1. ¹Department of Oncology, University of California at Los Angeles, Los Angeles, CA, USA
2. ²Department of Pathology, University of California at Los Angeles, Los Angeles, CA, USA
3. ³Department of Medicine, University of California at Los Angeles, Los Angeles, CA, USA

Received 25 March 2003; Accepted 28 May 2003.

TO THE EDITOR

Hairy cell leukemia (HCL) is a hematologic neoplasm characterized by infiltration of the bone marrow and spleen by abnormal lymphocytes with prominent hair-like projections. The disorder is rare, constituting roughly 2% of adult leukemias with 600–800 diagnoses per year in the United States. The median age at the time of diagnosis is 52 years with a male to female ratio of 4:1.¹ Given current therapy with purine nucleoside analogues, the prognosis of HCL is favorable, with progression-free survival estimated to be 70% at 4 years.²

Secondary malignancies have frequently been described in HCL, with an estimated two-fold risk for development of solid tumors among patients treated with purine analogues.³ A number of nonmalignant disorders have also been associated with HCL. There are also several reports suggesting an association with systemic immunologic disorders such as scleroderma, polymyositis and various vasculitides. Among the vasculitides, polyarteritis nodosa represents the most frequently associated disorder, with 18 reports in the medical literature to date. It has been suggested that a common membrane antigen occurs in leukemia cells and vascular endothelium, leading to a vasculitis via a crossreacting antibody.⁴ Four cases of scleroderma occurring with HCL have been previously described in the literature. A humoral or cell-mediated response initiated by leukemia cells is thought to be responsible. Alternatively, fibrosis triggered by hairy cells in the bone marrow may occur in an extramedullary distribution, mimicking the clinical symptoms of scleroderma.⁵ To date, two reports exist describing a concomitant diagnosis of HCL and sarcoidosis.^6,7 We now report a third case and suggest a possible association of the two disorders.

A 71-year-old patient developed tender skin nodules in the right lower extremity followed by bilateral ankle edema and then development of erythema nodosum in the left lower extremity in November 2000. A biopsy obtained 1 month later showed a noncaseating inflammatory infiltrate with multinucleated giant cells without evidence of caseating necrosis. A chest roentgenogram showed no lymphadenopathy. The patient underwent rheumatologic evaluation that showed a normal angiotensin converting enzyme level, a positive antinuclear antibody, and otherwise negative serologies. Bronchoscopic biopsy in February 2002 again showed noncaseating granuloma consistent with a diagnosis of sarcoidosis (Figure 1). The patient was then treated with high-dose corticosteroids although she soon developed generalized osteopenia and a pelvic fracture.

Figure 1.

Bronchoscopic biopsy specimen from the patient demonstrating presence of noncaseating granuloma consistent with sarcoidosis.

Full figure and legend (154K)

In May 2002, the patient developed asymptomatic leukopenia. Blood examination showed WBC 2.75 10⁹ with 43.4% segmented neutrophils, 38.8% lymphocytes and 15.3% monocytes, hemoglobin 12.9 g/dl and platelets 125 10⁹/l. Peripheral smear demonstrated reduced leukocytes with atypical lymphocytes displaying cytoplasmic projections (Figure 2). The bone marrow aspirate showed progressive maturation of all lineages with an M:E ratio of 0.6, erythroid preponderance with a myeloid left shift, and TRAP-stain positivity in numerous cells. Biopsy sections showed a variable cellularity (15–50%) with several aggregates of abnormal lymphoid cells with round, central nuclei and clear cytoplasm. Immunophenotyping was negative for bcl-1 and bcl-6 and positive for bcl-2, CD3, CD20 and CD68.

Figure 2.

Peripheral smear from the patient demonstrating presence of a hairy cell (arrow). Note the presence of fine, outward cytoplasmic projections.

Full figure and legend (109K)

Flow cytometry was performed on a peripheral blood specimen with gating for cells of lymphoid origin (approximately 25% of total cells). T cells comprised 70% of the specimen and B cells 14%. Table 1 summarizes the results of flow cytometry.

Table 1 - Flow cytometry results.

Full table

Shortly after the development of osteoporosis, corticosteroid therapy was tapered. Following the diagnosis of HCL, the patient received 1 week of intravenous chemotherapy with the purine analogue cladribine (2-chlorodeoxyadenosine, 0.09 mg/kg/day). A repeated TRAP stain on a peripheral smear in July 2002 was negative. The patient was seen in February 2003. Blood examination at this time showed continued remission, with WBC 4.08 10⁹ with 76.5% segmented neutrophils and 10.8% monocytes, hemoglobin 11.8 g/dl and platelets 210 10⁹/l.

Coexpression of B-cell marker CD19 with CD11c, CD25, CD103 and FMC7 as well as morphologic features on bone marrow and peripheral smear strongly suggest a diagnosis of hairy cell leukemia arising 18 months after the diagnosis of sarcoidosis. Two disorders in particular bear histologic resemblance to hairy cell leukemia (splenic lymphoma with villous lymphocytes, SL VL; and HCL-variant, HCL-V), but these typically lack expression of the aforementioned hairy cell surface antigens.⁸ In two previous reports of HCL with sarcoidosis, the diagnosis of HCL occurred 12 years following the diagnosis of sarcoidosis,⁶ and in the other patient the diagnoses were concurrent.⁷

In the prior reports, it was hypothesized that defects in T lymphocytes allowed for B-cell proliferation into both HCL and sarcoid. In the setting of sarcoidosis, T cells participate in antigen recognition and amplification of local cellular immune responses. This immune response is achieved through the expression of various cytokine mediators. Recent studies demonstrate the presence of cells of T helper 1 phenotype in organs affected by sarcoidosis. Cells of T helper 1 phenotype produce IL-2 and IFN-, leading to the induction of a nonspecific inflammatory response and subsequent granuloma formation.⁹ Several studies also implicate the role of the same cytokines in the development of HCL. More specifically, activation of hairy cells involves expression of the autoregulated IL-2 receptor (the CD25 surface antigen represents the -chain).¹⁰ Thus, alteration of the cellular microenvironment by the defective T-cells of sarcoidosis may lead to development of HCL.

The presence of an as yet unexplained oligoclonal T-cell proliferation in HCL also bears resemblance to sarcoidosis. In sarcoidosis, patients possess a wide variety of oligoclonal T-cell lines, theoretically representing responses to separate epitopes.⁹ This scenario gives rise to the possibility that certain epitopes that trigger the inflammatory changes of sarcoidosis may serve as exogenous stimuli for hairy cell activation. More likely, the presence of oligoclonal T cells in both diseases supports the role of local cytokine mediators released in sarcoidosis in the activation of hairy cells, as previously described.

Alhough not concordant with the temporal pattern described in one of the case studies referenced, the development of sarcoidosis as a sequela of HCL could also be considered. Several of the antigens restricted to hairy cells, such as CD11c and CD103, have been associated with the activation of both lymphoid and nonlymphoid cell types.⁸ This serves as a rationale for the frequently reported association of non-Hodgkin's lymphoma and HCL. Among the nonlymphoid lineages activated are natural killer (NK) cells, which participate directly in the nonspecific inflammatory response triggered by T cells.⁸ Thus, sarcoidosis may be induced either directly through hairy cell antigen-mediated activation of T cells or indirectly through promotion of the nonspecific inflammatory response.

A more concrete association of sarcoidosis and HCL will require examination of a larger number of patients. However, the interplay of these two disorders with respect to T-cell function and subsequent cytokine-mediated responses does appear to have a molecular basis. With further study, this interplay may serve to explain the relationship between numerous other lymphoproliferative and systemic immunologic diseases.