Real‐world KINDLE‐Latin America subset data on treatment patterns and clinical outcomes in patients with stage III non‐small‐cell lung cancer

Abstract Introduction Stage III non‐small‐cell lung cancer (NSCLC) management is challenging given the heterogeneous nature of the disease. The LATAM subset of the real‐world, global KINDLE study reported the treatment patterns and clinical outcomes for LATAM from the pre‐immuno‐oncology era. Methods The study was conducted in seven countries (Argentina, Chile, Colombia, Dominican Republic, Mexico, Peru and Uruguay) in stage III NSCLC (American Joint Committee on Cancer, 7th edition) diagnosed between January 2013 and December 2017. Retrospective data from patients' medical records (index date to the end of follow‐up) were collected. Summary statistics, Kaplan–Meier survival estimates and a two‐sided 95% confidence interval (CI) were provided. Cox proportional hazard model was used for univariate and multi‐variate analyses. Results A total of 231 patients was enrolled, the median age was 65.0 years (range 21.0–89.0), 60.6% were males, 76.6% had smoking history, 64.0% had adenocarcinoma and 28.7% underwent curative resection. Multiple treatment regimens (>25) were used; chemotherapy alone was the most common (24.8%). The overall median progression‐free survival (mPFS) and median overall survival (mOS) were 14.8 months (95% CI, 12.1–18.6) and 48.6 months (95% CI, 34.7 to not calculable). Significantly better mPFS and mOS were observed for stage IIIA with curative surgery and resectable tumours and stage IIIB with an Eastern Cooperative Oncology Group score of 0/1, female gender, resectable tumours, adenocarcinoma and curative surgery (p < 0.05). Conclusion Results show diversity in treatment practices and the corresponding clinical outcomes in stage III NSCLC. There is a need to streamline treatment selection and sequencing to decrease relapse rates after initial therapy.


| INTRODUCTION
Latin America (LATAM) is a diverse region with a population of about 600 million and more than 20 countries. 1 In 2020, an estimated 97,601 new cases of lung cancer (6.6% of total cancer cases) and 86,627 lung cancer deaths were reported in LATAM and the Caribbean. Lung cancer was ranked first in cancer-related deaths. 2 The number of new lung cancer cases and deaths in women is expected to almost double in LATAM, whereas in men it is estimated to increase by 50% in LATAM and the Caribbean by 2030. 3 Non-small-cell lung cancer (NSCLC) is the predominant subtype accounting for approximately 85% of all lung cancer cases. 4 According to the American Joint Committee on Cancer (AJCC) staging system (7th edition), stage III includes two subtypes, IIIA and IIIB. 5 In the recent AJCC edition (8th) published in 2017, stage IIIC was added to include locally advanced (LA) T3 and T4 tumours associated with N3 disease but without metastasis, reflecting their relatively worse prognosis compared with the prognosis of stage IIIB disease. 6 About one-third of all patients with NSCLC present with stage III LA disease. 7 The heterogeneous nature of stage III disease makes the management challenging and often warrants an integrative multi-modal treatment approach taking into account the demographic and clinical factors. Some stage III patients may benefit from surgery alone; however, for curative intent, (neo)adjuvant chemotherapy, concurrent chemoradiotherapy (cCRT) or sequential chemoradiotherapy (sCRT) is the standard of care (SoC). For patients with unresectable stage III disease, the SoC is platinumbased doublet regimens comprising cisplatin/etoposide or carboplatin/paclitaxel combinations with concurrent radiotherapy and durvalumab as consolidation therapy in patients who have not progressed after ≥2 cycles of cCRT. The interim results of the ADAURA trial demonstrate that osimertinib has a definite role in treating resectable epidermal growth factor receptor (EGFR)-mutated NSCLC after completion of curative-intent surgery and adjuvant chemotherapy. [8][9][10][11][12] The National Comprehensive Cancer Network (NCCN) 2021 guidelines now recommend adjuvant osimertinib for patients with completely resected EGFR mutation-positive stage IIB-IIIA NSCLC or highrisk stage IB-IIA NSCLC who received previous adjuvant chemotherapy or are ineligible to receive platinum-based chemotherapy. 12 More recently, IMpower010 showed a disease-free survival benefit with atezolizumab versus best supportive care after adjuvant chemotherapy in patients with resected stage II-IIIA NSCLC, with pronounced benefit in the subgroup whose tumours expressed programmed death ligand 1 (PD-L1) on 1% or more of tumour cells. 13 The treatment practices in LATAM vary from country to country. A retrospective analysis of patient medical records (all stages of NSCLC) in a study in Uruguay showed chemotherapy or radiotherapy as first-line treatment in 65.4% of cases 14 ; whereas in Peru, patients with stage III NSCLC received chemoradiation (CRT) and occasionally surgery, where radiotherapy was used in the management of early, LA and metastatic lung cancer. 15 Novel therapies including EGFR tyrosine kinase inhibitors (TKIs) have shown improvement in survival in patients with stage III NSCLC. [16][17][18][19][20] However, patients in LATAM have not benefitted considerably from the advent of molecularly targeted therapies due to lack of uniformity in the availability of novel therapies across and within countries. Thus, the incidence/mortality ratio for lung cancer is considerably high in developing countries in LATAM than the developed countries. 1 LATAM has a wide diversity in healthcare facilities, especially regarding the availability of specialised oncology care. Furthermore, there are limited existing databases or resources available that include information on diagnosis, treatment patterns and clinical outcomes of patients with stage III NSCLC. An understanding of real-world treatment patterns of NSCLC is needed to evaluate whether the advent of new treatment modalities is translated into clinical practice in LATAM and its impact on clinical outcomes. A multi-national real-world KINDLE study was conducted to characterise treatment patterns and survival outcomes of patients with stage III NSCLC in Asia, the Middle East and Africa and LATAM. 21 We report on treatment patterns and their associated clinical outcomes in patients with stage III NSCLC in the LATAM subset from the pre-immuno-oncology (pre-IO) era. The study results have been reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology checklist. 22

| METHODS
The LATAM subset of the non-interventional, multicentre KINDLE study was conducted in seven countries (Argentina, Chile, Colombia, Dominican Republic, Mexico, Peru and Uruguay) across 13 centres (Supplementary Figure S1). Retrospective data were collected for patients with stage III NSCLC. The study was conducted in accordance with the Declaration of Helsinki, International Council for Harmonisation, Good Clinical Practice, Good Pharmacoepidemiology Practice and the applicable legislation for noninterventional studies.

| Study population
The study population included adult patients aged ≥18 years, diagnosed with de novo LA stage III NSCLC (AJCC 7th edition) between January 2013 and December 2017 and with medical records available for at least 9 months from the date of initial diagnosis (index date). Patients with concomitant cancer at the time of diagnosis of stage III NSCLC or within 5 years before NSCLC diagnosis, except for non-metastatic, non-melanoma skin cancers or in situ or benign neoplasms were excluded as were patients with an initial diagnosis of stage I to II NSCLC and diagnosed with stage III disease at the time of relapse. Patients who participated in clinical trials were not specifically excluded from this study.

| Data collection
Extracted data from available medical records were entered in the electronic case report forms from the date of initial diagnosis to the end of the follow-up (defined as death, last available medical record or end of data collection, whichever occurred first). Baseline data included demographic characteristics (age, gender and body mass index) and smoking status, clinical characteristics (Eastern Cooperative Oncology Group [ECOG] performance status, histology, EGFR mutation status, PD-L1 status and stage as per AJCC 7th edition) and treatment patterns (treatment modality and line of treatment). The occurrence and date of disease progression were determined from patients' medical records such as imaging reports, pathology reports and oncologists' notes and statements on disease progression. Overall survival (OS) was defined as the time from stage III NSCLC diagnosis or time from treatment initiation until death from any cause. Real-world progression-free survival (PFS) was defined as the time from treatment initiation to documented disease progression or death due to any cause, whichever occurred first.

| Statistical analyses
Summary statistics were used to describe patient demographics, clinical characteristics and treatment modalities. Categorical variables were summarised using frequencies and percentages and continuous variables by descriptive statistics including median, minimum and maximum, with a 95% confidence interval (CI). The treatment patterns and their associated clinical outcomes were analysed for the overall cohort followed by further analyses in patients with stage IIIA/IIIB and resectable/unresectable disease. Median survival estimates (OS and PFS) were determined descriptively using the Kaplan-Meier survival curves. The median survival estimates were reported along with the two-sided 95% CI. A multi-variate Cox proportional hazards model and hazards ratio along with 95% CI were used to identify the significant effects of clinical and demographic factors on OS and PFS by controlling relevant demographic and clinical covariates affecting OS and PFS. A p value of <0.05 was considered statistically significant. SAS9.4 was used to perform the analyses.

| Sociodemographic and clinical characteristics
A total of 231 patients with stage III NSCLC from seven countries in the LATAM region participated in the KINDLE study. A total of 142 (61.5%) patients were alive at the time of data collection. A majority of patients were from Argentina (33.8%) followed by Chile (21.6%). The median (range) duration of follow-up for patients was 660 days (7-2404). The median (range) age of patients was 65.0 years (21.0-89.0), most were men (60.6%) and more than three-fourths (76.6%) were current or past smokers.   Figure 1 shows the most common treatment modalities administered under initial therapy. All treatment modalities are summarised in Supplementary Data Tables S1 and S2. Pooling the patients who underwent surgery (alone or with [neo] adjuvant) as part of initial therapy showed about onefourth of patients (58 [

| Recurrence and survival outcomes
After initial therapy, 67.0% (136/203) of patients relapsed (had disease progression or died). Among patients with stage IIIA and IIIB NSCLC, 70.8% (63/89) and 73.8%  Figure 2A,B). The mPFS was 30.3 months (95% CI, 18.0-NC) for patients with resectable NSCLC and 11.1 months (95% CI, 8.3-14.7) for patients with unresectable disease. The mOS was NC in patients with resectable NSCLC and was 36.5 months (95% CI, 19.9-48.0) for patients with unresectable disease. Survival outcomes based on treatment modality received in patients with resectable NSCLC showed mPFS of surgery + adjuvant chemotherapy to be significantly longer (p < 0.05) than that of either radiotherapy or chemotherapy alone. However, there was no significant difference in mOS between these groups. In the unresectable category, mPFS for cCRT and sCRT were significantly longer (p < 0.05) than that for chemotherapy alone or radiotherapy alone. However, there were no significant differences in mOS of cCRT and sCRT and other treatment modalities.
Survival outcomes based on treatment modality received in patients with stage IIIA showed mPFS of 9.0 months (95% CI, 3.2-19.1) with the most frequent modality of chemotherapy alone. The mPFS was the longest in patients who received cCRT (21.2 months; 95% CI, 2.7-NC) followed by patients who received targeted therapy (20.3 months, 95% CI, 8.3-32.3). In stage IIIB, mPFS in patients who received the most frequent modality of chemotherapy alone was 7.3 months (95% CI, 3.2-9.5). The mPFS was the longest in patients who received sCRT (17.5 months; 95% CI, 6.2-NC) followed by patients who received surgery + chemotherapy alone (14.1 months; 95% CI, NC to NC). In stage IIIA, mOS in 17 patients who received the most frequent modality of chemotherapy alone was 25.9 months (95% CI, 9.5-53.2). Nine patients who received sCRT had the longest mOS (70.9 months; F I G U R E 1 Frequent treatment modalities for stage III NSCLC initial Therapy. The frequency of the various initial treatment modalities used, are shown by stage (IIIA vs IIIB) and overall. cCRT, concurrent chemoradiotherapy; NSCLC, non-small-cell lung cancer; sCRT, sequential chemoradiotherapy. Note: The cut-off used for selection of treatment modalities is 5% in either of the groups.

| Logistic regression analysis
The effect of various sociodemographic and clinical characteristics on PFS and OS using univariate and multi-variate analyses by stage of NSCLC is shown in

| DISCUSSION
This is the first report of real-world data on management and survival outcomes for patients with stage III NSCLC in the LATAM region. The data collected predates the approval of durvalumab, which at the time of reporting is the only IO drug approved for the treatment of patients with unresectable/inoperable NSCLC. 23 The overall percentage of patients receiving initial therapy was similar in the KINDLE global study and the LATAM subset. The study results show heterogeneity in treatment patterns with >25 treatment modalities used as initial therapy. In spite of this diversity, about 50% of patients received chemotherapy alone, cCRT or sCRT as the most frequent treatment modality. In the KINDLE global study, cCRT was the most frequent modality followed by chemotherapy alone and sCRT. 21 About 30% of patients underwent surgery in the LATAM subset, whereas in the KINDLE global study, about 20% of patients underwent surgical resection with a much higher percentage of patients (about 75%) receiving adjuvant therapy compared with about 30% in the LATAM subset. Patients who underwent surgery were mostly those with stage IIIA NSCLC in both KINDLE global cohorts and the LATAM subset. Also, as per the current NCCN guidelines, surgery is not recommended for patients with stage IIIB NSCLC given the extent of the tumour. 12 Similarly, although chemotherapy alone was the most frequent modality in both stages IIIA and IIIB NSCLC, a greater percentage of patients with stage IIIB NSCLC received chemotherapy alone than those with stage IIIA disease. F I G U R E 2 (A) Kaplan-Meier plot of progression-free survival after initial therapy by disease stage. The Kaplan-Meier curve for progression-free survival by stage and overall is shown. CI, confidence interval; mPFS, median progression-free survival. mPFS for stage III overall: 14 Treatment patterns vary in different countries in LATAM. In Peru, patients with stage III disease received CRT and rarely surgery as part of multi-disciplinary care. This could be attributed to the scarcity of thoracic oncology surgeons in the country. 15 Results from a retrospective cohort study in Brazil in patients with stage IIIB NSCLC showed that 30.4% of patients received cCRT and 8.7% patients received sCRT as part of initial therapy. 24 Another study (IIIA 20.9%, IIIB 22.4%) showed that only about one-fourth of patients (24.6%) received curative surgery 25 that could be attributed to the unequal distribution of thoracic surgeons in the country. 26 There is heterogeneity in the treatment pattern within Brazil 27 where although CRT is largely the SoC for treatment of patients with LA NSCLC, results of the retrospective analysis in elderly patients (>65 years) with unresectable LA disease has shown that about 40% of patients received best supportive care and about 30% of patients each received definitive radiotherapy alone or CRT. 28 There are gaps in implementing the international and national guidelines given the fragmented healthcare system that impedes optimum care for patients. 29 There is a lack of uniform access to the healthcare system clubbed with diverse access to lung cancer education for the scientific community in the region. 1 About two-thirds of patients relapsed after initial therapy. Of the patients who received any second-line or thirdline treatment, chemotherapy alone was the preferred treatment in the second-line, whereas chemotherapy alone or targeted therapy was the preferred treatment in the third-line treatment. The relapse rate was similar in patients with stage IIIA and IIIB NSCLC, whereas it was about 25% higher in patients with unresectable disease than in those with resectable NSCLC similar to that observed in the KINDLE global cohort.
The mPFS of 14.8 months (95% CI, 12.1-18.6) observed in the LATAM subset is similar to the KINDLE global cohort results. The mOS of 48.6 months (95% CI, 34.7 to NC) is substantially higher than that observed for the global cohort (34.9 months, 95% CI, 32.0-38.0). The mPFS and mOS were higher in stage IIIA NSCLC than in stage IIIB disease. Similar observations were noted for patients with the resectable disease compared with unresectable NSCLC. Surgery as part of initial therapy was an independent predictor for PFS in stage IIIA, whereas an ECOG score of 0/1 was an independent predictor for PFS and OS in stage IIIB. Surprisingly, age >65 years showed improved PFS in both stages; however, this could be attributed to small sample size, selection bias, missing data and patient disease characteristics at study entry.
With the advent of immunotherapy, there has been a paradigm shift in the treatment of NSCLC. 30,31 Adjuvant osimertinib, a third-generation EGFR-TKI, has recently been demonstrated to provide longer disease-free survival in patients with resected stage I to III EGFR-mutated NSCLC over placebo. 32 A study in LATAM showed an overall response rate in EGFR-mutated patients who received TKI to be at 62.5% (95% CI, 50.0%-75.0%) with an mOS of 16.5 months. 33 Data from the phase 3 PACIFIC study demonstrated that durvalumab, a human monoclonal antibody, improved PFS and OS, irrespective of the biomarker status when used as consolidation therapy after platinum-based CRT in patients with stage III NSCLC. 31,34 Results from the LATAM subset in EGFR-mutated NSCLC patients (n = 175) who received TKIs showed mPFS  35 Although the study provides insights into treatment practices for stage III NSCLC in the LATAM region, the retrospective design limits the representativeness of the findings to the post-IO era. Also, being a real-world study, the data collection was limited to the availability of existing medical records. Indeed, many data points were missing from being reflective of a real-world study. With a low sample size, country-specific analyses for treatment patterns and survival outcomes were not conducted. Hence, future studies with a larger sample size are needed to facilitate country-level analyses considering the heterogeneity in the treatment patterns for NSCLC in the LATAM region. With the advent of newer therapies, the survival outcome is projected to change. Nevertheless, the data from this study will help in establishing baseline data and planning upcoming real-world studies in the LATAM region. Considering the role of precision treatment and targeted therapy in the management of NSCLC, the incorporation of molecular tumour boards comprising specialists involved in the treatment decisions is important for recommending genetic testing in this patient population. Implementing guideline-recommended novel treatment strategies such as targeted therapy and immunotherapy through a multi-disciplinary approach may homogenise treatment and improve clinical outcomes of patients with stage III NSCLC in the LATAM region.

| CONCLUSIONS
In the LATAM subset of the KINDLE global study in patients with stage III NSCLC, heterogeneity in treatment patterns and corresponding clinical outcomes observed in real-world settings were reported. The national and international guidelines were not uniformly followed in the region. The relapse rates after initial therapy were high, suggesting a need to streamline treatment selection and sequencing to decrease relapse rates after initial therapy.