Case Report

Top Case Report Methods Results Discussion Author & Article Info References

A 36-year-old woman with AIDS (CD4 cell count, 10 cells/mm³) previously treated for recurrent Pneumocystis carinii pneumonia, disseminated Mycobacterium avium complex infection, bacterial sinusitis, and folliculitis presented with a 1-week history of painful, erythematous, nodular skin lesions on her lower legs. Her sinusitis symptoms had also recurred. Examination revealed three tender, erythematous, scaly, crusted nodules, 0.5 cm to 1.0 cm in diameter. The patient did not have fever. She had minimal bilateral maxillary sinus tenderness and a small amount of purulent rhinorrhea. The patient's medications included clarithromycin, ethambutol, ciprofloxacin, and trimethoprim-sulfamethoxazole. Antiretroviral therapy had been stopped 2 months earlier because of drug intolerance. Because of a ß-lactam allergy, the patient began receiving an empirical 7-day course of oral clindamycin (150 mg every 6 hours) for presumptive bacterial sinusitis and possible furunculosis of the legs. After 1 week of therapy, her sinus symptoms and leg nodules resolved.

Within 1 week of stopping clindamycin therapy, the nodules recurred. On re-evaluation, eight painful, erythematous, tender, nodular lesions were found on her lower legs. These lesions were 1.0 cm to 3.0 cm in diameter, and some had dark, necrotic centers (Figure 1). Aspiration of a nodule produced a small amount of purulent material. A Gram stain of this aspirate revealed many neutrophils; monocytes; and large, beaded, gram-positive, bacillary organisms suggestive of mycobacteria (Figure 2, top). Standard cultures for aerobic and anaerobic bacteria, fungi, Bartonella species, and mycobacteria were negative. Treatment with clindamycin was resumed, and the skin lesions resolved gradually.

View larger version (58K): [in this window] [in a new window]   Figure 1. Recurrent erythematous nodules with necrotic centers on lower legs.

View larger version (101K): [in this window] [in a new window]   Figure 2. Top. Gram stain of nodule aspirate. Many beaded, gram-positive organisms are present. Bottom. Transmission electron µgraph of skin aspirate. Two microsporidia are visible in a portion of the aspirated material. Both have obvious clear external spaces corresponding to a chitinous coat. The circular coils of the polar tubules can be seen in each organism. Other sections showed similar organisms throughout the aspirated material (original magnification, x40 000).

After 1 additional month of therapy, with all skin nodules resolved, clindamycin therapy was discontinued. After 2 weeks, the lesions recurred, and additional nodules appeared on the patient's upper extremities. The patient described worsening pain underlying the affected areas of her lower legs. Another nodule was aspirated, and results of a Gram stain were identical to the previous results. Standard cultures were again done, and results were unrevealing. Multiple blood cultures were negative after prolonged incubation and subculturing. A punch biopsy of a skin nodule showed an intense, dermal, acute, and granulomatous reaction with myriad organisms present. Cultures of the skin biopsy specimen were negative. A bone scan showed several areas of uptake in both legs and arms that were consistent with multifocal osteomyelitis; many of the affected areas of bone were beneath the skin lesions. The patient declined to undergo bone biopsy. Treatment with oral clindamycin was resumed, this time at a higher dosage (300 mg every 6 hours). Clindamycin therapy was continued for 1 year with complete resolution of all skin lesions and bone scan findings. No substantial toxicity was related to the prolonged clindamycin therapy. The patient died 3 years later of other AIDS-related complications with no recurrence of the skin nodules or osteomyelitis.

Methods

Top Case Report Methods Results Discussion Author & Article Info References

In this patient, microsporidiosis was diagnosed by morphologic observation of tissue samples by light, fluorescent, and electron microscopy and correlation with microbiological analysis. Identification of species by polymerase chain reaction (PCR) was attempted on stained, aspirated material. Transmission electron microscopy was later done on retained Gram-stained glass slides.

Results

Top Case Report Methods Results Discussion Author & Article Info References

The Gram-stain appearance of the organisms was reminiscent of mycobacteria: a series of gram-positive beads against a clear halo of the organism. Ziehl-Neelsen and Hank stains for acid-fast bacteria did not stain the organisms. True dividing organisms were not present.

Sections of a punch biopsy specimen taken from a skin nodule and stained with hematoxylin-eosin showed an intense, predominantly histiocytic infiltrate in the lower dermis that extended into the subcutaneous fat. Most of the histiocytes had a foamy appearance, which on high magnification seemed to be caused by rod-shaped organisms. Gram stains of tissue revealed organisms with a mottled gram-positive appearance. These organisms were innumerable, were found free and in phagocytes, were ovoid- to rod-shaped, and ranged from 2 µm to 4 µm in length and 1 µm to 1.5 µm in diameter. A calcofluor white stain showed strong cell-wall fluorescence, implying the presence of chitin.

Transmission electron microscopy of previously Gram-stained material was done approximately 1 year after the patient's initial presentation. The electron µgraphs show organisms that are clearly identifiable as microsporidia. They have a clear exterior space, corresponding with the chitinous coat. The polar tubule is easily identifiable and consists of five to seven single coils (Figure 2, bottom). The precise number of nuclei could not be clearly ascertained. Analysis by PCR was unsuccessful, possibly because of persistence of crystal violet stain in the Gram-stained specimen. Additional paraffin-embedded tissue was not available for PCR.

Discussion

Top Case Report Methods Results Discussion Author & Article Info References

The finding of disseminated, nodular, cutaneous microsporidiosis in this patient with AIDS was unexpected. A search of the English-language medical literature by using MEDLINE and AIDSLINE did not identify other reports of this manifestation of microsporidial disease. The differential diagnosis of nodular skin lesions in patients with AIDS includes Kaposi sarcoma; bacillary angiomatosis; chronic varicella zoster; syphilis; and disseminated bacterial, mycobacterial, and fungal infections [5, 6]. Our findings show that microsporidia should be considered in the differential diagnosis of nodular skin disease in patients with AIDS. The association of skin lesions with adjacent osteomyelitis in our patient suggests that the organisms were locally invasive, penetrating from the dermis down into the underlying bone. Although the source of infection in our patient is unknown, the distribution and progression of the cutaneous and osseous disease over time favors hematogenous dissemination.

Microsporidia are obligate intracellular parasites. Many species have been described by using predominantly morphologic criteria. Four genera (Enterocytozoon, Encephalitozoon, Pleistophora, and Nosema) have been associated with most reported cases of human disease, usually in patients who are substantially immunocompromised. We could not identify the species of the organism that we found, but the morphologic appearance of the coiled polar tubule is most consistent with that of an Encephalitozoon-like species. The taxonomy of microsporidia is in a state of flux, leading to some inconsistencies in speciation. In addition, new genera are being reported to cause human disease; Trachipleistophora hominis, for example, was reported as a cause of myositis in a patient with AIDS [7]. Typical infections associated with AIDS include enteritis and keratitis [1]. Small numbers of disseminated visceral infections have been reported for Encephalitozoon species. None, however, has presented in the manner that we describe [2-4]. In most infections, the host cell for the microsporidian parasite is apparent (such as the enterocyte for E. bieneusi). In this case, the organisms were free in the dermis as well as in macrophages. Given the lack of a clearly defined host cell, it is interesting to speculate whether the macrophages themselves might serve as a host cell, especially given the probable hematogenous dissemination of the infection. Supporting evidence is found in a report of in vitro cultivation of Encephalitozoon intestinalis on monocyte-derived macrophages [8].

The difficulty of diagnosing cutaneous microsporidial infection is highlighted in this case report. The fact that different microsporidial species have variable staining characteristics makes electron microscopy an important diagnostic tool [9]. Such techniques as amplification of DNA or ribosomal RNA by PCR, along with tissue culture or staining with monoclonal antibodies, have been reported to be useful for confirming microsporidial infection [10]. In our case, no residual paraffin-embedded tissue was available for additional studies.

Therapy for microsporidial infections is problematic. The current therapeutic choice for enteritis caused by Encephalitozoon intestinalis is albendazole, but this agent does not provide adequate therapy in some patients [1]. Treatment of infections caused by other microsporidial species remains poorly defined. Although the use of clindamycin in other intracellular parasitic infections (such as toxoplasmosis) has precedent, no data are available on its use for the treatment of any microsporidial infections [11].

This case report highlights a new clinical presentation of a still incompletely understood class of organisms, notable for the diversity of its host species. Our finding of an Encephalitozoon-like species of microsporidia as the cause of disseminated nodular skin disease adds another consideration in the differential diagnosis of this condition in patients with AIDS. The therapeutic response described here strongly suggests a possible role for clindamycin; however, controlled studies, including in vitro susceptibility testing, must be undertaken to better assess the utility of this agent in therapy for microsporidial infections.

Current Mailing Addresses: Dr. Kester: Department of Immunology, Walter Reed Army Institute of Research, Washington, DC 20307-5100.

Dr. Turiansky: Dermatology Service, Walter Reed Army Medical Center, Washington, DC 20307-5001.

Dr. McEvoy: Department of Infectious and Parasitic Diseases Pathology, Armed Forces Institute of Pathology, Washington, DC 20307.