Editors' ChoiceCancer

Aging T cells portend poor outcome in follicular lymphoma

See allHide authors and affiliations

Science Translational Medicine  06 Mar 2019:
Vol. 11, Issue 482, eaaw7628
DOI: 10.1126/scitranslmed.aaw7628


Immune cell profiling identifies intratumoral T cell subsets associated with patient survival in follicular lymphoma.

Patients with follicular lymphoma (FL), the major form of non-Hodgkin lymphoma, have a wide range of survival times. There are no immunological markers to predict patient outcome. However, T cell diversity in FL has not been comprehensively characterized. Furthermore, although immune checkpoint inhibition with anti-PD-1 therapy achieved success in Hodgkin lymphoma, it has limited efficacy in FL, and the reasons behind this failture are unclear.

Yang et al. leveraged mass cytometry, a single-cell protein measurement technology, to profile intratumoral T cell subsets from a cohort of newly diagnosed FL patients by evaluating 33 T cell markers. Sophisticated computational algorithms clustered T cells into 12 different subsets. To ascertain if any of the T cell subsets and markers might predict patient outcome, the authors conducted careful and systematic association analyses. They made several notable findings. First, they discovered that increased naïve T cells, but not memory T cells, were correlated with better survival. Second, expression of PD-1 was primarily concentrated on various CD4+ T cell subsets, but not on CD8+ T cells. Lastly, the strongest predictors of patient survival turned out to be the expression of CD27 and CD28. Loss of either CD27 or CD28 is a signature of premature T cell aging. Patients whose intratumoral T cells lost either molecule had shorter survival. The authors further demonstrated that CD70+ lymphoma cells likely induced T cell aging, which compromised T cell proliferative capacity.

These data suggest that the lack of efficacy of anti-PD-1 therapy in FL may result from the low expression of PD-1 on CD8+ T cells, as CD8+ T cells are the primary anti-tumor T cells. The identification of CD27 and CD28 as predictors for patient outcome should improve patient stratification, but it needs to be verified with a larger cohort study. Moreover, biomarkers from blood are usually more useful because of its greater accessibility than tumor tissues. It will be interesting to assess whether CD27 and CD28 on blood T cells have similar predictive power in FL. Finally, premature aging in blood T cells is associated with autoimmunity. Since development of autoimmunity during solid tumor treatment is associated with favorable outcome, one may expect that loss of CD27 and CD28 would instead predict better outcome in FL. Whether this contradiction is due to different functions of blood T cell and intratumoral T cells or fundamental differences between solid tumors and hematological tumors awaits further investigation.

Highlighted Article

View Abstract

Navigate This Article