POLE -mutated endometrial carcinomas ( POLE mut EC) have the most favorable prognosis among all TCGA molecular subgroups. Though the consequent "ultramutated" phenotype is recognized as a trademark feature, hypermutated tumors (10-100 mutations/megabase), as well as those that are histologically low-grade, have been described. Herein, we investigate 19 POLE mut ECs from a single institution to determine whether morphologic and genomic differences exist between tumors with high tumor mutational burden (TMB-H: 10-100 mutations/megabase) and ultra-high TMB (TMB-UH: >100 mutations/megabase). All tumors were FIGO stage I, of endometrioid histotype, and no patients recurred (median follow-up of 90 mo). Six previously described POLE exonuclease domain mutations were detected, with the 2 most common being P286R (n=7) and V411L (n=6). Seven ECs were TMB-H (median 76 mutations/megabase), and 12 TMB-UH (median 187 mutations/megabase). TMB-UH tumors were more frequently high-grade, with intratumoral heterogeneity, serous-like nuclei, and bizarre nuclei. Using an integrated approach, multiple classifiers were observed in both cohorts (53% of all tumors); however, those associated with MMRd were exclusive to TMB-UH ECs. All TMB-H ECs had a dominant (>50%) POLE mutational signature, in contrast to only 5 TMB-UH tumors. The remaining TMB-UH ECs demonstrated mixed signatures associated with mismatch repair deficiency (MMRd, n=4) or nonspecific signatures (n=3). All POLE mut tumors were enriched in single-nucleotide variants, while insertions-deletions were rare (maximum of 1.7%). In both groups, all harbored PTEN mutations, with other commonly recurring mutations including PIK3CA, ATRX, BRCA2, FBXW7 , and NF1 . ARID1A mutations were not identified in any TMB-H ECs. Although our cohort is small, the morphology of POLE mut ECs appears to be influenced by TMB, a phenomenon not yet described in the evolving landscape of these tumors. Taken in combination with differences in underlying genomic signatures, these findings provide an interesting springboard for better understanding POLE mut EC pathobiology.

HER2 amplification in cervical cancer has been associated with worse clinical prognosis and a potential favorable response to HER2 inhibitors. Immunohistochemistry for the HER2 receptor is a universally accepted surrogate test for HER2 amplification, but no standardized scoring system currently exists for cervical carcinomas. In this study, we investigated HER2 overexpression in cervical squamous cell carcinoma and correlated it with HER2 amplification using fluorescence in situ hybridization (FISH) and molecular methods. Seventy-two cases of human papillomavirus-associated cervical cancer were retrospectively reviewed, and at least 2 representative tumor sections were stained for HER2. HER2 scoring was performed using the 2018 American Society of Clinical Oncology/College of American Pathologist breast cancer criteria, and cases with equivocal (2+) to positive (3+) expression were analyzed for HER2 amplification using FISH and next-generation sequencing. The average patient age was 50 yrs (range: 27-85 yr), with most patients being African American (73.6%) and diagnosed at FIGO stage I (65.3%). Nineteen (26.4%) had equivocal HER2 expression and 4 (5.5%) showed positive expression. Three of the 4 cases with positive expression had enough tumors for FISH, and all 3 were amplified. Three cases with equivocal expression showed HER2 polysomy on FISH, and none showed HER2 amplification. Late clinical stage, high tumor grade, and regional lymph node metastasis were significantly correlated with HER2 overexpression and HER2 amplification. Next-generation sequencing of the 3 HER2-amplified tumors showed amplification of various genes, including CD274, JAK2, BIRC3, and ERBB2, and a PIK3CA missense mutation. In summary, HER2 immunohistochemistry is a reliable predictive marker of HER2 amplification in cervical cancer.

POLE -mutated endometrial carcinomas ( POLE mut EC) have the most favorable prognosis among all TCGA molecular subgroups. Though the consequent “ultramutated” phenotype is recognized as a trademark feature, hypermutated tumors (10-100 mutations/megabase), as well as those that are histologically low-grade, have been described. Herein, we investigate 19 POLE mut ECs from a single institution to determine whether morphologic and genomic differences exist between tumors with high tumor mutational burden (TMB-H: 10-100 mutations/megabase) and ultra-high TMB (TMB-UH: >100 mutations/megabase). All tumors were FIGO stage I, of endometrioid histotype, and no patients recurred (median follow-up of 90 mo). Six previously described POLE exonuclease domain mutations were detected, with the 2 most common being P286R (n=7) and V411L (n=6). Seven ECs were TMB-H (median 76 mutations/megabase), and 12 TMB-UH (median 187 mutations/megabase). TMB-UH tumors were more frequently high-grade, with intratumoral heterogeneity, serous-like nuclei, and bizarre nuclei. Using an integrated approach, multiple classifiers were observed in both cohorts (53% of all tumors); however, those associated with MMRd were exclusive to TMB-UH ECs. All TMB-H ECs had a dominant (>50%) POLE mutational signature, in contrast to only 5 TMB-UH tumors. The remaining TMB-UH ECs demonstrated mixed signatures associated with mismatch repair deficiency (MMRd, n=4) or nonspecific signatures (n=3). All POLE mut tumors were enriched in single-nucleotide variants, while insertions-deletions were rare (maximum of 1.7%). In both groups, all harbored PTEN mutations, with other commonly recurring mutations including PIK3CA, ATRX, BRCA2, FBXW7 , and NF1 . ARID1A mutations were not identified in any TMB-H ECs. Although our cohort is small, the morphology of POLE mut ECs appears to be influenced by TMB, a phenomenon not yet described in the evolving landscape of these tumors. Taken in combination with differences in underlying genomic signatures, these findings provide an interesting springboard for better understanding POLE mut EC pathobiology.

Highly conserved transport protein particle (TRAPP) complexes regulate subcellular trafficking pathways. Accurate protein trafficking has been increasingly recognized to be critically important for normal development, particularly in the nervous system. Variants in most TRAPP complex subunits have been found to lead to neurodevelopmental disorders with diverse but overlapping phenotypes. We expand on limited prior reports on TRAPPC6B with detailed clinical and neuroradiologic assessments, and studies on mechanisms of disease, and new types of variants. We describe 29 additional patients from 18 independent families with biallelic variants in TRAPPC6B. We identified seven homozygous nonsense (n = 12 patients) and eight canonical splice-site variants (n = 17 patients). In addition, we identified one patient with compound heterozygous splice-site/missense variants with a milder phenotype and one patient with homozygous missense variants. Patients displayed non-progressive microcephaly, global developmental delay/intellectual disability, epilepsy and absent expressive language. Movement disorders including stereotypies, spasticity and dystonia were also observed. Brain imaging revealed reductions in cortex, cerebellum and corpus callosum size with frequent white matter hyperintensity. Volumetric measurements indicated globally diminished volume rather than specific regional losses. We identified a reduced rate of trafficking into the Golgi apparatus and Golgi fragmentation in patient-derived fibroblasts that was rescued by wild-type TRAPPC6B. Molecular studies revealed a weakened interaction between mutant TRAPPC6B (c.454C>T, p.Q152*) and its TRAPP binding partner TRAPPC3. Patient-derived fibroblasts from the TRAPPC6B (c.454C>T, p.Q152*) variant displayed reduced levels of TRAPPC6B as well as other TRAPP II complex-specific members (TRAPPC9 and TRAPPC10). Interestingly, the levels of the TRAPPC6B homologue TRAPPC6A were found to be elevated. Moreover, co-immunoprecipitation experiments showed that TRAPPC6A co-precipitates equally with TRAPP II and TRAPP III, while TRAPPC6B co-precipitates significantly more with TRAPP II, suggesting enrichment of the protein in the TRAPP II complex. This implies that variants in TRAPPC6B may preferentially affect TRAPP II functions compared to TRAPP III functions. Finally, we assessed phenotypes in a Drosophila TRAPPC6B-deficiency model. Neuronal TRAPPC6B knockdown impaired locomotion and led to wing posture defects, supporting a role for TRAPPC6B in neuromotor function. Our findings confirm the association of damaging biallelic TRAPPC6B variants with microcephaly, intellectual disability, language impairments, and epilepsy. A subset of patients also exhibited dystonia and/or spasticity with impaired ambulation. These features overlap with disorders arising from pathogenic variants in other TRAPP subunits, particularly components of the TRAPP II complex. These findings suggest that TRAPPC6B is essential for brain development and function, and TRAPP II complex activity may be particularly relevant for mediating this function.

View Supplementary Video 1View Supplementary Video 2View Supplementary Video 3