Genetic Markers Associated with Progression in Early Mycosis Fungoides

Abstract

Background

Mycosis fungoides is a rare, but potentially devastating malignancy. It classically presents with cutaneous patches and plaques and can progress to tumors on the skin with lymph node, blood, and visceral involvement. While most patients with MF have a relatively benign disease course, a subset of patients will develop progressive disease that is often fatal.

Objective

The aim of this study is to identify genetic markers in early MF limited to the skin (stages IA-IIA) that distinguish those patients who will have progressive disease from those who will not, so that early appropriate treatment may be instituted.

Methods

The study includes eighteen patients who were diagnosed with early stage MF at the time of biopsy and had follow up to determine which patients developed progressive disease. RNA was extracted from skin biopsy specimens and analyzed for expression of CD3, FOXP3, IFNγ, IL-4, IL-13, KIR3DL2, MICB, PLS3, and STAT4 by quantitative real-time polymerase chain reaction.

Results/Conclusions

Reduced expression of FOXP3 and STAT4 and increased expression of IL-4 relative to CD3 expression levels were significantly associated with MF progression. Further studies will be needed to fully assess the usefulness of these genetic markers to predict disease progression and guide treatment options in patients diagnosed with early MF.

Introduction

Mycosis fungoides (MF) is the most common form of cutaneous T-cell lymphoma (CTCL). It classically progresses from patches to infiltrated plaques and ultimately tumors on the skin, often with extracutaneous involvement.¹ Although 80–90% of those affected with early MF have a favorable prognosis (73–97% survival rate at 5 years), 10–20% of patients can progress onto more advanced disease with a marked increase in mortality (26% survival at 5 years).^2–4

One of the difficulties with this malignancy is the inability to reliably predict which patients with clinically early MF will develop progressive disease versus those who will not. It is thought that as MF progresses, it becomes more Th2 polarized by accumulating malignant T-cells that secrete Th2 cytokines such as IL-4⁵, IL-10, and IL-13.⁶ Additionally, decreased numbers of FOXP3⁺ regulatory T-cells in skin lesions of MF have been reported to be associated with a worse prognosis.^7–9 As MF progresses FOXP3 may also increase as it has also been shown that FOXP3 can be expressed on transformed malignant T-cells.¹⁰ More recent studies have identified the expression of the microRNA processing protein, Dicer¹¹, and the microRNA let-7a¹² as being independent prognostic indicators of survival. Additionally, increased expression of Th17 genes¹³ and also elevated serum IgE, large Pautrier’s micoabscesses, and a dermal infiltrate composed of numerous hyperchromatic or large atypical cells and reduced proportion of CD8+ cells have been associated with an increased risk of disease progression.¹⁴

Several genes are known to be aberrantly expressed by malignant T-cells in CTCL and may serve as genetic markers of tumor burden including the overexpressed MIC-B, KIR3DL2, and PLS3/T-plastin.¹⁵ A reduction of STAT4 has also been associated with malignant T-cells.¹⁶ Accordingly, we hypothesize that patients with early stage MF who are prone to have more aggressive disease will have increased mRNA expression of tumor-associated markers (KIR3DL2, MIC-B, PLS3), decreased FOXP3 and STAT4 expression, and increased Th2 cytokines compared to patients with less aggressive disease.

Materials and Methods

This retrospective study utilized the CTCL tissue repository at Johns Hopkins School of Medicine and was approved by the Johns Hopkins Institutional Review Board. Eighteen subjects were selected based on the diagnosis of MF limited to the skin (Stage IA-IIA) and established, long-term follow up (median 44.5 months). Patients were grouped into disease “progressors” (n=8) and “non-progressors” (n=10) according to an increase in TNMB status during follow-up with a single Dermatologist (E.V.). RNA was extracted from each of the lesional skin samples and quantitative real-time polymerase chain reaction (RT-PCR) was performed to identify expression levels of FOXP3, IFNγ, IL-4, IL-13, KIR3DL2, MIC-B, PLS3, and STAT4. CD3 expression was also measured as a control for Tcell infiltrate, and the expression of the aforementioned genes were evaluated relative to CD3 expression.

RNA isolation

Four to 6 mm skin biopsy specimens obtained from patch or plaque lesions were snap frozen in liquid nitrogen, and stored at −80°C. The biopsies were thawed, ground, and immediately placed into Trizol (Invitrogen, Grand Island, NY) for RNA extraction. RNA was isolated as per manufacturer’s protocol (Invitrogen, Grand Island, NY).

cDNA preparation and quantitative real-time polymerase chain reaction

Determination of relative gene expression was performed by quantitative RT-PCR using an ABI 7500 real time PCR system (Applied Biosystems, Foster City, CA). cDNA was prepared from the isolated RNA using random hexamers to prime reverse transcription (Ready- to-Go You-Prime First Strand Beads; GE Healthcare, United Kingdom) following the manufacturer’s protocol. Multiplexed quantitative determination was completed in triplicate wells using approximately 1:30 of the cDNA per well and primer/probe sets for the FAM labeled target genes and VIC labeled endogenous reference gene (GAPDH) with the standard ABI chemistry and reagents. Relative transcripts were determined by the formula: 1/2(CTtarget - CTcontrol). Primers for CD3, FOXP3, IFNγ, IL-4, IL-13, KIR3DL2, MIC-B, PLS3, and STAT4 were obtained from ABI Corporation (Foster City, CA). For gene expression that was too low to register a recorded value by quantitative RT-PCR, a value of 0.05 lower than the lowest recorded value was assigned. As we utilized non-parametric statistical analyses, these data were grouped in the lowest category and the lack of a true value did not affect outcomes. IL-13 transcripts were rarely detected in any tissues samples and were therefore were not included in our analysis.

Statistical analysis

Statistical analysis was performed with SPSS 20.0 for Windows (SPSS, Inc., Chicago, IL). The Mann Whitney U test was performed to evaluate for significant associations between single continuous variables and clinical outcome. Fishers’ exact test and logistic regression were utilized for evaluation of dichotomous variables and association with clinical outcome.

Results

Demographic and Clinical Data

Clinical characteristics, disease stage, and disease progression are summarized in Table 1. Of the 18 patients studied, the male to female ratio was 1.6:1 (male= 11, female= 7). Two patients (11.1%) were diagnosed at stage IA, 11 (61.1%) were diagnosed at stage IB, and 5 (27.8) at stage IIA. After diagnosis, 8 of the 18 (44%) progressed to more advanced disease. 6 patients developed tumors, 1 became erythrodermic, and 1 progressed directly to lymph node involvement. The increased age of patients in the non-progressor group (median, 66.5 years, range 51 to 81) was not statistically different from patients in the progressor group (median, 61.5 years, range, 32 to 69; p= 0.102). There was also a non-significant increase in females in the progressor group (50%) as compared to the non-progressor group (30%, p=0.35). There was no statistically significant difference in initial stage between the progressors or non-progressors (Stage IA 12.5% vs 10%, p=0.71; Stage IB 50% vs 70% p=0.35; Stage IIA, 37.5% vs 20%, p=0.38).

Table 1.

Study Population

ID#	Progression*	Sex	Age	Stage	TNMB
1	No	F	67	IB	T2bN0B0xM0
2	No	M	62	IB	T2bN0B0xM0
3	No	M	51	IA	T1bN0B0xM0
4	No	M	68	IB	T2bN0B0xM0
5	No	M	65	IIA	T2bN1xB0aM0
6	No	F	76	IB	T2bN0B0xM0
7	No	M	81	IIA	T2bNXB0xM0
8	No	F	68	IB	T2bN0B0bM0
9	No	M	53	IB	T2bN0B0aM0
10	No	M	66	IB	T2bN0B0xM0
11	Yes	M	65	IB	T2bN0B0xM0
12	Yes	F	40	IB	T2bN0B0aM0
13	Yes	M	58	IIA	T2bN1aB0aM0
14	Yes	F	69	IIA	T2bN0B0xM0
15	Yes	F	67	IIA	T2bN1aB0xM0
16	Yes	M	65	IB	T2bN0B0xM0
17	Yes	M	49	IA	T1bN0B0xM0
18	Yes	F	32	IB	T2bN0B1aM0

^*Defined as increasing in disease stage over the course of follow-up (median 44.5 months); M= male; F= female.

Decreased expression of either STAT4 or FOXP3 are associated with progressive MF

mRNA expression of CD3, FOXP3, IFNγ, IL-4, IL-13, KIR3DL2, MIC-B, PLS3, and STAT4 in the skin specimens was evaluated by quantitative RT-PCR to determine the association between gene expression and disease progression in MF. We utilized CD3 expression as a surrogate marker to control for T-lymphocyte infiltrate size and compared gene expression relative to CD3 expression. There was no significant difference in CD3 expression between the progressors and non-progressors. We identified a reduction in average expression of both STAT4 and FOXP3 in the skin from patients with eventual disease progression compared to those with stable disease [STAT4: 0.03 +/− SEM 0.007 vs. 0.065 +/− SEM 0.013 (p=0.05); FOXP3: 0.093 +/− SEM 0.039 vs. 0.267 +/− SEM 0.076 (p=0.05)] (Table 2 and Fig. 1). There was no association between the expression levels of the other genes (IFNγ, IL-4, IL-13, KIR3DL2, MIC-B, PLS3) relative to CD3 expression between progressors and non-progressors (Table 2).

Table 2.

Gene Expression in Progressive and Non-Progressive Mycosis Fungoides

Gene expression relative to T-cell infiltrate (CD3)	Mean Relative mRNA Expression		p-Value
	Progressors (n=8)	Non-progressors (n=10)
FOXP3/CD3	0.093 (SEM 0.039)	0.267 (SEM 0.076)	p=0.05
IFNγ/CD3	0.089 (SEM 0.057)	0.112 (SEM 0.057)	p=0.86
IL4/CD3	0.00013 (SEM 0.00009)	0.00017 (SEM 0.00016)	p=0.08
KIR3DL2/CD3	0.052 (SEM 0.026)	0.051 (SEM 0.025)	p=1.00
MICB/CD3	0.032 (SEM 0.013)	0.041 (SEM 0.008)	p=0.11
PLS3/CD3	82.57 (SEM 36.81)	180.62 (SEM 76.23)	p=0.68
STAT4/CD3	0.03 (SEM 0.007)	0.065 (SEM 0.013)	p=0.05

Figure 1. STAT4 and FOXP3 are decreased in progressive MF.

qRT-PCR expression of STAT4 and FOXP3 relative to CD3 in MF progressors and non progressors (*p=0.05).

The signature of elevated IL-4 and decreased STAT4 and FOXP3 is associated with progressive MF

For further evaluation of these potential markers, each gene was assigned a cut-off value, converting them into dichotomous variables. There was a significant association of disease progression with FOXP3/CD3 expression of < 0.1 (p=0.02, Table 3). Similarly, a relative expression of IL-4/CD3 > 1 ×10⁻⁵ was also associated with disease progression (p=0.03, Table 3). Although not statistically significant, STAT4/CD3 < 0.05 was associated with disease progression. When utilizing a combination of the genes most associated with MF progression in our data, we found the expression signature of FOXP3/CD3 < 0.1, STAT4/CD3 < 0.05, and IL-4/CD3 > 1 ×10⁻⁵ together was significantly associated with disease progression in MF (p=0.04; Table 3).

Table 3.

Progressive Disease Associated with the Expression of Specific Genes

Tumor Marker	Cut-off Value	p-value
FOXP3/CD3	<0.1	p=0.02
STAT4/CD3	<0.05	p=0.06
IL4/CD3	>1 ×10−5	p=0.03
FOXP3/CD3, STAT4/CD3, IL-4/CD3	as above	p=0.04

Through the assignment of a cut-off value, dichotomous variables were generated from the gene expression data.

Discussion

While most patients diagnosed with early stage MF will have an indolent disease course, there is a subset of patients who will experience progressive disease with an associated increased mortality. Currently, it is not possible to accurately predict what course an individual patient will follow, and given this widely variable spectrum of disease outcome, improved predictors of MF progression would be helpful to identify those requiring more intensive therapy. We hypothesized that the development of progressive MF would be associated with an altered expression of several proposed tumor-associated markers previously identified in MF (KIR3DL2, MIC-B, and PLS3),¹⁵ polar Th cytokines (IFN γ, IL-4, IL-13)^{5, 6} and decreased STAT4¹⁶ and FOXP3 (previously shown to be associated with poorer outcome in MF).^7–9

In this study, we identified decreased expression of STAT4 and FOXP3, relative to CD3 expression, in patients with early stage MF who had eventual progression of disease. Through the assignment of cut-off values for FOXP3, STAT4, and IL-4 all relative to CD3, we were able to identify an expression signature of three genes (FOXP3/CD3 < 0.1, STAT4/CD3 < 0.05, and IL-4/CD3 > 1 ×10⁻⁵) that shows significant association with progressive disease.

This mRNA expression signature of a decrease in FOXP3 and STAT4 and an increase in IL- 4 is consistent with known mechanisms of tumorigenesis. For example, FOXP3 is the most specific available marker for regulatory T-cells (T reg) and has been shown to be present in various neoplasms, including MF.⁹ However, the role of FOXP3 in MF and CTCL is quite complex.¹⁷ T reg cells in early MF infiltrates appear to have a beneficial effect on the disease, perhaps by directly inhibiting neoplastic T-cells.^7,9 However, under some circumstances such as large cell transformation, neoplastic T-cells may acquire T reg-like properties including expression of FOXP3, and this can be deleterious through suppression of the anti-tumor immune response.¹⁰ In this study we observed that diminished FOXP3 mRNA expression in lesional skin of clinically early MF was associated with subsequent progression of disease. This suggests that the MF infiltrate of progressors contains fewer non-neoplastic FOXP3⁺ cells than non-progressors. Thus the inferred decrease of T reg cells, as measured by decreased FOXP3, may account for the more aggressive disease course found in patients with progressive MF ⁸. Although more studies examining this theory and the prognostic value of FOXP3 expression in early MF are needed, these data offer additional support for assessing mRNA expression of FOXP3 in the evaluation of early MF patients.

Similarly, a reduction in both expression and phosphorylation of STAT4 has been identified in CTCL. ^{16, 18–20} STAT proteins are cytoplasmic transcription factors involved in cytokine signaling.²¹ Because expression of STAT4 is activated by IL-12 and required for Th1 differentiation, ^22–24 it is suggested that the dysregulation of the STAT4 signal transduction pathway in MF results in a shift to Th2 polarization and more aggressive disease.^{19, 20} When looking at patients with a leukemic variant of CTCL, Sézary Syndrome, STAT4 protein is markedly decreased,^{18, 20} suggesting that decreased STAT4 denotes more progressive disease.

Interleukin 4 (IL-4) is a well-known cytokine produced by Th2 cells, which induces Th2 cell proliferation.²⁵ In MF, the transition from a Th1 to a Th2 cytokine profile is thought to mark progression to more advanced disease.^{5, 26,27} Thus, IL-4 has been implicated as an early indicator of disease progression and a predictor of advanced MF.⁵ This concept of increasing Th2 polarization as MF progresses is disputed, however. Recently, IL-4 was shown to be decreased in late-stage MF when compared to normal skin, although an increase in IL-4 was still observed in early MF.⁶ The use of IL-4 as a biomarker in early MF is promising, yet further studies are needed to clarify the utility of IL-4 in predicting disease progression.

Lastly, it is interesting to note that there was a non-significant trend with female gender and decrease in age in those whose MF progressed from early stage disease. These data are in keeping with the findings of Sun, et al, who reported increased aggressivity of MF disease in younger females.²⁸ Additional studies to support younger age and female gender bias towards early MF progression to more aggressive disease are needed.

Conclusion

Early MF can have a widely divergent clinical course, and those with progressive disease have a markedly worse prognosis with a 5-year survival of 26% for advanced disease². Therefore, it is essential that we identify markers for those with early stage MF that will help predict who will go on to develop advanced disease. In this study we show that decreased mRNA expression of FOXP3/CD3 and STAT4/CD3 and increased IL-4/CD3 can be useful as potential early markers of eventual clinical progression. Given our small cohort, additional studies and evaluation will be needed to further clarify the role of these markers. It is also imperative that we continue the identification of other possible markers early in disease, in order to accurately diagnose patients with greater risk of morbidity and steer appropriate treatment.