Cutavirus Infection in Primary Cutaneous B- and T-Cell Lymphoma

Abstract

This virological study examines the association of cutavirus DNA with different types of cutaneous B-cell lymphoma and cutaneous T-cell lymphoma in lesional skin samples.

For decades, infectious agents, in particular viruses, have been postulated as potential antigenic stimulators in the pathogenesis of primary cutaneous lymphomas, leading to clonal expansion of T-cells and malignant transformation.¹ However, all of the so far suspected viruses including retroviruses, herpesviruses, and polyomaviruses have failed to reveal a consistent association with both cutaneous B-cell lymphoma (CBCL) and cutaneous T-cell lymphoma (CTCL).¹ Recently, a novel parvovirus named cutavirus was discovered in human fecal samples and in 4 lesional biopsy specimens of patients with mycosis fungoides (MF).² Therefore, we initiated a retrospective virological study to screen for the presence of cutavirus DNA in a large number of lesional skin samples from patients with different types of CBCL and CTCL.

Methods

A total of 189 paraffin-embedded biopsy specimens from 130 patients with CBCL or CTCL were analyzed (Table). All tumors were classified according to the current World Health Organization/European Organization for Research and Treatment of Cancer classification system for cutaneous lymphomas.³ Samples were analyzed by real-time polymerase chain reaction (PCR) for the presence of cutavirus DNA using primers 5′-AACCAAACACACCGAACCAG-3′ and 5′-TGAAAAGGCTTACCTCTTTTGG-3′ with locked nucleic acid probe No. 5 (Roche) and PCR conditions as described.⁴ The specificity of the PCR was confirmed by sequence analysis. Cutavirus DNA load was defined as viral DNA copies per β-globin gene copy. The study was approved by the ethics review board of the University Witten/Herdecke.

Table. Detection of Cutavirus DNA in Primary Cutaneous B- and T-Cell Lymphomas^a.

Lymphoma Typesb	Samples Analyzed, No.	Cutavirus DNA-Positive Samples, No. (%) [95% CI]	Patients Analyzed, No.c	Cutavirus DNA-Positive Patients, No. (%) [95% CI]	Cutavirus DNA Load, Median (IQR)d
Primary cutaneous T-Cell and NK cell lymphomas
Mycosis fungoides	104	6 (5.8) [2.7-12.0]	71	6 (8.5) [3.9-17.2]	19.4 (7.5-55.6)
Mycosis fungoides variants and subtypes
Folliculotropic MF	8	0	6	0	NA
Pagetoid reticulosis	2	0	2	0	NA
Sézary syndromee	15	0	12	0	NA
Primary cutaneous CD30-positive lymphoproliferative disorders
Primary cutaneous anaplastic large cell lymphoma	19	0	10	0	NA
Lymphomatoid papulosis	6	0	6	0	NA
Subcutaneous panniculitis-like T-cell lymphoma	1	0	1	0	NA
Primary cutaneous peripheral T-cell lymphoma, not otherwise specified (NOS)	3	0	3	0	NA
Primary cutaneous γ/δ T-cell lymphoma	2	0	1	0	NA
Aggressive epidermotropic CD8-positive cytotoxic T-cell lymphoma	1	0	1	0	NA
Primary cutaneous CD4-positive small/medium T-cell lymphoproliferative disorder	9	0	4	0	NA
Primary cutaneous B-cell lymphomas
Primary cutaneous follicle center lymphoma	10	0	6	0	NA
Primary cutaneous diffuse large B-cell lymphoma, leg type	6	0	4	0	NA
EBV-positive diffuse large B-cell lymphoma, not otherwise specified	2	0	2	0	NA
Primary cutaneous intravascular large B-cell lymphomae	1	0	1	0	NA
Total	189	6 (3.2) [1.5-6.8]	130	6 (4.6) [2.1-9.7]	19.4 (7.5-55.6)

Abbreviations: IQR, interquartile range; MF, mycosis fungoides; NA, not applicable; NK, natural killer.

^aLymphoma types not available for cutavirus DNA analysis include granulomatous slack skin, adult T-cell leukemia/lymphoma (HTLV⁺), extranodal NK/T-cell lymphoma, nasal type, primary cutaneous acral CD8⁺ T-cell lymphoma, primary cutaneous marginal zone B-cell lymphoma, EBV-positive mucocutaneous ulcer, hematological neoplasms of precursor cells, and blastic plasmacytoid dendritic cell neoplasm.

^bAccording to the current World Health Organization/European Organization for Research and Treatment of Cancer classification of cutaneous lymphomas.³

^cSeventy-one patients were male and 59 were female; their mean age was 63.8 years (range 8-91).

^dCutavirus DNA load was defined as cutavirus DNA copies per β-globin gene copy.

^eEntities that can present as primary cutaneous lymphomas yet often already show disseminated disease at the time of initial diagnosis.

Results

Cutavirus DNA was found in 6 (3.2%) of 189 lymphoma biopsies and in 6 (4.6%) of 130 patients (Table). Cutavirus was exclusively detected in patients with MF and was absent in all patients with other types of lymphoma (6 of 71 [8.5%; 95% CI, 3.9%-17.2%] vs 0 of 59 [0%; 95% CI, 0%-6.1%], P = .03; Fisher exact test) (Table). All cutavirus-positive patients were male and had early stage MF (IA or IB). In the cutavirus-positive samples, viral DNA loads ranged between 1.3 and 85.0 cutavirus DNA copies per β-globin gene copy (Table).

Discussion

All previous attempts to identify an infectious origin in cutaneous lymphoma have yielded conflicting results.¹ Several human parvoviruses have been isolated recently, among them cutavirus in 2016.^2,5 Some of them have been detected in cancer tissue (eg, parvovirus B19 in thyroid carcinomas, bocavirus in lung and colorectal cancers).^2,5 In a recent study, cutavirus DNA was found in 4 (23.5%) of 17 CTCL samples, whereas 31 non-CTCL samples (eg, eczema, parapsoriasis, skin cancer) were PCR negative. In situ hybridization revealed only few cutavirus-positive cells located in the upper spinal layer of 2 CTCL samples, but not in the infiltrating neoplastic T cells.² In an ongoing study of our group, cutavirus DNA was found in 4% of skin swabs of over 200 healthy individuals, which is similar to the cutavirus DNA positivity rate found in the present study. These observations argue against a pathogenic role of cutavirus in cutaneous lymphoma. Possibly, cutavirus is similar to β human papillomaviruses and some polyomaviruses, a constituent of the human skin microbiome. In line with this, parvovirus B19 is still detectable in skin biopsy samples years after initial infection.² Nevertheless, we cannot exclude a role of cutavirus in a small subset of MF cases.

Our findings should be interpreted in light of the limitations of the study. Although a relatively large number of samples were analyzed, the number of rare lymphoma subtypes was small for most tumor groups. For some lymphoma subtypes no biopsy samples were available. Moreover, in most patients only a single lesional specimen was available, possibly underestimating the true prevalence of cutavirus. By analyzing a large number of samples, our observations argue against a pathogenic role of cutavirus in most primary cutaneous lymphomas.