Real World Analysis of Small Cell Lung Cancer Patients: Prognostic Factors and Treatment Outcomes

In this observational study, we assessed treatment patterns and prognostic factors in patients with small cell lung cancer (SCLC) in a large state-mandated healthcare organization in Israel. Methods: All incident cases with histologically confirmed SCLC who initiated systemic anti-cancer treatment between 2011 and 2017 were identified. Treatment patterns and overall survival (OS) were evaluated for each line of therapy. Results: A total of 235 patients were identified (61% male, median age 64 years, 95% ever smokers, 64% had extensive stage). The first-line treatment was platinum–etoposide regimen for 98.7% of the cohort. The second and third-line regimen were given to 43% and 12% of patients, respectively. Mean OS for extensive and limited stage patients was 9.1 and 23.5 months respectively. In a multivariable model, increased risk for mortality was observed among patients with an ECOG performance status (PS) of 2 compared to a PS of 0–1 for the extensive stage patients (Hazard ratio (HR) = 1.63, 95% confidence ratios (CI): 1.00–2.65); and for males compared to females for the limited stage patients (HR = 2.17; 95% CI: 1.12–4.20). Regarding all 2nd line patients in a multivariable model incorporating relevant confounding factors, demonstrated a significantly better outcome with platinum–based regimens compared to topotecan. Median survival after initiation of 2nd line in platinum-sensitive patients was longer (p = 0.056) for those re-challenged with platinum–based regimen (n = 7): 6.8mo (6.1-not reported (NR)), compared with those switched to a different treatment (n = 27): 4.5 mo (2.6–6.6) for extensive stage patients, and a non-significant difference was also observed for limited stage patients. Conclusion: To our knowledge, this is one of the largest real-world studies of SCLC patients. OS for SCLC patients was similar to that reported in clinical trials. PS for extensive stage patients and sex for limited stage patients were significant correlates of prognosis. Re-challenge of the platinum–based doublet was associated with longer OS compared to switching treatment in extensive stage patients.

members. Such registries are continuously updated, and they identify patients via automatic search formulas, as opposed to being dependent upon active reporting by physicians [46][47][48].
In addition to the automatic data collection, data that were not available in the main database such as disease staging, imaging results, metastases and eastern cooperative oncology group performance status (PS), were manually extracted by trained data extractors from individual PDF files of patient data in the MHS medical health records.

Study Population
In this retrospective cohort study we identified MHS members aged 18 years or above with a SCLC diagnosis, based on the national cancer registry (which uses diagnosis data from the Israel Ministry of Health, linked to cancer medication approvals and pathology reports from MHS) [49] or with a SCLC diagnosis code in the MHS electronic medical records, and confirmed by manual review of pathology reports. To be included in the study patients had to have received at least one systemic chemotherapy regimen.
Patients were included if they received the first systemic treatment for SCLC between 1 January 2011 and 31 December 2017. Index date was set as the date of initiation of first systemic treatment. Patients with less than one year of healthcare registration in MHS before index date were excluded.
Data was collected up to June 2018 to allow for a minimum follow-up of 6 months.

Study Variables
Demographic and clinical data collected included age at index date, sex, socioeconomic status, district, comorbid conditions, weight and smoking. Socioeconomic status was categorized into quartiles based on the poverty index of the member's enumeration area, as defined by 2008 Israeli National Census [50]. The poverty index is based on several parameters including household income, educational level, crowding and car ownership.
Co-morbidities at baseline were identified using MHS registries (for diabetes mellitus, hypertension, chronic obstructive pulmonary disease, cardiovascular disease, chronic kidney disease and osteoporosis).
Drug purchases and smoking data were cross-linked between the automatic extraction from the database and additional manually extracted data.

Treatment Patterns
Treatment lines were defined according to the sequence of dispensed medications, with information captured both from pharmacy database (for medications approved by MHS), and from hospital medical records (including information on medications provided by private insurance and clinical studies). Addition of a new drug to a current regimen was considered a new treatment line, and cessation of a medication from a combination regimen (likely due to tolerance issues) was considered the same line.

Statistical Analysis
Baseline descriptive characteristics were compared using t tests for continuous variables and chi-square tests for discrete variables.
OS was assessed using all-cause mortality data from the National Insurance Institute, with Kaplan-Meier and Cox proportional hazards regression methods, and plots of expected survival were generated from L1 and L2 treatment initiation. All analyses were performed for extensive and limited stage patients separately, and subdivided into platinum-resistant and platinum-sensitive patients. Platinum sensitivity was defined as an interval of ≥3 months from the end of L1 treatment to the beginning of L2 treatment. Relapse free interval was defined as the interval between last dose of L1 treatment and initiation of L2 therapy (split by 0-3 months, 3-6 months, 6-9 months and 9-12 months).
All analyses were conducted using IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp, and a p-value < 0.05 was considered statistically significant.
The study was approved by the local ethics review board of Bayit Balev Hospital in Israel.
Among those that received L2 therapy and were resistant to L1 platinum-basedtherapy (initiated L2 therapy ≤3 months from the cessation of L1 therapy), all patients (n = 28) were switched to a different L2 treatment and did not receive re-challenge with L1 platinum-based regimen: 19 received topotecan (OS = 3.5, 2.1-6.3) and 4 received PD-1/PD-L1 inhibitor treatment (OS = 5.6, 0.66-NR, p = 0.43) for the difference between the treatment options (data not shown). Median survival after initiation of 2nd line in platinum-sensitive patients was longer (p = 0.056) for those re-challenged with platinum-regimen (n = 7): 6.8mo (6.12-NR), compared with those switched to a different treatment (n = 27): 4.5 mo (2.6-6.6) for extensive stage patients (Figure 2). Of these, 22 received topotecan, 3 received PD-1/PD-L1 inhibitor therapy and 2 received other treatment. Demographic and clinical variables were similar for platinum-sensitive patients between those re-challenged and switched (data not shown).

Figure 2.
Overall survival from L2 treatment initiation for patients sensitive to L1 therapy (initiated L2 ≥3 months after cessation of L1 therapy), for re-challenge of L1 therapy or switch, split by extensive (n = 34) and limited (n = 33) stage.

Limited Stage Disease
A total of 85 (36%) patients with limited stage disease were eligible for this analysis and consisted of 40% females, 98% ever smokers, 66% 0-1 PS and 0% brain metastases at index date (Table 1). Mean OS was 23.5 months (95% CI: 19.4-27.9, Table 2). An additional factor that significantly correlated with better OS in the univariate analysis was female sex (29.6 vs. 21.5 months, p = 0.03, Table 2). In a multivariable model for all-cause mortality, males had significant increase in mortality with a HR of 2.17 (95% CI: 1.12-4.20, Table 3).
Relapsed limited stage patients were further stratified by those that were platinum-resistant (n = 6) and platinum-sensitive (n = 33) to L1 therapy and demographic and clinical characteristics were similar between the groups.
Among those that received L2 therapy and were resistant to L1 platinum-basedtherapy (initiated L2 therapy ≤3 months from the cessation of L1 therapy), all patients (n = 6) were switched to a different L2 treatment and did not receive re-challenge with L1 platinum-based treatment (data not shown).
Median survival after initiation of L2 for platinum-sensitive patients was numerically longer though results were non-significant between those re-challenged with L1 platinum-based regimen (n = 20, OS = 9.1 mo (7.4-18.8)) and those switched to a different therapy (n = 13, OS = 6.4 mo (5.1-NR)), for the difference between the treatment options ( Figure 2). Demographic and clinical variables were similar for platinum-sensitive patients between those re-challenged and switched (data not shown).
Additionally, a multivariable model considering of all L2 patients (focusing on OS from initiation of L2 therapy) incorporated age, sex, stage (at initial diagnosis), initial PS, brain metastasis, platinum sensitivity to L1 and treatment regimen at L2, found that extensive stage patients had a HR for death of 2.05 (95% CI: 1.19-3.53) as compared to limited stage, and patients that received treatment with topotecan had a HR of 2.67 (95% CI: 1.47-4.85) as compared to those re-treated with platinum based regimen (data not shown).
A further sensitivity analysis examined OS from initiation of L2 therapy stratified by the relapse free interval from end of L1 therapy to initiation of L2 therapy (split by 0-3 months, 3-6 months, 6-9 months and 9-12 months). No difference in OS was observed between the different relapse free intervals.
We further focused on 39 patients (16.6% of the study cohort) who survived for at least 24 months from initiation of L1 treatment. Of these, 80% were limited stage, 54% were female and 72% had 0-1 PS. Of the 8 patients with initial extensive stage disease who demonstrated long-term survival, 5 (63%) were female, 7 (88%) had 0-1 PS, 6 (75%) received radiation during L1 treatment (who had excellent response of mostly advanced localized disease), and 7 (88%) were brain metastases free.

Discussion
Our study reports patient characteristics and OS for a consecutive cohort of unselected SCLC patients. To the best of our knowledge, this is the first real-world study of SCLC patients with relatively recent data, including some patients that received immunotherapy as L2 therapy. In our study, only 43% of patients received L2 therapy and even fewer (12%) went on to receive L3 therapy.
Median OS from L1 treatment initiation was 11.8 months, similar to previous studies [51][52][53][54], a result that has not significantly changed over the past 20 years. OS for extensive stage patients was 9.1 months (95% CI: 8.4, 10.3), comparable to a recent study which reported 10.7 months (95% CI: 9.3, 11.8) [44] in a real-world setting, as well as the control arm of a clinical study which reported 10.3 mo (95% CI: 9.3, 11.3) [12] Predictors of survival included better PS for extensive stage patients, and female sex for limited stage patients, and were mostly similar to those reported in previous studies [55][56][57][58]. Unlike studies which identified female sex as positive prognostic factors for all patients, in our cohort the benefit was seen only for the limited disease patients who dominated the long-term survivors. Other prognostic factors reported in the literature include PS at the time of disease recurrence for all patients [59] and sensitivity to L1 therapy [31,60].
Median OS for patients from L2 treatment was 4.5 months for extensive stage and 8.7 months for limited stage, similar to other studies [31,61] showing the important prognostic significance of stage at initial diagnosis. We found when limiting the analysis to platinum-sensitive patients, those re-treated with platinum-based chemotherapy had a longer numerical survival than those switched to treatment with topotecan single agent for both extensive and limited stage, but did not reach statistical significance, similar to a previously published study [31]. A randomized phase III prospective trial including 180 platinum sensitive patients were treated at relapse with platinum-based regimen or other chemotherapy. Those re-treated with platinum-based therapy had better outcomes than those that were switched [62]. The non-statistical significance in our study could be due to the small numbers in the sub-cohorts. L2 therapy is generally much less effective than initial treatment making it more challenging to draw conclusions on optimal treatment. Finally, we compared OS from initiation of L2 treatment, by relapse free interval from end of L1 treatment to initiation of L2 treatment, however did not find any difference in OS between the groups, similar to another study [63].
OS for SCLC has not improved over the last few decades, and in recent years immunotherapy has become a potential new treatment for SCLC patients. In 2019 the US Food and Drug Administration approved atezolizumab in combination with L1 platinum doublet chemotherapy for extensive stage disease, and recently durvalumab was approved in a similar setting, based on improvement in OS and progression free survival [12,64]. Although the numbers were small, we found that those switched to immunotherapy had a longer numeric median OS than those that received platinum-basedor topotecan therapy for extensive stage. These results need to be interpreted with caution but may be seen as supportive for the role of immunotherapy in SCLC treatment.
Long-term survivors were more likely to be female, have limited stage and 0-1 PS, at baseline and [31] received up-front radiotherapy similar to reports in the literature [65,66].
This study has several strengths including high quality data obtained from the MHS digital database and comprehensive review of patient medical records, long follow up and relatively recent data. Limitations include the retrospective nature of this study with a lack of data as to treatment decisions and patient treatment preference in L2.

Conclusions
OS for SCLC patients in a real-world setting was found to be similar to those reported in clinical trials. Factors significantly associated with prognosis included PS for extensive stage patients and sex for limited stage patients. Re-challenge of the platinum-based doublet in relapsing patients who were platinum sensitive was associated with longer OS compared to switching to topotecan treatment in extensive stage patients. In this cohort, long term (24 months or more) survivors, were associated mostly with female limited stage patients, with low PS and upfront combined chemotherapy-irradiation approach. Funding: The study was funded by Abbvie. The design and study conduct were performed in collaboration of AbbVie with Maccabi Healthcare Services as well as participation of the interpretation of data, review, and approval of the publication. Gabriel Chodick, Varda Shalev, Nava Siegelmann-Danieli, and Sarah Sharman Moser do not have any conflict of interest. Jair Bar has served as a consultant to AbbVie, and has received research funding and speaker fees from AbbVie. Keren Ofek, Inna Kan, Nikhil Khandelwal, and Raanan Cohen are employees of AbbVie and own AbbVie stock. The authors wish to acknowledge and thank Maccabi Healthcare Services and AbbVie for overall support and forming REAL WORLD research collaboration in order to generate and develop the data for this abstract.

Conflicts of Interest:
The authors declare no conflict of interest.