Study of T cell Receptor g Gene Clonality in Early Mycosis Fungoides

Mycosis fungoides (MF) is the most common type of cutaneous T cell lymphoma. The lesions of MF can be divided into patch, plaque and tumor stages as the disease progresses. Most cases are detected in early stages, when it is most difficult to distinguish the skin lesions from common inflammatory skin diseases. Although several pathologic criteria for early diagnosis have been suggested, none are 100% sensitive, and each criterion can vary in severity. The reading of pathologic slides is also subject to interobserver variability. Patients are often followed for long periods, with a definite diagnosis reached only after repeat biopsies. However, the earlier MF is diagnosed, the better the chance for a cure. Therefore, other adjunct diagnostic tools such as detection of TCR gene rearrangement are studied.

The purpose of this study was to evaluate monoclonal TCR γ gene rearrangement as a method for early detection of MF. We also compared the association between detection of T-cell monoclonality and histologic paramenters for diagnosis of MF. In addition, we investigated serial biopsies of patients before and after treatment to see whether results of the test would be helpful in following patients’ clinical course.

We retrospectively collected skin biopsy specimens from patients diagnosed with MF in the past 8 years in the Department of Dermatology of National Taiwan University Hospital. After excluding uninterpretable and poor quality slides, 20 specimens from 9 patients were included in the study. DNA was extracted from formalin-fixed, paraffin sections, confirmed with an internal control, and then subjected to two cycles of multiplex PCR with different primers (Mix 1 and Mix 2) to detect monoclonal TCR γ gene rearrangement. We compared the results of PCR with the histologic parameters, in attempt to find the histologic difference between PCR-positive and PCR-negative groups. The histologic diagnoses were categorized as diagnostic, consistent, suggestive, and non-diagnostic.

The PCR was positive in 53% (8/15) of specimens in the patch stage, 100% (2/2) in the plaque stage, and 100% (3/3) in the tumor stage. The test was thus positive in 59% (9/17) of cases of early MF (that is, patch and plaque stages). The PCR was positive in 50% (3/6) specimens considered pathologically diagnostic, 50% (1/2) in the consistent group, 71% (5/7) in the suggestive group, and 50% (1/2) in the non-diagnostic group. Analyzing the seven pathologic parameters, we found that in PCR-negative specimens, inflammation was more severe and dermal fibrosis was commonly present, although only the latter differed significantly between PCR positive and negative groups (p = 0.01). We used laser capture microdissection to change the ratio of monoclonal T cells and reactive T cells to see if it was helpful in increasing the sensitivity of detection. We found that this method was helpful in detecting epidermal clonal T cells. In one patient, we futher cloned and sequenced the PCR product of a lesion that responded completely to treatment and found that the original T cell clone still existed, although in a decreased amount (5/12). The clinical significance of the minimal residual disease remains to be clarified.

The results of PCR are more helpful in the pathologically consistent group, the group that presents the greatest diagnostic dilemma. The PCR is more likely to be negative when there is moderate to severe inflammation and particularly with dermal fibrosis. We suggest the ratio of monoclonal to reactive T cells is the most important factor in detection. Dermal fibrosis is probably related to chronicity and dependent on the degree of inflammation. The sequences of monoclonal T cell gene rearrangement are patient-specific and can thus serve as a marker for following the disease.