COMMENTARYSee related article on page 614The Current State and Future of Clonality

Studies in Mycosis Fungoides

Dan Jones and Madeleine DuvicDepartments of Hematopathology and Dermatology; UT MDAnderson Cancer Center, Houston,Texas, USA

In this issue, Dereure and colleagues (p. 614) present a study com-paring the clonality patterns in the skin and peripheral blood ofpatients with mycosis fungoides (MF) using polymerase chain re-action (PCR) analysis of the T-cell receptor (TCR) gamma gene(Dereure et al, 2003). The authors found a good concordance be-tween the size of the dominant ampli¢ed clone in involved bloodand skin biopsy specimens but a relatively poor correlation be-tween clinical response and the persistence of detectable TCR-gamma clones. This ¢nding is likely related to the fact thatcurrent MF therapies are rarely curative and the sensitivity of thePCR method easily detects residual disease that may not be clini-cally apparent. The authors also detected TCR-gamma clonal re-arrangements in the blood that appear unrelated to the dominantskin tumor clone, in a subset of MF patients.This study demonstrates some of the problems inherent in the

use of the TCR-gamma PCR assay as a method for staging andfollowing outcomes in MF patients. The work also highlights theimportance of acknowledging the limitations of clonality assaysin evaluating T-cell disorders and the need to de¢ne the role andrange of future uses of molecular testing. Below, we summarizethe several di¡erent goals for undertaking clonality studies in MFand how other methodologies might be used to approach them.

USE OF CLONALITY STUDIES TO DIFFERENTIATEBENIGN FROM MALIGNANT CUTANEOUS T-CELL

DISORDERS

In a number of studies published over the last 15 years, it has beenshown that dominant T-cell clones can be detected by eitherTCR Southern blot or PCR-based methods in entities (e.g., lym-phomatoid papulosis, Mucha-Habermann disease or psoriasis)that had been generally regarded as benign or in£ammatory innature (Weiss et al, 1986). In contrast, clinically obvious T-cell tu-mors can sometimes lack demonstrable clonal TCR gene rearran-gements by either Southern blot or PCR analysis (Weiss et al,1988; Vega et al, 2002). These studies show that the presence orabsence of T-cell clonality cannot be exactly equated with theclinical categorization of lesions as benign or malignant.There are also technical problems with all current TCR clon-

ality assays. TCR-beta or -delta Southern blot analyses requireadequate amounts of high-quality DNA and will fail to detectclonal T-cell populations when they represent less than 1^5% ofthe cells analyzed. TCR-gamma PCR assays can show false-positive ‘‘pseudoclonal’’ ampli¢ed peaks and will fail to detect T-cell clones that have undergone point mutation within the TCRprimer-binding sequences. Nonetheless, with good-quality mate-rial,TCR-gamma PCR and TCR-beta Southern blot have a con-cordance rate of 80^90% in most laboratories. Thus, concedingthe limitations above, routine TCR clonality studies on skinbiopsies can be highly useful in con¢rming a clinical and patho-logic suspicion of cutaneous T-cell lymphoma (Dadej et al, 2001).

USE OF CLONALITY STUDIES TO STAGE MF PATIENTSAND PREDICT PROGRESSION

A number of studies have suggested that TCR clonality assaysmight be used to stage MF patients at extracutaneous sites. How-ever, in a large study using TCR-gamma PCR analysis, Delfau-Larue and colleagues convincingly demonstrated that peripheralblood samples commonly show ampli¢ed PCR products thatare di¡erent from the dominant skin clone, thus giving mislead-ing staging results (Delfau-Larue et al, 2000). Furthermore, in thatstudy, nearly all MF patients with a common clone detected byPCR analysis had clinically evident erythroderma and tumorcells identi¢ed on review of the peripheral blood smear. Thus,TCR molecular studies contributed little useful clinical informa-tion. In this regard, we have found that £ow cytometric analysisof peripheral blood, even with a limited T-cell panel, is highlye¡ective in demonstrating and quantifying small numbers ofcirculating tumor cells (Washington et al, 2002). Therefore, webelieve that £ow cytometry is a clearly superior technology forstaging MF patients at extracutaneous sites.The nature of the peripheral blood TCR-gamma clones which

are di¡erent from the dominant MF clone have not been deter-mined in most studies. However, their increased detection fre-quency in older patient suggests that they may be coming fromthe benign oligoclonal and clonal cytotoxic T-cell large granularlymphocyte (T-LGL) expansions that are relatively common inthe blood of elderly patients (Khan et al, 2002). They may alsorepresent antigen-restricted T-LGL populations responding tothe MF tumor cells or to various therapeutic interventions. Thefrequent presence of such T-LGL proliferations in blood can thuscomplicate any T-cell clonality assay unless cell separations aredone prior to analysis. Given the low abundance of T-LGLs inlymph node,T-cell clonality assays may be more useful in di¡er-entiating benign dermatopathic lymphadenitis from early MFinvolvement, especially in those cases where £ow cytometricstudies cannot be done.Comparison of TCR-gamma PCR results at several involved

skin sites or serial analysis of skin biopsies over the course of dis-ease may prove more useful in demonstrating a systemically dis-seminated T-cell clone. Our group has demonstrated that earlyMF may show signi¢cant clonal heterogeneity at di¡erent skinsites (Vega et al, 2002). Furthermore, we showed that demonstra-tion of an identical clonal gene rearrangement in multiple biopsyspecimens at the time of diagnosis can help predict patients whowill show clinical progression.

USE OF CLONALITY STUDIES TO IDENTIFYPATHOGENETIC FACTORS IN MYCOSIS FUNGOIDES

Many cases of mycosis fungoides arise from preexisting in£amma-tory dermatoses. Speci¢c antigenic and host immune interactions

0022-202X/03/$15.00 � Copyright r 2003 by The Society for Investigative Dermatology, Inc.

ix

are believed to drive clonal expansion in the early stages of lym-phomagenesis (Jackow et al, 1997). The response of any givenT-cell to a speci¢c antigen is largely encoded by the precise se-quence of the complementarity-determining regions (CDR) onthe expressed T-cell receptor chains, which arise from di¡erentialjoining of speci¢c V-D-J gene segments during recombination.As such, we believe that the future direction of clonality studiesin MF will be directed to determining the speci¢c TCR se-quences utilized by the neoplastic clone. This approach will giveinformation on both the presence of clonality and the speci¢cvariable-region utilized that may indicate di¡erent etiologies as-sociated with speci¢c MF subtypes.Given that most neoplastic and reactive T-cell populations ex-

press the alpha/beta TCR, clonality assays for this purpose needto be focused on those genes rather than the TCR-gamma gene,which is only rarely expressed as a result of out-of-frame recom-binations during T-cell development. PCR analysis of either theTCR-a or TCR-b is complicated by the large numbers of vari-able region gene segments in these loci. Nonetheless, some studiesexamining patterns of TCR-b rearrangement in MF by PCRmethods have been published (Assaf et al, 2000). PCR assays alsohave the advantage of permitting subsequent direct sequencing ofthe major ampli¢ed product(s) to allow con¢rmation of theV- D- and J-segments utilized. This approach allows more carefulanalysis of the exact CDR sequence and its possible role in bind-ing common antigen targets (Jackow et al, 1997).However, given the di⁄culties of a PCR-based approach, di-

rect detection of the speci¢cTCR-a/b protein present on the sur-face of MF cells using variable region-speci¢c antibodies mayturn out to be a better strategy. This approach is feasible giventhe fact that nearly all T-cell tumors continue to express surfaceTCR at all stages. Given the di¡ering etiologies and variable clin-ical course of MF,TCR-a/b clonality studies are likely to give the

most useful information both in establishing the diagnosis and inproviding pathogenetic and predictive information.

REFERENCES

Assaf C, Hummel M, Dippel E, et al: High detection rate of T-cell receptor betachain rearrangements in T-cell lymphoproliferations by family speci¢c poly-merase chain reaction in combination with the GeneScan technique andDNA sequencing. Blood 96:640^646, 2000

Dadej K, Gaboury L, Lamarre L, Petorin C, Seguin C, Cadotte M, Gorska-Flipot I:The value of clonality in the diagnosis and follow-up of patients with cuta-neous T-cell in¢ltrates. Diagn Mol Pathol 10:78^88, 2001

Delfau-Larue MH, Laroche L,Wechsler J, et al: Diagnostic value of dominant T-cellclones in peripheral blood in 363 patients presenting consecutively with a clin-ical suspicion of cutaneous lymphoma. Blood 96:2987^2992, 2000

Dereure O, Balavoine M, Salees M-T, Candon-Kerlau S, Clot J, Guilhou J-J, EliaouJ-F: Correlations between clinical, histological, blood and skin PCR outcomein patients treated for mycosis fungoides. J Invest Dermatol, 614^617

Jackow CM, Cather JC, Hearne V, Asano AT, Musser JM, Duvic M: Associationof erythrodermic cutaneous T-cell lymphoma superantigen-positive Staphylo-coccus aureus and oligoclonal T-cell receptor V beta gene expansion. Blood 89:32^40, 1997

Khan N, Shari¡ N, Cobbold M, et al: Cytomegalovirus seropositivity drives theCD8 T cell repertoire toward greater clonality in healthy elderly individuals.J Immunol 169:1984^1992, 2002

Vega F, Luthra R, Medeiros LJ, Dunmire V, Lee SJ, Duvic M, Jones D: Clonal het-erogeneity in mycosis fungoides and its relationship to clinical course. Blood100:3369^3373, 2002

Washington LT, HuhYO, Powers LC, Duvic M, Jones D: A stable aberrant immuno-phenotype characterizes nearly all cases of cutaneous T-cell lymphoma in bloodand can be used to monitor response to therapy. BMC Clin Pathol 2:5, 2002

Weiss LM, Picker LJ, Grogan TM,Warnke RA, Sklar J: Absence of clonal beta andgamma T-cell receptor gene rearrangements in a subset of peripheral T-celllymphomas. AmJ Pathol 130:436^442, 1988

Weiss LM, Wood GS, Trela M, Warnke RA, Sklar J: Clonal T-cell populations inlymphomatoid papulosis. Evidence of a lymphoproliferative origin for a clini-cally benign disease. N Engl J Med 315:475^479, 1986

x JONES AND DUVIC THE JOURNAL OF INVESTIGATIVE DERMATOLOGY