A, Macular lesion and surface of detached retina overlying choroidal melanoma in the left eye. Reprinted with permission from Irvine et al.12 B, Peripapillary lesion (small arrow) and "histo" spots (large arrows) in the right eye.
A, Nasal aspect of peripapillary lesion. The choroid is infiltrated with chronic inflammatory cells. The outer retinal layers adhere to the inflamed choroid; the retinal pigment epithelium is focally absent (hematoxylin-eosin, original magnification ×50). B, A vascularized scar lies between the outer layers of the macular retina and the inflamed choroid (hematoxylin-eosin, original magnification ×50). Reprinted with permission from Irvine et al.12
A, Small midperipheral choroidal lesion composed of focal collections of chronic inflammatory cells. The retinal pigment epithelium (RPE) exhibits minor irregularities (hematoxylin-eosin, original magnification ×50). B, Midperipheral punched-out lesion. The scarred retina adheres to the choroid. Inflammatory cells are sparse. The RPE is absent within the scar (hematoxylin-eosin, original magnification ×50). Reprinted with permission from Irvine et al.12
Microdissection of macular and midperipheral lesions. Arrows in the left and middle set of photographs (labeled samples 1 and 2) mark the sites chosen for microdissection before and after biopsy. The excised tissue samples are depicted at the right.
Gel electrophoresis of samples 1 and 2, negative control sample (N), and positive control sample (P). The dark lines in lanes 1, 2, and P demonstrate migration of each DNA product to the same level in the gel.
Spencer WH, Chan C, Shen DF, Rao NA. Detection of Histoplasma capsulatum DNA in Lesions of Chronic Ocular Histoplasmosis Syndrome. Arch Ophthalmol. 2003;121(11):1551-1555. doi:10.1001/archopht.121.11.1551
To evaluate choroidal lesions in histological sections from the enucleated eye of a patient with chronic ocular histoplasmosis syndrome for the presence of Histoplasma capsulatum DNA.
Laser-capture microdissection was used to procure cells from macular and midperipheral choroidal lesions in a deparaffinized hematoxylin-eosin–stained section prepared from the enucleated left eye of a patient with an ipsilateral choroidal melanoma and bilateral chronic histoplasmosis syndrome. The captured cells were initially subjected to polymerase chain reaction (PCR) amplification using a pair of primers unique to each end of the nucleotide sequences that are complementary to the DNA known to flank the internal transcribed spacer regions of the ribosomal RNA genes of H capsulatum. This product was then reamplified using a second set of internally situated nested primers. The results were compared with a positive control sample of H capsulatum DNA and with a negative microdissected sample from noninflamed choroid in the same slide.
Products of H capsulatum DNA were identified in both samples of microdissected tissue and the positive control. They were absent in the negative control.
The observations provide molecular biological evidence linking the chronic choroidal lesions to earlier infection by H capsulatum.
THE CHRONIC ocular histoplasmosis syndrome is typically manifested, on ophthalmoscopic examination, by the presence of multifocal posterior, midperipheral, and peripheral choroidal lesions that vary in size from pinpoint to 2 mm in diameter and appear quiescent, atrophic, or scarred ("histo" spots). Larger lesions are often seen adjacent to the optic disc and in the macula. The condition is characteristically bilateral and is seen in adults with a positive result of a histoplasmin skin test who may also exhibit x-ray evidence of pulmonary hilar calcification. Some patients have a history of Histoplasma capsulatum infection manifested by mild flulike symptoms or pneumonia, often accompanied by fever, lymphadenopathy, hepatomegaly, splenomegaly, and leukopenia resulting from bone marrow involvement. The term presumed ocular histoplasmosis was suggested by investigators of the chronic syndrome who were unable to identify the characteristic dimorphic fungus in sections containing the chronic lesions and to grow histoplasma organisms from fresh ocular tissue before fixation. Inability to fulfill the Koch postulate, despite clinical evidence suggesting a cause-and-effect connection to previous acute histoplasma infection, led to speculation that the chronic ocular lesions may represent an immune response to the residua of an earlier infection by H capsulatum.1,2 This presumption has subsequently been reinforced by persuasive epidemiological linkage data and histopathologic investigations of acute lesions in an animal model, in which the Histoplasma organisms disappeared from the lesions within a few weeks after they were inoculated into the eye.3- 5 In addition, Histoplasma-like organisms have been observed in eye specimens from several patients with acute or chronic lesions.6- 11
We have retrieved archived microslides prepared from the enucleated eye of a patient with bilateral presumed chronic ocular histoplasmosis syndrome to determine whether DNA characteristic of H capsulatum could be identified in these chronic lesions. The clinical manifestations of this patient's ocular lesions as well as their light- and electron-microscopic appearance and fluorescent antibody reactivity were initially described in 1976.12 We now recapitulate the clinical history and the gross and light-microscopic findings pertinent to the present study and append the results of molecular analysis of microdissected tissue obtained from macular and midperipheral choroidal lesions in this specimen. Integration of these findings with those observed in the original study reinforces the hypothesis that the chronic lesions represent the residual effects of previous infection by H capsulatum.
A white man, born April 18, 1925, in western Pennsylvania, moved to Michigan in 1953, when he first noted blurred central vision in his left eye. Examination revealed a left subacute central chorioretinal lesion and several midperipheral lesions described as inactive. The right eye exhibited 4 small, inactive paramacular lesions. No abnormal results of a general physical examination were found. He was treated by means of tuberculin desensitization for 6 months and, subsequently, with corticotropin. In April 1956, the patient experienced diminution of vision in his right eye, attributed to a new lesion in the macula. Results of toxoplasmosis and histoplasmosis skin tests were positive, and he was treated with a combination of pyramethamine and sulfadiazine, as well as foreign protein and cortisone acetate.
From 1957 (when he moved to California) to 1975, both eyes exhibited periodic mild fluctuating posterior and anterior uveitis, and a new left peripapillary lesion became apparent. A left temporal choroidal melanoma, initially observed in 1973, had slowly enlarged and, in 1975, it underwent further growth that led to temporal retinal detachment. The eye was enucleated in July 1975. Before surgery, the histoplasmin complement fixation test result was negative, with a titer of less than 1:2. The precipitin test result was also negative, but the histoplasmin skin test result was strongly positive (erythema and induration from elbow to wrist). Despite this strong response, the skin test was not followed by exacerbation of the fundus lesions. Examination of each eye shortly before enucleation revealed the presence of bilateral peripapillary, macular, midperipheral, and peripheral lesions (Figure 1).
Gross examination of the enucleated eye demonstrated a superior temporal pigmented choroidal melanoma measuring 11 mm at its base and elevated 5 mm. The overlying retina was flatly detached. The macula and peripapillary areas appeared pale and scarred. In addition, there were multiple smaller focal lesions, ranging in size from pinpoint to 2 mm in diameter, that also appeared pale or scarred. These lesions were mostly located posterior to the equator, but a few were seen anteriorly. Some of the smaller lesions had a mottled irregular appearance, whereas others were more sharply demarcated and white.
Results of light-microscopic examination of the anterior segment of the eye were unremarkable except for scattered lymphocytes and plasma cells in the iris stroma and anterior portion of the ciliary body. Superior and temporal to the macula, a choroidal melanoma composed of type B spindle cells elevated the intact Bruch's membrane, and there was serous separation of the overlying retina.
The peripapillary lesion was characterized by a moderately dense choroidal infiltrate composed of lymphocytes, plasma cells, and scattered histiocytes associated with partial destruction of the retinal pigment epithelium (RPE) and loss of the outer retinal layers (Figure 2A). No organisms were identified and there was no evidence of subretinal neovascularization.
The macular portion of the choroid exhibited a similar round cell infiltrate. Bruch's membrane was partially fragmented and incorporated within a vascularized scar that included degenerated RPE cells. The outer retinal layers were absent, but the inner retina was well preserved (Figure 2B). No organisms were identified.
The midperipheral and peripheral lesions exhibited a spectrum of histological changes ranging from small (<1 mm) focal accumulations of choroidal chronic inflammatory cells with minor irregularities in the overlying RPE, to slightly larger (1.0-1.5 mm) lesions in which the RPE exhibited localized irregularity or degeneration (Figure 3A). In some lesions, the RPE was focally absent and there were adhesions between the outer portions of the retina and the choroidal inflammatory focus (Figure 3B). All lesions were well delineated, and the surrounding choroid appeared normal. Organisms were not identified in any of the lesions, and no inflammatory cells were observed in the inner retina or vitreous. The number and composition of the inflammatory cells in these multifocal choroidal lesions varied. Most of the small midperipheral and peripheral lesions were nongranulomatous. However, some of the posterior lesions contained histiocytes adjacent to or on the inner surface of Bruch's membrane; these were classified as nonnecrotizing granulomas.
We used a laser-capture microscope (Pix Cell II; Arcturus, Mountain View, Calif) for light-microscopic observation of microdissected biopsy specimens from the macular lesion and a midperipheral lesion in a formalin-fixed, deparaffinized, 6-µm-thick hematoxylin-eosin–stained section from which the coverslip had been removed (Figure 4, samples 1 and 2). The samples were immediately placed in proteinase K–enriched DNA extraction buffer. A pair of primers (P3 and 2R8) unique to each end of the nucleotide sequences and complementary to the DNA known to flank the internal transcribed spacer regions of the ribosomal RNA genes of H capsulatum isolates were prepared similar to the method described by Jiang et al13 (Table 1). We then used polymerase chain reaction (PCR) to amplify the extracted DNA that had been digested overnight at 37°C. After 35 replication cycles, a second 40-cycle amplification of the first PCR product was performed using an internally situated second set of nested primers (F5 and 2R5) (Table 1). The resultant DNA product from each sample was then electrophoresed on a 15% polyacrylamide gel and compared with the DNA product prepared from a control sample positive for H capsulatum (catalog 8136; American Type Culture Collection, Manassas, Va) and a negative control sample prepared from an uninflamed portion of the choroid (Figure 5). We did not attempt to clone or sequence the DNA products.
Identification of the molecular signature of H capsulatum in the choroidal lesions of this specimen long after recognizable remnants of the organism have disappeared adds credence to the evidence obtained by previous investigators that indicts immunogenic residua of this organism as the likely direct or indirect cause of persistent inflammation in focal lesions of the chronic ocular histoplasmosis syndrome. The immunogenic residua include intracellular, fungal cell wall, and DNA components. Different immunodominant cell wall antigens have been identified that induce cell-mediated and humoral responses.14 A similar immune response is also triggered by the DNA components by virtue of the presence of unmethylated cytidine 5′-phosphate guanosine motifs in the microbes.15 Such DNA components can activate antigen-presenting cells and stimulate immune cells to produce proinflammatory cytokines.15 Although cell wall components or intact organisms could not be detected in the present case, the molecular studies revealed the presence of the mycotic DNA. The latter, by its immunogenic potentials, can perpetuate focal choroidal inflammation.
When Irvine and coworkers12 initially studied this specimen, indirect immunofluorescence studies showed the presence of Ig G and complement C3, suggesting that the immune complex deposits were most likely composed of Histoplasma antigens and the patient's IgG. The presence of H capsulatum DNA in the choroidal lesions is consistent with these findings.
The persistent inflammatory response in the chronic choroidal lesions might explain why new lesions, or fluctuations in the appearance of existing lesions, appear to arise de novo during the extended clinical course of this disease. Inflammatory foci limited to the choroidal stroma (Figure 3A) may not initially be clinically visible, but with the passage of time, they can become evident and change in appearance when late secondary degenerative changes occur in the overlying RPE and retinal adhesions form (Figure 3B). Loss of the RPE and the absence of inflammatory cells in the inner retina and vitreous may also account for the pale quiescent clinical appearance of these lesions. Similarly, persistence of inflammation in response to H capsulatum DNA could account for the first clinical appearance of the macula lesion in this patient's right eye 3 years after his initial ophthalmoscopic examination.
Corresponding author: William H. Spencer, MD, Department of Ophthalmology, California Pacific Medical Center, 2340 Clay St, San Francisco, CA 94115 (e-mail: Spencew@sutterhealth.org).
Submitted for publication February 11, 2003; final revision received April 14, 2003; accepted July 2, 3003.
This study was supported in part by the Pacific Vision Foundation, San Francisco, Calif.
We thank Denice A. Barseness, CRA, for her assistance in preparing the illustrations.
The fundus photographs and photomicrographs depicted in Figure 1, Figure 2, and Figure 3 were not white balanced when they were taken 28 years ago. These images have been digitally altered to rectify fading and changes in color of the background that have occurred with the passage of time. No other alterations have been made in the content of the illustrations.