Challenges and State of the Art: The intermediate stage patient

Staging: How to precisely determine mediastinal lymph node involvement

A key determinant for which type of treatment can be offered to patients with NSCLC is their intrathoracic (mediastinal) nodal status. If the disease has not spread to the ipsilateral mediastinal nodes, subcarinal (N2) nodes, or both, and the patient is otherwise considered fit for surgery (i.e., ‘functional resectability’), resection is often the treatment of choice.The planning of treatment therefore critically depends on accurate staging of the disease. Positron-emission tomography–computed tomography (PET-CT) is increasingly available and used by lung cancer multidisciplinary teams for staging the mediastinum. The non-invasive nature of PET-CT defines one of its major advantages; however, it might be suboptimal for the detection of malignancy in normal-sized lymph nodes, as well as to rule out malignancy in patients with coexisting inflammatory or infectious diseases.

The definition of stage III disease changed with the introduction of the 8th TNM classification. For example, further subdivision of stages III and IV has resulted in two stages (i.e., IIIC, IVA), as Markus Dietlein, MD, Department of Nuclear Medicine, University Hospital of Cologne, pointed out: Stage IIIC includes patients with T3 or T4 tumors and N3 nodal status, but M0, whereas in stage IVA, patients with M1a and M1b are grouped, irrespective of their T and N stages. Survival curves of patients with stages IIIC and IVA overlap, while the prognosis for patients with stages IIIA and IIIB is significantly better (Fig. 2; [1]). Determination of the nodal status (N0–3) is therefore of the utmost importance regarding the prognosis of a patient, as survival differs significantly between all neighboring categories (Fig. 3; [2]).

Detection of mediastinal lymph-node metastasis is a prerequisite for accurate staging of stage III NSCLC, which is in turn required for individualized, stage-adapted therapy. There is currently no single modality for accurate characterization of enlarged mediastinal lymph nodes as benign or malignant. In addition to 18F-fluorodeoxyglucose (18F-FDG), the radioactively labeled nucleoside 18F-fluorothymidine (18F-FLT) has recently been introduced as a tracer in PET, as it represents a proliferation marker in contrast to the metabolic nature of 18F-FDG. In a prospective study, both techniques were investigated in parallel in 70 consecutive patients with mediastinal lymphadenopathy detected on CT or chest radiographs [3]. Nodal uptake of the respective tracers was determined by calculation of the maximum standardized uptake (SUVmax) with each of the tracers. The results of PET-CT were compared with histopathology of the lymph nodes.

In nine patients with NSCLC, the 18F-FDG SUVmax and 18F-FLT SUVmax of the lymph nodes with pathologically detected tumor infiltrations were 6.7 and 3.9, respectively, while in those without nodal infiltration, these were 6.4 and 3.7, respectively. Either of the tracers alone did not therefore characterize the nodal status as malignant or benign (p > 0.05), but the 18F-FDG tracer appeared to be taken up more avidly by suspicious lesions. These results suggest that reliable determination of the state of the mediastinal nodes is not possible based on SUVmax values alone.

Figure 2: Overall survival of patients with NSCLC by clinical stage according to the 8th edition of the TNM classification. Mod. according to [1]. MST, median survival time in months. Survival is weighted by type of database submission: registry versus other.

Figure 2: Overall survival of patients with NSCLC by clinical stage according to the 8th edition of the TNM classification. Mod. according to [1]. MST, median survival time in months. Survival is weighted by type of database submission: registry versus other.

In a retrospective series that compared PET-CT and pathological results at surgery for 200 N2 lymph nodes in 64 patients with NSCLC, logistic regression demonstrated significant linear association between PET-CT sensitivity and time from scanning to surgery (p = 0.031), but not for the specificity. In patients scanned < 9 weeks prior to pathological sampling, PET was significantly more sensitive (64 % at < 9 weeks, 0 % at ≥ 9 weeks, p = 0.013) and more accurate (94 % at < 9 weeks, 81 % at ≥ 9 weeks, p = 0.007). No differences in specificity were seen. Thus, the authors recommended that if a PET-CT scan was taken more than 9 weeks previously, and the detection of N2 nodes would alter management decisions, re-staging of the mediastinum is advisable [4].

In a cohort study that included 938 patients with NSCLC staged as T1/T2 by CT and N0/N1 by PET, a model was developed to predict the risk of N2 lymph nodes. Among six risk variables, only N1 stage detected by PET was significantly associated with higher probability of pathological N2 stage (p < 0.001) in the multivariate analysis. While the potential impact of prediction models like this one on outcome remains unclear, further development and validation of similar models might enable physicians to reduce the frequency of invasive staging procedures, and thereby the associated risk and cost for lung cancer patients with low probability of pN2 disease [5]. 

To test the suggestion that endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is not necessary when mediastinal lymph nodes are PET-CT negative, an analysis was performed on 167 patients with N0 (n = 115) and N1 (n = 52) lung cancer and no involvement of the mediastinum at PET-CT. The probability of clinically relevant up-staging by EBUS-TBNA in patients judged as N0 and N1 according to PET-CT was 6.0 % overall; however, this was only 0.9 % in patients originally classified as N0 by PET-CT, but 17.3 % in patients originally classified as N1 by PET-CT. The risk of overlooking N2 or N3 disease after both PET-CT and EBUS-TBNA was 10.4 % [6].

To explore predictors for false-negative N2 diagnosis in PET-CT, the data of clinically (PET-CT) N0 patients who had subsequently been operated on were analyzed retrospectively. In a training set of 284 patients, the false-negative rate was 8.5 %, with these tumors appearing predominantly in subcarinal and right lower paratracheal lymph nodes. A higher SUVmax of the primary tumor was a unique independent risk factor for occult N2 NSCLC (odds ratio, 0.88; 95 % CI, 0.81–0.96; p = 0.003). A cut-off threshold of 2.6 for SUVmax discriminated patients into low risk and high risk for occult N2 nodes (1.0 % vs. 12.5 %; p = 0.001). This correlation was confirmed in a test set of 151 patients (9.3 % with N2 overall, 4 % with low, and 11.9 % with high SUVmax). Thus, in patients with SUVmax of the primary tumor ≥ 2.6, there is a level of risk of N2 disease that should not be ignored. These patients might be candidates for mediastinoscopy [7].

A multicenter study investigated 18F-FDG-PET predictors of mediastinal malignancy that could minimize inter-center variability and improve the selection of the subsequent staging procedures. Here, 121 NSCLC patients were staged by 18F-FDG-PET and EBUS-NA, and they subsequently underwent therapeutic surgery with systematic nodal dissection as the gold standard. Ninety-four (72 %) of these patients had ≥1 hypermetabolic spots in the mediastinum. The variability between hospitals of 18F-FDG-PET measures in terms of the mean SUVmax of the primary tumor and the median SUVmax of the highest hypermetabolic spots in the mediastinum was statistically significant (p = 0.016, p < 0.001, respectively), although significance was lost when the ratio or the difference between the ­SUVmax in the mediastinum and the primary tumor were chosen as the parameters. The SUVmax mediastinum/tumor ratio showed high accuracy under ROC analysis (AUC, 0.77; 95 % CI, 0.68–0.85; p <  0.001), and high predictive power for mediastinal malignancy with a ratio of 0.4 as cut-off (OR, 6.62; 95 % CI, 2.98–14.69). The sensitivities and negative predictive powers obtained by clinical staging using EBUS-NA ranged between 57 % and 92 % after 18F-FDG-PET, and increased with tumors > 3 cm in diameter and/or SUVmax mediastinum/tumor ratio > 0.4 [8].

In an approach to create a reliable method for interpretation of visible mediastinal lymph nodes from visual assessment of PET images, a standardized windowing (threshold: 2 × liver SUVmean) was introduced to assess the lymph node uptake using a four-step score (1: LN uptake ≤ mediastinal blood pool structures [MBPS]; 2: MBPS < LN < liver; 3: liver ≤ LN < ‘black’; 4: LN appears ‘black’). When used by three independent readers with varying levels of experience, this score was reliable for identification of 54 of 278 lymph nodes as malignant when using an optimal cut-off of > 3 for defining malignancy. All three readers achieved comparable levels of accuracy with no differences between subgroups of patients (e.g., hilar vs. mediastinal lymph nodes, adenocarcinoma vs. squamous cell carcinoma, grading G1/2 vs. G3/4). Thus, by applying unified windowing, highly accurate and robust lymph node assessment is achievable through introduction of this score [9].

In a Cochrane analysis of 45 trials that assessed the diagnostic accuracy of integrated PET-CT for diagnosing N2 disease in patients with suspected resectable NSCLC and used pathology as the reference standard, the authors come to the conclusion that the accuracy of PET-CT is not sufficient as the sole source of guidance for management of these patients. They recommended adherence to National Institute for Health and Care Excellence (NICE) guidance on this topic, where PET-CT is used to guide clinicians to the next step, as either a biopsy, or where nodes are negative and small, directly to surgery. An apparent difference in PET-CT accuracy estimates between scanner types, NSCLC subtypes, 18F-FDG dose, and country of origin of study, along with the general variability of the results, suggested that large centers should actively monitor their accuracy [10]. 

In another meta-analysis that included eight studies with 654 patients, the diagnostic performance of dual-time-point PET-CT was compared with single-time-point imaging for the detection of mediastinal nodal metastases in patients with NSCLC. Dual-time-point PET-CT performed a little better than single-time-point imaging. Due to the small sample size and large heterogeneity, however, current evidence does not justify implementation of dual-time-point imaging in routine PET protocols for mediastinal lymph node staging of NSCLC [11].

Many patients with NSCLC have positive mediastinal lymph nodes on preoperative PET, but do not have mediastinal involvement after surgery. The prognostic significance of this discordance was assessed in a study of 547 patients, of whom 105 (19 %) were PET positive in the mediastinum prior to surgery. There were no significant differences between PET-positive and PET-negative patients in terms of 5-year risk of local recurrence, patterns of local failure, risk for distant metastases, and OS. Also in multivariate analysis, a false-positive PET was not significant for local recurrence (HR, 1.00; p = 1.00), distant metastases (HR, 0.82; p = 0.42), or OS (HR, 1.08; p = 0.62). Thus, pathologic staging remains the standard to determine the N2 status of patients with NSCLC [12].

In summary, Dietlein stated that interpretation of mediastinal PET-CT cannot be improved by building the ratio of SUVmediastinum:SUVtumor, by scoring of mediastinal uptake, and by dual-time-point PET acquisition. Clearly, a time from PET-CT scan to surgery of more than 9 weeks decreases the sensitivity to detect N2 stage by PET-CT.

Figure 3: Overall survival of patients with NSCLC by nodal status according to the 8th edition of the TNM classification. Modified from [2]. MST, median survival time in months.

Figure 3: Overall survival of patients with NSCLC by nodal status according to the 8th edition of the TNM classification. Modified from [2]. MST, median survival time in months.

How to integrate chemotherapy, radiotherapy and surgery

Stage III NSCLC includes a very heterogeneous group of patients with differences in localization and extent of disease. Many aspects of their treatment remain controversial, as Karolina Jablonska, MD, Department of Radio­oncology, University Hospital of Cologne, pointed out – the more so, because the definition of stage III disease has changed with the introduction of the new TNM classification. For example, differences in terms of survival between the newly defined stages IIIC and IVA are becoming blurred (see Fig.2). Clinical trials that investigate treatments in specific patient populations can often be limited by recruitment of heterogeneous patient populations, inadequate power to detect small differences in therapeutic outcome, missing randomization, or limited duration of follow-up. Up-front consultation as a multidisciplinary tumor board and determination of resectability is therefore mandatory in every case.

Treatment options for stage III NSCLC following, for instance, the European Society for Medical Oncology and the National Comprehensive Cancer Network guidelines [13, 14] usually consist of multimodal therapies that combine surgery, chemotherapy, and radiotherapy, including: 

  • Disease with limited extent is usually resected and treated by adjuvant chemotherapy or radiochemotherapy. For adjuvant chemotherapy, an OS benefit of 4 % to 5 % after 5 years has been shown for patients with N1 or N2 stage, as well as for those with N0 and tumor size > 4 cm [15]. 
  • Patients with initial stage I or II and up-staged pathologically to N2 after surgery can also receive postoperative chemotherapy and radiotherapy. 
  • The same applies to patients where the resection cannot be performed with maximal radicality (R1/2). 
  • Patients with extensive disease can receive neoadjuvant radiochemotherapy followed by surgery or preoperative chemotherapy and postoperative radiotherapy. The optimal timing of these interventions has not been established and remains controversial. Although one meta-analysis showed that preoperative chemotherapy can improve outcomes in patients with stage IB–IIIA NSCLC [16], differences between preoperative and postoperative chemotherapy were confirmed in a large meta-analysis with data from more than 10,000 patients [17]. 
  • Patients with unresectable disease (T4, N2 or N3) and who are sufficiently fit are treated with definitive concurrent chemoradiation as the preferred option. In the latter situation, based on the recently published data of the PACIFIC trial, a consolidation treatment with the PD-L1 antibody durvalumab can be recommended as soon as it is approved for this indication [18]. 
  • A further meta-analysis investigated trials in which definitive radiochemotherapy was compared with surgery in patients with mainly stage IIIA N2 disease. In the patients treated with induction chemotherapy, there was no difference in terms of OS between surgical resection and definitive radiochemotherapy (HR, 0.92; [19]). However, in the surgical arms, there was a trend towards excess early mortality (within the first 6 months of follow-up) and an advantage in comparison to definitive radiochemotherapy thereafter (HR, 0.78; 95 % CI, 0.63–0.98). With respect to PFS, no significant differences were found, although in the largest of the trials analyzed, there was an advantage for the surgical arm (HR, 0.77; 95 % CI, 0.62–0.96). The authors of the meta-analysis concluded by saying: “Currently, based on the finding of a comparable outcome in survival in the randomized trials, the safer approach of radiochemotherapy remains the preferred approach in many institutions. Surgery may represent a good treatment choice within a multimodality treatment program for patients in good condition and up-front potentially resectable tumors provided that patients will be treated by an expert team incorporating all disciplines of thoracic oncology, ensuring a high level of expertise.”
  • Finally, frail patients who are not resectable and who cannot tolerate aggressive definitive treatment can receive sequential chemotherapy and radiotherapy, or radiotherapy alone.


  1. Goldstraw P et al. The IASLC Lung Cancer Staging Project: Proposals for revision of the TNM stage groupings in the forthcoming (eighth) edition of the TNM Classification for Lung Cancer. J Thorac Oncol 2015; 11: 39-51. 
  2. Asamura H et al. The International Association for the Study of Lung Cancer Lung Cancer Staging Project: Proposals for the revision of the N descriptors in the forthcoming 8th edition of the TNM Classification for Lung Cancer. J Thorac Oncol. 2015; 10: 1675–84.
  3. Rayamajhi SJ et al. 18F-FDG and 18F-FLT PET-CT imaging in the characterization of mediastinal lymph nodes. Ann Nucl Med 2016; 30: 207-16.
  4. Booth K et al. The mediastinal staging accuracy of 18F-fluorodeoxyglycose positron emission tomography-computed tomography in non–small-cell lung cancer with variable time intervals to surgery. Ulster Med J 2013; 82: 75-81.
  5. Farjah F et al. A prediction model for pathologic N2 disease in lung cancer patients with a negative mediastinum by positron emission tomography. J Thorac Oncol. 2013; 8: 1170-80.
  6. Naur TMH et al. Endobronchial ultrasound-guided transbronchial needle aspiration for staging of patients with non–small-cell lung cancer without mediastinal involvement at positron emission tomography-computed tomography. Respiration 2017; 94: 279-84.
  7. Lin JT et al. Association of maximum standardized uptake value with occult mediastinal lymph node metastases in cN0 non–small-cell lung cancer. Eur J Cardiothorac Surg 2016; 50: 914-9.
  8. Serra Fortuny M et al. FDG-PET parameters predicting mediastinal malignancy in lung cancer. BMC Pulm Med 2016; 16: 177.
  9. Rogasch JM et al. Standardized visual reading of 18F-FDG-PET in patients with non–small-cell lung cancer scheduled for preoperative thoracic lymph node staging. Eur J Radiol 2016; 85: 1345-50.
  10. Schmidt-Hansen M et al. PET-CT for assessing mediastinal lymph node involvement in patients with suspected resectable non–small-cell lung cancer. Cochrane Database Syst Rev. 2014 Nov 13; (11): CD009519. DOI: 10.1002/14651858.CD009519.pub2.
  11. Shen G et al. Diagnostic value of dual-time-point 18F-FDG PET-CT versus single-time-point imaging for detection of mediastinal nodal metastasis in non–small-cell lung cancer patients: a meta-analysis. Acta Radiol 2015; 56: 681-7.
  12. Tandberg DJ et al. Are discordant positron emission tomography and pathological assessments of the mediastinum in non–small-cell lung cancer significant? J Thorac Cardiovasc Surg 2013; 146: 796-801.
  13. Postmus PE et al. Early and locally advanced non–small-cell lung cancer (NSCLC): ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2017; 28 (Suppl 4): iv1-iv21.
  14. NCCN Guidelines, version 2.2018 Non–small-cell lung cancer.
  15. Artal Cortes A et al. Adjuvant chemotherapy in non–small-cell lung cancer: state of the art. Transl Lung Cancer Res 2015; 4: 191-7.
  16. NSCLC Meta-Analysis Collaborative Group. Preoperative chemotherapy for non–small-cell lung cancer: a systematic review and meta-analysis of individual participant data. Lancet 2014; 383: 1561-71.
  17. Lim E et al. Preoperative versus postoperative chemotherapy in patients with resectable non–small-cell lung cancer: systematic review and indirect comparison meta-analysis of randomized trials. J Thorac Oncol 2009; 4: 1380-8.
  18. Antonia SJ et al. Durvalumab after chemoradiotherapy in stage III non–small-cell lung cancer. N Engl J Med 2017; 377: 1919-29.
  19. Pöttgen C et al. Definitive radiochemotherapy versus surgery within multimodality treatment in stage III non–small-cell lung cancer (NSCLC) – a cumulative meta-analysis of the randomized evidence. Oncotarget 2017; 8: 41670-8.