How blasts get into the skin—and how to get rid of them Free

In this issue of Blood, Penter et al¹ used bulk and single-cell transcriptome profiling to provide insight into the mechanisms associated with skin tropism and the specific microenvironment of leukemia cutis.

A profound understanding of cancer biology is crucial for the development of new treatment strategies for high-risk malignancies, such as extramedullary manifestation of acute myeloid leukemia (eAML). The prognosis is especially dire for those with eAML at relapse after allogeneic stem cell transplantation (alloSCT).

Leukemia cutis is one of the most frequent forms of eAML. However, limited data are available to date to explain the mechanisms behind leukemia infiltration of the skin. In vitro data have shown activation of the rat sarcoma (RAS)-MAPK/ERK pathway as a consequence of nucleophosmin 1 overexpression. In line with these findings, gene mutations directly related to the RAS pathway have been identified in cutaneous infiltration by AML.² Further proposed mechanisms include the overexpression of chemokine receptors³ and differential expression of cellular adhesion molecules.⁴ Penter et al have now convincingly demonstrated a genetic cause of cutaneous tropism in AML by showing marked differences among extramedullary and bone marrow blasts in terms of the expression of 8 homing receptor molecules. As proposed in the study, beyond increasing our understanding of the pathophysiology of eAML, these findings might also offer a potential therapeutic approach against leukemia cutis by blocking adhesion molecules, such as ICAM-1 (CD54), using monoclonal antibodies. Furthermore, systematic exploration of the described gene expression studies in eAML samples from other sites might identify patterns or characteristics of leukemia infiltration in different tissues.

Penter et al also demonstrated that various immune escape mechanisms are present in cutaneous manifestations of AML following a transplantation. Intensive research in patients with medullary AML relapse after alloSCT have revealed a broad spectrum of mechanisms by which myeloid blasts evade the surveillance of the allogeneic immune system.⁵ Genomic loss of mismatched HLA antigens after haploidentical alloSCT was the first identified evasion strategy of AML cells. Accordingly, leukemic blasts became undetectable by alloreactive donor T cells. As a consequence, the use of donor lymphocyte infusion or a second alloSCT from the same donor would not have the expected graft vs leukemia benefit for cases in which genomic loss at relapse was dedected.⁶ Impaired HLA expression caused by epigenetic downregulation of classical HLA II genes and genes involved in antigen presentation were detected later on after AML relapse in the HLA-identical transplant setting. Reversal was demonstrated after specific epigenetic treatments.^7,8 The detection of decreased HLA II expression in leukemia cutis extends the potential therapeutic consequences to extramedullary manifestations.

Aberrant immune checkpoint expression, characterized by an exhausted T-cell phenotype, represents an alternative immune escape strategy that has been observed in medullary AML relapse after alloSCT.⁹ Penter et al again extended these observations to extramedullary leukemia at the gene expression level and found this phenomenon to be significantly more prominent in leukemia cutis than in medullary AML. This finding adds to the understanding of the particular sensitivity of cutaneous AML to checkpoint inhibitor treatment that was observed earlier in clinical trials.¹⁰ Remarkably, in broader studies of hematologic relapse after a transplant, the 2 mechanisms of decreased HLA expression and T-cell exhaustion have been found to be almost mutually exclusive. Transferring this concept, the findings by Penter et al support the idea of detailed, multiomic profiling of both leukemia and its microenvironment in patients with posttransplant relapse, including extramedullary leukemia, to determine the predominant immune escape mechanism for each individual patient.

In summary, the work by Penter et al demonstrates that leukemia cutis, in particular when occurring after alloSCT, seems to be a consequence of a complex interplay of particular features on both the leukemic side and the local immune effector cells. These include tissue-related tropism of the blasts, characteristic (organ-specific) immune infiltration, and mechanisms to escape from the immune surveillance after alloSCT. The identified changes might represent a step forward on the way toward specific strategies for the prevention and treatment of extramedullary manifestations of AML and toward tailored treatment of posttransplant relapse based on the identified individual pathophysiology.

Conflict-of-interest disclosure: The author declares no competing financial interests.