasis stations

• N1: ipsilateral hilar lymph node metastasis.
• N2: ipsilateral mediastinal lymph node metastasis (N2a: metastasis in a single station within the mediastinum; N2b: metastasis in multiple stations within the mediastinum).
• N3: contralateral hilar, mediastinal, and other distant lymph node metastases.

N staging based on size and anatomical location

• N1: ipsilateral hilar lymph node metastasis (N1a: maximum short axis of lymph nodes <1.00 cm; N1b: maximum short axis of lymph nodes 1.00–1.60 cm; N1c: maximum short axis of lymph nodes >1.60 cm).
• N2: ipsilateral mediastinal lymph node metastasis (N2a: maximum short axis of lymph nodes <1.00 cm; N2b: maximum short axis of lymph nodes 1.00–1.60 cm; N2c: maximum short axis of lymph nodes >1.60 cm).
• N3: contralateral hilar, mediastinal, and other distant lymph node metastases (N3a: maximum short axis of lymph nodes <1.00 cm; N3b: maximum short axis of lymph nodes 1.00–1.60 cm; N3c: maximum short axis of lymph nodes >1.60 cm).

N staging based on fusion and anatomical location

• N1: ipsilateral hilar lymph node metastasis (N1a: no lymph node fusion observed; N1b: lymph node fusion observed).
• N2: ipsilateral mediastinal lymph node metastasis (N2a: no lymph node fusion observed; N2b: lymph node fusion observed).
• N3: contralateral hilar, mediastinal, and other distant lymph node metastases (N3a: no lymph node fusion observed; N3b: lymph node fusion observed).

Due to the low number of advanced NSCLC patients in stage N1, this study primarily focused on patients in stages N2 and N3 (based on the ninth edition of the TNM staging system) for statistical research. Figure 5 displays the time-dependent ROC curves for the three staging methods, showing that the staging method based on lymph node size and anatomical location (Size.N.stage) has superior performance in predicting prognosis compared to the methods based on anatomical location and metastasis stations (Location.N.stage) and fusion and anatomical location (Fusion.N.stage), especially in long-term predictions beyond 24 months [3-year AUC of 0.651 (95% CI: 0.535–0.767)]. The DCA indicates that incorporating size into the N staging yields better long-term clinical net benefits (Figures S4,S5). In the combined treatment cohort, the new N-staging system showed significantly higher AUC values compared to the conventional system at all time points. For example, at 6 months, the AUC was 0.807 for the new system vs. 0.649 for the conventional system. At 36 months, the AUC was 0.781 for the new system vs. 0.469 for the conventional system. In the immunotherapy-alone cohort, although the sample size was smaller, the new N-staging system also demonstrated promising results. At 36 months, the AUC was 0.906 for the new system, indicating strong long-term predictive performance.

Figure 5 Time-dependent ROC curves. Fusion.N.stage: N staging based on fusion and anatomical location; Location.N.stage: N staging based on anatomical location and mediastinal metastasis stations; Size.N.stage: N staging based on size and anatomical location. AUC, area under the curve; CI, confidence interval; ROC, receiver operating characteristic.

Discussion

In this study, we evaluated the size, number, and fusion of metastatic lymph nodes in patients with advanced NSCLC who received immunotherapy as first-line treatment. We concluded that the size and fusion of metastatic lymph nodes are independent prognostic factors. Specifically, lymph node size (regardless of whether the nodes are classified as N1, N2, or N3) demonstrated stronger predictive performance for prognosis than anatomical location. In the combined treatment cohort (n=301), the new staging system showed superior performance compared to the conventional system, with AUC values of 0.807 vs. 0.649 at 6 months and 0.781 vs. 0.469 at 36 months. Even in the smaller immunotherapy-alone cohort (n=38), the new system demonstrated promising results, particularly in long-term prediction (AUC =0.906 at 36 months). Although there was no statistical significance in immunotherapy-alone cohort, this may be limited by the small number of patients in the monotherapy group. This indicates that the new lymph node staging system has excellent performance in predicting prognosis in advanced NSCLC patients who commonly receive chemotherapy combined with immunotherapy.

Regional lymph nodes are pivotal in orchestrating the initiation and sustenance of the host’s antitumor immune response (11,12). Loss of their function can disrupt the tumor immune cycle, impairing immune surveillance and reducing the effectiveness of immunotherapy (13,14). In the era of immunotherapy, the role of lymph nodes is increasingly valued. Recent research has highlighted the heterogeneity of metastatic lymph nodes within the same anatomical region (15). For example, Guo et al. found that the number and ratio of positive lymph nodes following lymph node dissection and pulmonary resection were independent predictors of OS, regardless of anatomical location (N1 or N2) (16). Similarly, Xu et al. analyzed the Surveillance, Epidemiology, and End Results (SEER) database and identified optimal cutoffs for lymph node staging (0, 1–3, and ≥4), which improved prognosis prediction (17). Katsumata et al. also confirmed that the number of metastatic lymph nodes is a prognostic factor in a Japanese cohort (18). However, assessing the degree of lymph node invasion in advanced NSCLC is challenging, as patients in late stages are typically not candidates for surgical resection, limiting access to pathological samples. Minimally invasive techniques like EUS/EBUS, video-assisted mediastinoscopy, and video-assisted mediastinoscopic lymph node biopsy are valuable for obtaining tissue samples for pathological examination (19). However, these methods primarily assess visibly enlarged metastatic lymph nodes in the mediastinum, and they are not suitable for evaluating small or non-enlarged nodes (e.g., N1 lymph nodes with a maximum short axis <1 cm). Thus, imaging studies remain the primary method for evaluating lymph node involvement in these patients.

Cross-sectional imaging studies generally rely on morphological characteristics to distinguish benign from malignant lymph nodes, with key metrics such as the long-to-short axis ratio, internal structure, and maximum short axis. While these metrics can vary depending on lymph node location, a maximum short axis greater than 1.00 cm is most commonly used as a cutoff for malignancy (20). For thoracic surgeons, a maximum short axis exceeding 3.00 cm in a single mediastinal node is considered a contraindication for surgery. However, there is a lack of large-scale studies to further refine the staging criteria for nodes between 1.00 and 3.00 cm. In our study, we identified 1.60 cm as the optimal threshold for predicting prognosis, showing strong predictive performance. This threshold could serve as a practical and convenient tool for clinical staging.

Additionally, we analyzed the total number of lymph nodes with a maximum short axis greater than 1.00 cm. However, this variable did not show significant prognostic value in multivariate regression analysis, possibly due to the limitations of imaging techniques in accurately counting metastatic lymph nodes. While imaging studies can be somewhat subjective and less sensitive, the maximum short axis of the largest lymph node and the fusion of lymph nodes are relatively stable indicators. Although the composite score incorporating lymph node size and number showed some advantages at certain time points, the overall improvement in prognostic performance was limited compared to lymph node size alone. Future studies should explore optimizing the weighting of the composite score or incorporating additional factors such as lymph node fusion to better capture metastatic burden in advanced NSCLC.

We acknowledge the limitations of this study, including the potential subjectivity in lymph node assessment and the reduced sensitivity of imaging methods. The relatively small size of our validation cohort may limit the generalizability of our findings. Future studies with larger, multi-center datasets are needed to confirm the robustness of our proposed N-staging system. Additionally, validating our findings in surgical cohorts could provide complementary insights into lymph node characteristics in early-stage NSCLC patients. Lymph node characteristics, such as size and fusion, may also reflect tumor aggressiveness and immune microenvironment features. Future studies incorporating biomarkers such as PD-L1 and tumor mutational burden (TMB) could provide deeper insights into how lymph node burden influences response to immunotherapy.

Conclusions

The size of metastatic lymph nodes serves as a reflection of tumor metastatic burden and demonstrates superior performance in predicting the prognosis of advanced NSCLC. This finding warrants further investigation through larger-scale clinical studies, potentially informing revisions to the current TNM staging system.

Acknowledgments

We express our sincere gratitude to all individuals who participated in and supported this research.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2201/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2201/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2201/prf

Funding: This study was supported by Jinling Hospital Management Project grants 22LCYY-XH2 (H.L.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2024-2201/coif). All authors report that this study was supported by Jinling Hospital Management Project grants 22LCYY-XH2 (H.L.). The authors have no other 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. This retrospective study was approved by the Ethics Committee of Jinling Hospital, Affiliated Hospital of Medical School, Nanjing University (No. 2024DZKY-049-01). Patient data were de-identified prior to analysis, and the need for informed consent was waived by the ethics committee.

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. Miller KD, Nogueira L, Mariotto AB, et al. Cancer treatment and survivorship statistics, 2019. CA Cancer J Clin 2019;69:363-85. [Crossref] [PubMed]
2. Chen W, Zheng R, Baade PD, et al. Cancer statistics in China, 2015. CA Cancer J Clin 2016;66:115-32. [Crossref] [PubMed]
3. Sung H, Ferlay J, Siegel RL, et al. Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin 2021;71:209-49. [Crossref] [PubMed]
4. Gao S, Li N, Wang S, et al. Lung cancer in People’s Republic of China. J Thorac Oncol 2020;15:1567-76. [Crossref] [PubMed]
5. Mamdani H, Matosevic S, Khalid AB, et al. Immunotherapy in Lung Cancer: Current Landscape and Future Directions. Front Immunol 2022;13:823618. [Crossref] [PubMed]
6. Detterbeck FC, Woodard GA, Bader AS, et al. The Proposed Ninth Edition TNM Classification of Lung Cancer. Chest 2024;166:882-95.
7. Lee HY, Choi HJ, Park KJ, et al. Prognostic significance of metastatic lymph node ratio in node-positive colon carcinoma. Ann Surg Oncol 2007;14:1712-7. [Crossref] [PubMed]
8. Vinh-Hung V, Verschraegen C, Promish DI, et al. Ratios of involved nodes in early breast cancer. Breast Cancer Res 2004;6:R680-8. [Crossref] [PubMed]
9. Rusch VW, Asamura H, Watanabe H, et al. The IASLC lung cancer staging project: a proposal for a new international lymph node map in the forthcoming seventh edition of the TNM classification for lung cancer. J Thorac Oncol 2009;4:568-77.
10. Riquet M, Assouad J, Bagan P, et al. Skip mediastinal lymph node metastasis and lung cancer: a particular N2 subgroup with a better prognosis. Ann Thorac Surg 2005;79:225-33. [Crossref] [PubMed]
11. Chiappetta M, Leuzzi G, Sperduti I, et al. Lymph-node ratio predicts survival among the different stages of non-small-cell lung cancer: a multicentre analysis†. Eur J Cardiothorac Surg 2019;55:405-12. [Crossref] [PubMed]
12. Ding X, Hui Z, Dai H, et al. A Proposal for Combination of Lymph Node Ratio and Anatomic Location of Involved Lymph Nodes for Nodal Classification in Non-Small Cell Lung Cancer. J Thorac Oncol 2016;11:1565-73. [Crossref] [PubMed]
13. Spitzer MH, Carmi Y, Reticker-Flynn NE, et al. Systemic Immunity Is Required for Effective Cancer Immunotherapy. Cell 2017;168:487-502.e15. [Crossref] [PubMed]
14. Rahim MK, Okholm TLH, Jones KB, et al. Dynamic CD8(+) T cell responses to cancer immunotherapy in human regional lymph nodes are disrupted in metastatic lymph nodes. Cell 2023;186:1127-1143.e18. [Crossref] [PubMed]
15. Köhler C, Klemm P, Schau A, et al. Introduction of transperitoneal lymphadenectomy in a gynecologic oncology center: analysis of 650 laparoscopic pelvic and/or paraaortic transperitoneal lymphadenectomies. Gynecol Oncol 2004;95:52-61. [Crossref] [PubMed]
16. Guo Q, Hu S, Wang S, et al. Comparative analysis of methodologies for predicting overall survival in patients with non-small cell lung cancer based on the number and rate of resected positive lymph nodes: A study based on the SEER database for 2010 through 2019. Clin Respir J 2023;17:1145-57. [Crossref] [PubMed]
17. Xu L, Si H, Su H, et al. The number of metastatic lymph nodes is more predictive of prognosis than location-based N stage for nonsmall cell lung cancer: a retrospective cohort study. Int J Surg 2023;109:4126-34. [PubMed]
18. Katsumata S, Aokage K, Ishii G, et al. Prognostic Impact of the Number of Metastatic Lymph Nodes on the Eighth Edition of the TNM Classification of NSCLC. J Thorac Oncol 2019;14:1408-18.
19. Bousema JE, Dijkgraaf MGW, van der Heijden EHFM, et al. Endosonography With or Without Confirmatory Mediastinoscopy for Resectable Lung Cancer: A Randomized Clinical Trial. J Clin Oncol 2023;41:3805-15. [Crossref] [PubMed]
20. Glazer GM, Gross BH, Quint LE, et al. Normal mediastinal lymph nodes: number and size according to American Thoracic Society mapping. AJR Am J Roentgenol 1985;144:261-5. [Crossref] [PubMed]