This retrospective cohort study enrolled 291 patients aged over 65 years diagnosed with cancer at Yonsei Cancer Center between March 2018 and March 2019. A CGA test was performed on patients diagnosed with cancer and scheduled for chemotherapy. That is, patients for whom chemotherapy was not planned were not enrolled. Of these, 21 patients were diagnosed with SCLC and completed CGA prior to initiation of treatment. The interval between the date of CGA after diagnosis of SCLC was approximately one month. Their survival was followed until August 2021. Apart from the CGA, ECOG PS was also measured. We aim to analyze the CGA domain associated with OS in older SCLC patients. Ethical approval for this study was obtained from the Institutional Review Board (IRB) of Yonsei University’s Health System (IRB number: 4-2021-1348) and adhered to the tenets of the Declaration of Helsinki. Informed consent was waived by the IRB of Yonsei University’s Health System for this study as all the data was obtained from medical records.

The Cox regression test was used for univariate and multivariate analyses using the Cox proportional hazards model to evaluate the prognostic value of CGA domains. The log-rank test was used to evaluate survival differences between the groups according to CGA scores. Cumulative survival rates were calculated by the Kaplan-Meier method. Confidence intervals (CIs) at the 95% level were calculated, and statistical significance was considered with p values less than 0.05. All analyses were performed using SPSS for Windows version 25 (IBM Co., Armonk, NY, USA).

All participants underwent the CGA by trained geriatric nurses before the starting of chemotherapy. The CGA consisted of the following domains: basic items, physical activity, functional status, frailty, nutritional status, cognition, and depression. Basic items consisted of medical history including comorbidity, medications being taken, Timed Up and Go (TUG) test, grip strength, and lifestyle habits such as smoking. Comorbidity was evaluated by Charlson’s comorbidity index (CCI) [6], summing up the data regarding myocardial infarction, congestive heart failure, peripheral vascular disease, cerebral vascular disease, dementia, chronic pulmonary disease, rheumatologic disease, peptic ulcer disease, mild liver disease, diabetes, and so on. Since we only enrolled the SCLC patients, all patients scored 6 points in CCI, and 9 points were set as cut-offs for analysis [6,9]. The number of medications was analyzed by dividing five drugs as a cut-off value [10]. The clinical significance of sarcopenia is increasingly recognized as a component of cancer cachexia syndrome, and with this increasing awareness, a recent international consensus has established sarcopenia as the major diagnostic criterion for cancer cachexia [11]. In general, to evaluate sarcopenia, skeletal muscle mass is measured by computed tomography (CT) scan. However, not all patients had CT scans at a specific spinal level, so for ease of evaluation, arm and calf circumferences were measured. This has already been verified through other studies as an index to evaluate sarcopenia [12–14]. Since muscle mass itself may not indicate reliable muscle function, we further measured the TUG test and grip strength. In the TUG test, the patient was observed and timed while standing up from the chair, walking 3 m, turns walking back, and sitting down again [15]. Patients who were able to complete the task in less than 12.6 seconds were considered to have good mobility [16]. The cut-off value of handgrip strength was based on the consensus of the Asian Working Group for Sarcopenia (AWGS): 26.0 kg for male and 18.0 kg for female patients, respectively [17]. It is necessary to define the operational definition of sarcopenia by developing an index that combines the arm and calf circumferences, TUG test results, and grip strength values. However, in this study, the effect of only each domain on the OS were analyzed. Additionally, body mass index (BMI) was also measured. The level of frailty was assessed by the FRAIL scale, which was translated into Korean from Morley et al.’s FRAIL scale [18,19]. This self-reported five-item tool includes fatigue (F), resistance (R), ambulation (A), illnesses (I), and loss of weight (L). The total score (0–5) can be categorized into three: robust (0), pre-frail (1–2), frail (3–5). Functional status was evaluated by assessing activities of daily living (ADL) [20] and Korean Instrumental Activities of Daily Living (IADL) [21]. Six items of ADL included dressing, eating, walking on a corridor, toilet use, bathing, fecal, and urinary continence. Older adults with at least one dependency in ADL category were considered as ADL dependent. IADL included eight items which were shopping, housekeeping, ability to handle finances, ability to prepare food, traveling via car or public transportation, doing laundry, ability to use the telephone, and medication use. Older adults with at least one dependency in IADL category were assessed as IADL dependent. Nutritional status was assessed by the Mini Nutritional Assessment short-form (MNA-SF) [22]. It consists of six questions: decreased food intake, weight change, exercise capacity, medical history, cognitive impairment, and BMI. The MNA-SF score (range 0–14) of over 12 was evaluated as normal nutritional status, a score of 8–11 as a risk of malnutrition, and a score of 0–7 as a stage of malnutrition. Cognitive function was assessed using the Korean version of the Mini-Mental Status Examination for Dementia Screening (MMSE-DS). This 30-item tool indicated decrease in cognitive capacity with lower score [23]. A score below 17 denotes impairment, a score from 17 to 24 denotes mild impairment, and a score of 25 or more denotes normal in cognitive capacity [24]. The Korean version Short-Form of the Geriatric Depression Scale (GDSSF-K) was used to assess depression. This 15-item tool (range, 0–15) indicated the risk of depression when the score is more than 5 points [25].

Of the 21 older SCLC patients, 20 were males, and one was female (Table 1). The median age of the patients was 75 years old (range, 67 to 85). Nineteen patients were initially at the extensive stage; two patients were at the limited stage. Of the 21 patients, 20 (95.24 %) patients had ECOG PS 0 or 1. Treatment of SCLC was classified as combined chemoradiation (sequential or concurrent), chemotherapy alone, and best supportive care. The recommended chemotherapy for SCLC in this study was carboplatin-etoposide. Nine patients underwent combined chemoradiation therapy including radiation to the lung, and only one of them received radiation to the metastatic lesion of the pelvic bone. Nine patients received chemotherapy alone, one patient was not treated due to spondylitis and the other patient was due to a urinary tract infection. Only one patient received prophylactic cranial irradiation. Six patients with brain metastases underwent the whole brain radiation therapy, and 15 patients were without brain metastases. The median OS was 9.3 months (range, 1.53 to 36.30) (Table 1).

Results regarding the relationship between the CGA domains and OS are summarized in Table 2. The effect on OS according to ECOG PS 0, 1, and 2 did not show any significant difference in univariate analysis. The degree of frailty also did not affect OS. The median arm circumference of the patients was 24 cm and the median calf circumference was 30.5 cm, which also did not affect OS. Seven patients with BMI over 25 kg/m² were defined as obese, six patients with BMI over 23 kg/m² were overweight, and eight patients with BMI less than 23 kg/m². Differences by BMI were also insignificant. There were eight patients with normal TUG test, five patients with prolonged TUG test, and eight patients who did not perform the test. The TUG test also did not affect OS. A minimum score of 6 points on CCI was the baseline score for patients diagnosed with SCLC, and 10 patients with a total score of over 9 points, including other chronic diseases, did not affect OS. Ten patients were taking more than five drugs, and this also did not affect OS. There were six patients with current smokers, 14 patients with ex-smokers, and one patient with never-smoker. Smoking history did not affect OS. Eleven patients with a depression-related GDS score of over 5 and 11 patients with IADL dependency were measured, but none of them significantly affected OS. The MMSE defined cognitive impairment based on a score of less than 24, with only one patient scoring less than 24 and all other patients being normal. Therefore, MMSE did not affect survival. Finally, in univariate analysis, ADL and nutritional status significantly affected the survival of SCLC patients (p = 0.017, p = 0.025, respectively). There were seven patients with ADL dependency, 11 patients with normal nutritional status with an MNA-SF score of over 12, eight patients with 8–11 points, and two patients with 0–7 points. It showed that the better the nutritional status, the longer the survival. In multivariate analysis, adjusted for ECOG PS, sex, and age, only nutritional status had a significant effect on OS. Those with an MNA-SF score of over 12 had significantly longer survival than those with an MNA-SF score of less than 12 (median survival 5.63 vs. 15.50; hazard ratio, 0.17; 95% CI, 0.05 to 0.57; p = 0.004) (Fig. 1).