Toxoplasmosis in an Adult Receiving Long-Term Corticosteroid Therapy

Ocular Don H.

Nicholson, MD, Eugene B. Wolchok, MD

death of a 58-year-old who developed ocular toxoplasmosis while receiving long-term systemic corticosteroid therapy permitted correlation of early histopathologic lesions with their clinical counterpart recorded on fundus photographs. A widespread, paravascular, whitish, retinal opacification dominated the initial clinical picture. These lesions were represented histologically by focal zones of inner retinal necrosis located adjacent to arteries and veins. No associated inflammatory cell infiltrate was present, and numerous viable free Toxoplasma organisms were consistently present at the interface between necrotic and healthy retina. Electron microscopy of formaldehyde-fixed tissue provided a useful technique for confirming the identity of the infecting organism in the absence of serologic or culture data. \s=b\ Sudden

woman

(Arch Ophthalmol 94:248-254, 1976)

Opportunistiincreasingly c

infections by orga¬ nisms of low virulence have become an frequent problem in patients whose immune mechanisms are compromised natu¬ rally by disease or iatrogenically by Submitted for publication Dec 16, 1974. From the Bascom Palmer Eye Institute, Department of Ophthalmology, University of Miami School of Medicine, Miami, Fla (Dr Nicholson) and the University Hospital of Jacksonville, Jacksonville, Fla (Dr Wolchok). Read in part at the New England Ophthalmo-

logical Society meeting, Boston, May 18, 1974, and at the Bascom Palmer Eye Institute Resident's Day meeting, Miami, June 14, 1974. Reprint requests to Bascom Palmer Eye Institute, Box 875, Biscayne Annex, Miami, FL 33152 (Dr Nicholson).

immunosuppressive or cytotoxic ther¬ apy. Clinical and histopathologic documentation of Candida endoph¬ thalmitis1 and cytomegalic inclusion retinitis2 in such patients has firmly

established two types of unusual in¬ fections that the ophthalmologist should consider when evaluating pos¬ terior uveitis in a patient with sup¬ pressed immune mechanisms. This case reminds us that ocular tox¬ oplasmosis should also be considered in this clinical setting. REPORT OF A CASE The patient was a 58-year-old woman with an unexplained and enigmatic sys¬ temic illness. She was in good general health until June 1971, when she developed what was thought to be viral pneumonia associated with anemia, eosinophilia, and splenomegaly. In October 1971, recurrent fever and hepatosplenomegaly prompted rehospitalization, and with a tentative diag¬ nosis of myeloproliferative disease of un¬ known type, splenectomy and liver biopsy were performed. Both liver and spleen con¬ tained extensive areas of fibrosis, focal in¬ flammation, and necrosis. No evidence of myeloid metaplasia or active granulomatous disease was present. She was read¬ mitted to the hospital in March 1972, with persistent fever and weight loss. With the presumed diagnosis of sarcoidosis, treat¬ ment with 16 to 24 mg of systemically ad¬ ministered methylprednisolone per day was begun and continued until the time of her death one and a half years later. On two occasions (June 1972 and May 1973), exacerbations of systemic symptoms ne¬ cessitated more intensive corticosteroid therapy (up to 150 mg of methyl-

Downloaded From: http://archopht.jamanetwork.com/ by a University of Pennsylvania User on 06/14/2015

prednisolone per day) for as long as three weeks before the dosage was tapered to the maintenance level of 16 to 24 mg per

day. Ocular symptoms were first noted in January 1973, nine months after systemic steroid therapy was begun. She developed

what appeared to be an extensive paravascular retinal inflammation associated with granulomatous anterior uveitis in the left eye. The condition progressed over a period of two and one half months to opaci¬ fication of the vitreous and total loss of light perception in spite of three 40 mg and two 80 mg sub-Tenon capsule methylprednisolone injections. During this time the right eye remained entirely normal. She had no old chorioretinal scars in either fundus at any time. In September 1973, eighteen months af¬ ter institution of systemic steroid therapy and six months after loss of vision in the left eye, she developed a similar anterior and posterior uveitis in the right eye. Aqueous ray and cells were accompanied by scattered keratic precipitates and cells in the vitreous. Treatment with an 80 mg sub-Tenon capsule methylprednisolone in¬ jection cleared the vitreous and permitted visualization of fundus lesions (Fig 1) that were identical to those initially seen in the left eye. Extensive areas of focal, whitish, paravascular, retinal opacification paral¬ leled both arteries and veins. The few vitreous cells that remained appeared ran¬ domly distributed, with no particular pred¬ ilection for the areas of paravascular reti¬ nal opacification. In spite of a second sub-Tenon cap¬ sule injection with 80 mg of methyl¬ prednisolone, visual acuity decreased to 20/100, and in late September 1973, a three-week course of treatment with 25 mg

of methotrexate twice weekly was insti¬ tuted in an attempt to arrest the inexor¬ able progress of the uveitis that was presumed to be associated with systemic sarcoidosis. No improvement was noted, and methotrexate therapy was discontin¬ ued. She died suddenly Oct 4, 1973, and no autopsy was performed. The right eye was obtained for histopathologic study five weeks after the clinical photograph in Fig¬ ure 1 was made.

OCULAR PATHOLOGIC FINDINGS Gross Examination.—The

right eye obtained several hours after death and fixed in 4% formaldehyde solution (10% formalin). The globe was of normal size, and no external abnormalities were evident. Removal of the superior calotte disclosed mul¬ tiple focal areas of cheesy white, paravascular, retinal necrosis occur¬ ring along the course of both arteries and veins in all quadrants (Fig 2). Scattered intraretinal hemorrhages were present, and the vitreous was grossly clear. was

Light Microscopy.—The cornea was except for a few isolated in¬

normal

flammatory cells adherent to endo¬ thelium. No inflammatory cells were present in the iris, and the ciliary body contained sparsely scattered accumulations of lymphocytes and plasma cells. The collapsed vitre¬ ous contained a moderate number of mononuclear cells and plasma cells. The site of most extensive histo¬ pathologic change was the retina. Se¬ rial sections were prepared to study the area documented on the fundus photograph in Figure 1. These sec¬ tions demonstrated areas of inner retinal necrosis involving nerve fiber and ganglion cell layers adjacent to both arteries and veins (Fig 3). At the margin of each focus of paravascular retinal necrosis, spherical cysts were scattered that were generally 20µ to 40µ in diameter, and that contained 5 to 30 basophilic, 2µ to 3µ diameter particles consistent with the mor¬ phologic appearance of Toxoplasma gondii. In addition, the interface between intact and necrotic inner retina contained free organisms of elliptical or spherical configuration that usually measured 2µ 6µ (Fig 4). No intraretinal in¬ flammatory response was present falciform,

Fig 1.—Composite fundus photograph of taken five weeks before patient's death.

optic disc and

around the areas of paravascular ne¬ crosis. Although the outer retinal lay¬ ers were well preserved in most areas of the superonasal vessel distribution, two foci of full thickness necrosis of sensory retina and pigment epithe¬ lium were present. A collection of lymphocytes and plasma cells infil¬ trated the choroid only in areas of full thickness retinal necrosis, and at no point did the choroidal inflammatory

infiltrate

assume

a

granulomatous

character. The central section of the globe demonstrated more extensive full thickness retinal necrosis, extending in some sections from the nasal mar¬ gin of the optic disc to the ora ser¬ rata. At irregular intervals in the necrotic nasal retina were isolated islands of retinal cell nuclei that re¬ tained more normal tinctorial charac¬ teristics, and surrounding these resid¬ ual islands were scattered encysted organisms (Fig 5). In areas of total retinal necrosis, the encysted orga¬ nisms lost their basophilia (and pre¬ sumably their viability), becoming homogeneous, eosinophilic, and PAS

positive. Encysted organisms could also be identified in pigment epithelial cells and within cells comprising a thin layer of fibrous metaplasia between totally necrotic pigment epithelium and necrotic retina. Although not as

upper nasal vessels,

right eye,

organisms surrounding necrosis, encysted or¬ ganisms were present in histologi¬ cally normal portions of the retina, and rare clusters of free organisms numerous as

foci of retinal

were

ited

also present in areas that exhib¬ evidence of tissue necrosis

no

(Fig 6).

In the temporal retina, a more nor¬ mal architectural pattern was pre¬ served. However, two focal areas of

paravascular inner retinal necrosis associated with free and encysted or¬ ganisms (Fig 7) appeared identical to the areas identified along the supero¬ nasal vessels. A single focus of full thickness retinal necrosis, with its as¬

sociated minimal choroidal inflamma¬ tory infiltrate, occupied a one-disc diameter area in the paramacular zone (Fig 8). The Bruch membrane was intact and contained equatorial drusen and some basophilic stippling at the nasal margin of the optic nerve.

The nasal half of the optic nerve atrophie, and both cysts and lib¬ erated organisms were present in the prelaminar, laminar, and anterior retrolaminar portions of the nerve (Fig 6). No organisms were present in the 2 mm of orbital optic nerve exam¬ ined, and no inflammatory cells were present in the subarachnoid space. The sclera was normal. was

Downloaded From: http://archopht.jamanetwork.com/ by a University of Pennsylvania User on 06/14/2015

Electron

Microscopy.—Formalde-

and free organisms were identified (Fig 9 and 10) and their ul¬ trastructural identity with Toxo¬ plasma gondii3 was verified in a written communication from Shef¬ field (August 1974).

encysted

COMMENT

Fig 2.—Postmortem appearance of posterior pole and inferior retina (top) and superior calotte (bottom), right eye. Multiple white paravascular foci of retinal necrosis and scat¬ tered intraretinal hemorrhages are evident.

hyde-fixed, paraffin-embedded tissue from the superior nasal paravascular zone of retinal necrosis shown in Fig¬ ure 3 was rehydrated and postfixed in osmium tetroxide. The tissue

was

in graded alcohols and reembedded in epoxy resin. Thin sec¬ tions were stained with uranyl ni¬ trate and lead citrate and examined with the electron microscope. Both

dehydrated

Downloaded From: http://archopht.jamanetwork.com/ by a University of Pennsylvania User on 06/14/2015

Disseminated toxoplasmosis, with a particular predilection for the central nervous system, has been reported in association with several diseases and treatment modalities that compro¬ mise host immune mechanisms.4'8 Un¬ fortunately, lack of autopsy infor¬ mation precludes conclusions about either the initial disease process or the extent of systemic dissemination of toxoplasmosis at the time of death in our patient. Retrospective review of the tissue obtained at the time of splenectomy in 1971 (liver, spleen, ac¬ cessory spleen, lymph node) reveals no changes characteristic of sarcoido¬ sis or myeloproliferative disease, and the lymph node does not exhibit the changes that have been described as characteristic for disseminated tox¬ oplasmosis.9 Dr. Joseph Eggleston re¬ viewed these slides and concluded that the underlying process may be in the spectrum of malignant lymphoproliferative disorders. Previous reports of ocular toxo¬ plasmosis in immunosuppressed adults have been presented by Cogan and Smith10 by Reeh.11 The clinical de¬ scription of the fundus lesion in Reeh's patient is strikingly similar to that seen in our case. He described a man with generalized malignant lymphoma of the giant follicular type who developed bilateral panuveitis five years after institution of sys¬ temic anti-inflammatory and cytotox¬ ic therapy and external irradiation. When the vitreous had cleared some¬ what with intensive steroid therapy, the ophthalmologist noted in both eyes, "large, massive, whitish, exuda¬ tive lesions which at first appeared in a vascular distribution with para¬ vascular cuffing and exudation." Clin¬ icians in the United States are well acquainted with the clinical charac¬ teristics of Toxoplasma retinochoroid¬ itis as a focal necrotizing retinitis with underlying choroidal reaction and overlying vitreous cellular exu-

Fig 3 (right).—Histologie section from midportion of area depicted in clinical photo¬ graph (Fig 1). Focal zones of inner retinal necrosis (arrows) are located adjacent to

each of vessels in this section

(hematoxy¬ 70).

lin-eosin, original magnification

date, frequently occurring at the margin of a hyperpigmented chorio¬ retinal scar. However, in this country toxoplasmosis is not usually consid¬ ered in the differential diagnosis of inflammatory or necrotizing retinal disease that is distributed along the course

of retinal vessels. For many European litera¬

years, reports in the ture have implicated

toxoplasmosis as

for retinal periphlebitis or Eales disease.1214 These reports gen¬ erally assume that the vascular lesion represents an inflammatory cell infil¬ trate adjacent to retinal vessels. Our histologie observations suggest that these lesions may in some cases rep¬ resent paravascular retinal necrosis without any associated inflammation. However, the clinical appearance of the paravascular retinal necrosis seen in our patient bears only a superficial resemblance to the exudative appear¬ ance of periphlebitis retinae. Perkins' review of 31 cases of Eales disease did not confirm an association with in¬ creased frequency of positive sero¬ logie findings for Toxoplasma, and he noted that "retinal vessels may cer¬ tainly be affected during the course of a toxoplasmic chorioretinitis but the clinical picture is quite distinct from that of Eales's disease."15 The early clinical picture described by Cogan and Smith1" in the less se¬ verely affected eye of their patient corresponds more closely to the con¬ ventional picture of Toxoplasma reti¬ nochoroiditis: "an old healed chorio¬ retinal lesion near the macula and an active equatorial lesion" with subse¬ quent development of freshly active equatorial lesions. In that patient, vi¬ sual acuity in the right eye improved one cause

Fig 4—Higher magnification of top left vessel in Fig 3. One cyst and wave-front of liber¬ organisms lie at interface (arrows) of intact and necrotic retina. Three additional cysts are present in adjacent intact nerve fiber layer. V indicates vitreous cavity and L blood vessel lumen (hematoxylin-eosin, original magnification x 700). ated

Fig 5 (right).—Nasal retina in area of fullthickness retinal necrosis. Islands of re¬ maining retinal nuclei are surrounded by Toxoplasma cysts (arrows) of various sizes. Retinal pigment epithelium (rpe) is necrotic. Black dots at level of pigment epithelium are melanin granules, not liber¬ ated Toxoplasma organisms (hematoxylin-

eosin, original magnification

440).

Downloaded From: http://archopht.jamanetwork.com/ by a University of Pennsylvania User on 06/14/2015

6 (left).—Encysted (triangles) and lib¬ erated (arrows) Toxoplasma organisms in prelaminar optic nerve tissue. Minimal tis¬ sue necrosis is present (hematoxylin-eo¬

Fig

sin, original magnification

700).

from 20/200 to 20/70 after substitu¬ tion of pyramethamine (Daraprim) and sulfadiazine therapy for cyclo¬ phosphamide when enucleation of the blind left eye disclosed the correct eti¬ ologic diagnosis. In the rare instance where disseminated toxoplasmosis af¬ fecting the central nervous system is diagnosed during life, anti-Toxoplasma therapy may be effective.4·8 Cogan and Smith's patient is of further interest because the initial reason for institution of systemic corticosteroid and cyclophosphamide therapy was for treatment of uveitis, rather than for treatment of a malig¬

neoplasm. Experimental study of the effect of immunosuppression on Toxoplasma in¬ fection has focused primarily on Tox¬ oplasma encephalitis. In their study of immunosuppression and Tox¬ oplasma encephalitis in the hamster model, Frenkel et al concluded that necrotizing multifocal encephalitis arising after institution of immuno¬ suppressive therapy is more likely to nant

Fig 7.—Focal area of paravascular inner retinal necrosis from temporal retina. Liberated organisms are at interface of intact and necrotic retina (arrows) (hematoxylin-eosin, original magnification 440).

result from recrudescence of chronic latent infection than from dis¬ semination or potentiation of a new primary infection.16 Presumably, host immune responses that promote maintenance of encysted forms and contain viable organisms released by cyst rupture can to some degree be nullified by immunosuppression.17 Why in many patients with ocular toxoplasmosis might corticosteroids exert a beneficial effect, whereas in this case and those described by Co¬ gan and Smith,10 Reeh,11 and, possi¬ bly, Crawford18 might corticosteroids have been deleterious? Frenkel19 sug¬ gests that the answer lies primarily in the timing of cyst rupture with re¬ spect to establishment of the immunosuppressed state. If cyst rupture

Fig 8 (left).—Paramacular focus of fullthickness retinal necrosis. Mild degree of choroidal inflammation is present only where necrosis extends to outer retinal layers and pigment epithelium (hematoxy¬ lin-eosin, original magnification

Downloaded From: http://archopht.jamanetwork.com/ by a University of Pennsylvania User on 06/14/2015

70).

Fig 9.—Cyst

from paraffin-embedded tissue in superior calotte

and liberation of viable organisms oc¬ curs in the retina prior to immuno¬ suppression, then immune-inactivation of liberated organisms occurs in association with an intense in¬ flammatory response that is pre¬ sumed to reflect the normal hyper¬ sensitivity of a previously exposed host. Institution of corticosteroid

(original magnification

therapy at this stage will suppress the inflammatory response after

cellular

immune-inactivation of viable orga¬ nisms has already occurred. If, on the other hand, cyst rupture occurs in a host whose immune mechanisms are already impaired, then the normal cellular response to infection may not be mobilized and immune-inactiva-

14,000).

tion of liberated organism fail, per¬ mitting viable organisms to infect additional retinal cells. Increase in corticosteroid dosage at this stage would then further suppress the feeble host immune response and po¬ tentiate infection. Rupture of addi¬ tional cysts in a patient already being treated with an immunosuppressive

Downloaded From: http://archopht.jamanetwork.com/ by a University of Pennsylvania User on 06/14/2015

Fig 10—Higher magnification of encysted organism. Anterior polar ring (P), double-layered plasma membrane, conoid (C), paired organelles, and posteriorly placed nucleus (N) are characteristic of many coccidian parasites, but white spher¬ ules with hazy borders, possibly representing amylopectin granules, permit positive ultrastructural identification of Tox¬ oplasma organisms (H.G. Sheffield, written communication, August 1974) (original magnification 37,000). dose of corticosteroids for

Toxoplasma retinochoroiditis, for example, might vitiate the salutary steroid effect on inflammation by permitting viable organisms to multiply and infect unchecked. Another possible explana¬ tion, advanced by O'Connor,-" is that the specific mode of steroid adminis¬ tration—sub-Tenon capsule injection

of a repository preparation—may pro¬ duce progressive deterioration in pa¬ tients with active retinal toxoplas¬ mosis who are not receiving concomi¬ tant antimicrobial therapy. This study was supported in part by Florida Lions Eye Bank, Ine, and by Burroughs Wellcome

Company.

Mr E. Barry Davis was responsible for the electron micrographs. Joseph C. Eggleston, MD, Department of Pathology, Johns Hopkins Hospi¬ tal and University, reviewed the biopsy material obtained at laparotomy. Dr Harley G. Sheffield of the Laboratory of Parasitic Diseases, National Institutes of Health, reviewed the electron mi¬ crographs. Dr J. K. Frenkel of the University of Kansas Medical Center provided helpful sugges¬ tions for revision of the manuscript. Ms Nicole Todaro assisted with French language trans¬ lations.

References 1. Edward JE Jr, Foos RY, Montgomerie JZ, et al: Ocular manifestations of Candida septicemia: Review of seventy-six cases of hema-

togenous Candida endophthalmitis. Medicine 53:47-75, 1974. 2. Wyhinny GJ, Apple DJ, Guastella FR, et al:

Adult cytomegalic inclusion retinitis. Am J Ophthalmol 76:773-781, 1973. 3. Sheffield HG, Melton ML: The fine structure and reproduction of Toxoplasma gondii. J Parasitol 54:209-226, 1968. 4. Vietzke WM, Gelderman AH, Grimley PM, et al: Toxoplasmosis complicating malignancy: Experience at the National Cancer Institute. Cancer 21:816-827, 1968. 5. Cohen SN: Toxoplasmosis in patients receiving immunosuppressive therapy. JAMA 211:657-660, 1970. 6. Ghatak NR, Poon TP, Zimmerman HM: Toxoplasmosis of the central nervous system in the adult: Light and electron microscopic study of the three cases. Arch Pathol 89:337-348,1970.

7. Dubin HV, Courter MH, Harrell ER: Toxoplasmosis: A complication of corticosteroid and cyclophosphamide-treated lupus erythematosus. Arch Dermatol 104:547-550, 1971. 8. Carey RM, Kimball AC, Armstrong D, et al: Toxoplasmosis: Clinical experiences in a cancer hospital. Am J Med 54:30-38, 1973. 9. Dorfman RF, Remington JS: Value of lymph-node biopsy in the diagnosis of acute acquired toxoplasmosis. N Engl J Med 289:878-881,

1973. 10. Cogan DG, Smith TR: Toxoplasmosis. Presented at the joint meeting of the European Ophthalmic Pathology Society and Verhoeff Society, April 25-28, 1971. 11. Reeh MJ: Toxoplasmosis in an immunosuppressed host. Presented at the Verhoeff Society meeting, Washington, DC, May 1, 1973. 12. Fran\l=c;\oisJ: La toxoplasmosis oculaire. Bull Soc Belge Ophtalmol 7:165-232, 1961. 13. Offret G, Campinchi R, Pouliquen Y: P\l=e'\riphl\l=e'\bites r\l=e'\tiniennes au cours de la tox-

Downloaded From: http://archopht.jamanetwork.com/ by a University of Pennsylvania User on 06/14/2015

oplasmose

uveale. Bull Soc

Ophtalmol Fr Suppl

4:852-864, 1961.

14. Blanck C: A propos d'un cas de p\l=e'\riphl\l=e'\biter\l=e'\tiniennetoxoplasmique. Bull Soc Oph-

talmol Fr 73:559-562, 1973. 15. Perkins ES: Ocular toxoplasmosis. Br J Ophthalmol 57:1-17, 1973. 16. Frenkel JK, Nelson B, Arias-Stella J: Immunosuppression and toxoplasmic encephalitis. Human Pathol, to be published. 17. Shimada K, O'Connor GR, Yoneda C: Cyst formation by Toxoplasma gondii (RH strain) in vitro: The role of immunologic mechanism. Arch Ophthalmol 92:496-500, 1974. 18. Crawford JB: Toxoplasma retinochoroiditis. Arch Ophthalmol 76:829-832, 1966. 19. Frenkel JK: Toxoplasmosis: Mechanisms of infection, laboratory diagnosis and management. Curr Top Pathol 54:28-75, 1971. 20. O'Connor GR: Manifestations and management of ocular toxoplasmosis. Bull NY Acad Med 50:192-210, 1974.

Ocular toxoplasmosis in an adult receiving long-term corticosteroid therapy.

Sudden death of a 58-year-old woman who developed ocular toxoplasmosis while receiving long-term systemic corticosteroid therapy permitted correlation...
19MB Sizes 0 Downloads 0 Views