Patient data, lung cancer specifications and treatment characteristics are summarized in Table 1. This study included 118 patients, all aged ≥ 80 years (median age 82 years, range 80–91 years), with pathologically proven NSCLC at the time of SBRT. Among these, 73 (61.9%) patients were males, and the remaining 45 were females. The median ECOG performance status of all patients was 1 (range 0−2). The age-adjusted Charlson Comorbidity Index (4 index points added for an age ≥ 80 years) ranged from 4−12, with the majority of patients (79%) exhibiting a lower comorbidity score between 4 and 6. Our cohort consisted of 44.9% of patients without any smoking history. Smokers and former smokers had a median of 22 pack years (range 3–65). The vast majority (86.4%) of patients were classified as unsuitable for surgery by the interdisciplinary tumour board. The remaining 16 patients (13.6%) rejected the recommended surgery and subsequently presented for SBRT. Among all patients, 83.9% were treated upon detection of one early-stage primary tumour (n = 99).

A total of 15 patients (12.7%) were treated for two primary tumours (10 synchronous vs. 5 metachronous), and 4 patients (2.4%) received SBRT for three primary lung cancers (2 synchronous vs. 2 synchronous as well as metachronous). Metachronously diagnosed MPLC patients were treated at a mean interval of 24.0 months (SD 17.4 months).

To summarise, a total of 118 patients with 141 lung primaries were treated with SBRT. Pathology revealed 70.9% adenocarcinoma (n = 100) and 29.1% squamous cell carcinoma (n = 41)..

The median PTV was 37.4 (range 5.4−186.0) cm³. Dose prescriptions included 3 × 15 Gy (n = 72; 51.1%), 4 × 12 Gy (n = 7; 5.0%), 10 × 6 Gy (n = 25; 17.7%), and 6 × 8 Gy (n = 37; 26.2%). The median BED10 at the PTV periphery was 112.5 Gy (range 86.4–112.5 Gy).

The median follow-up after SBRT was 47 months (range 3–169 months). The median OS was 42 months (95% CI: 36.2 to 47.8 months). Among all the patients who underwent SBRT, the one-year LC rate, PFS and OS were 99.1% (95% CI: 94.0 to 100.0%), 80.9% (72.5 to 87.0%) and 89.7% (82.5 to 94.0%), respectively, and the two-year LC rate, PFS and OS were 96.2% (90.1 to 98.6%), 67.0% (57.4 to 74.9%) and 74.7% (65.4 to 81.1%), respectively. Five years after SBRT, the five-year PFS and OS rates were 86.4% (71.8 to 93.8%), 24.7% (14.5 to 35.6%) and 30.2% (19.4 to 41.7%), respectively. The CSS rate was 97.3% (91.9 to 99.1%) at one year, 88.6% (80.3 to 93.6%) at two years and 76.6% (61.4 to 86.4%) at 5 years. Kaplan-Meier plots for LC, PFS, OS and CSS are provided in Fig. 1. The clinical outcomes of the MPLC subgroup are summarised in Table 2. In addition, a cancer-specific survival rates reported in international studies are summarised and compared to our data in Table 3.

Kaplan-Meier curves showing (a) local control (141 SBRT-treated NSCLC) (b) progression-free survival, (c) overall survival, and (d) cancer-specific survival for 118 patients aged ≥ 80 years (grey shaded area: 95% confidence intervals)

In total, seven patients experienced tumour recurrence (out of 141) at the SBRT-treated site, with a median time to recurrence of 26 months (range 14–47 months). Among these seven patients, three died with local and distant tumour progression, one patient experienced local progression but died because of fulminant SCLC, one patient was lost to follow-up one year after local recurrence, and two patients experienced local progression without distant metastases and died due to cerebral haemorrhage.

A total of 66 patients died, but only 17 of whom had progressive tumour disease. Death of the remaining 49 patients were due to the following: other malignancies (4), heart failure (1), pulmonary embolism (1), acute myocardial infarction (3), gastrointestinal bleeding (3), cerebral haemorrhage (2), urosepsis (3), severe pneumonia (10), and unknown causes but no history of tumour progression (22).

In the univariate analysis, the prognostic factors significantly associated with reduced PFS were age (HR = 1.12, 95% CI: 1.02–1.22, p = 0.012; see Table 4) and a higher Charlson comorbidity index (≥ 7 vs. 4–6: HR = 2.23, 95% CI: 1.29–3.85, p = 0.004). The multivariate analysis confirmed that both factors were significantly associated with reduced PFS: age (HR = 1.13, 95% CI: 1.03–1.23, p = 0.008) and a higher Charlson Comorbidity Index (≥ 7 vs. 4–6, HR = 2.33, 95% CI: 1.34–4.03, p = 0.003).

In the univariate analysis, both age (HR = 1.16, 95% CI: 1.06–1.27, p = 0.010) and a higher Charlson Comorbidity Index (≥ 7 vs. 4–6, HR = 2.34, 95% CI: 1.33–4.11, p = 0.003; see Table 5) were associated with shorter OS. In the multivariate analysis, both age (HR = 1.17, 95% CI: 1.07–1.28, p < 0.001) and a higher Charlson Comorbidity Index (≥ 7 vs. 4–6, HR = 2.44, 95% CI: 1.38–4.32, p = 0.003) were confirmed as prognostic factors significantly associated with impaired OS.

The Cox proportional hazards model for CSS did not identify any significant predictors (see Table 6). No significant association with local control was observed in the univariate analysis for histology (adenocarcinoma vs. squamous cell carcinoma, p = 0.272), PTV (p = 0.395) or BED₁₀ (> 100 Gy vs. <100 Gy, p = 0.846).

SBRT was well tolerated, despite the advanced age of the patients and the high rate of comorbidities, and all patients completed their SBRT courses as planned. None of the patients experienced SBRT-induced lung toxicity ≥ grade 3 until the last follow-up visit. In total, 19 patients (16.1%) developed grade 2 pneumonitis requiring steroids. The median time to onset of pneumonitis was 4.4 months (range 1.0–7.1 months) after the completion of SBRT. Fisher’s exact test revealed that the frequency of pneumonitis did not differ significantly between patients treated for one NSCLC and patients treated for MPLC (p = 0.162, one-tailed). In 12 patients (10.2%), rib fractures were detected by CT imaging during oncological follow-up. These patients did not report any symptoms caused by their rib fractures, and they did not present to clinics with pain. The fractures were already consolidated in all patients at the time of diagnosis. None of these patients had previously been treated for MPLC. No other SBRT-associated toxicities ≥ grade 3 (e.g., oesophagitis, fatigue, stenosis or haemorrhage) were detected.