Deep learning for dual detection of microsatellite instability and POLE mutations in colorectal cancer histopathology.

In the spectrum of colorectal tumors, microsatellite-stable (MSS) tumors with DNA polymerase ε (POLE) mutations exhibit a hypermutated profile, holding the potential to respond to immunotherapy similarly to their microsatellite-instable (MSI) counterparts. Yet, due to their rarity and the associated testing costs, systematic screening for these mutations is not commonly pursued. Notably, the histopathological phenotype resulting from POLE mutations is theorized to resemble that of MSI. This resemblance not only could facilitate their detection by a transformer-based Deep Learning (DL) system trained on MSI pathology slides, but also indicates the possibility for MSS patients with POLE mutations to access enhanced treatment options, which might otherwise be overlooked. To harness this potential, we trained a Deep Learning classifier on a large dataset with the ground truth for microsatellite status and subsequently validated its capabilities for MSI and POLE detection across three external cohorts. Our model accurately identified MSI status in both the internal and external resection cohorts using pathology images alone. Notably, with a classification threshold of 0.5, over 75% of POLE driver mutant patients in the external resection cohorts were flagged as "positive" by a DL system trained on MSI status. In a clinical setting, deploying this DL model as a preliminary screening tool could facilitate the efficient identification of clinically relevant MSI and POLE mutations in colorectal tumors, in one go.

Deep learning-based phenotyping reclassifies combined hepatocellular-cholangiocarcinoma.

Primary liver cancer arises either from hepatocytic or biliary lineage cells, giving rise to hepatocellular carcinoma (HCC) or intrahepatic cholangiocarcinoma (ICCA). Combined hepatocellular- cholangiocarcinomas (cHCC-CCA) exhibit equivocal or mixed features of both, causing diagnostic uncertainty and difficulty in determining proper management. Here, we perform a comprehensive deep learning-based phenotyping of multiple cohorts of patients. We show that deep learning can reproduce the diagnosis of HCC vs. CCA with a high performance. We analyze a series of 405 cHCC-CCA patients and demonstrate that the model can reclassify the tumors as HCC or ICCA, and that the predictions are consistent with clinical outcomes, genetic alterations and in situ spatial gene expression profiling. This type of approach could improve treatment decisions and ultimately clinical outcome for patients with rare and biphenotypic cancers such as cHCC-CCA.

Intentional R1 resection of liver metastases: A new treatment paradigm for patients with advanced colorectal cancer based on a propensity score-Matched case-control analysis.

BACKGROUND: Clinical outcomes of colorectal cancer (CRC) patients after an incomplete microscopic (R1) resection of liver metastases may not differ from those following a microscopically margin negative (R0) resection, when the latest is not feasible because of anatomic issues. We aimed at comparing the clinical outcomes of CRC patients with an intentional R1 or with a R0 resection of liver metastases. METHODS: All patients with advanced in CRC and liver metastases consecutively treated by liver resection between February 2005 and January 2019 at in the department of Digestive and Hepatobiliary Surgery of Henri Mondor University Hospital (Créteil, France) were included in this retrospective case-control study. Overall survival (OS) and event-free survival (EFS) were compared between patients who underwent an intentional (pre-operative decision) R1 resection (iR1) to those who had a R0 resection of liver metastases. To account for confounding, comparison between the 2 groups was performed after adjustment using propensity score analysis. RESULTS: Twenty-six CRC patients treated by iR1 resection of liver metastases were compared to 98 patients treated by R0 resection. Median OS reached 39 months [95% confidence interval (CI): 25-67] and 63 months [95% CI: 52-76] in the iR1 and R0 groups, respectively. After adjustment by inverse probability of treatment weighting, patients' OS and EFS did not differ significantly between the iR1 and R0 groups (hazard ratio (HR): 1.19 [0.54-2.62] and 1.67 [0.93-3.03]), respectively. CONCLUSION: iR1 resection of liver metastases in advanced CRC patients is an acceptable therapeutic strategy, when R0 resection is not feasible.

Evaluation of two new highly multiplexed PCR assays as an alternative to next-generation sequencing for IDH1/2 mutation detection.

IDH1 and IDH2 somatic mutations have been identified in solid tumors and blood malignancies. The development of inhibitors of mutant IDH1 and IDH2 in the past few years has prompted the development of a fast and sensitive assay to detect IDH1R132 , IDH2R140 and IDH2R172 mutations to identify patients eligible for these targeted therapies. This study aimed to compare two new multiplexed PCR assays - an automated quantitative PCR (qPCR) on the PGX platform and a droplet digital PCR (ddPCR) with next-generation sequencing (NGS) for IDH1/2 mutation detection. These assays were evaluated on 102 DNA extracted from patient peripheral blood, bone marrow and formalin-fixed paraffin-embedded tissue samples with mutation allelic frequency ranging from 0.6% to 45.6%. The ddPCR assay had better analytical performances than the PGX assay with 100% specificity, 100% sensitivity and a detection limit down to 0.5% on IDH1R132 , IDH2R140 and IDH2R172 codons, and a high correlation with NGS results. Therefore, the new highly multiplexed ddPCR is a fast and cost-effective assay that meets most clinical needs to identify and follow cancer patients in the era of anti-IDH1/2-targeted therapies.

Nestin as a diagnostic and prognostic marker for combined hepatocellular-cholangiocarcinoma.

BACKGROUND & AIMS: Combined hepatocellular-cholangiocarcinoma (cHCC-CCA) is a rare primary liver cancer (PLC) associated with a poor prognosis. Given the challenges in its identification and its clinical implications, biomarkers are critically needed. We aimed to investigate the diagnostic and prognostic value of the immunohistochemical expression of Nestin, a progenitor cell marker, in a large multicentric series of PLCs. METHODS: We collected 461 cHCC-CCA samples from 32 different clinical centers. Control cases included 368 hepatocellular carcinomas (HCCs) and 221 intrahepatic cholangiocarcinomas (iCCAs). Nestin immunohistochemistry was performed on whole tumor sections. Diagnostic and prognostic performances of Nestin expression were determined using receiver-operating characteristic curves and Cox regression modeling. RESULTS: Nestin was able to distinguish cHCC-CCA from HCC with AUCs of 0.85 and 0.86 on surgical and biopsy samples, respectively. Performance was lower for the distinction of cHCC-CCA from iCCA (AUCs of 0.59 and 0.60). Nestin, however, showed a high prognostic value, allowing identification of the subset of cHCC-CCA ("Nestin High", >30% neoplastic cells with positive staining) associated with the worst clinical outcome (shorter disease-free and overall survival) after surgical resection and liver transplantation, as well as when assessment was performed on biopsies. CONCLUSION: We show in different clinical settings that Nestin has diagnostic value and that it is a useful biomarker to identify the subset of cHCC-CCA associated with the worst clinical outcome. Nestin immunohistochemistry may be used to refine risk stratification and improve treatment allocation for patients with this highly aggressive malignancy. LAY SUMMARY: There are different types of primary liver cancers (i.e. cancers that originate in the liver). Accurately identifying a specific subtype of primary liver cancer (and determining its associated prognosis) is important as it can have a major impact on treatment allocation. Herein, we show that a protein called Nestin could be used to refine risk stratification and improve treatment allocation for patients with combined hepatocellular carcinoma, a rare but highly aggressive subtype of primary liver cancer.

High prevalence of unusual KRAS, NRAS, and BRAF mutations in POLE-hypermutated colorectal cancers.

Exonucleasic domain POLE (edPOLE) mutations, which are responsible for a hypermutated tumor phenotype, occur in 1-2% of colorectal cancer (CRC) cases. These alterations represent an emerging biomarker for response to immune checkpoint blockade. This study aimed to assess the molecular characteristics of edPOLE-mutated tumors to facilitate patient screening. Based on opensource data analysis, we compared the prevalence of edPOLE mutations in a control group of unselected CRC patients (n = 222) vs a group enriched for unusual BRAF/RAS mutations (n = 198). Tumor mutational burden (TMB) and immune infiltrate of tumors harboring edPOLE mutations were then analyzed. In total, 420 CRC patients were analyzed: 11 edPOLE-mutated tumors were identified, most frequently in microsatellite (MMR)-proficient young (< 70 years) male patients, with left-sided tumors harboring noncodon 12 KRAS mutation. The prevalence of edPOLE-mutated tumors in the control vs the experimental screening group was, respectively, 0.45% (n = 1) vs 5.0% (n = 10). Among the 11 edPOLE-mutated cases, two had a low TMB, three were hypermutated, and six were ultramutated. EdPOLE-mutated cases had a high CD8+ tumor-infiltrating lymphocyte (TIL) infiltration. These clinicopathological and molecular criteria may help to identify edPOLE mutations associated with a high TMB in CRC, and improve the selection of patients who could benefit from immunotherapy.

Immune Profiling of Combined Hepatocellular- Cholangiocarcinoma Reveals Distinct Subtypes and Activation of Gene Signatures Predictive of Response to Immunotherapy.

PURPOSE: Combined hepatocellular-cholangiocarcinoma (cHCC-CCA) is a rare malignancy associated with an overall poor prognosis. We aimed to investigate the immune profile of cHCC-CCA and determine its impact on disease outcome. EXPERIMENTAL DESIGN: We performed a multicenter study of 96 patients with cHCC-CCA. Gene expression profile was analyzed using nCounter PanCancer IO 360 Panel. Densities of main immune cells subsets were quantified from digital slides of IHC stainings. Genetic alterations were investigated using targeted next-generation sequencing. RESULTS: Two main immune subtypes of cHCC-CCA were identified by clustering analysis: an "immune-high" (IH) subtype (57% of the cases) and an "immune-low" (IL) subtype (43% of the cases). Tumors classified as IH showed overexpression of genes related to immune cells recruitment, adaptive and innate immunity, antigen presentation, cytotoxicity, immune suppression, and inflammation (P < 0.0001). IH cHCC-CCAs also displayed activation of gene signatures recently shown to be associated with response to immunotherapy in patients with HCC. Quantification of immunostainings confirmed that IH tumors were also characterized by higher densities of immune cells. Immune subtypes were not associated with any genetic alterations. Finally, multivariate analysis showed that the IH subtype was an independent predictor of improved overall survival. CONCLUSIONS: We have identified a subgroup of cHCC-CCA that displays features of an ongoing intratumor immune response, along with an activation of gene signatures predictive of response to immunotherapy in HCC. This tumor subclass is associated with an improved clinical outcome. These findings suggest that a subset of patients with cHCC-CCA may benefit from immunomodulating therapeutic approaches.

Idylla MSI test as a new tool for microsatellite instability detection in sebaceous tumours and keratoacanthomas.

AIM: Sebaceous tumours and keratoacanthomas can be associated with mismatch repair (MMR) deficiency and thus microsatellite instability (MSI). In such tumours, MSI phenotype could be an argument to search for an underlying Muir-Torre syndrome (MTS). MTS has been recognised as a variant of Lynch syndrome, characterised by a deficiency of the MMR proteins. In Lynch syndrome, the sensitivity and specificity of the techniques used to detect MSI is well described, which is not the case for skin tumours. In our hands, immunohistochemistry is a sensitive and specific method to detect MMR deficiency in those tumours. Contrasting with tumours of Lynch spectrum, sensitivity and specificity of molecular methods has not been extensively studied. This study aimed at evaluating two molecular methods to detect MSI phenotype in MTS associated tumours: a commonly used pentaplex PCR using Bethesda markers and the fully automated method using the Idylla MSI assay. METHODS: A comparison between PCR, and Idylla was performed on 39 DNA extracted from cutaneous tumours. Immunohistochemistry was used as the gold standard to calculate sensitivity and specificity of both molecular techniques. RESULTS: Concordant results were found in 32 cases (82%) with pentaplex PCR and in 36 cases (92%) with Idylla. The sensitivity of pentaplex PCR to detect MSI phenotype was 76% whereas Idylla sensitivity was 90%. CONCLUSION: Idylla is more performant than PCR, for the detection of MSI in MTS-associated tumours and is a reliable additional technique to help detecting MTS in these tumours.

Detection of microsatellite instability in a panel of solid tumours with the Idylla MSI Test using extracted DNA.

AIM: During the last few years, determination of microstatellite instability (MSI) status has become a routine part of clinical practice, essentially to detect Lynch syndrome. Recently, MSI testing has increased with the development of immunotherapy and has expanded to a large panel of solid tumours. The aim of our work was to evaluate a fully automated system developed by Biocartis, the Idylla MSI Test, which performs an MSI analysis within 150 min. METHODS: A comparison between pentaplex PCR, immunohistochemistry and Idylla MSI Test was performed in 53 colorectal carcinoma samples, 7 small intestine adenocarcinomas, 15 duodenal and pancreatic adenocarcinomas, 16 gastric tumours, 15 endometrial adenocarcinomas, 5 ovarian carcinomas and 4 cases of urinary tract tumours using extracted DNA. Limit-of-detection (LOD) experiment was also done using a commercial DNA known to harbour MSI phenotype. RESULTS: The overall sensitivity was 94% and the overall specificity was 100%. Two invalid and three false-negative results were observed. Our experiments showed that the amount of DNA loaded into the cartridge was decisive and should be superior to 25 ng. LOD comprised between 4% and 8%. CONCLUSION: Overall, we have demonstrated that the Idylla MSI Test is a rapid and valid option to detect MSI phenotype which can be used in a large panel of solid tumours.

Predictive molecular pathology in the time of coronavirus disease (COVID-19) in Europe.

AIMS: Lung cancer predictive biomarker testing is essential to select advanced-stage patients for targeted treatments and should be carried out without delays even during health emergencies, such as the coronavirus (COVID-19) outbreak. METHODS: Fifteen molecular laboratories from seven different European countries compared 4 weeks of national lockdown to a corresponding period in 2019, in terms of tissue and/or plasma-based molecular test workload, analytical platforms adopted, number of cases undergoing programmed death-ligand1 (PD-L1) expression assessment and DNA-based molecular tests turnaround time. RESULTS: In most laboratories (80.0%), tissue-based molecular test workload was reduced. In 40.0% of laboratories (6/15), the decrease was >25%, and in one, reduction was as high as 80.0%. In this instance, a concomitant increase in liquid biopsy was reported (60.0%). Remarkably, in 33.3% of the laboratories, real-time PCR (RT-PCR)-based methodologies increased, whereas highly multiplexing assays approaches decreased. Most laboratories (88.9%) did not report significant variations in PD-L1 volume testing. CONCLUSIONS: The workload of molecular testing for patients with advanced-stage lung cancer during the lockdown showed little variations. Local strategies to overcome health emergency-related issues included the preference for RT-PCR tissue-based testing methodologies and, occasionally, for liquid biopsy.

The BCL-2 family protein inhibitor ABT-737 as an additional tool for the treatment of EBV-associated post-transplant lymphoproliferative disorders.

Post-transplant lymphoproliferative disorders (PTLD) and Burkitt's lymphoma (BL) are B-cell malignancies strongly associated with Epstein-Barr virus (EBV) infection. In these lymphoproliferative disorders, EBV infection induces an increase in the expression of the anti-apoptotic protein BCL-2. Given its chemoprotective effect, BCL-2 constitutes an attractive target for new therapeutic strategies for EBV-positive B-cell malignancies. Here, we show that ABT-737, a small inhibitor of BCL-2, BCL-X(L), and BCL-w, strongly induced apoptosis in vitro in EBV-positive lymphoblastoid cell lines (which is a model for PTLD), whereas BL was less sensitive. ABT-737 reduced tumor growth and increased the overall survival of mice in a xenograft model of PTLD but had no effect on BL xenograft mice. ABT-737 combined with a low dose of cyclophosphamide, a major component of the conventional CHOP chemotherapy regimen for BL patients, reduced tumor growth during treatment but failed to improve the overall survival of BL xenograft mice. By contrast, the combination of ABT-737 and rituximab, one of the main options for the treatment of PTLD, was highly efficient and induced approximately 70% remission in PTLD xenograft mice. These results suggest that the use of agents targeting BCL-2, either alone or in combination with other conventional drugs, represents a novel promising approach for post-transplant EBV-positive B lymphoproliferative disorders.

Evaluation of the Idylla system to detect the EGFRT790M mutation using extracted DNA.

INTRODUCTION: During the last few years, detection of epidermal growth-factor-receptor (EGFR)-activating mutations has become a routine part of clinical practice because of their importance in choosing the optimal treatment strategy for non-small-cell lung cancers (NSCLCs). The emergence of third-generation EGFR-tyrosine-kinase inhibitors required the implementation of sensitive methods to detect the subclonal EGFRT790M mutation. Clinical implications make it essential to rapidly search for the T790M mutation, which is a real challenge for laboratories. The aim of this study was to compare performances of next-generation sequencing (NGS), one of the most frequently used molecular biology methods, and Idylla EGFR-Mutation Assay (henceforth Idylla), a fully automated real-time polymerase chain reaction (PCR) that is increasingly used in pathology laboratories, to detect the EGFRT790M mutation using DNA. METHODS: This retrospective study used 47 DNA samples extracted from NSCLC biopsies that previous NGS identified as: 29 harboring EGFR and T790M resistance mutations, 11 EGFR-activating mutation without T790 M and 7 wild-type EGFR. EGFRT790M limit-of-detection (LOD) experiments used a commercial DNA known to harbor that mutation. RESULTS: Idylla detected primary EGFR-activating mutations and the T790 M mutation in 97.5 % and 65.5 % of the cases, respectively. The results of this retrospective analysis and LOD experiments showed that the Idylla should only be used to detect EGFR mutations in samples with > 25 ng of DNA and > 10 % tumor cells. CONCLUSIONS: Idylla was able to rapidly detect EGFR-activating mutations but detecting subclone mutations, like T790M, with < 25 ng of good-quality DNA or < 10 % tumor cells (variant allele frequency below the assay's validated LOD) was not always reliable.

Defining signatures of peripheral T-cell lymphoma with a targeted 20-marker gene expression profiling assay.

Peripheral T-cell lymphoma comprises a heterogeneous group of mature non-Hodgkin lymphomas. Their diagnosis is challenging, with up to 30% of cases remaining unclassifiable and referred to as "not otherwise specified". We developed a reverse transcriptase-multiplex ligation-dependent probe amplification gene expression profiling assay to differentiate the main T-cell lymphoma entities and to study the heterogeneity of the "not specified" category. The test evaluates the expression of 20 genes, including 17 markers relevant to T-cell immunology and lymphoma biopathology, one Epstein-Barr virus-related transcript, and variants of RHOA (G17V) and IDH2 (R172K/T). By unsupervised hierarchical clustering, our assay accurately identified 21 of 21 ALK-positive anaplastic large cell lymphomas, 16 of 16 extranodal natural killer (NK)/T-cell lymphomas, 6 of 6 hepatosplenic T-cell lymphomas, and 13 of 13 adult T-cell leukemia/lymphomas. ALK-negative anaplastic lymphomas (n=34) segregated into one cytotoxic cluster (n=10) and one non-cytotoxic cluster expressing Th2 markers (n=24) and enriched in DUSP22-rearranged cases. The 63 TFH-derived lymphomas divided into two subgroups according to a predominant TFH (n=50) or an enrichment in Th2 (n=13) signatures. We next developed a support vector machine predictor which attributed a molecular class to 27 of 77 not specified T-cell lymphomas: 17 TFH, five cytotoxic ALK-negative anaplastic and five NK/T-cell lymphomas. Among the remaining cases, we identified two cell-of-origin subgroups corresponding to cytotoxic/Th1 (n=19) and Th2 (n=24) signatures. A reproducibility test on 40 cases yielded a 90% concordance between three independent laboratories. This study demonstrates the applicability of a simple gene expression assay for the classification of peripheral T-cell lymphomas. Its applicability to routinely-fixed samples makes it an attractive adjunct in diagnostic practice.

Resistance mechanisms to osimertinib in EGFR-mutated advanced non-small-cell lung cancer: A multicentric retrospective French study.

OBJECTIVES: The understanding of histo-molecular mechanisms associated with resistance to osimertinib is a critical step to define the optimal treatment strategy in advanced EGFR-mutated Non-Small-Cell-Lung-Cancer (NSCLC). MATERIALS AND METHODS: We performed a multicentric retrospective analysis on a cohort of consecutive patients treated with osimertinib for an advanced EGFR-mutated NSCLC and collected histo-molecular data from plasma and tumor samples at the time of progression. Next-generation sequencing (NGS) was performed for all samples. Best Overall Response Rate (ORR), Progression Free Survival (PFS), Overall Survival (OS) and data on treatment post-progression efficacy were also collected. RESULTS: Two-hundred and twenty-six patients were included from 9 Academic French Hospitals between April 2015-October 2018. Osimertinib was given in second-line or more in 219 patients (97%). Best ORR was 52% and best central nervous system ORR was 56%. Median PFS and OS were 9.5 months (IQR 4.0-17.2) and 24 months (IQR 12.4-NR) respectively. At the time of analysis, 150 patients (66%) had tumor progression. Among them, 73 contributive samples (56 tumor biopsies) were available. The most frequent molecular alterations were C797S mutation (n = 9 (13%)) and MET amplification (n = 8 (11%)). Histologic transformation occurred in 5 patients (9% of tumor biopsies). In T790M + NSCLC, loss of T790 M occurred in 68% of cases. Median PFS and OS with treatment beyond progression were 6.0 months (IQR 2.0-10.4) and 15.1 months (IQR 6.7-NR) respectively and longer in case of osimertinib continuation beyond progression. CONCLUSION: We confirmed the efficacy of osimertinib in patients with advanced EGFR mutation positive NSCLC. At progression, the most frequent molecular alterations were MET amplification and C797S mutation.

Isolated 5' Signals Are an Atypical Pattern To Be Considered as Positive for ALK Rearrangement: A Brief Report of Three Cases and Review of the Literature.

Anaplastic lymphoma kinase (ALK) rearrangement is reported in 3% to 8% of patients with lung adenocarcinoma and can be detected by fluorescent in situ hybridization (FISH) or indirectly by immunohistochemistry. In FISH assay, isolated 5' signal (loss of 3' signal) is usually considered negative. We report three young nonsmoking patients with stage IV lung adenocarcinoma. Strong ALK expression in tumor cells detected by immunohistochemistry was observed in all cases, but FISH revealed an isolated 5' signal pattern. Massive parallel "next-generation" sequencing was performed in two patients and confirmed ALK rearrangement. The three patients were treated and responded to crizotinib after 14, 10, and 31 months.

Multiple Ways to Detect IDH2 Mutations in Angioimmunoblastic T-Cell Lymphoma from Immunohistochemistry to Next-Generation Sequencing.

Angioimmunoblastic T-cell lymphoma (AITL) is a peripheral T-cell lymphoma associated with chemoresistance and a poor prognosis. Various nonsynonymous mutations in the R172 residue of IDH2 are present in 20% to 30% of AITL patients. In addition to their diagnostic value, these mutations are potentially targetable, especially by isocitrate dehydrogenase (IDH) 2 inhibitor, and therefore their identification in a routine setting is clinically relevant. However, in AITL, the neoplastic cells may be scarce, making the identification of molecular anomalies difficult. We evaluated the diagnostic value of different methods to detect IDH2 mutations in formalin-fixed, paraffin-embedded tumor samples. Immunohistochemistry with an anti-IDH2 R172K antibody, Sanger sequencing, high-resolution melting PCR, allele-specific real-time quantitative PCR, and next-generation sequencing (NGS) were applied to biopsy specimens from 42 AITL patients. We demonstrate that the IDH2 R172K antibody is specific to this amino acid substitution and highly sensitive for the detection of the IDH2R172K variant, the most frequent substitution in this disease. In our study, NGS and allele-specific real-time quantitative PCR displayed a good sensitivity, detecting 96% and 92% of IDH2 mutations, respectively, in contrast to Sanger sequencing and high-resolution melting PCR, which showed a significantly lower detection rate (58% and 42%, respectively). These results suggest that a combination of immunohistochemistry and AS-PCR or NGS should be considered for the identification of IDH2 mutations in AITL in a routine setting.

Expression of TFH Markers and Detection of RHOA p.G17V and IDH2 p.R172K/S Mutations in Cutaneous Localizations of Angioimmunoblastic T-Cell Lymphomas.

Skin biopsies of 41 angioimmunoblastic T-cell lymphoma patients were retrospectively analyzed for the expression of follicular helper T-cell (TFH) markers, Epstein-Barr virus (EBV), and the presence of RHOA (p.G17V) and IDH2 (p.R172K/S) mutations using allele-specific polymerase chain reaction. We categorized cases into 4 distinctive patterns: (1) low-density lymphocytic perivascular infiltrates (n=11), (2) dense perivascular infiltrates with atypical cells and occasional inflammatory cells (n=13), (3) diffuse infiltrates reminiscent of angioimmunoblastic T-cell lymphoma (n=4), or (4) other aspects (n=13). Two EBV and 2 plasmacytoid lymphoproliferative disorders were seen. We observed variable expression of TFH markers (CD10 [50%], BCLB6 [84%], PD1 [94%], CXCL13 [84%], and ICOS [97.5%]), and EBV B-blasts (26%). A TFH phenotype was identified in 82% and 73%, respectively, of cases with the most challenging patterns 1 and 2. TFH markers and EBV can thus help for diagnosis and are detected in samples with low-density infiltrates. We found RHOA G17V and IDH2 R172K/S mutations in the skin in 14/18 (78%) and 3/16 (19%) cases, respectively. The RHOA G17V mutation was identified in a proportion of biopsies with patterns 1 and 2, which represent a diagnostic challenge. The RHOA G17V mutation was detected both in the skin and lymph node (LN) biopsies in 7/9 (64%) cases, and in only the skin or the LN of 1 sample each. The frequency of RHOA G17V mutation was similar to that reported in LNs. It may represent a sensitive diagnostic marker in the skin, helpful in cases with low-density infiltrates.