Care of patients with non-small-cell lung cancer stage III – the Central European real-world experience

The present study was prospective, observational, non-interventional, multicentric, multinational and registry based. The study had been approved by ethics committees of participating centres and was performed in accordance with the Declaration of Helsinki. Study entry criteria were patients of any age, histological and/or cytological diagnosis of NSCLC stage III according to 7^th edition of the TNM classification3 and signed written informed consent. The study allowed enrolling patients who have been treated between March 2014 and March 2017. Follow-up of patients continued until February 2018.

Registration of all data was fully anonymous and performed in an electronic case report form (CRF [eCRF]) by qualified personnel. Patient identification was in the responsibility of each investigator. The following data were collected: age, gender, race, smoking status, PS, weight loss, date of NSCLC diagnosis, histology and mutational status of tumors, staging procedures, method of mediastinal lymph-node examination, TNM and tumor stage. Data collected on therapeutic procedures were primary therapy, surgical procedures (date, extent, completeness of resection, information on repeated resections), thoracic radiotherapy (date, dose, fractions, technique, energy), prophylactic cranial irradiation, and chemotherapy (dates of start and end, number of cycles, cytotoxic drugs). Combined modality therapy was assessed according to investigator statement and/or dates of overlapping therapies. Type of best treatment response, date of recurrence or progression, and dates of death were collected.

Descriptive statistics and frequency tables were used to characterize the sample data set. Statistical significances of differences for categorical and continuous parameters were assessed by means of Fisher’s exact test and Mann-Whitney test, respectively. Overall survival (OS) was defined as the time from treatment initiation until death of any cause. Progression-free survival (PFS) was defined as the time from treatment initiation until first documented progression and/or death of any cause. Patients without an event were censored at the time of last follow-up visit. Patients lost to follow-up were counted as interval-censored observations. The assumption was made that patients were lost to follow-up due to their treatment failure and that, therefore, no information is available about them. Interval set (interval between visits) was 6 months.15 PFS and OS were estimated by Kaplan-Meier analysis and 95% confidence intervals (95% CI) were provided for all point estimates. Statistical significance of differences in survival between the subgroups was assessed by means of the log-rank test. The multivariable Cox proportional hazard model was used to evaluate the effects of potential prognostic factors on survival measures. Hazard ratios (HR) were complemented with 95% confidence intervals and supported with significance levels. Overall response rate (ORR) was defined as the sum of complete response rate (CR) and partial response rate (PR) and disease control rate as ORR plus stable disease rate (SD). All statistical tests were performed at the significance level of P = 0.05.

A total of 617 patients were enrolled but 34 patients were excluded due to various violations of the study protocol. Thus the study population comprises 583 patients from 16 centres of seven Central European countries: eight centres from Czech Republic (269 patients), two centres from Serbia (109 patients), two centres from Hungary (48 patients) and one centre each from Slovenia (53 patients), Latvia (43 patients), Lithuania (38 patients) and Austria (23 patients). The date for final analysis was February 19, 2018. Minimum follow-up since initial diagnosis was 11 months. Patient characteristics are shown in Table 1: 53.3% aged ≥

65 years, 32.2% females, 7.2% never-smokers, 20.6% weight loss (≥ 10% within prior three months), 83% PS WHO 0–1. Stages IIIA and IIIB were diagnosed in 321 (55.1%) and 262 (44.9%) patients, respectively. Females, patients with weight loss and patients with poor PS were slightly more frequent among patients with stage IIIB. Squamous cell carcinomas were diagnosed in 309 (53.0%) patients, adenocarcinomas in 220 (37.7%) patients, NSCLC not otherwise specified in 36 (6.2%) patients, and other types (e.g. adeno-squamous, large cell) in 18 (3.1%) patients. Results of molecular analyses were documented for 150 (25.7%) patients (data not shown). EGFR mutations were detected in 14/142 (9.9%) patients, KRAS mutations in 11/31 (35.5%) patients and ALK aberrations in 2/88 (2.3%) patients (data not shown).

Diagnostic and staging procedures are summarized in Table 2. Widely used procedures were chest x-ray (96%), chest CT scan (97%), bronchos-copy (89%) and upper abdominal CT scan (67%). PET-CT or PET were performed in 28% of patients. Brain imaging by means of CT or MRI was done in 20% of patients. EBUS or EUS were performed in 14% of patients, CT-guided biopsy in 9%, video-assisted thoracoscopic surgery (VATS) in 3%, and mediastinoscopy in 2%. Other procedures included diagnostic thoracotomy in 5%, radical thoracic surgery with gain of histology in 1%, ultrasonog-raphy (other than abdominal) (1%) and extracranial MRI examinations. The frequencies of the procedures showed great variations between centres with particularly great variations for PET or PET-CT (0–78%), brain imaging (0–91%) and EBUS/EUS (0–71%) (Table 2).

Stage IIIA and st age IIIB were diagnosed in 55.1% and 44.9% of patients, respectively (Table 1). Subgroups based on T and N descriptors are shown in Table 3. Stages T1–T3 were found in 340 patients (58.3%) mostly having N2 or N3 nodal stage in 301 (51.6%) patients. T4N0, T4N1 and T4N2 tumours were found in 33 (5.7%), 26 (4.5%) and 129 (22.2%) patients, respectively. T3N3 and T4N3 tumours were diagnosed in 29 (5%) and 55 (9.4%) patients, respectively. T3–4N3 tumours (stage IIIC according to 8^th TNM classification) were diagnosed in 84 (14.4%) patients. N2 or N3 lymph nodes were found in 485 (83.2%) patients. Histopathological involvement of mediastinal lymph nodes was confirmed in 172 (29.5%) patients by surgery (96 patients), transbronchial biopsy (58 patients), mediastinoscopy (11 patients) or other types of biopsy (10 patients) (data not shown).

Therapeutic modalities are shown in overview for the total study population in Figure 1, for patients with stage IIIA and those with IIIB in detail in Table 4. Time from diagnosis to treatment initiation ranged from 0 to 369 (median 23) days. Combined therapies of any type were delivered to 326 (55.9%) patients. With regard to a single treatment modality only, surgery was delivered to 1.5% thoracic radiotherapy to 11.8%, chemotherapy to 25.7% and best supportive care (BSC) to 5.0% of patients of the total study population. Prophylactic cranial irradiation was not delivered to any patient.

Figure 1

Surgery was performed in 135 (23.2%) patients (119 IIIA, 16 IIIB) and resulted in radical tumour resection in 121 patients and explorative surgery or non-radical resections in 14 patients (Table 4 and Figure 1). The types of surgery included lobectomy (77 patients), bi-lobectomies (6 patients), pneumonectomy (32 patients) and atypical radical resections (6 patients) (Table 4). Surgery was delivered as single modality (nine patients), in combination with chemotherapy (77 patients) or radiotherapy (eight patients) or as trimodality therapy (41 patients). Adjuvant chemotherapy was about twice as frequent as pre-operative chemotherapy among patients undergoing surgery.

Thoracic radiotherapy was delivered to 318 (54.6%) patients with similar proportions among stage IIIA and stage IIIB (Table 4). Thoracic radiotherapy was delivered as single modality (69 patients), as definitive chemoradiotherapy (200 patients) and as trimodality therapy (41 patients). Among patients undergoing definitive chemoradiotherapy, the sequential administration was much more frequent than the concurrent one (146 versus 54 patients). Among patients undergoing trimodality therapy, concurrent chemoradiotherapy was delivered in 13 patients (Table 4).

Chemotherapy as the most frequent treatment modality was delivered to 468 (80.3%) patients with similar percentages among IIIA and IIIB (Table 4). Chemotherapy was the only treatment modality in 150 patients. Carboplatin was used in 43.6% and cisplatin in 37.4% of patients (data not shown). Platins were combined with vinorelbine (33%), gemcitabine (22%), etoposide (11%) or paclitaxel (11%). Pemetrexed was used in 7.7% of patients in stage IIIB and 2.5% in stage IIIA. Other drugs including docetaxel, vinblastine and bevacizumab were each used in less than 2% of patients. No relevant differences in chemotherapy protocols between stage IIIA and IIIB patients were seen (data not shown).

At a median follow-up time of 30 months, 154 (26.4%) patients were alive, 295 (50.6%) patients were dead and 134 (23%) patients were lost to follow-up. Among 334 patients evaluable for overall best response, complete remission, partial remission, stable disease and progressive disease were seen in 27.5%, 35.9%, 24.9% and 11.7% patients, respectively. Information on progression during follow-up was available for 338 patients, while no information was available for the other patients due to death (171 patients) or loss to follow-up (74 patients). Eighty-two (24.3%) patients were without progression and 256 (75.7%) patients had progression, either local (40.2%), distant (26.6%) or both (8.9%). Progression after end of treatment was more frequent among patients with stage IIIB than those with stage IIIA (91.2% versus 81.6%).

OS and PFS are shown in Table 5, Table 6 and Figure 2. Median OS was 16.8 months and the 3-year OS rate was 21%. Median PFS was 11.2 months and the 3-year PFS rate was 11.5%. OS and PFS were longer among patients with IIIA than among those with IIIB and the corresponding hazard ratios were 0.70 (95% CI 0.58–0.85; p < 0.001) and 0.71 (95% CI 0.59–0.85; p < 0.001), respectively (Table 5; Table 6). Among patients with stage IIIA, median OS was 20 months and the 3-year OS rate was 27%. The corresponding values for patients with stage IIIB were 14.4 months and 13.5%, respectively. Among patients with T3–4/N3 tumours who are classified as stage IIIC based on the 8^th edition of the TNM classification, survival outcome (median OS 11.6 months; 3-year OS rate 9.4%) was worse than the outcome of patients with stage IIIB based on the 7^th TNM classification.

Figure 2

OS and PFS were longer among patients with PS 0–1 and no weight loss (good prognostic group) than among those with PS 2–3 and/or weight loss ≥ 10% (poor prognostic group) (Table 5; Table 6; Figure 3). These differences were observed in the

Figure 3

total study population, in IIIA patients and in IIIB patients. Among the good prognostic group of IIIB patients, the survival outcome was similar to the survival outcome of IIIA patients.

Additional outcome data are summarized in Tables 5, 6 and 7 and Figures 3 and 4. Univariate analysis demonstrated longer OS for females, never-smokers, good PS, stage IIIA, mediastinal node verification, and for those with PET/CT or PET staging exam. In contrast, OS was not different between patients with adenocarcinomas and those with squamous cell carcinomas. Including these variables into the multivariate survival proportional hazard model, the differences in OS remained significant for females versus males, no weight loss versus weight loss, PS 0–1 versus 2–3, stage IIIA versus IIIB, and pathological verification of mediastinal nodes versus no verification. EGFR mutation status was not significant in the multivariate analysis but trend for better survival was found in the univariate model (Table 7). OS was also assessed in various treatment subgroups. Significant survival benefits were seen in patients undergoing surgery compared to patients without surgery (HR = 0.43; 95% CI 0.33–0.56; p-value < 0.001) and in patients receiving combined treatment modalities compared to patients with a single treatment modality (Table 5, hazard ratios not shown). The best OS outcome was shown for patients receiving trimodality therapy (Table 5). A significant OS difference between concurrent and sequential chemoradiotherapy could not be demonstrated. Time from diagnosis to initiation of treatment start had no impact on outcome (analysis by quartiles, p = 0.585).

Figure 4