In this single-center, cross-sectional retrospective study, we analyzed 106 patients with extracranial OMD who were initially identified using PET imaging and screened for the presence of BMs using concurrent brain MRI scans. The primary cancers predominantly originated either from skin or lung and pleura, each accounting for 45% of patients, respectively. In 45% of patients, the disease state transitioned from OMD to PMD upon accounting for BMs, illustrating the dynamic nature of metastatic progression in this patient cohort. In this cohort of patients with a low extra-cranial tumor burden and presence of brain metastases, the oligometastatic state remained significantly associated with OS, with a median OS of 28 months compared to 10 months in the OMD vs. PMD group. These results indicate that the total number of metastases and total-body OMD status remains an important prognostic factor in patients with BMs.

In the landscape of phase II OMD trials, the inclusion or exclusion of patients with BMs has varied, leading to a nuanced evaluation of the prognostic significance of BMs in OMD patients. Notably, the NSCLC phase II trial by De Ruysscher et al. (2012) and the study by Gomez et al. (2016) allowed OMD patients with BMs, yet the percentages of such cases in the two studies varied, with 43.9% and 27% of patients exhibiting BMs, respectively [21, 22]. Interestingly, despite the prevalence of OMD patients with BMs in both studies, neither trial could establish prognostic significance for OS or PFS. Similarly, the NSCLC phase II trial by Iyengar et al. (2018) included patients with treated BMs, but the presence of BMs did not achieve statistical significance for OS or PFS either [17]. In contrast, while the SABR-COMET trial by Palma et al. (2019) allowed for the presence of BMs, patients with one to three BMs or a dominant BM requiring surgical decompression was excluded, resulting in only 4% of patients exhibiting BMs [15]. The trial by Wang et al. (2022) with EGFR-mutated OMD NSCLC patients deliberately excluded patients with BMs, highlighting the different approaches in study design [23]. Furthermore, the NRG-BR001 phase I and II/III trials on OMD breast cancer patients by Chmura et al. (2021 and 2022) also strictly excluded patients with evidence of BMs [24, 25]. The variability in the inclusion and exclusion criteria across these OMD studies underscores the challenges in assessing the prognostic significance of BMs in OMD patients, as the limited number of cases in some trials precludes a robust analysis of this important factor.

The inconsistency in BMs as an important factor in OMD trials persists in ongoing research efforts. Despite the significance of the OMD status with BMs as a potentially important prognostic factor for OS, the exclusion of BMs remains a notable practice in numerous ongoing clinical trials targeting OMD patients. A snapshot of the current landscape, as of February 2024, gleaned from ClinicalTrials.gov, reveals a lack of uniformity in trial design [26]. Out of 41 RCTs focusing on OMD across diverse primary tumors, only 20 (48.8%) trials allowed the inclusion of patients with BMs, while seven (17.0%) trials explicitly excluded them. Adding to the complexity, 14 (34.1%) studies implemented precise criteria for distant metastases, such as permitting only those exclusively located in regional lymph nodes, for instance. This disparity in the approach to tackling BMs in ongoing OMD trials mirrors the inconsistency observed in previously conducted studies, further underscoring the challenge of establishing a cohesive understanding of the impact of BMs on OMD patient outcomes.

Exploring the distribution of BMs within the context of OMD reveals interesting patterns that could potentially impact their prognostic significance. Our study suggests that, similar to the bimodal distribution observed in extracranial OMD and as recently published by our research group [27], the distribution of BMs in OMD patients may also demonstrate a bimodal pattern. In our study, nearly 50% of OMD patients with BMs presented with either one or two intracranial lesions, while roughly 30% exhibited more than five BMs. The observation that many OMD patients have one of two BMs resonates with findings from the study by De Ruysscher et al. (2012), where the brain emerged as the most common distant metastasis site, with a notable majority of patients harboring only a single BM [21]. This bimodal distribution prompts consideration of potential implications for prognostic assessments in OMD patients with BMs. The distinctive patterns in the number of intracranial lesions may contribute to the heterogeneity observed in patient outcomes, suggesting that identifying the presence of BMs as a crucial prognostic factor in OMD could be influenced by the specific distribution and number of brain lesions. Further investigation into these nuanced aspects may enhance our understanding of the prognostic power of BMs in OMD, refining strategies for patient risk stratification and tailored therapeutic interventions.

Is it a strength of our study to be grounded in a large initial dataset of 7,000 oncological PET scans, However, our study’s single-center and retrospective design introduced potential bias and limited the generalizability of our findings to more diverse populations. While the initial dataset was extensive, our study’s final inclusion criteria narrowed it down to only 106 patients. This reduction significantly impacted our study’s statistical power. Additionally, the diagnostic tools used – namely PET and brain MRIs – although currently the best available options, are not infallible in terms of sensitivity and specificity for detecting metastases. This inherent limitation introduced a measure of error into our data, which must be considered when interpreting the results. This was further magnified by the fact that the constrained sample size prevented differentiation between patients with or without prior intracranial cancer treatment, including radiotherapy. Consequently, this led to the presence of ablated BMs alongside active ones, making them difficult to distinguish on MRI.

In conclusion, almost half of extracranial OMD patients were re-classified to PMD when BMs were considered. Our study confirmed that BMs in OMD patients with a low total tumor burden remained a significant prognostic factor for OS. However, no significant correlation between treatment types for the two groups and risk factors for transitioning from OMD to PMD states were found. To expand on this research, future studies should leverage multi-center, longitudinal data spanning multiple years, allowing for more nuanced differentiation between patients and exploration of the behavior of BMs from various primary cancers.