Are brain metastases a special issue in POLE-mutated endometrial cancer? a case report series.

This patient was initially treated with surgery for an endometrioid EC grade 3 FIGO (2009) stage IB with LVSI. In the retrospective molecular analysis a POLE -EDM (p.P286R) was found. MMR proteins, p53 and estrogen receptors were regularly expressed, L1CAM was negative. Very shortly after the diagnosis brain metastases were detected with MRI and were surgically resected one month after the primary surgery. The patient underwent chemotherapy and unfortunately recurred with brain metastasis 6 months later. This cerebral metastasis was again resected, and adjuvant whole brain radiation was performed. A second relapse to the brain occurred 9 months later and was also treated with radiotherapy. The patient died 4.5 years after the primary diagnosis.

Molecular classification of endometrial cancers (ECs) into four subgroups led to better risk stratification and was promptly integrated in treatment guidelines. Even in the 2023 renewed International Federation of Gynecology and Obstetrics (FIGO) staging system for EC molecular, properties like abnormal p53 expression or polymerase epsilon exonuclease domain mutation ( POLE -EDM) are directly affecting tumor stage ( 1 – 3 ). POLE -mutated ( POLE mut ) ECs are usually high-grade, ultramutated cancers most commonly of endometrioid histology. These cancers come with an excellent prognosis, a fact that has been underscored in large cohorts and meta-analyses ( 4 – 6 ). The good prognosis of POLE mut tumors is maintained even in the presence of properties otherwise considered as high-risk such as substantial lymph vascular space invasion (LVSI), TP53 mutation, high tumor grade, or aggressive histology ( 7 , 8 ).Therefore, recent treatment recommendations insistently favor de-escalation of adjuvant treatment in early POLE -mutated EC irrespective of other additional risk factors. While the overall prognosis is excellent, most of the existing literature is focused on early stage disease and data of large cohorts of advanced POLE mut ECs is currently unavailable due to the rarity of these cases. One key factor to explain their excellent prognosis is thought to be caused by the high tumor mutational burden (TMB), resulting in the expression of abundant neoantigens, which make tumor cells highly visible for the host’s immune system ( 9 ). Metastasis to the brain is a rare event in EC. Incidence is reported between 0.3% and 1.16%, but brain metastases are more common in advanced and recurrent cancers with an incidence of 3% and 6%, respectively. On the other hand, only 0.7%–0.84% of occurring brain metastases are estimated to originate from a primary cancer of the endometrium. The two ways of hematogenous metastatic spread to the brain are either via the vena cava, the heart, and the aorta or alternatively through the paravertebral (Baston) venous plexus ( 10 – 12 ). The endometrioid histologic subtype is the most common subtype accounting for 53.3% of brain metastases derived from ECs. Among endometrioid carcinomas that metastasize to the brain, approximately 80% are of high-grade histology. Non-endometrioid subtypes appear disproportionately represented, given that they account for the remaining 47.7% of brain metastases, despite constituting only about 20% of ECs overall ( 13 , 14 ). Furthermore, 70% of affected patients have additional extracranial metastases, most commonly in the lung, bone or liver. Data on the frequency of brain metastasis in the now routinely recognized molecular subgroups of EC is currently still missing.

Two additional patients with brain metastases from POLE mut EC were identified in a retrospective cohort study of patients who underwent treatment at the Norwegian Radium Hospital, Oslo University Hospital. The cohort is described elsewhere. Retrospective molecular classification was performed on archived tissue. Allele-specific PCR for the five hotspot POLE -EDMs (P286R, V411L, S297F, A456P, S459F) was performed instead of POLE sequencing. IHC was performed to detect aberrations in MMR protein and p53 expression ( 5 ).

The presented cases represent ECs with two very distinct and rare characteristics namely a pathological POLE -EDM and the development of brain metastases in their course ( Table 2 ). To our knowledge, association between these two events has not been comprehensively investigated so far. Clinical presentation, timeline of treatment and outcomes of the presented cases of POLE mut endometrioid carcinoma with brain metastases. POLE -EDM, polymerase epsilon exonuclease domain mutation; dMMR, mismatch repair deficient; pMMR, mismatch repair proficient; ER, estrogen receptor. Stasenko et al. reported two cases of patients with brain metastases from POLE mut endometroid EC. One Patient initially presented with FIGO (2009) Stage IB disease and had a recurrence with brain metastases 20 months after primary therapy. After a follow-up of 42 months this patient had no evidence of disease. The other patient presented with FIGO stage IV disease and subsequently progressed with brain metastases. This patient died after 33 months of initial diagnosis ( 4 ). The excellent prognosis of POLE mut ECs has been shown repeatedly in large cohorts. Nevertheless, advanced, primary metastatic as well as local and distant recurrences have been reported in this molecular subtype. An important question regarding these cases is whether special properties of the cancer itself (e.g., specific POLE mutations, a particular genomic or epigenetic context) or patient-related attributes (e.g., host’s immune competence), are responsible for the unfavorable disease dissemination. Generally, brain metastases either primary or in the recurrent setting are associated with an extremely poor prognosis in all cancers. Median survival after the diagnosis of brain metastasis from ECs is described to be 6.8 months ( 13 ). However, patients reported here survived substantially longer and responded very well to various treatment approaches. It seems that even in the setting of primary advanced tumor spread and recurrent disease the prognosis of POLE mut ECs appears to remain far superior to tumors of other molecular subgroups. Although no reliable data on the incidence of brain metastases in POLE mut EC do exist, it remains speculative whether the brain is not representing a predestinated localization for tumor spread in POLE mut EC. In order to protect the brain against fatal immune mediated damages from destructive inflammation, the local brains’ immune response is thought to be significantly attenuated in comparison to peripheral tissue ( 15 , 16 ). Therefore, POLE mut EC despite its high visibility to the host’s immune system by the large number of neoantigens presented could take advantage of the locally reduced immune surveillance in spreading to the brain. The influence of the molecular subtype of EC on the potential for CNS spread has yet to be investigated. Even in cancer types with known brain tropism, no universally applicable molecular risk profile has been found that is valid across various cancers. Breast and lung cancers that metastasize to the brain were shown to utilize the cell adhesion molecule L1CAM for migration along capillaries ( 17 ). Although, in EC L1CAM overexpression is closely related to poor outcome as well in endometrioid as in serous subtypes no overexpression was revealed in the herein described EC showing CNS spread ( 18 ). Generalization based on this report should be limited due to the small sample size ( n = 3) and the lack of systematic evaluation in large cohorts. This is a serious risk for selection bias. Nonetheless, given the rarity of this metastatic pattern in POLE mut EC, our case series provides clinically relevant insights and raises questions for possible further research. A large multicenter evaluation or an international registry would be needed to comprehensively describe the dissemination pattern of this subgroup of EC and investigate whether the brain is a preferred region of metastases. In the context of this research, potential confounding molecular alterations occurring together with pathogenic POLE mutations explaining brain tropism could be identified.

Despite efforts to de-escalate adjuvant treatment at least for early POLE mut EC as recommended in the recent guidelines, it remains open how these reported uncommon courses from at least two early stage cases will influence the discussion about under- and over-treatment of POLE mut EC. However, currently available data is insufficient to abandon general de-escalating strategies ( 2 , 19 ). Relative to comparable cases of EC from other molecular subtypes, our report clearly shows that POLE mut EC with brain metastases nonetheless appear to exhibit better treatment responses and are obviously related to a more favorable clinical outcome. The identification of reliable factors able to predict such unusual systemic dissemination in POLE mut EC should be an urgent goal of the future research on this molecular subtype.