In this study, we evaluated the prognoses and factors influencing the outcomes of patients with advanced IMA in relation to different treatment modalities. The results indicated that patients with advanced IMA had a comparatively favorable OS, with a median OS of 32.0 months. Importantly, LAT was identified as a significant predictor of improved survival outcomes in multivariate analysis. However, conventional chemotherapy, targeted therapy, and immunotherapy showed no significant effect on survival.

The efficacy of LAT has been predominantly assessed in patients with NSCLC presenting with oligo-metastasis (OM). OM has been reported in 21% of resected NSCLC cases, ranging from pathological stage IA–IIIB (Hishida et al. 2016) to 25% of patients with stage IV NSCLC (Parikh et al. 2014). The introduction of LAT aimed to enhance the prognosis in these scenarios, defined as the application of radical therapy (such as surgery and radiotherapy) to all technically feasible tumor sites capable of influencing disease progression while maintaining acceptable toxicity levels (Na and Kim 2023). A meta-analysis examining the role of LAT in NSCLC with OM, encompassing 1,750 patients from 20 studies, revealed pooled odds ratios for OS and PFS of 3.492 (95% CI, 2.612–4.699; p < 0.001) and 3.743 (95% CI, 2.586–5.419, p < 0.001), respectively, favoring LAT (Rim et al. 2023). Similarly, our study demonstrated a significant survival benefit associated with LAT in patients with advanced IMA, with an HR of 0.119. Moreover, in our study cohort, 39.4% (n = 13/33) of the patients received LAT for OM, a proportion higher than that observed in the general NSCLC population. This discrepancy could be attributed to the distinctive characteristics of IMA, which are characterized by aerogenous spread with intrapulmonary metastasis and fewer occurrences of lymph node and distant metastases, thus rendering LAT more feasible (Lee et al. 2016; Shim et al. 2015). Therefore, LAT is a promising treatment modality, particularly for IMA.

In our study, surgery was the preferred modality in 92.3% (n = 12/13) of cases among various LAT modalities. A previous meta-analysis of 54 studies on the treatment of NSCLC with OM found no significant difference in OS between surgery and radiotherapy (p = 0.65) (Schanne et al. 2019). However, another meta-analysis focusing on LAT for NSCLC with OM, including 499 patients from eight studies, revealed a lower HR of 0.33 (95% CI, 0.22–0.48) for PFS in LAT cases incorporating surgery, compared to 0.55 (95% CI, 0.36–0.83) for those without surgical intervention (Zhang et al. 2021). Critical pathological parameters, such as tumor cell spread size, invasive size, and mucin spread size, which are known to correlate with poor prognosis (Saito et al. 2020), are not easily identifiable through the radiological means. Given the lower incidence of lymph node metastasis and the challenge of encompassing pathological parameters within radiation fields during radiotherapy for IMA, surgery may represent a viable modality to improve prognosis and provide feasibility for patients with IMAs and OMs. Nonetheless, further studies are imperative to validate these findings and elucidate the full extent of its efficacy.

Unfortunately, our study showed that conventional chemotherapy had no significant effect on patient survival. This aligns with the findings of Cha et al., indicating limited survival improvements with chemotherapy (p = 0.667) (Cha et al. 2016), possibly because of the protective role of extracellular mucus expressing MUC1 and MUC5AC in fostering chemoresistance (Lakshmanan et al. 2015; Saito et al. 2020). Thus, novel approaches targeting the mucinous nature of the IMA have shown promise. In a rat patient-derived xenograft model, pretreatment with a mucolytic therapy comprising bromelain and N-acetylcysteine significantly augmented drug delivery, improved chemosensitivity, and mitigated mucinous tumor growth (Dilly et al. 2021). Hence, the integration of mucolytics into chemotherapeutic regimens has the potential to become an effective strategy. Combining cisplatin/pemetrexed with bevacizumab has also shown prolonged survival, ranging from 12.6 to 30 months, which is likely attributed to increased drug delivery (Sun et al. 2018). Furthermore, our study showed no significant benefit of targeted therapy on survival. Although targetable mutations, such as EGFR mutations, are rare in IMA. KRAS mutations could offer opportunities for targeted therapy, with G12D and G12V mutations accounting for 42.0% and 31.0%, respectively (Xu et al. 2023). Clinical trials investigating the efficacy of the KRAS G12D inhibitor MRTX1133 are underway (Zeissig et al. 2023). Moreover, neuregulin-1 fusion, the second most frequent genetic alteration in IMA, has identified afatinib, an irreversible erb-b2 receptor tyrosine kinase family inhibitor, as a promising therapeutic option (Rosas et al. 2021).

Previous studies have reported low expression of PD-L1 (≥ 1%) in patients with IMA compared with that of those with NIMA (48.1% vs. 9.7%; p = 0.001), along with diminished CD8 + infiltration (≥ 10%) (81.5% vs. 35.5%; p < 0.001) (Xu et al. 2021). These findings suggest a poor response rate to immune checkpoint inhibitors in patients with IMA. However, in advanced IMA, we observed a relatively high proportion of PD-L1 expression (50%), which did not significantly affect survival. Interestingly, Jang et al. reported similar PD-L1 expression levels (42.4%) in advanced IMA but observed better survival rates with immunotherapy compared with those of non-immunotherapy treatments (undefined vs. 17.0 months; p < 0.001) (Jang et al. 2021). This discrepancy might be attributed to differences in co-mutations, such as KRAS, which were more prevalent in the study by Jang et al. (91.7%) than in ours (60%) (Jang et al. 2021). Peng et al. demonstrated that KRAS mutations are positively associated with the immunotherapy response, which might be linked to high PD-L1 expression, an inflammatory tumor microenvironment, and enhanced tumor immunogenicity (Peng et al. 2022). Additionally, differences in the immunotherapy regimen, especially in the combination of conventional chemotherapy and immunotherapy, were noted (3.0% in our study vs. 5.1% in Jang et al.’s study) (Jang et al. 2021). IMA also exhibit unique inhibitory immune checkpoints, such as VTCN1, suggesting potential targets for overcoming immune resistance (Guo et al. 2017). Further research is needed to validate our results and improve treatment strategies by focusing on optimized combination therapies with immunotherapy and targeting the tumor microenvironment.

Our study had some limitations. First, the number of patients was relatively small, owing to the rarity of IMA. This limited sample size reduces the generalizability of our findings and may impact the statistical power, particularly in subgroup analyses. Second, its retrospective design from a single institution may have introduced selection bias. LAT might have been more frequently administered to patients with favorable prognostic factors, such as lower tumor burden. However, we addressed this potential limitation by performing a multivariate analysis that adjusted for these factors, thereby strengthening the validity of our findings. Third, not all patients in our cohort underwent mutational profiling or PD-L1 IHC assays, which could have limited the efficacy of therapies based on these profiles. Despite these limitations, this study is the first to offer significant insights into the use of LAT in the treatment and prognosis of patients with IMAs. Further largescale prospective studies are required to address these limitations.