Early CTCL diagnosis, a (miR)age no more?

Early and accurate diagnosis of cutaneous T-cell lymphoma (CTCL) has long been an elusive target for clinicians, pathologists, and patients alike. In this issue of Blood, Ralfkiaer and colleagues show that microRNAs (miRNAs), a family of single-stranded regulatory RNA molecules, are a powerful new tool for distinguishing CTCL from benign dermatoses.¹ 

CTCL is a heterogeneous family of primary extranodal T-cell non-Hodgkin lymphomas whose incidence and prevalence rates appear to have increased over the past few decades but are likely still underestimated because of insensitive diagnostic criteria.²  The most common types (∼ 70%) are Mycosis Fungoides (MF), initially characterized by erythematous patches and thick scaly plaques in non–sun-exposed areas of the skin, and Sezary syndrome (SS), a less common but more aggressive leukemic T-cell disorder associated with generalized erythroderma and lymphadenopathy.³  Although other less common types of CTCL exist, CTCL is often used interchangeably with MF/SS.

Early skin lesions in CTCL are nonspecific and result from cell-cell interactions and cytokine cross-talk between neoplastic CD4⁺ T cells and the cutaneous microenvironment. At this stage, the neoplastic CD4⁺ T cells are outnumbered by reactive CD8⁺ T cells and by skin-resident innate effectors cells, leading to a nonspecific inflammatory infiltrate, which is very difficult to distinguish from benign dermatitis. With progression, the CD4⁺ T-cell infiltrate becomes heavier and more atypical, and gradually acquires a vertical growth pattern, leading to the classic tumoral (fungating) lesions described by Alibert in 1806.³  When tumor lesions, large cell transformation, and extracutaneous dissemination occur, in advanced stage CTCL, the diagnosis is obvious, but the horse is out of the barn.⁴ 

Most patients with early-stage CTCL discover what ailment afflicts them only after endless doctors' visits and multiple skin biopsies. Having generally applied some form of over-the-counter topical steroid by the time they present to the doctor's office, they are typically told to stop, hang on, and come back in 2 to 3 weeks for a biopsy, which too often returns inconclusive (the dreaded atypical T-cell infiltrate). The journey to diagnosis takes on average 5 to 6 years. While a delayed diagnosis of CTCL may not negatively affect survival, there are times when it certainly does. It is always inefficient, unnecessarily postpones counseling, and frustrates patients.

Currently, the diagnosis of early CTCL relies primarily on the triad of histopathology, immunophenotype, and T-cell receptor (TCR) gene rearrangement analysis, in the presence of compatible skin lesions.^3,5  When epidermotropism, atypia, a high CD4:CD8 ratio, and clear loss of pan T-cell markers, such as CD7 and CD26, are all present, the diagnosis of CTCL is easy, especially if buttressed by a monoclonal TCR gene rearrangement. In most cases, however, findings are vague, diagnosis is deferred, and patients bounce from office to office, befuddled and discouraged. Therefore, the identification of biomarkers that accurately predict the difference between CTCL and benign inflammatory skin diseases (BDs) has a significant impact.

Here, Ralfkiaer et al describe how, using a relatively large sample, they identified a specific and consistent microRNA (miR) signature in the lesional skin of patients with CTCL. Recent studies have suggested several miR signatures in the diagnosis and prognosis of CTCL.^6-8  To date, however, findings have been inconsistent and there is not a single genetic marker or method capable of diagnosing CTCL in clinically overlapping subgroups.

Ralfkiaer and colleagues used 63 patients with CTCL and 85 patients with BDs (psoriasis, dermatitis) or healthy individuals. Conventional, formalin-fixed, paraffin-embedded (FFPE) samples were used to perform miR microarray analysis and identify the most induced or repressed miRs in CTCL compared with the control group. Through systematic statistical analysis of the 27 most deregulated miRs, the authors further identified the 3 most up-regulated (miR-326, miR-663b and miR-711) and the 2 most down-regulated miRs (miR-203 and miR-205) as successful candidates for further testing. They then confirmed the robustness of the signature by performing expression analysis on both the training and validation sample sets. Intriguingly, they also demonstrated that the 5 miR classifiers were effective in separating inflammatory skin disease from peripheral T-cell lymphoma (PTCL) secondarily affecting the skin. However, in the absence of more detailed clinical and pathologic data on the PTCL cohort, it is difficult to draw any conclusions about the general validity of the classifier beyond CTCL.

The highlight of this study is the unexpected observation that the 5-miRNA classifiers not only discriminated benign from malignant samples in skin xenografts on laboratory mice, but that the majority of the miRNAs were not affected by drug treatment. If confirmed, this would be excellent news for patients who will then be spared weeks of itching and disease flares before the diagnostic biopsy. The translational implications of this work are also very significant. An accurate and reproducible instrument to identify the earliest stages of CTCL will provide a much needed opportunity to study the key molecular events that kick off its development and shed some light on the real incidence and prevalence of CTCL, which remain a subject of debate.

A number of questions, of course, remain. To what degree will the application of the miR classifiers improve diagnostic sensitivity in CTCL cases that are histopathologically and immunophenotypically borderline? Will the classifiers help resolve the all too common dilemma of the atypical T-cell infiltrate with a monoclonal TCR rearrangement that is not CTCL? Finally, what is going to be the best assay for routine clinical use: RT-PCR, or an immunohistochemistry method? As we congratulate Ralfkiaer and collegues for advancing the field with their important work, we look forward to the next set of studies aimed at answering these questions.