Se ' zary syndrome is an aggressive and disseminated form of cutaneous T-cell lymphoma associated with dismal prognosis in which the histone deacetylase inhibitor romidepsin has shown remarkable activity as a single agent. However, clinical responses to romidepsin are typically transient, highlighting the need for more effective therapies. In this study, we show synergistic antilymphoma effects of romidepsin in combination with mechlorethamine, an alkylating agent, in cutaneous T-cell lymphoma cell lines and primary samples with strong antitumor effects in an in vivo model of Se ' zary syndrome. Mechanistically, gene expression profiling points to abrogation of Jak/signal transducer and activator of transcription (STAT) signaling as an important mediator of this interaction. Consistently, the combination of mechlorethamine plus romidepsin resulted in down regulation of STAT5 phosphorylation in romidepsin-sensitive cell lines and primary Se ' zary syndrome samples, but not in romidepsin-resistant tumors. Moreover, in further support of Jak/STAT signaling as a modulator of romidepsin activity in cutaneous T-cell lymphoma, treatment with romidepsin in combination with Jak inhibitors resulted in markedly increased therapeutic responses. Overall, these results support a role for romidepsin plus mechlorethamine in combination in the treatment of cutaneous T-cell lymphoma and uncover a previously unrecognized role for Jak/STAT signaling in the response to romidepsin and romidepsin-based combination therapies in Se ' zary syndrome.

To the Editor: Worldwide reports describe cutaneous findings, including maculopapules, pseudo-chilblain, vesicles, urticaria, livedo, and multisystem inflammatory syndrome, as manifestations of coronavirus disease 2019 (COVID-19).1Recalcati S. Cutaneous manifestations in COVID-19: a first perspective.J Eur Acad Dermatol Venereol. 2020; 34: e212-e213Crossref PubMed Scopus (913) Google Scholar, 2Galván Casas C. Català A. Carretero Hernández G. et al.Classification of the cutaneous manifestations of COVID-19: a rapid prospective nationwide consensus study in Spain with 375 cases.Br J Dermatol. 2020; 183: 71-77Crossref PubMed Scopus (939) Google Scholar, 3de Masson A. Bouaziz J.D. Sulimovic L. et al.Chilblains are a common cutaneous finding during the COVID-19 pandemic: a retrospective nationwide study from France.J Am Acad Dermatol. 2020; 83: 667-670Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar, 4Jia J.L. Kamceva M. Rao S.A. et al.Cutaneous manifestations of COVID-19: a preliminary review.J Am Acad Dermatol. 2020; 83: 687-690Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar, 5Shulman S.T. Pediatric coronavirus disease-2019–associated multisystem inflammatory syndrome.J Pediatr Infect Dis Soc. 2020; 9: 285-286Crossref PubMed Scopus (24) Google Scholar Here, we report on the cutaneous findings observed in hospitalized COVID-19–positive patients at Columbia University Irving Medical Center by the adult dermatology consultation service between March 25, 2020, and May 1, 2020. Cases were included if COVID-19 was most likely associated with or contributed to skin findings observed after other potential causes were excluded. Although previous studies on the cutaneous manifestations of COVID-19 have focused primarily on outpatients and those with mild disease,2Galván Casas C. Català A. Carretero Hernández G. et al.Classification of the cutaneous manifestations of COVID-19: a rapid prospective nationwide consensus study in Spain with 375 cases.Br J Dermatol. 2020; 183: 71-77Crossref PubMed Scopus (939) Google Scholar,3de Masson A. Bouaziz J.D. Sulimovic L. et al.Chilblains are a common cutaneous finding during the COVID-19 pandemic: a retrospective nationwide study from France.J Am Acad Dermatol. 2020; 83: 667-670Abstract Full Text Full Text PDF PubMed Scopus (145) Google Scholar this study provides insights into the cutaneous manifestations of severe disease. This study was approved by the institutional review board at Columbia University Irving Medical Center. Nine patients developed signs of acral ischemia, including duskiness, necrosis, and bulla (Fig 1, A). Eight of these patients with acral ischemia required admission to the intensive care unit, and the location of ischemia included fingers, toes, ears, and genital skin (Supplemental Table I available via Mendeley at https://data.mendeley.com/datasets/nywrhw3d8y/1). Two patients with severe lesions had a confirmed deep vein thrombosis, highlighting the potential for thrombotic events (cases 7 and 9). Three patients developed livedo racemosa, 1 of whom had livedo as a presenting sign of COVID-19 and 2 of whom developed livedo during their hospitalization. One patient developed vesicles in association with his livedoid eruption (case 10). All 3 patients required intensive care unit admission and had evidence of coagulopathy (cases 10 to 12) (Fig 1, B). Two patients developed evidence of bleeding diatheses. One, a critically ill man in his 60s (case 14), developed petechiae, with perivascular lymphocytic infiltrate with prominent red blood cell extravasation on biopsy. He was mildly thrombocytopenic (114,000 per μL) and had mild coagulopathy (prothrombin time/international normalized ratio 14.7 seconds/1.2, activated partial thromboplastin time 36.2 seconds, and D-dimer 3.95 μg/mL). The other patient, a man in his 70s, developed purpura fulminans with areas of necrosis and superficial desquamation on the chest and axilla (case 13). Laboratory data were consistent with disseminated intravascular coagulation, and ultimately, the patient died. Erythema multiforme–like lesions were observed in a woman in her 80s 1 week following discharge after treatment for COVID-19–associated pneumonia (case 15). She had no other risk factors for development of erythema multiforme–like lesions, including active herpes simplex virus infection, and received no new medications in the last 3 days of her hospitalization or on discharge. No pseudo-chilblain (COVID toes) was observed in the hospitalized patients evaluated. Pseudo-chilblain may be more strongly associated with mild disease or present after acute illness has resolved. In summary, cutaneous manifestations in hospitalized COVID-19 patients are varied and are an important part of this potentially life-threatening illness. Findings in critically ill patients may differ from those in outpatients with mild disease. Limitations of our study include that cases were limited to a single institution, there was a lack of histology on the majority of cases, and there was inability to establish the pathophysiologic role of severe acute respiratory syndrome coronavirus 2 in the reported skin diseases. Additionally, less severe skin findings may not have required dermatology consultation. Further large-scale cohort studies with classification of histology are necessary to better describe both the frequency and etiology of these findings.

WWW.MDEDGE.COM/DERMATOLOGY An 82-year-old man presented with acute abdominal pain and distension as well as an abdominal rash of 4 months’ duration that was expanding despite treatment with topical miconazole. He had a history of melanoma and bladder cancer treated with cystoprostatectomy. He previously was diagnosed with candidiasis of his urostomy and was taking oral fluconazole. Physical examination revealed a large, welldemarcated, erythematous, smooth plaque covering the entire abdomen, scrotum, penis, inguinal folds, and bilateral upper thighs, with several satellite plaques and firm nodules clustered around the umbilicus. An 8-mm punch biopsy of a periumbilical nodule was performed.

The potential role of viruses as oncogenic triggers in cutaneous T-cell lymphoma (CTCL) pathogenesis is a subject of ongoing investigation. CTCL occurs with an increased incidence in immunosuppressed patients (Nikolaou et al., 2015Nikolaou V. Papadavid E. Economidi A. Marinos L. Moustou E. Karampidou K. et al.Mycosis fungoides in the era of antitumour necrosis factor-alpha treatments.Br J Dermatol. 2015; 173: 590-593Crossref PubMed Scopus (23) Google Scholar, Pomerantz et al., 2010Pomerantz R.G. Campbell L.S. Jukic D.M. Geskin L.J. Posttransplant cutaneous T-cell lymphoma: case reports and review of the association of calcineurin inhibitor use with posttransplant lymphoproliferative disease risk.Arch Dermatol. 2010; 146: 513-516Crossref PubMed Scopus (22) Google Scholar, Wilkins et al., 2006Wilkins K. Turner R. Dolev J.C. LeBoit P.E. Berger T.G. Maurer T.A. Cutaneous malignancy and human immunodeficiency virus disease.J Am Acad Dermatol. 2006; 54: 189-206Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar). Spectratyping studies have shown depleted T-cell receptor diversity in CTCL patients similar to that seen in patients with advanced HIV, and Ingenuity Pathway Analysis (Qiagen, Hilden, Germany) shows increased expression of genes critical to host viral response (Yawalkar et al., 2003Yawalkar N. Ferenczi K. Jones D.A. Yamanaka K. Suh K.Y. Sadat S. et al.Profound loss of T cell receptor repertoire complexity in cutaneous T-cell lymphoma.Blood. 2003; 102: 4059-4066Crossref PubMed Scopus (140) Google Scholar). However, no consistent association between a viral pathogen and CTCL has been established (Mirvish et al., 2013Mirvish J.J. Pomerantz R.G. Falo Jr., L.D. Geskin L.J. Role of infectious agents in cutaneous T-cell lymphoma: facts and controversies.Clin Dermatol. 2013; 31: 423-431Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). The development of high-throughput sequencing (HTS) has provided powerful new tools for pathogen discovery, yet HTS studies have not identified viral sequences in CTCL (Dereure et al., 2013Dereure O. Cheval J. Du Thanh A. Pariente K. Sauvage V. Claude Manuguerra J. et al.No evidence for viral sequences in mycosis fungoides and Sezary syndrome skin lesions: a high throughput sequencing approach.J Invest Dermatol. 2013; 133: 853-855Abstract Full Text Full Text PDF PubMed Scopus (17) Google Scholar, Dulmage et al., 2015Dulmage B.O. Feng H. Mirvish E. Geskin L. Black cat in a dark room: the absence of a directly oncogenic virus does not eliminate the role of an infectious agent in cutaneous T-cell lymphoma pathogenesis.Br J Dermatol. 2015; 172: 1449-1451Crossref PubMed Scopus (10) Google Scholar, Lee et al., 2012Lee C.S. Ungewickell A. Bhaduri A. Qu K. Webster D.E. Armstrong R. et al.Transcriptome sequencing in Sezary syndrome identifies Sezary cell and mycosis fungoides-associated lncRNAs and novel transcripts.Blood. 2012; 120: 3288-3297Crossref PubMed Scopus (71) Google Scholar). One explanation for failure is that viral burden in samples may be below the threshold for detection in complex backgrounds of host nucleic acid. To address this possibility, we used a positive selection method for HTS with enhanced sensitivity (Briese et al., 2015Briese T. Kapoor A. Mishra N. Jain K. Kumar A. Jabado O.J. et al.Virome capture sequencing enables sensitive viral diagnosis and comprehensive virome analysis.MBio. 2015; 6 (e01491–15)Crossref PubMed Scopus (209) Google Scholar). The Virome Capture Sequencing Platform for Vertebrate Viruses (VirCapSeq-VERT; Roche, Pleasanton, CA) is a positive-selection probe-based method that targets all 207 known vertebrate viruses, allowing for a 100- to 10,000-fold increase in the number of viral reads over traditional HTS, a degree of sensitivity comparable to targeted real-time PCR. Over 90% of genome recovery can be achieved with only 100 viral copies in 50 ng of whole-blood nucleic acid. Superior to real-time PCR, however, VirCapSeq-VERT can detect viruses that differ from known sequences by 40% (Briese et al., 2015Briese T. Kapoor A. Mishra N. Jain K. Kumar A. Jabado O.J. et al.Virome capture sequencing enables sensitive viral diagnosis and comprehensive virome analysis.MBio. 2015; 6 (e01491–15)Crossref PubMed Scopus (209) Google Scholar). We used VirCapSeq-VERT to search for viral sequences in mononuclear cells derived from the peripheral blood of leukemic CTCL patients with Sézary syndrome. In total, 27 million paired-end reads were generated using Illumina MiSeq, with an average of 2.25 million paired-end reads per sample. An average of 11% of the reads were removed after filtration; 74% of the reads were removed after host subtraction. Overall, 49,349 sequences showed partial homology to 23 unique viral species including human endogenous retroviruses (HERVs) H and K, human herpesvirus (HHV) 4 and 5, HIV, human T-lymphotropic virus-1 (HTLV-1), Lassa virus, Luna virus, bovine viral diarrhea virus, and Ngari virus (Table 1).Table 1Potential viral species counts from MegaBLASTViral SpeciesSample IDMaximum Sequence Length (nt)1Maximum sequence length and contig depth are in sample 1.Maximum Contig Depth (Number of Reads)1Maximum sequence length and contig depth are in sample 1.Significance123456BVDV8,264108,844—14—1,837108Library preparation artifactHERV-H/env598———9101452Proviral sequencesHERV-H/env62158————HERV-K60199641455287514Proviral sequencesHHV-424—————1453Low abundance/prevalenceHHV-512—————HIV-110—————844Nonfunctional pol polymerase protein geneHTLV-1153282869221452Proviral sequencesLassa virus118——26201451Likely misannotationLuna virus3,9123,8043,6731113723,53343226Likely misannotationNgari virus332026—18161452Likely misannotationAbbreviations: BVDV, bovine viral diarrhea virus; contig, contiguous sequence; HERV, human endogenous retrovirus; HHV, human herpes virus; HTLV-1, human lymphotropic virus 1; ID, identification; nt, nucleotide.1 Maximum sequence length and contig depth are in sample 1. Open table in a new tab Abbreviations: BVDV, bovine viral diarrhea virus; contig, contiguous sequence; HERV, human endogenous retrovirus; HHV, human herpes virus; HTLV-1, human lymphotropic virus 1; ID, identification; nt, nucleotide. Bovine viral diarrhea virus sequences are common in bovine serum products used in sample collection and thus were excluded from further analysis. HERV-H/env62 and HERV-H/env59 sequences were present in 2 of 6 and 3 of 6 samples, respectively, matching to a proviral copy. The HHV-4 and HHV-5 sequences were present in 1 of 6 samples tested; 36 reads matched to the partial HHV genome. HIV-1 sequences were present in 1 of 6 samples, partially matching to a nonfunctional pol protein gene. Short sequences with partial homology to Lassa, Ngari, and Luna viruses were present in 4 of 6, 5 of 6, and 6 of 6 samples, respectively. Detailed analysis indicated that these sequences were 100% homologous to human genomic sequences (accession number KF478765.1 for Lassa, KJ716849.1 for Ngari, and AB972431.1 and KX121618.1 for Luna). Thus, the reference sequences likely represent misannotations. HERV-K and HTLV-1 sequences were present in 6 of 6 samples. The HERV-K sequences comprised partially coding proviral sequences, and the HTLV-1 sequences matched to defective proviral sequences, both showing high similarity to the host genome. We conducted this study using VirCapSeq-VERT to search for viral sequences in patients with CTCL. VirCapSeq-VERT's ability to detect all known vertebrate viruses offers the most sensitive detection methods compared with conventional HTS and targeted reverse transcription PCR. No substantive coding sequences for viral pathogens or unknown viruses, or evidence for active infection, were obtained. All six Sézary syndrome samples expressed partial, noncoding sequences for HERV-K and HTLV-1, but these sequences had low read counts and contiguous sequence (contig) depths, and were thus insufficient to be considered as positive results. Nevertheless, their presence is of interest, because both viruses have been previously implicated in CTCL. HTLV-1's role in CTCL has been debated at length (Mirvish et al., 2013Mirvish J.J. Pomerantz R.G. Falo Jr., L.D. Geskin L.J. Role of infectious agents in cutaneous T-cell lymphoma: facts and controversies.Clin Dermatol. 2013; 31: 423-431Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). HERVs are expressed in all human tissues but have been implicated in cancer pathogenesis, and increased expression of HERVs has been seen in CTCL patients (Fava et al., 2016Fava P. Bergallo M. Astrua C. Brizio M. Galliano I. Montanari P. et al.Human endogenous retrovirus expression in primary cutaneous T-cell lymphomas.Dermatology. 2016; 232: 38-43Crossref PubMed Scopus (17) Google Scholar, Maliniemi et al., 2013Maliniemi P. Vincendeau M. Mayer J. Frank O. Hahtola S. Karenko L. et al.Expression of human endogenous retrovirus-w including syncytin-1 in cutaneous T-cell lymphoma.PLoS One. 2013; 8: e76281Crossref PubMed Scopus (42) Google Scholar). Sequencing of healthy control patient samples with matched demographics was not available for this pilot study for direct comparison, but further investigation may elucidate the significance of these partial noncoding sequences and whether these viruses could play a genuine role in CTCL pathogenesis. Six Sézary syndrome patients diagnosed according to the World Health Organization-European Organization for Research and Treatment of Cancer criteria were enrolled after receiving Columbia University Medical Center Institutional Review Board approval. Written informed patient consent was obtained from all patients (5 women, 1 man; age range = 58–82 years; 3 Caucasian, 2 Hispanic, and 1 African American). Whole blood from each patient was collected. Peripheral blood mononuclear cells were isolated via density gradient centrifugation (Ficoll-Plaque; Millipore Sigma, St. Louis, MO), and plasma was isolated via centrifugation. Total nucleic acid (TNA) was extracted from the plasma and peripheral blood mononuclear cells using NucliSENS easyMAG (BioMérieux, Marcy l'Etoile, France). TNAs were processed according to Center for Infection and Immunity standard operating procedures for VirCapSeq-VERT. Briefly, the VirCapSeq-VERT probe library was added to the TNAs. Positively selected and enriched sequences underwent PCR amplification and sequencing using the Illumina (San Diego, CA) MiSeq platform. The 150-nucleotide–long paired end reads were generated (an average of more than 2 million reads per sample). HTS data were analyzed using the Center for Infection and Immunity bioinformatics viral discovery pipeline: the de-multiplexed fastq files were filtered for low-quality and low-complexity reads, adaptors were removed, reads were trimmed, and short reads were discarded. These pre-processed reads were subjected to computational subtraction against human reference databases from the National Center for Biotechnology Information (NCBI) to remove host background sequences. The databases used for host subtraction include human genomic, ribosomal, and mitochondrial sequences. The host-subtracted reads were then assembled de novo using MIRA (Bastien Chevreux, Lexington, MA) assembler, version 4.0. Contigs and singletons were annotated using homology search program BLAST+ from NCBI (Silver Spring, MD). The sequences were screened for highly similar sequences using megaBLAST against the GenBank nucleotide database. To ensure that mutant viral sequences were not discarded, sequences with low/no homology to the GenBank viral nucleotide database were reblasted using Blastx against the GenBank database. Contigs and reads were then mapped to reference genomes identified from BLAST using Geneious, version 6.0.6 (Biomatters INC, Newark, NJ). The sequence data supporting the results of this article are available in NCBI under accession number PRJNA415045 with biosample accession numbers SAMN07812478–83. Mary Elizabeth Anderson: http://orcid.org/0000-0003-1364-1248 Christina Chung Patrone: http://orcid.org/0000-0001-8906-7460 The authors state no conflict of interest. We thank Lisa Keller and Saky Yakas for their philanthropic support of cutaneous lymphoma research at Columbia University.

Cutaneous T-cell lymphoma (CTCL) is a clinically and genetically heterogeneous disease characterized by the infiltration of malignant CD4+ T lymphocytes in the skin. Mechlorethamine (nitrogen mustard, NM), an alkylating agent, and romidepsin, a histone deacetylase inhibitor, are two FDA-approved monotherapies for CTCL. In vitro analysis showed synergism of the drug combination in CTCL cell lines and primary samples from patients with Sézary Syndrome. We performed next generation RNA sequencing to elucidate the mechanism underlying the effects of this drug combination in CTCL. Three CTCL cell lines and CD4+ cells cultured from six primary samples were each treated with vehicle, 0.25μM NM, 1nM romidepsin, and the two drugs combined for 24 hours. Total RNA was extracted from the treated cells and was sequenced on an Illumina HiSeq 2500. Differential expression analysis, principal component analysis, KEGG pathway analysis, and gene set enrichment analysis (GSEA) were performed. We found a significantly larger overlap of genes enriched by romidepsin and the combination compared to NM and the combination (p<0.00001). A number of KEGG pathways were uniquely enriched with combination therapy compared to monotherapy, most notably downregulation of the JAK/STAT pathway in the cell lines. This result was supported by GSEA of primary samples, which showed significant downregulation of genes downregulated by JAK inhibitors and shRNA-JAK2 knockdown, and downregulation of genes upregulated by IL2 treatment. Constitutive activation of the JAK/STAT pathway has been identified previously in CTCL and our gene expression profiling data identifies this pathway as a promising target of the NM-romidepsin combination.

In the current era of checkpoint inhibitors, some patients with metastatic melanoma have shown a significant improvement in survival. However, optimization of immunotherapy is an ongoing effort. Monocyte-derived dendritic cell (MODC) vaccines have been shown in clinical trials to be safe and capable of inducing tumor-specific immunity as well as occasional objective clinical responses. Here, we conducted a three-arm pilot clinical study in 15 patients with metastatic melanoma to evaluate three types of MODC vaccines, differing only by strategies of tumor antigen delivery. MODCs were isolated from each patient and loaded with patients’ own melanoma cells as sources of antigens. Antigen loading was achieved ex vivo by fusing melanoma cells with MODCs, co-culturing melanoma cells with MODCs, or by pulsing MODCs with melanoma cell lysates. The vaccines were then injected into superficial lymph nodes using high-resolution ultrasound guidance. Primary end points included delayed-type hypersensitivity responses and positive ELISpot result, which measures interferon-γ production. Five of 15 patients achieved delayed-type hypersensitivity responses and six of 15 patients had positive ELISpot results. We demonstrated that the vaccines were safe and well-tolerated by all patients and produced immunological responses in all arms. Although MODC vaccine monotherapy has limited efficacy, combining this vaccine with other immunotherapies, such as checkpoint inhibitors, to engage multiple components of the immune system may be an effective and viable future approach.

Recently, several groups have conducted deep sequencing studies of small cohorts of cutaneous T-cell lymphoma (CTCL) genomes to identify disease driving mutations and therapeutic targets (Choi et al., 2015Choi J. Goh G. Walradt T. Hong B.S. Bunick C.G. Chen K. et al.Genomic landscape of cutaneous T cell lymphoma.Nat Genet. 2015; 47: 1011-1019Crossref PubMed Scopus (287) Google Scholar, da Silva Almeida et al., 2015da Silva Almeida A.C. Abate F. Khiabanian H. Martinez-Escala E. Guitart J. Tensen C.P. et al.The mutational landscape of cutaneous T cell lymphoma and Sézary syndrome.Nat Genet. 2015; 47: 1465-1470Crossref PubMed Scopus (272) Google Scholar, McGirt et al., 2015McGirt L.Y. Jia P. Baerenwald D.A. Duszynski R.J. Dahlman K.B. Zic J.A. et al.Whole-genome sequencing reveals oncogenic mutations in mycosis fungoides.Blood. 2015; 126: 508-519Crossref PubMed Scopus (154) Google Scholar, Prasad et al., 2016Prasad A. Rabionet R. Espinet B. Zapata L. Puiggros A. Melero C. et al.Identification of gene mutations and fusion genes in patients with Sézary syndrome.J Invest Dermatol. 2016; 136: 1490-1499Abstract Full Text Full Text PDF PubMed Scopus (60) Google Scholar, Ungewickell et al., 2015Ungewickell A. Bhaduri A. Rios E. Reuter J. Lee C.S. Mah A. et al.Genomic analysis of mycosis fungoides and Sézary syndrome identifies recurrent alterations in TNFR2.Nat Genet. 2015; 47: 1056-1060Crossref PubMed Scopus (209) Google Scholar, Wang et al., 2015Wang L. Ni X. Covington K.R. Yang B.Y. Shiu J. Zhang X. et al.Genomic profiling of Sézary syndrome identifies alterations of key T cell signaling and differentiation genes.Nat Genet. 2015; 47: 1426-1434Crossref PubMed Scopus (232) Google Scholar, Woollard et al., 2016Woollard W.J. Pullabhatla V. Lorenc A. Patel V.M. Butler R.M. Bayega A. et al.Candidate driver genes involved in genome maintenance and DNA repair in Sézary syndrome.Blood. 2016; 127: 3387-3397Crossref PubMed Scopus (77) Google Scholar). The low incidence rate of CTCL (Korgavkar et al., 2013Korgavkar K. Xiong M. Weinstock M. Changing incidence trends of cutaneous T-cell lymphoma.JAMA Dermatol. 2013; 149: 1295-1299Crossref PubMed Scopus (136) Google Scholar) makes it difficult to conduct such investigations; therefore, combining multiple small cohorts may increase the statistical power for identifying rarer mutations and key disease pathways (Park et al., 2017Park J. Yang J. Wenzel A.T. Ramachandran A. Lee W.J. Daniels J.C. et al.Genomic analysis of 220 CTCLs identifies a novel recurrent gain-of-function alteration in RLTPR (p.Q575E).Blood. 2017; 130: 1430-1440Crossref PubMed Scopus (95) Google Scholar) and may uncover genomic types of CTCL, which may guide future therapies via selective targeting of mutation-specific patient populations. For many cancers, integrated genomic datasets are accessible via centralized portals such as cBioPortal, but a similar resource for CTCL is not available. Here we have collected, selected, and re-analyzed individual datasets whenever raw data were available to generate high-quality, homogeneous data, and we present the compilation and analysis of this integrated dataset in CTCL, including matching genomic mutations and gene copy number profiling for all cases. Most importantly, our database can be expanded as more CTCL studies are published. We compiled genomic data of 139 CTCL cases from seven sequencing studies of mycosis fungoides and Sézary syndrome (SS) (Supplementary Table S1 online). We identified 11,520 single nucleotide mutations and 1,248 insertions/deletions in 121 SS cases (Supplementary Table S2 online), and 1,774 single nucleotide mutations and 37 insertions/deletions in 18 mycosis fungoides cases (Supplementary Table S3 online). In parallel, gene copy number analysis resulted in 682,298 and 60,067 genomic regions in SS and mycosis fungoides cases (Supplementary Tables S4 and S5 online), providing additional information and confirming consistency with previous studies (Lin et al., 2012Lin W.M. Lewis J.M. Filler R.B. Modi B.G. Carlson K.R. Reddy S. et al.Characterization of the DNA copy-number genome in the blood of cutaneous T-cell lymphoma patients.J Invest Dermatol. 2012; 132: 188-197Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar) (Supplementary Figure S1 online). To identify significantly mutated genes in the entire database, we took a Poisson statistics-based approach (Crescenzo et al., 2015Crescenzo R. Abate F. Lasorsa E. Tabbo F. Gaudiano M. Chiesa N. et al.Convergent mutations and kinase fusions lead to oncogenic STAT3 activation in anaplastic large cell lymphoma.Cancer Cell. 2015; 27: 516-532Abstract Full Text Full Text PDF PubMed Scopus (306) Google Scholar), taking into account the effective length of a gene (Nei and Gojobori, 1986Nei M. Gojobori T. Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions.Mol Biol Evol. 1986; 3: 418-426PubMed Google Scholar). Briefly, Poisson probabilities of observing a given number of mutations or more were calculated and corrected for multiple hypothesis testing by the Benjamini-Hochberg method. We found 125 significantly mutated genes with a corrected P-value <0.05 (Supplementary Table S6 online). Twenty of these genes were known oncogenes in the COSMIC cancer gene census. Most frequently mutated oncogenes were TP53 (19%), PLCG1 (10%), CARD11 (7%), DNMT3A (6%), FAS (6%), POT1 (6%), RHOA (3%), KIT (5%), and tumor necrosis factor receptor 2 (TNFRSF1B) (4%). The complete genomic metadata including the identified frequently mutated genes are available for download from the Dryad Digital Repository (https://www.datadryad.org/). One of the important problems we were able to address, which was not possible with the data from smaller CTCL cohorts, was mutual exclusivity of mutations and disrupted pathways in CTCL. Mutual exclusivity provides critical information about the genomic basis of cancer formation. Mutual exclusivity of mutations within the same pathway may indicate that disruption of that pathway alone is sufficient to trigger tumorigenesis. In addition, mutual exclusivity of mutations among different pathways could be the result of negative selection, as cancer cells could not survive with concurrent mutations in multiple pathways. Importantly, such mutual exclusivity among different pathways may signify subtypes of a tumor, each of which results from a distinct mechanism of tumorigenesis (Kim et al., 2017Kim Y.A. Madan S. Przytycka T.M. WeSME: uncovering mutual exclusivity of cancer drivers and beyond.Bioinformatics. 2017; 33: 814-821Crossref PubMed Scopus (8) Google Scholar). We took advantage of the combined dataset to study mutual exclusivity of mutations in CTCL. We found that mutations within the NFkB pathway genes PLCG1, CARD11, and TNFRSF1B were mutually exclusive in SS (Figure 1a). Mutations in the KIT gene were exclusive from the three NFkB pathway genes in all but two SS cases. KIT is known to activate PLCG1, which contributes to activation of the NFkB pathway, which may explain this finding. Furthermore, mutations in p53 were mutually exclusive from the three NFkB/KIT genes (Fisher P-value = 0.03). This exclusivity held true when an expanded set of 90 genes annotated as NFkB pathway genes were tested (KEGG ID:04064) (Fisher P-value = 0.05) (Supplementary Figure S2 online). Using these data we were able to classify CTCL by underlying pathogenic pathways. Nineteen percent of SS cases in the entire cohort had mutations in p53; the remaining nonmutant p53 cases could be further divided into NFkB/KIT-mutant (28% of cases) or NFkB/KIT-normal (53% of cases) (Figure 1a). Because p53 is a major tumor suppressor, either a disrupting mutation or a copy loss may be sufficient to disrupt the normal p53 pathway and contribute to malignant phenotype in CTCL. Thus, we evaluated if disrupting mutations in p53 were mutually exclusive with p53 gene copy loss in SS. We and others observed p53 gene deletion on chromosome 17p to be a common finding in SS, which is detected in 40% of SS cases. However, nearly equal proportion of p53-mutant and nonmutant SS cases had one p53 copy loss (43% and 40%, respectively; Fisher P-value = 0.82). Thus, although p53 is the most frequently mutated gene in CTCL, dysregulation of this pathway is not sufficient to trigger cancer progression. These data support previous studies that showed that p53 alteration alone was not associated with disease prognosis of SS (Gros et al., 2017Gros A. Laharanne E. Vergier M. Prochazkova-Carlotti M. Pham-Ledard A. Bandres T. et al.TP53 alterations in primary and secondary Sézary syndrome: a diagnostic tool for the assessment of malignancy in patients with erythroderma.PLoS One. 2017; 12: e0173171Crossref PubMed Scopus (10) Google Scholar). On the contrary, p53 gene status may affect overall survival in patients with advanced mycosis fungoides, suggesting differences in the pathophysiology of the progression of these two diseases (Wooler et al., 2016Wooler G. Melchior L. Ralfkiaer E. Rahbek Gjerdrum L.M. Gniadecki R. TP53 gene status affects survival in advanced mycosis fungoides.Front Med (Lausanne). 2016; 3: 51PubMed Google Scholar). To further evaluate this finding, we performed survival analysis on 39 SS cases with clinical data available. We found no significant difference in overall survival between SS cases with or without a p53 genomic mutation (Figure 1b). When both genomic mutation and gene copy loss were considered, no difference in overall survival was seen between p53-altered and nonaltered SS cases (Figure 1c). Similarly, mutations in NFkB/KIT did not confer a significant difference in overall survival (Figure 1d). When SS cases were stratified by the total number of mutations, there was a difference in overall survival between the SS cases with the most mutations (poor prognosis) and those with the fewest mutations (better prognosis, Figure 1e). In summary, we created an integrated genomic dataset in CTCL and successfully used this tool to show that there are mutual exclusive mutations affecting p53 or NFkB/KIT genes and pathways. Remarkably, the cases that did not carry p53 or NFkB/KIT abnormalities did not have any significantly mutated genes using the Poisson method, suggesting that other aberrant genomic features such as those implicated in transcription and epigenetic regulation may cooperate with or enhance the damaging effects of genetic mutations. Further prognostic studies utilizing clinicopathologic characteristics, survival data, and integrated genomic datasets expanded from this report may uncover these interactions. These findings have direct implications for future diagnosis and therapy design in CTCL through selective targeting of patient-specific mutations, because CTCL is a particularly heterogeneous malignancy and "one size" of therapy does not fit all patients. We believe that this dataset, as it continues to expand and offers more statistical power, would be a very useful tool in uncovering previously indiscernible relevant relationships, revealing affected pathways, and inferring possible genomic classification of CTCL. The authors state no conflict of interest. We thank Teresa Palomero and Adolfo Ferrando for insightful discussions and for providing genomic data. We thank Paul Khavari, David Wheeler, Madeleine Duvic, and Sean Whittaker for providing genomic data. Download .pdf (1.51 MB) Help with pdf files Supplementary Figures 1 and 2 Download .xlsx (22.63 MB) Help with xlsx files Supplementary Tables 1–6

In the current era of checkpoint inhibitors, patients with metastatic melanoma have demonstrated a significant improvement in survival rates for the first time. However, optimization of immunotherapy is an ongoing effort. Dendritic cell (DC) vaccines have been used in clinical trials, in which they have been shown to be safe and capable of inducing tumor-specific immunity. We conducted a three-arm pilot clinical study in 15 patients with metastatic melanoma to evaluate three types of DC vaccines, differing only by strategies of tumor antigen delivery. DCs and melanoma cells were isolated from each patient, incubated with GM-CSF and IL4, and evaluated for maturation markers prior to antigen loading. Using patients’ own tumor cells as sources of antigens, we created DC vaccines by fusing melanoma cells with DCs, by co-culturing tumor cells with DCs, or by pulsing DCs with tumor cell lysates ex vivo. The vaccines were then injected into the superficial lymph nodes using high-resolution ultrasound guidance. Primary endpoints included delayed-type hypersensitivity (DTH) response and positive ELISpot, a measure of tumor-specific interferon-alpha production. We demonstrate that the vaccines were safe and well tolerated by all patients. DCs loaded with tumor antigens by co-culture produced more immunological responses than did fusion- or lysate-loaded DC vaccines. DC vaccines loaded by co-culture also showed longer survival in this cohort, though our study was not powered to demonstrate survival advantage. Combining this vaccine with other immunotherapies such as checkpoint inhibitors, in order to engage multiple components of the immune system, may be an effective and viable future approach.

Sézary syndrome (SS) is a type of cutaneous T cell lymphoma (CTCL) characterized by leukemic involvement. Its pathogenesis is poorly understood; however, investigators have hypothesized that a viral pathogen may play a role. Efforts to implicate specific viruses have yielded inconsistent results. We used a cutting edge viral detection technique, Virome Capture Sequencing Platform for Vertebrate Viruses (VirCapSeq-VERT), to search for viral sequences in peripheral blood mononuclear cells (PBMCs) and plasma taken from patients with SS. Blood was collected from six SS patients. Total nucleic acids from plasma and PBMCs were extracted and sequenced with Illumina MiSeq, then compared against human nucleotide and viral nucleotide and protein databases. Contigs and reads were mapped to reference genomes to assess genome coverage and depth. Sequences from human endogenous retrovirus–K and human T-lymphotropic virus 1 were detected across all 6 samples. However, these lacked contiguity depth, making the likelihood of a directly incorporated oncogenic virus remote. Hypothetically, these viruses may play a relevant role at an early disease stage before being cleared. Future studies are needed to explore this possibility.

Combination regimens are the mainstay of treatment for hematologic malignancies, but there are currently no FDA-approved combination therapies for cutaneous T-cell lymphoma (CTCL), and few combinations have been studied systematically in vitro. We hypothesize that romidepsin, an HDAC inhibitor, allows a more open chromatin structure that can provide better access to the alkylating effects of mechlorethamine in malignant T cells. We have previously reported more than additive effects of romidepsin and mechlorethamine. Here, we formally evaluated synergy of these two drugs by examining their combined effect in 4 CTCL cell lines, SeAx, HH, Hut78, and Hut102, and 6 primary samples from patients with Sézary Syndrome. Single-agent dose response curves were generated using the cell lines to determine the approximate IC50 for each drug at 24, 48, 72, and 96 hours. The cell lines were treated at fixed ratios of the drugs using Chou and Talalay's median effect method to calculate isobolograms and combination indices (CI) to determine if the drugs are synergistic. Flow cytometry for apoptosis and cell viability was also performed to characterize the combined effects of the drug compared to single agents and controls. Our results showed synergism of romidepsin and mechlorethamine in all 4 cell lines, and 5 of 6 patient samples showed enhanced cytotoxic effects of the combination in malignant cells compared to single agents. The synergistic effects of romidepsin and mechlorethamine in CTCL/Sézary Syndrome make a strong argument to test this drug combination in clinical trials.

Mycosis Fungoides and Sézary Syndrome, the two most common types of Cutaneous T-Cell Lymphoma (CTCL), present many management challenges for dermatologists. Here, we provide a comprehensive review of up-to-date literature, guidelines, and expert clinical insights. We highlight the updates in the World Health Organization Classification of Cutaneous Lymphomas; we summarize the epidemiology, including a recently observed stabilization of increasing incidence of CTCL in the past decade and increased incidence in males, blacks, and veterans; we also provide the most recent updates on prognostic factors for CTCL. Utilization of Next-Generation Sequencing and other novel technologies has shed light on pathogenic mechanisms of CTCL, including immune dysregulation, antigen stimulation, and genomic alterations. CTCL management still remains a significant challenge due to lack of standardization of therapies for every stage of the disease. We provide a straightforward approach to clinical evaluation, diagnostic workup via immunophenotyping and molecular studies, staging guidelines, and select treatment strategies in Mycosis Fungoides and Sézary Syndrome. CTCL patients require individualized, holistic, and multidisciplinary care, for whom addressing management in different skin types and prioritizing quality of life issues are essential. J Drugs Dermatol. 2017;16(5):405-412. .