Lung cancer is the most commonly diagnosed cancer worldwide and the leading cause of cancer-related deaths representing a serious medical condition, especially among men, with nearly 2.5 million cases and 1.8 million deaths each year [18, 19]. For early-stage NSCLC, surgery is the preferred treatment option [20, 21]. However, when surgery is not feasible, SBRT is an established alternative, leading to a median overall survival of more than 65 months and high rates of local control in medically inoperable NSCLC, according to recent studies [22, 23]. Histopathological confirmation of lung cancer is recommended before treatment. However, many patients cannot undergo a biopsy due to severe health issues, small tumor size, or challenging tumor locations [24]. In such cases, clinical and imaging features including positron-emission tomography (PET)/computed tomography (CT) imaging and risk scores guide treatment without requiring biopsy confirmation [25]. Notably, the use of empiric lung SBRT has significantly increased over the last decade [24, 26, 27]. Even within the same healthcare system, the use of lung SBRT without a preceding biopsy varies widely, ranging from 0 to 61% across medical centers [28]. The consensus-based recommendation in the S3 guidelines of the German Guideline Program in Oncology recommends offering stereotactic ablative radiotherapy directly if an interdisciplinary consensus determines that the risk of biopsy is too high [29]. Several studies have explored the outcomes of empiric SBRT for presumed early-stage lung cancer. However, treatment-related side effects and detailed evaluations of changes in lung function within this highly vulnerable patient population remain inadequately addressed. Only a few comprehensive studies have examined the results of empiric SBRT for lung tumors [27, 28]. According to the US Veterans Affairs health care system, 2221 patients treated with SBRT for cT1–T2a cN0 cM0 NSCLC were evaluated, of which 330 (14.9%) were treated empirically, similar to the utilization rate of our study [28]. Interestingly, no significant differences were found between the histopathologically confirmed and the assumed NSCLC groups regarding age, gender, tobacco consumption, and comorbidities. Their study found that patients in the pathological confirmation group were more likely to have T2 tumors (p < 0.05) and reside within the histoplasmosis belt compared to the nonpathological confirmation group (p < 0.05), while overall survival was negatively associated with T2a tumor stage, and CCI, and lung cancer-specific survival was adversely affected by T2a tumor stage, lack of PET scan use, and being in the pathological confirmation group.

Furthermore, overall survival was similar in both groups, with 34 months compared to 37 months (p = 0.29). Notably, fractionation concepts were not reported and may vary based on tumor location, size, medical center, and treatment date. Lung function testing was also not reported [28]. Another cohort from the Netherlands assessed 591 patients, revealing a significantly higher rate of empiric SBRT utilization at 65% [27]. SBRT concepts were reported, delivering 60 Gy in 3–8 fractions (BED10 ranging from 105–180 Gy). Furthermore, the outcomes for histopathologically confirmed versus presumed NSCLC were similar, with a median overall survival of 39.2 months compared to 40.2 months, respectively. Severe treatment-related toxicity (grade 3+) was rarely observed in the entire cohort. A total of 18 patients (3%) developed grade 3 + RP in the combined cohort; however, no significant difference was found or reported between the two cohorts. Patients with a clinical diagnosis had significantly smaller tumors (p < 0.001) and worse pulmonary function (p = 0.025) than those with a pathological diagnosis. However, overall survival, regional and distant control, and treatment-related side effects showed no significant differences between the groups [27]. A Japanese study of 173 patients with an empiric SBRT utilization rate of 33% found a similar 3‑year overall survival rate of 54% compared to 57% (p = 0.48) but reported no treatment-related grade 4 or 5 radiation pneumonitis [30]. A Chinese multicenter study evaluated 90 patients after propensity score matching between biopsy-proven and clinically diagnosed early-stage NSCLC, of which 45 (50%) patients were treated empirically [31]. Regarding outcomes, the 3‑year local control and overall survival were 92.6%/93% and 84.4%/88.8%, the 5‑year local control and overall survival (OS) rates 85.5%/89.8%, and 63.2%/76.1%, respectively. Relevant treatment-related toxicity (grade 2+ respiratory disorders) occurred in 4 patients (9%) in the cohort without histopathological confirmation. Lung function was not included in routine follow-up for the study. In a small investigation by Haider et al., 23 patients with clinically diagnosed early NSCLC were examined [32]. The median OS was 30.2 months, and severe toxicity (grade 3+) was not reported. In fact, 2 patients (8.7%) experienced grade 1–2 acute side effects, while 3 patients (13%) had grade 1–2 late side effects. However, results and rates of lung function were not reported [32].

The CCI and the necessity for LTOT emerged as significant predictors of overall survival (OS) in our study cohort. With a median CCI of 5 (range: 2–13), our findings underscore the considerable comorbidity burden among these patients, emphasizing the challenges of treating this vulnerable population. Patients with lower comorbidity scores (CCI ≤ 5) exhibited significantly improved OS (p = 0.011) compared to those with CCI > 5, reaffirming the prognostic importance of comorbidities in this context. Moreover, LTOT prior to SBRT was associated with shorter OS (p = 0.02) and a relatively high incidence of grade 2–3 respiratory disorders. Notably, only 3 (14%) of the patients who developed grade 2+ respiratory disorders did not require LTOT before SBRT, reflecting the adverse effect of severely diminished pulmonary function on treatment-related toxicity. Interestingly, other studies examining empiric SBRT have not documented the prevalence of LTOT in their patient cohorts [28, 30, 31]. Our findings underscore the importance of comprehensive pretreatment assessments, which include thorough evaluations of comorbidities and respiratory support requirements, to enhance patient stratification and optimize management strategies. COPD was identified as a possible risk factor for severe treatment-related toxicity, including radiation pneumonitis, with nearly all patients who developed grade 2+ respiratory disorders having a diagnosis of COPD (21 patients, 95.46%). Furthermore, patients requiring LTOT prior to SBRT were at a significant increased risk of severe treatment-related toxicity. Notably, all patients who experienced grade 3+ respiratory disorders required LTOT. Therefore, involving these patients in a shared decision-making process is crucial, as it ensures they are well-informed about the potential risks associated with treatment and the importance of subsequent follow-up care.

Our study observed a significantly higher incidence of respiratory disorders such as RP within 6 months after SBRT compared to the rates reported in the literature reviewed by Berman et al. Specifically, 21 patients in our cohort experienced grade 2+ respiratory disorder, with a concerning 19 cases classified as grade 3. Importantly, no grade 4 or 5 respiratory disorders events occurred. In contrast, the studies cited by Berman et al. demonstrated lower RP rates: Sakanaka et al. observed 5.4% grade 2 RP, Yoshitake et al. reported 2.3% grade 2 RP, and Inoue et al. reported 8.8% grade 2, 5.3% grade 3, and 1.8% grade 5 RP, Wang et al. also reported lower rates with 8% grade 2 and 8% grade 3 RP [9]. A crucial limitation in comparing our findings directly is the lack of detailed patient characteristics, such as LTOT or COPD, within the studies reviewed by Berman et al. While our analysis suggests LTOT and COPD as potential contributing factors to the observed discrepancy, further investigation is warranted to elucidate the specific factors influencing RP development following SBRT.

In addition to LTOT and COPD, Dover et al. emphasized tumor location as a potential risk factor for other toxicities, particularly when the tumor is centrally located. Central or ultracentral tumors can increase the risk of esophageal toxicity [33]. In our cohort, 4 patients experienced dysphagia. Three patients had grade 1 dysphagia, and 1 patient had grade 2 dysphagia. Notably, the patient with grade 2 dysphagia had a peripheral tumor, while 2 of the grade 1 dysphagia cases were associated with central tumors and 1 with an ultracentral tumor. Several studies have investigated the use of SBRT with a BED10 < 90 Gy in order to reduce the risk of treatment-related toxicity, particularly in elderly or frail patients with significant comorbidities and limited life expectancy [34, 35]. These studies demonstrated that acceptable local control can still be achieved with reduced dose regimens in selected patient populations.

Nevertheless, it is important to note that a BED10 ≥ 100 Gy remains the established standard of care for medically fit patients, as it is associated with improved tumor control outcomes [20]. However, in clinical reality, individualized treatment approaches are often necessary to balance efficacy and safety.

In our study, where many patients had compromised general health and were treated with heterogeneous dose schedules—including some with BED10 < 90 Gy—we observed local recurrences in 10 lesions, resulting in a local control rate of 96.6%, 92.3%, 87.1% after 1‑, 2‑, and 3‑years, respectively. While these rates are slightly lower than those reported in studies using higher BEDs, they still demonstrate meaningful local disease control, particularly considering the frailty of the treated population. Our findings support the notion that in selected high-risk patients, SBRT with reduced BED may offer a reasonable compromise between efficacy and tolerability.

Importantly, if clinically justified, a biopsy attempt should always be performed. However, not all patients will undergo the procedure due to life-threatening risks. A multidisciplinary tumor board recommendation for empiric SBRT should be made collaboratively by thoracic surgeons, radiologists, pulmonologists, and radiation oncologists. Additionally, after SBRT, all patients should be encouraged to pursue routine clinical follow-up, including lung function testing. According to current literature, the decrease in DLCO (%) is the most sensitive parameter for the early detection of RP.

Therefore, all patients should undergo routine testing to identify those at risk and start treatment early to prevent severe or life-threatening complications. While providing valuable insights into the treatment-related toxicity and outcomes of empiric SBRT in presumed early-stage NSCLC patients, our study has several limitations. The rate of empiric therapy matches what is reported in the literature, but the relatively small sample size and potential selection bias must be taken into account. Additionally, the retrospective design and the limited number of complete pulmonary function tests due to the frailty and comorbidities of the study population restrict the statistical power of our analysis and limit the generalizability of our findings to the broader population of patients undergoing empiric SBRT. Despite these significant limitations, we believe that our study contributes to the clinical evidence of empiric SBRT in patients without histopathological confirmation of early-stage NSCLC, identifies high-risk patients, and supports patient stratification and personalized treatment. The focus of the current study is clinical outcome and longitudinal pulmonary function analysis/changes. A detailed follow-up analysis with extensive dissection of patterns of relapse analysis is in preparation.