Case Number : Case 2766 - 11 February 2021 Posted By: Saleem Taibjee

B-cell lymphoma pattern.  However, there is quite a lot of collagen dissection, so I would also wonder about leukaemia cutis or metastasis.  Clinical correlation and immunohistochemistry crucial.

Primary cutaneous low grade B-cell lymphoma particularly marginal zone Vs CD4+ small/medium sized T-cell LPD (Lymphoproliferative Disorder) depending on further IHC, light chain and clonality results.

59 minutes ago, Saman Fatah said:

Primary cutaneous low grade B-cell lymphoma particularly marginal zone Vs CD4+ small/medium sized T-cell LPD (Lymphoproliferative Disorder) depending on further IHC, light chain and clonality results.

Agree. I have the same impression. Slightly favor Marginal Zone Lymphoma

Some immunos to lead the way:

Mixed T- and B-cell infiltrate with apparent predominance of T-helper cells.  Is this a pseudolymphoma?

Mmm...CD4 strongly positive. CD56,  CD123?

Maybe CD4+ small/medium lymphoproliferative disorder?

Any immunos for follicular helper T-cell markers especially PD-1 and evidence of light chain restriction if performed? 

Primary cutaneous CD4 + Small/Medium T cell Lymphopriferative disorder

CD56 data

Many thanks to Eduardo Calonje for his expert opinion in this case, and for also arranging TCR gene rearrangement studies.

This did show clonality (TCRB DB-JB: Clonal).

There was no evidence of light chain restriction on Kappa/Lambda. However, PD-1 in particular was quite impressive (ICOS and CXCL13 less so). The CD123 also highlighted a significant subpopulation of cells. However, this case would not be typical of plasmacytoid dendritic cell tumour, which in my experience tends to show a more monotonous infiltrate (rather than mixed).

See immunostains below.

Hence the favoured diagnosis was indeed primary cutaneous CD4 + small/medium T cell lymphoproliferative disorder.

Eduardo also drew my attention to this recent case series. Beltzung F, et al. Primary Cutaneous CD4+ Small/Medium T-Cell Lymphoproliferative Disorders. Am J Surg Pathol 2020;44:862–872). In their series, the authors found PD1 and ICOS had quite similar expression rates (median=20% and 30%, respectively), but were not expressed by the same lymphocytes: PD1 was mostly expressed by atypical medium-to-large cells, whereas ICOS was also expressed by small lymphocytes. ICOS sometimes (n=14) showed a double staining intensity, with a stronger expression on the medium and large lymphocytes. The distribution of PD1+ and ICOS+ lymphocytes was also different: PD1+ cells tended to form clusters or “rosettes” around large lymphocytes in 63.8% of the cases. ICOS+ lymphocytes were more often sparse (86.0%) than grouped in clusters or in “rosettes” (14.0%). The proportion of CXCL13+ lymphocytes was low (median=5% [1 to 10]).

BW Saleem