PracticeUpdate Conference Series: IID 2018

" …the results are exciting because we show inherent differences in the microenvironment early in the disease course, when patients present clinically very similarly. "

He continued, “On that same note, the results are exciting because we show inherent differences in the microenviron- ment early in the disease course, when patients present clinically very similarly.” “The results are very encouraging in terms of risk stratification of cutaneous T-cell lymphoma,” Mr. Seger added. “We are in the process of validating these findings in a much larger patient cohort. Additionally, we are interested in the significance of the proximity of these cytotoxic T cells to malignant immune cells. We are utilizing gene expression profiling and multispec- tral immunofluorescence to identify these close proximity cells and assess their functional capacity.” Benign T cells as drivers of visible inflammation P ablo Augusto Vieyra-Garcia, PhD, MSc, of the Medical University of Graz in Austria, explained that skin lesions of mycosis fungoides contain both malignant and benign infiltrating T cells. Dr. Vieyra-Garcia told Elsevier’s PracticeUpdate , “Understanding the sig- nificance of inflammatory signals at play, and the predominance of the malignant clone in tissue, is the main focus of our efforts.” Dr. Vieyra-Garcia and colleagues set out to study mycosis fungoides skin before and after psoralen and ultraviolet A ther- apy using high-throughput T-cell receptor sequencing, gene expression profiling, and immunostaining. Though all patients improved clinically, 30% of patients experienced improve- ment of clinical inflammation (Composite Assessment of Index Lesion Severity, modified Severity Weighted Assessment Tool) despite the continued presence of many malignant T cells in skin. Improved Composite Assessment of Index Lesion Severity scores correlated with turnover of benign T-cell clones as measured by High-throughput T cell receptor sequencing (R=0.8) but not with reduction in malignant cells (R=0.3). NanoString profiling demonstrated that benign T cell-associated genes differed markedly before and after treatment.

Benign T cells were strongly associated with a CD4/Th2 signature (CCL18, CSF1) before but with a CD8/Tc1 (CXCL9, 10, 11) signature after treatment. In all, 25 of 29 downregulated genes in untreated mycosis fungoides vs healthy skin mapped to the posttreatment CD8 cell signature. This mapping suggested that psoralen and ultraviolet A recruits a CD8 population present in healthy skin but absent from untreated mycosis fungoides. The CD8 signature also correlated with clearance of malignant cells. Changes in score on the Composite Assessment of Index Lesion Severity were correlated to changes in score across the data- set to identify genes that drive clinical inflammation. Reduced clinical inflammation was linked to loss of the pretreatment benign T-cell signature and acquisition of the post- treatment benign T-cell signature, but was unrelated to the malignant T cell sig- nature. This observation suggested that benign T cells drive visible inflammation in mycosis fungoides. OX-40/OX-40L interactions are known to drive chronic T cell stimulation. Malignant T cells expressed OX-40L and benign T cells expressed OX-40 before, but not after, psoralen and ultraviolet A. Dr. Vieyra-Garcia concluded that malig- nant and benign Th2-infiltrating T cells create an inflammatory circuit that is dis- rupted when psoralen and ultraviolet A induces Th2 cells to leave and replaces them with CD8 T cells capable of killing malignant T cells. "We are very excited with our findings,” Dr. Vieyra-Garcia said. “Together, they clearly indicate that benign T cells not only play an important role in the onset of this disease but they are also implicated in therapeutic mechanisms of well- established treatments like psoralen and ultraviolet A.” “We have many additional questions await- ing answers,” he noted, “that will bring us closer to providing better, personalized care of these patients.”

The number of FoxP3-positive regulatory T cells, programmed death 1-positive T cells, and total CD4-positive T cells were not associated with improved survival. Ratios between these cell types were also examined for potential associations with prognosis. An increased ratio of CD4-positive T cells to FoxP3-positive T cells was associated with worse pro- gression-free survival (HR 1.02, CI 1.0–1.03, P = .01), but this result did not remain sig- nificant in multivariate analysis. Mr. Seger’s group is in the process of confirming this finding in a larger cohort. Mr. Seger concluded that CD8-positive T cell density in the lesional skin of patients with cutaneous T-cell lymphoma is predictive of disease progression and may be a useful adjunct in risk-stratifying early-stage patients. “The results carry tremendous clinical sig- nificance,” Mr. Seger noted. “Using initial skin biopsies and staining for common cellular markers, we are using a technique that can easily be implemented into clin- ical practice.”

www.practiceupdate.com/c/68300

IID 2018 • PRACTICEUPDATE CONFERENCE SERIES 13