- Email: [email protected]
Ocular Toxoplasmosis Misdiagnosed as Cytomegalovirus Retinopathy in Immunocompromised Patients
Ocular Toxoplasmosis Misdiagnosed as Cytomegalovirus Retinopathy in Immunocompromised Patients Brad S. Elkins, MD/ Gary N. Holland, MD/ E. Mitchel 0premcak, MD/ James P. Dunn, Jr., MD,3 Douglas A. Jabs, MD,3 William H. Johnston, MD, l William R. Green, MD 3 Background: Cytomegalovirus (CMV) and Toxoplasma gondii both cause necrotizing retinopathy in immunosuppressed hosts. Because of the high prevalence of serum antibodies to these agents in the general population and the risks associated with retinal biopsies, diagnosis of these infections is usually based on clinical findings alone , but the two infections can be confused w ith one another because of similar clinical features. Accurate diagnosis is critical, however, because both diseases are treatable but require different medical therapies . Methods: Case histories were reviewed for five immunosuppressed patients with necrotizing retinopathy, which was initially diagnosed incorrectly as CMV retinopathy but was subsequently found to be toxoplasmic retinochoroiditis . Correct diagnosis was based on retinal biopsy (2 cases) or rapid response to antiparasitic drug therapy (3 cases) . Factors were sought that might help differentiate toxoplasmic retinochoroiditis from CMV retinopathy at presentation. Results: In all cases , the character of retinal opacification (densely opaque, th ick) and the appearance of lesion borders (smooth, nongranular) was different from that typically seen with CMV ret inopathy. These cases also were characterized by prominent anterior chamber and vitreous inflammatory reactions (four of five cases) and relative lack of retinal hemorrhage. One patient subsequently developed CMV retinopathy; characteristics of the two lesions in the same eye highlighted the differences between these two infections. Conclusion: Clinicians should consider ocular toxoplasmosis as a cause of necrotizing retinopathy in immunosuppressed patients and consider an empiric course of antiparasitic therapy for lesions with features described in this report. Ophthalmology 1994; 10 1:499- 507
Originally received: June II , 1993. Revision accepted: October I, 1993. 1 UCLA Ocular Inflammatory Disease Center, Jules Stein Eye Institute, a nd Department of Ophthalmology, UCLA School of Medicine, Los Angeles. 2 Department of Ophthalmology, Ohio State Un iversity College of Medicine, Columbus.
3 Wilmer Ophthalmological Institut e, Johns Hopkins University School of Medicine, Baltimore.
Supp orted in part by the Skirball Foundation, Los Angeles, California; Research to Prevent Blindness, Inc., New York , New York; NIH grant EY08057 , Bethesda, Maryland (Dr. Holland) ; Kricker Foundation, Columbus , Ohio (Dr. Oprerncak ); NIH/ NCI grant 4POICA 15396, Bethesda, Maryland (Drs. Dunn and Jabs); the E. A. Baker Foundation for the Prevention of Blindn ess, Canadia n National Institute for the Blind, Toronto, Ontario, Canada (Dr. John ston). Dr. Johnston is a Frances Howard Goldw yn Fellow in Cornea -External Ocular Diseases and Uveitis. Dr. Jabs is a Research to Prevent Blindness Olga Wiess Keith Scholar. Reprint requests to Ga ry N. Holland, MD, Jutes Stein Eye Institute, 100 Stein Plaza, UCLA, Los Angeles, CA 90024-7003.
499
Ophthalmology
Volume 101, Number 3, March 1994
Ocular toxoplasmosis in immunocompromised hosts presents a special challenge to ophthalmologists in terms of diagnosis and management. The clinical appearance may be very different from that seen in otherwise healthy adults . Active lesions can be bilateral and multifocal. They usually do not arise at the borders of pre-existing retinochoroidal scars, suggesting that disease is acquired or that lesions arise from organisms newly disseminated to the eye from nonocular sites of disease. I Among patients with acquired immune deficiency syndrome (AIDS), the appearance of these lesions can be confused with cytomegalovirus (CMV) retinopathy, which is a much more common problem. Without treatment, both toxoplasmic retinochoroiditis and CMV retinopathy in immunosuppressed patients will destroy extensive areas of the retina. Ocular toxoplasmosis responds rapidly to pyrimethamine and sulfadiazine or other antiparasitic therapy, 1-4 and CMV retinopathy usually responds to antiviral therapy.' Treatments for both may be associated with serious complications, however. Therefore, accurate differentiation between toxoplasmic retinochoroiditis and CMV retinopathy is critical , both to preserve vision and to avoid the unnecessary use of toxic medications. We present five cases of toxoplasmic retinochoroiditis in patients who were immunosuppressed by a variety of causes. All were initially misdiagnosed as having CMV retinopathy. A review of findings identified clinical features that can help the practitioner distinguish between these two infections at presentation.
Case Reports Case 1. A 44-year-old woman with a several year history of systemic lupus erythematosus reported decreased vision in her right eye that was associated with retrobulbar pain. At the time , she was taking prednisone 40 mg every other day. Three months earlier, azathioprine therapy had been discontinued because of bone marrow suppression. Plaquenil therapy had been discontinued 3 years earlier because of abnormal FarnsworthMunsell 100 hue testing, despite a normal fluorescein angiogram, electro-oculogram, and fundus photographs. Visual acuity was 20/25 in the right eye and 20/25 in the left. The right eye showed mild conjunctival injection , keratic precipitates, and I+ cells in the anterior chamber. There was a single focus of perivascular retinal necrosis in the inferonasal periphery without hemorrhage. Review of the photographs taken 3 years earlier showed no pre-existing lesions in the area of active disease. The left eye was normal . Intravenous ganciclovir therap y (5 mg/kg twice daily) was started for presumed CMV retinopathy. Thre e weeks later, her visual acuity was 20/70 in the right eye and 20/25 in the left. In the right eye, there were posterior synechia e, 2-3+ cells in the anterior chamber, and a 3+ vitreous inflammatory reaction. The area of necrosis had enlarged , and there was inflammatory sheathing of retinal vessels adjacent to the lesion (Fig I). Ocular toxopla smosis was believed to be a more likely cause of the lesion, and ganciclovir was discontinued. Oral pyrimethamine (25 mg once weekly) and sulfadiazine (500 mg 4 times daily) were started . The sulfadiazine was later switched to c1indamycin (750 mg 4 times daily) because ofnau-
500
sea, vomiting, and diarrhea. Six weeks after beginning antiparasitic therap y, her visual acuity was 20/50 in the right eye. There were fewer keratic precipitates , I+ cells in the anterior chamber, I+ vitreous inflammatory reaction , and decreased retinal opacification. The borders of the lesion were becoming scarred. The left eye remained normal. The pyrimethamine was discontinued , but the patient was maintained on oral c1indamycin (150 mg 4 times daily). Seven months later, visual acuity was 20/30 in the right eye. There were no cells in the anterior chamber or vitreous, and the retinal lesion appeared inactive. The c1indamycin was discontinued, but the patient was maintained on low dose prednisone to control the systemic lupus erythematosus. Four years later, her visual acuity was 20/30 in the right eye, and there was an atrophic retinochoroidal scar (Fig 2). Clinical signs of intracranial toxoplasmosis never developed. Case 2. A 36-year-old man underwent bone marrow transplantation for chronic myelogenous leukemia. Results of ophthalmic examination before transplantation were normal. Graft-versus-host disease developed, which was treated with a combination of methylprednisolone , azathioprine, cyclosporine, and psoralens ultraviolet A therapy. Aspergillus sp sinusitis also developed, which was treated with intravenous amphotericin B. One year later, the patient presented with a I-day history of "slight haze" over his right eye. Visual acuity was 20/25 in the right eye and 20/ 20 in the left. The right eye had trace anterior chamber cells, a 1-2+ vitreous inflammatory reaction, and a dense, yellow- white curvilinear infiltrate posterior to an area of atroph ic retina in the superonasal quadrant (Fig 3). The left fundus was normal. Blood cultures were positive for CMV. A diagnosis of CMV retinopathy was made, and intravenous ganciclovir therapy (5 rug/kg twice daily) was started . The immunosuppre ssive medications could not be tapered because of chroni c graft-versus-host disease. After 6 weeks of ganciclovir therapy, visual acuit y was 20/ 40 in the right eye and 20/20 in the left. There was still a 1-2+ vitreous inflammatory reaction in the right eye, and the retinal lesion had advanced posteriorly (Fig 4). Ganciclovir resistance was considered, and intravenous foscarnet therapy (60 mg/kg 3 times daily) was added. Two weeks later, visual acuity was 20/ 400 in the right eye and 20/20 in the left. The vitreous inflammatory reaction increased . Several days later, visual acuity was counting fingers in the right eye and 20/20 in the left, and an inferior visual field defect was noted in the right eye. The retinal lesion obscured the optic nerve head. The creatinine level had risen to 2.9 mg/dl, possibly as a result of foscarnet toxicity. Retinal and vitreous biopsies were performed. Toxoplasma gondii cysts were identified in the retina and vitreous (Fig 5), and immunohistochemical stains for CMV and varicella-zoster virus were negative. The patient had no clinical signs of intracranial toxoplasmosis. Antipa rasitic therapy using oral trimethoprim (160 mgj/sulfametho xazole (800 mg) twice daily was started. Three days later, there was a decrease in the vitreous inflammatory reaction and retinal opacification, which continued to resolve over a 3-week period of therapy. Visual acuity was not determined because of a gas bubble placed in the eye during the biopsy procedure. Soon thereafter, the patient died of sequelae of his graft-versus-host disease. Case 3. A 38-year-old man underwent bone marrow transplantation for chronic myelogenous leukemia. Acute graft-versus-host disease subsequently developed, which was treated with methylprednisolone (24 mg twice daily). Results of ophthalmic examin ation showed visual acuity of 20/50 in the right eye and 20/25 in the left. There was focal scarring of the right macula
Elkins et al . Ocular Toxoplasmosis
Top left, Figure 1. C ase 1, right eye. Untreated toxopl asmic ret inochoroiditis in the inferonasal periphery. The patient was being treated for systemic lupus erythematosu s. Top right, Figure 2. Case 1, right eye. The same port ion of the retin a shown in Figure 1, photographed 4 years later. There is an atrophic retinochoroidal scar in the area of the previous toxoplasmic infection . Center left, Figure 3. Case 2, right eye. Untreated toxoplasmic retinochoroiditis in the superonasal qu adrant. The patient had graft-versus-host disease. The border of the lesion is densely opaque; ant erior to the op acity, the retina is atroph ic. Center right, Figure 4. Case 2, right eye. The same area of the ret ina sho wn in Figure 3, photographed 6 weeks later . The border of the untreated lesion has advan ced posteriorly. Bottom left, Figure 5. Light micro graph of a biop sy specimen shown in Figures 3 an d 4 sho ws Toxoplasma gondii cysts (arrows) (hematoxylin-eosin; ori ginal magnification , X400). Bottom right, Figure 6. Case 3, left eye. Untreated toxoplasmic retinochoroiditis in th e inferotemporal region . The patient had graft-versus-ho st disease. There is scant retinal hemorrhage in the lesion . .
501
Ophthalmology
Volume 101, Number 3, March 1994
due to previous trauma. Results of examination of the left eye were normal. He had positive CMV blood cultures and serum antibodies to varicella-zoster virus and herpes simplex virus. Nine months later, chronic graft-versus-host disease developed , and a retinal lesion in the periphery of the left eye was discovered. Visual acuit y was 20/20 in the left eye. A tentative diagnosis ofviral retinitis was made , and oral corticosteroid dosage was tapered to 12.5 mg twice daily. Visual field constriction was eventuall y noted, and increased opacification of the retinal lesion was found on examination. The patient was treated for presumed CMV retinopathy with intravenous ganciclovir (5 mg/ kg twice daily), which was subsequently changed to intravenous foscamet (60 rug/kg 3 times daily) because ofleukopenia. Despite this therapy, the retinal lesion extended posteriorly (Fig 6). There was an oval area of retinal necrosis with thickened, densely opaque borders, a central atrophic area, and scant hemorrhage adjacent to a vessel that coursed through the area of necrosis. He had fine keratic precipitates, with I+ anterior chamber cells and a I+ vitreous inflammatory reaction. Intravenous acyclovir (IO mg/kg 3 times daily) was started because of the possibility of varicella-zoster virus retinopathy, but the lesion continued to worsen. Three months after development of the retinopathy and approximately 3 weeks after acyclovir was started , an endoretinal biopsy was performed. Tox oplasma gondii cysts were identified by electron microscopy (Fig 7). No viral particles were seen, and viral cultures were negative. The patient was treated with oral pyrimethamine (100 mg daily) and oral clindamycin (400 mg 3 times daily). Sulfadiazine therapy was discontinued after 3 days because of rash. Two weeks later, mental status changes developed, and results of computed tomography (CT) of the head showed probable intracranial toxoplasmosis . The retinal lesion resolved completely after 8 weeks of treatment, but the patient's vision remained poor because of a dense posterior subcapsular cataract. Mental status remained altered, and a brain biopsy was performed, which confirmed intracranial toxoplasmosis. The patient died shortly thereafter. Case 4. Decreased vision and severe pain developed in the right eye of a 44-year-old homosexual man with AIDS. Retinal necrosis without hemorrhage was noted in the right macula. His CD4 lymphocyte count was 20/mm 3• The patient was given intravenous ganciclovir (5 rng/kg twice daily) for presumed CMV retinopathy. Six weeks later, pain, redness, and decreased vision still existed. There was no resolution of retinal opacification, and the lesion had enlarged (Fig 8). Visual acuity was counting fingers (eccentric fixation) in the right eye and 20/2 0 in the left. There was marked ciliary flush, 3+ cells in the anterior chamber, extensive posterior synechiae , and a 3+ vitreous inflammatory reaction. A diagnosis of presumed toxoplasmic retinochoro iditis was made, and antiparasitic therapy with oral pyrimethamine (50 mg daily after a 200-mg loading dose) and oral sulfadiazine (4 g daily) was begun. The sulfadiazine was replaced with oral c1indamycin (600 mg three times daily) because of intolerance. The redness and pain resolved rapidly, and the patient was given oral tetracycline (500 mg 3 times daily). Ten weeks later, visual acuity remained unchanged because of macular destruction. There were only I+ cells in the anterior chamber, and the retinal lesion was inactive (Fig 9). There was no clinical or radiologic evidence of intracranial toxoplasmosis. Case 5. A 35-year-old homosexual man with AIDS presented with floaters in his right eye. Visual acuity was 20/20 in the right eye and 20/20 in the left. There were two areas of retinal necrosis without hemorrhage, superior and inferior to the optic nerve head. His CD4 lymphocyte count was 61/mm 3•
502
Figure 7. C ase 3. Retinal biops y with T oxoplasma gondii cysto zoite with micronemes (arrow), rhoptries (arr owh eads), and polysaccharide granules (asterisk) (original magnification , X28,OOO).
In the left eye, there were cotton-wool spots only. Intravenous ganciclovir (5 mg/kg twice daily) was started for presumed CMV retinopathy. One month later, the patient complained of a red eye. Visual acuit y had decreased to 20/50 in the right eye. There were 2+ cells in the anterior chamber with keratic precipitates and a 3+ vitreous inflammatory reaction. The areas of retinal necrosis had enlarged . (Fig 10). Visual acuity decreased further to counting fingers at 2 feet over the next 2 weeks. A diagnosis of presumed toxoplasmic retinochoroiditis was made, and therapy was changed to oral pyrimethamine (50 mg daily). Two weeks later, visual acuity had improved to 20/40 in the right eye. Anterior chamber cells and vitreous inflammatory reaction had diminished , and the retinal lesions were lessopaque . After 6 weeks of antiparasitic therapy, visual acuity was 20/25 in the right eye. He was maintained on oral pyrimethamine (25 mg daily) (Fig II). There was no clinical or radiologic evidence of intracranial toxoplasmosis . Approximately I year later, the patient experienced blurred vision in the left eye. Visual acuit y was 20/ 25 in the right eye and 20/15 in the left. Results of examination showed quiet anterior chambers in both eyes. In the right eye, there was an area ofgranular retinal opacity adjacent to the inferior temporal border of the old, inacti ve toxoplasmosis scar (Fig 12). In the left eye, there was a similar lesion in the macula. The CD4 lymphocyte count had decreased to 14/mm 3 • Cytomegalovirus retinopathy was diagnosed, and intravenous ganciclovir therapy was started (5 mg/kg twice daily), with resolution of the retinal opacification in both eyes (Fig 13). Cytomegalovirus retinopathy reactivated three times but responded to increased doses of antiviral medications on each occasion . The toxoplasmosis scars rema ined inactive. Th e patient died of aspergillosis.
Results Patient characteristics and ocular findings are summarized in Table I. All five patients were immunosuppressed. One
Elkins et al . Ocular Toxoplasmosis
Top left, Figure l:l. Case 4, right eye. Untreated toxoplasmic retinochoroiditis in the macula. There is a marked vitreous intlammatory reaction overlying the lesion. The patient had the acquired immune deficiency syndrome. Top right, Figure 9. Case 4, right eye. The same area shown in Figure 8, photographed 3 months later, while the patient was receiving antiparasitic therapy. There is an inactive retinochoroidal scar with subretinal fibrosis. Center left, Figure 10. Case 5, right eye. Two peripapillary foci of untreated toxoplasmic retinochoroiditis. Center right, Figure 11. Case 5, right eye. The same area shown in Figure 10, 5 months later, while the patient was receiving maintenance antiparasitic therapy. The patient had the acquired immune deficiency syndrome. Bottom left, Figure 12. Case 5, right eye. Fundus montage shows a new focus of untreated cytomegalovirus retinopathy (arrow) adjacent to the inferotemporal border of the retinochoroidal scar shown in Figure 11. The scar is inactive. The photograph was taken 10 months after Figure 11 was taken. Bottom right, Figure 13. Case 5, right eye. The same area shown in Figure 12, photographed 1 year after Figure 11 was photographed. The patient was receiving intravenous ganciclovir. The area of cytomegalovirus retinopathy appears inactive.
503
Ophthalmology
Volume 101, Number 3, March 1994
.;:;; '"
o
-.~ E ~ c~
C
o
Z
C
o Z
..'" 0. 0 ><
bB u
.f:
o Z
+ ......
o
o Z
Z
+ ......
c o
0
o
0
C C
ZZ
ZZ
+
N
o
(J)
o
+ +
"'" "'" ~
0
:>-Z
+ +
"'" N
00
00
II
o
[f)
•
'" 0
~z
504
II
c
o
was being treated with prednisone for systemic lupus erythematosus and had received azathioprine, which had been discontinued because of bone marrow suppression before development of ocular disease. Two patients had graft-versus-host disease, which was diagnosed after allogeneic bone marrow transplantation for chronic myelogenous leukemia , and were being treated with immunosuppressive drugs. Two patients had AIDS, as defined by the Centers for Disease Control." Before presentation, none of the patients had a clinically apparent nonocular Toxoplasma gondii infection, although one patient was found to have intracranial toxoplasmosis on subsequent evaluation, which included CT scanning. None showed any evidence of a pre-existing chorioretinal scar. Toxoplasmic retinochoroiditis was unilateral in all five patients and was unifocal in four of five patients. One patient had two discrete lesions. All patients had anterior chamber cells ranging from trace to 3+, as defined by Hogan et al,? at the time toxoplasmic retinochoroiditis was diagnosed. Two of the five patients had posterior iris synechiae. All patients had vitreous cells, with four of the five having a 2 or 3+ inflammatory reaction , as defined by Nussenblatt et al." Inflammatory reactions were least severe in the patients with graft-versus-host disease; in one patient (case 3), the reactions in both the anterior chamber and vitreous were believed to be no more severe than those seen in patients with AIDS and CMV retinopathy. Toxoplasmic retinochoroiditis lesions were characterized by full-thickness necrosis, a thick and densely opaque yellow- white appearance , and well-demarcated, smooth borders (Figs l , 3, 4, 6, 8, lO). None of the borders had the dry granular appearance associated with CMV retinopathy. There was no retinal hemorrhage in four patients; in the fifth patient, there was a trace of hemorrhage in one spot only (Fig 6). All patients were diagnosed initially with CMV retinopathy . The time between introduction of ganciclovir therapy and the switch to antiparasitic therapy ranged from 3 weeks to 3 months (median, 6 weeks). Toxoplasma gondii infection was confirmed by biopsy in two patients. The others were presumed to have toxoplasmic retinochoroiditis on the basis of rapid clinical improvement with antiparasitic therapy. Response to antiparasitic therapy was good in all patients. Anterior chamber and vitreous inflammatory reactions decreased noticeably within 3 days to 3 weeks. Retinal lesions became less opaque over a similar time period, and evolved into inactive, atrophic retinochoroidal scars within 2 to 6 months in those patients who survived. Two patients had improvement of their visual acuities: from 20/70 to 20/30 in one case and from counting fingers to 20/25 in the other. The other three patients had no change in their visual acuities. One died of an unrelated infection soon after his retinal biopsy; one had a dense posterior subcapsular cataract that was not removed; and one had destruction of the macula from the infection. The patient described in case 5 had toxoplasmic retinochoroiditis and CMV retinopathy in the same eye, but
Elkins et al . Ocular Toxoplasmosis in different locations and at different times. The toxoplasmic lesions were similar to those described for other patients in this series and was associated with a marked inflammatory reaction (Fig 10). In contrast, the CMV infection was characterized by a dry, granular opacification of the retina that was much less dense than the previous toxoplasmic lesion and had little associated inflammatory reaction in the vitreous (Fig 12).
Discussion Cytomegalovirus retinopathy is the most common infection of the eye in patients with AIDS. It is believed to occur in 20% to 25% of patients," Although less common, toxoplasmic retinochoroiditis is another well-documented infection in patients with AIDS. I - 3 It probably accounts for only I% to 3% of retinal infections associated with AIDS, even in populations with a high prevalence of Toxoplasma gondii infection.? In patients who are immunosuppressed for reasons other than AIDS, the prevalence of CMV retinopathy is only 1% to 5%.10-12 The prevalence of toxoplasmic retinochoroiditis in patients who are immunosuppressed for reasons other than AIDS is not known, although there are numerous case reports in the literature.P:" In our experience, toxoplasmic retinochoroiditis may be as common as CMV retinopathy in immunosuppressed patients without AIDS. Because of emphasis in the recent medical literature on CMV as the leading cause of necrotizing retinopathy in patients with AIDS, ophthalmologists may fail to consider other causes of necrotizing retinopathy in immunocompromised patients. Several clinical factors distinguished these cases of toxoplasmic retinochoroiditis from the typical features of CMV retinopathy in immunocompromised patients. They included a prominent anterior chamber and vitreous inflammatory reaction, absence of retinal hemorrhage, and opacified retina that appeared thicker and more densely white-yellow. Borders were better defined and smoother in contour than the irregular, granular borders associated with CMV retinopathy. Prominent intraocular inflammatory reactions should suggest a diagnosis of ocular toxoplasmosis, since only minimal inflammation is seen in cases of CMV retinopathy. Conversely, there may be a spectrum to the amount of inflammation seen in immunosuppressed patients with ocular toxoplasmosis; therefore, a lack of inflammation does not rule out toxoplasmic retinochoroiditis, as in our case 3. Inflammatory signs were most prominent in the two patients with AIDS in this series, although other investigators have found little intraocular inflammation in patients with AIDS and ocular toxoplasmosis." Nevertheless, in our overall experience, patients with AIDS and toxoplasmic retinochoroiditis have much more severe inflammatory reactions associated with their retinal infections than do patients with CMV retinopathy. They can present with red, painful eyes and may develop pos-
terior synechiae, which rarely occurs in patients with CMV retinopathy. Cytomegalovirus retinopathy occurs only in patients with severe degrees of immunosuppression. Reported median CD4 lymphocyte counts in patients with AIDS and CMV end-organ disease have varied from 0 to 29 mm' with a range from 0 to 42 mm 3 •5, 17, 18 The median CD4 count associated with ocular toxoplasmosis in immunocompromised patients is not known, but in patients with AIDS and central nervous system toxoplasmosis, it has been reported to be 50 mrn' (range, 0 to 730 mm 3) . 19 Although these ranges overlap, it is likely that patients with AIDS and ocular toxoplasmosis will frequently have CD4 counts higher than those seen in patients with CMV retinopathy. Although conclusions cannot be drawn from the two patients in this study (CD4 counts of 20 and 61 mm:'), determination of CD4 counts may help to differentiate these two diseases in some cases; in particular, retinopathy in a patient with AIDS and a CD4 count higher than 50 mm' is statistically unlikely to be CMV retinopathy, and toxoplasmic retinochoroiditis should be considered. 17,18 In most cases, antibody titers are of little help in differentiating these infections. The prevalence of anti-CMV and anti-Toxoplasma gondii antibodies in healthy adults in the United States may be approximately 40% or higher for each. 2o-23 Anti-CMV antibodies are present in at least 95% of patients with AIDS,22,23 and anti-Toxoplasma gondii antibodies are present in up to 74% of patients with AIDS.22 Lack of detectable antibodies makes the diagnosis of either CMV retinopathy or toxoplasmic retinochoroiditis very unlikely, but both have been reported in patients with negative serologies.r':" Blood and urine cultures are also not helpful, since there is a poor correlation between CMV viremia, viuria, and active CMV disease in patients with AIDS. 25 Response to empiric therapy may help distinguish between these two infections. Lack of response to antiviral therapy points the clinician toward a diagnosis of ocular toxoplasmosis, since 80% to 100% of patients with CMV retinopathy are believed to have at least some response to ganciclovir or foscarnet therapy.V' In our experience, all patients with ocular toxoplasmosis respond to pyrimethamine and sulfadiazine or other antiparasitic drugs. The best medical therapy for ocular toxoplasmosis in immunosuppressed patients has not been determined; a variety of regimens appear to be effective.':" Retinal biopsy will confirm a diagnosis in lieu of a therapeutic drug trial. Biopsy specimens should come from the active border of lesions. It is important to distinguish between these two infections. Both can be successfullycontrolled in many patients with appropriate medications, and therefore early and accurate diagnosis may allow preservation of vision. Furthermore, the development of ocular infections by either Toxoplasma gondii or CMV have important implications for the overall health of the patient. Autopsy studies have shown that all patients with CMV retinopathy have tissueinvasive infections in other parts of the body, and treating
505
Ophthalmology
Volume 101, Number 3, March 1994
CMV retinopathy appears to control concurrent, nonocular CMV infections." Ocular toxoplasmosis may be the first manifestation of multifocal disease that includes intracranial infection. The reported prevalence of concurrent cerebral toxoplasmosis in patients with AIDS and ocular toxoplasmosis ranges from 29% to 50%.1 ,2 Therefore, CT scanning of the head should be considered when a diagnosis ofocular toxoplasmosis is made. In cases where a diagnosis of ocular toxoplasmosis is uncertain, a ringenhancing lesion on CT scan of the head may also help to support the diagnosis, although a negative CT scan does not rule it out. Scanning may also allow diagnosis of an intracranial lesion before it becomes clinically apparent. Toxoplasmosis and CMV infections can coexist in immunosuppressed patients, as occurred in case 5, and may in fact be present concurrently, at rates higher than would be expected from chance alone." lt has been hypothesized that Toxoplasma gondii may carry CMV into the eye, or there may be some synergism between the agents." Concurrent CMV retinopathy and toxoplasmic retinochoroiditis in patients with AIDS have been reported previously.' Whether these cases reflect some interaction between the agents or whether it simply reflects the high prevalence of CMV retinopathy in patients with AIDS has not been determined. It is interesting to note, however, that in case 5, CMV retinopathy arose at the border of the previous Toxoplasma gondii infection. Perhaps the CMV was simply better able to invade the retina at this previous site of injury. It is important to remember that the appearance of toxoplasmic retinochoroiditis in immunosuppressed patients is very different from that in immunocompetent patients, in whom the appearance is characterized by a discrete focus of intense retinal inflammation adjacent to a chorioretinal scar. Also, it should be remembered that not all cases oftoxoplasmic retinochoroiditis in immunocompromised patients will have the features described herein. There can be a spectrum offindings, which include single discrete lesions in one or both eyes, or multifocal discrete lesions, instead of diffuse areas of retinal necrosis. 1 The cases presented in this series fall into the category that can be most easily confused with CMV retinopathy, and therefore represent an important subset of patients. Acknowledgment Some aspects of patient evaluations were performed by Brian P. Conway, MD, Charlottesville, Virginia, Susan Lawrence, MD , Lancaster, California, and R. Scott Hitt, MD, Beverly Hills, California.
References I. Holland GN, Engstrom RE Jr, Glasgow BJ, et al. Ocular
toxoplasmosis in patients with the acquired immunodeficicncy syndrome. Am J Ophthalmol 1988;106:653-67. 2. Cochereau-Massin I, LeHoang P, Lautier-Frau M, et al. Ocular toxoplasmosis in human immunodeficiency virusinfected patients. Am J Ophthalmol 1992;114:130-5. 3. Gagliuso DJ, Teich SA, Friedman AH, Orellana J. Ocular toxoplasmosis in AIDS patients. Trans Am Ophthalmol Soc 1990;88:63-88.
506
4. Engstrom RE Jr, Holland GN, Nussenblatt RB, Jabs DA. Current practices in the management of ocular toxoplasmosis. Am J Ophthalmol 1991;111:601-10. 5. Studies of Ocular Complications of AIDS Research Group, in collaboration with the AIDS Clinical Trials Group. Mortality in patients with the acquired immunodeficiency syndrome treated with either foscarnet or ganciclovir for cytomegalovirus retinitis. N Eng) J Med 1992;326:213-20. 6. Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded survcillance case definition for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep 1992;41:1-19. 7. Hogan MJ, Kimura SJ, Thygeson P. Signs and symptoms of uveitis. I. Anterior uveitis. Am J Ophthalmol I959;47(no.5, pt.2): 155-76 . 8. Nussenblatt RB, Palestine AC, Chan C-C, Roberge F. Standardization ofvitreal inflammatory activity in intermediate and posterior uveitis. Ophthalmology 1985;92:467-71. 9. Holland GN. Retinal and choroidal infections. In: Ryan SJ, Lewis H, eds. Medical and Surgical Retina: Advances, Controversies, and Management. St. Louis:Mosby Year Book [in press]. 10. Porter R, Crombie AL, Gardner PS, Uldall RP. Incidence of ocular complications in patients undergoing renal transplantation. Br Med J 1972;3:133-6. II. Astle IN, Ellis PP. Ocular complications in renal transplant patients. Ann Ophthalmol 1974;6:1269-74. 12. Oberman AE, Chatterjee SN. Ocular complications in renal transplant recipients. West J Med 1975;123:184-7. 13. Nicholson DH , Wolchok EB. Ocular toxoplasmosis in an adult receiving long-term corticosteroid therapy. Arch Ophthalmol 1976;94:248-54. 14. Hoerni B, VallatM, Durand M, Pesme D. Ocular toxoplasmosis and Hodgkin's disease. Arch Ophthalmol 1978;96:62-3. 15. Yeo JH, Jakobiec FA, Iwamoto T, et al. Opportunistic toxoplasmic retinochoroiditis following chemotherapy for systemic lymphoma. A light and electron microscopic study. Ophthalmology 1983;90:885-98. 16. Bottoni F, Gonnella P, Autelitano A, Orzalesi N. Diffuse necrotizing retinochoroiditis in a child with AIDS and toxoplasmic encephalitis. Graefcs Arch Clin Exp Ophthalmol 1990;228:36-9. 17. Pcrtel P, Hirschtick R, Phair J, et al. Risk of developing cytomegalovirus retinitis in persons infected with the human immunodeficiency virus. J Acquir Immune Defic Syndr 1992;5:1069-74. 18. Kuppermann BD, Petty JG, Richman DD, et al. Correlation between CD4+ counts and prevalence of cytomegalovirus retinitis and human immunodeficiency virus-related noninfectious retinal vasculopathy in patients with acquired immunodeficiency syndrome. Am J Ophthalmol 1993;115: 575-82. 19. Porter SB, Sande MA. Toxoplasmosis of the central nervous system in the acquired immunodeficiency syndrome. N Eng) J Med 1992;327:1643-8. 20. McCabe R, Remington JS. Toxoplasmosis: the time has come [editorial]. N Eng) J Mcd 1988;318:313-15. 21. Kean BH, Kimball AC, Christenson WN. An epidemic of acute toxoplasmosis. JAMA 1969;208:1002-4. 22. Quinn TC, Piot P, McCormick JB, et al. Serologic and immunologic studies in patients with AIDS in North America and Africa. The potential role of infectious agents as cofactors in human immunodeficiency virus infection . JAMA 1987;257:2617-21.
Elkins et al . Ocular Toxoplasmosis 23. Jordan Me. Cytomegalovirus infections. In: Hoeprich PD, Jordan MC, eds. Infectious Diseases: A Modern Treatise of Infectious Processes, 4th ed. Philadelphia: JB Lippincott, 1989; chap. 83. 24. Luft BJ, Brooks RG, Conley FK, et al. Toxoplasmic encephalitis in patients with acquired immune deficiency syndrome. JAMA 1984;252:913-17. 25. Neuwirth J, Gutman I, Hofeldt AJ, et al. Cytomegalovirus retinitis in a young homosexual male with acquired immunodeficiency. Ophthalmology 1982;89:805-8.
26. Zurlo JJ, O'Neill D, Polis MA, et al. Lack of clinical utility of cytomegalovirus blood and urine cultures in patients with HIV infection. Ann Intern Med 1993; 118: 12-17. 27. Morinelli EN, Dugel PD, Lee M, et al. Opportunistic intraocular infections in AIDS. Trans Am Ophthalmol Soc 1992;90:97-109. 28. Ryning FW, Mills J. Pneumocystis carinii, Toxoplasma gondii, Cytomegalovirus and the compromised host. West J Med 1979;130:18-34.
507
Originally received: June II , 1993. Revision accepted: October I, 1993. 1 UCLA Ocular Inflammatory Disease Center, Jules Stein Eye Institute, a nd Department of Ophthalmology, UCLA School of Medicine, Los Angeles. 2 Department of Ophthalmology, Ohio State Un iversity College of Medicine, Columbus.
3 Wilmer Ophthalmological Institut e, Johns Hopkins University School of Medicine, Baltimore.
Supp orted in part by the Skirball Foundation, Los Angeles, California; Research to Prevent Blindness, Inc., New York , New York; NIH grant EY08057 , Bethesda, Maryland (Dr. Holland) ; Kricker Foundation, Columbus , Ohio (Dr. Oprerncak ); NIH/ NCI grant 4POICA 15396, Bethesda, Maryland (Drs. Dunn and Jabs); the E. A. Baker Foundation for the Prevention of Blindn ess, Canadia n National Institute for the Blind, Toronto, Ontario, Canada (Dr. John ston). Dr. Johnston is a Frances Howard Goldw yn Fellow in Cornea -External Ocular Diseases and Uveitis. Dr. Jabs is a Research to Prevent Blindness Olga Wiess Keith Scholar. Reprint requests to Ga ry N. Holland, MD, Jutes Stein Eye Institute, 100 Stein Plaza, UCLA, Los Angeles, CA 90024-7003.
499
Ophthalmology
Volume 101, Number 3, March 1994
Ocular toxoplasmosis in immunocompromised hosts presents a special challenge to ophthalmologists in terms of diagnosis and management. The clinical appearance may be very different from that seen in otherwise healthy adults . Active lesions can be bilateral and multifocal. They usually do not arise at the borders of pre-existing retinochoroidal scars, suggesting that disease is acquired or that lesions arise from organisms newly disseminated to the eye from nonocular sites of disease. I Among patients with acquired immune deficiency syndrome (AIDS), the appearance of these lesions can be confused with cytomegalovirus (CMV) retinopathy, which is a much more common problem. Without treatment, both toxoplasmic retinochoroiditis and CMV retinopathy in immunosuppressed patients will destroy extensive areas of the retina. Ocular toxoplasmosis responds rapidly to pyrimethamine and sulfadiazine or other antiparasitic therapy, 1-4 and CMV retinopathy usually responds to antiviral therapy.' Treatments for both may be associated with serious complications, however. Therefore, accurate differentiation between toxoplasmic retinochoroiditis and CMV retinopathy is critical , both to preserve vision and to avoid the unnecessary use of toxic medications. We present five cases of toxoplasmic retinochoroiditis in patients who were immunosuppressed by a variety of causes. All were initially misdiagnosed as having CMV retinopathy. A review of findings identified clinical features that can help the practitioner distinguish between these two infections at presentation.
Case Reports Case 1. A 44-year-old woman with a several year history of systemic lupus erythematosus reported decreased vision in her right eye that was associated with retrobulbar pain. At the time , she was taking prednisone 40 mg every other day. Three months earlier, azathioprine therapy had been discontinued because of bone marrow suppression. Plaquenil therapy had been discontinued 3 years earlier because of abnormal FarnsworthMunsell 100 hue testing, despite a normal fluorescein angiogram, electro-oculogram, and fundus photographs. Visual acuity was 20/25 in the right eye and 20/25 in the left. The right eye showed mild conjunctival injection , keratic precipitates, and I+ cells in the anterior chamber. There was a single focus of perivascular retinal necrosis in the inferonasal periphery without hemorrhage. Review of the photographs taken 3 years earlier showed no pre-existing lesions in the area of active disease. The left eye was normal . Intravenous ganciclovir therap y (5 mg/kg twice daily) was started for presumed CMV retinopathy. Thre e weeks later, her visual acuity was 20/70 in the right eye and 20/25 in the left. In the right eye, there were posterior synechia e, 2-3+ cells in the anterior chamber, and a 3+ vitreous inflammatory reaction. The area of necrosis had enlarged , and there was inflammatory sheathing of retinal vessels adjacent to the lesion (Fig I). Ocular toxopla smosis was believed to be a more likely cause of the lesion, and ganciclovir was discontinued. Oral pyrimethamine (25 mg once weekly) and sulfadiazine (500 mg 4 times daily) were started . The sulfadiazine was later switched to c1indamycin (750 mg 4 times daily) because ofnau-
500
sea, vomiting, and diarrhea. Six weeks after beginning antiparasitic therap y, her visual acuity was 20/50 in the right eye. There were fewer keratic precipitates , I+ cells in the anterior chamber, I+ vitreous inflammatory reaction , and decreased retinal opacification. The borders of the lesion were becoming scarred. The left eye remained normal. The pyrimethamine was discontinued , but the patient was maintained on oral c1indamycin (150 mg 4 times daily). Seven months later, visual acuity was 20/30 in the right eye. There were no cells in the anterior chamber or vitreous, and the retinal lesion appeared inactive. The c1indamycin was discontinued, but the patient was maintained on low dose prednisone to control the systemic lupus erythematosus. Four years later, her visual acuity was 20/30 in the right eye, and there was an atrophic retinochoroidal scar (Fig 2). Clinical signs of intracranial toxoplasmosis never developed. Case 2. A 36-year-old man underwent bone marrow transplantation for chronic myelogenous leukemia. Results of ophthalmic examination before transplantation were normal. Graft-versus-host disease developed, which was treated with a combination of methylprednisolone , azathioprine, cyclosporine, and psoralens ultraviolet A therapy. Aspergillus sp sinusitis also developed, which was treated with intravenous amphotericin B. One year later, the patient presented with a I-day history of "slight haze" over his right eye. Visual acuity was 20/25 in the right eye and 20/ 20 in the left. The right eye had trace anterior chamber cells, a 1-2+ vitreous inflammatory reaction, and a dense, yellow- white curvilinear infiltrate posterior to an area of atroph ic retina in the superonasal quadrant (Fig 3). The left fundus was normal. Blood cultures were positive for CMV. A diagnosis of CMV retinopathy was made, and intravenous ganciclovir therapy (5 rug/kg twice daily) was started . The immunosuppre ssive medications could not be tapered because of chroni c graft-versus-host disease. After 6 weeks of ganciclovir therapy, visual acuit y was 20/ 40 in the right eye and 20/20 in the left. There was still a 1-2+ vitreous inflammatory reaction in the right eye, and the retinal lesion had advanced posteriorly (Fig 4). Ganciclovir resistance was considered, and intravenous foscarnet therapy (60 mg/kg 3 times daily) was added. Two weeks later, visual acuity was 20/ 400 in the right eye and 20/20 in the left. The vitreous inflammatory reaction increased . Several days later, visual acuity was counting fingers in the right eye and 20/20 in the left, and an inferior visual field defect was noted in the right eye. The retinal lesion obscured the optic nerve head. The creatinine level had risen to 2.9 mg/dl, possibly as a result of foscarnet toxicity. Retinal and vitreous biopsies were performed. Toxoplasma gondii cysts were identified in the retina and vitreous (Fig 5), and immunohistochemical stains for CMV and varicella-zoster virus were negative. The patient had no clinical signs of intracranial toxoplasmosis. Antipa rasitic therapy using oral trimethoprim (160 mgj/sulfametho xazole (800 mg) twice daily was started. Three days later, there was a decrease in the vitreous inflammatory reaction and retinal opacification, which continued to resolve over a 3-week period of therapy. Visual acuity was not determined because of a gas bubble placed in the eye during the biopsy procedure. Soon thereafter, the patient died of sequelae of his graft-versus-host disease. Case 3. A 38-year-old man underwent bone marrow transplantation for chronic myelogenous leukemia. Acute graft-versus-host disease subsequently developed, which was treated with methylprednisolone (24 mg twice daily). Results of ophthalmic examin ation showed visual acuity of 20/50 in the right eye and 20/25 in the left. There was focal scarring of the right macula
Elkins et al . Ocular Toxoplasmosis
Top left, Figure 1. C ase 1, right eye. Untreated toxopl asmic ret inochoroiditis in the inferonasal periphery. The patient was being treated for systemic lupus erythematosu s. Top right, Figure 2. Case 1, right eye. The same port ion of the retin a shown in Figure 1, photographed 4 years later. There is an atrophic retinochoroidal scar in the area of the previous toxoplasmic infection . Center left, Figure 3. Case 2, right eye. Untreated toxoplasmic retinochoroiditis in the superonasal qu adrant. The patient had graft-versus-host disease. The border of the lesion is densely opaque; ant erior to the op acity, the retina is atroph ic. Center right, Figure 4. Case 2, right eye. The same area of the ret ina sho wn in Figure 3, photographed 6 weeks later . The border of the untreated lesion has advan ced posteriorly. Bottom left, Figure 5. Light micro graph of a biop sy specimen shown in Figures 3 an d 4 sho ws Toxoplasma gondii cysts (arrows) (hematoxylin-eosin; ori ginal magnification , X400). Bottom right, Figure 6. Case 3, left eye. Untreated toxoplasmic retinochoroiditis in th e inferotemporal region . The patient had graft-versus-ho st disease. There is scant retinal hemorrhage in the lesion . .
501
Ophthalmology
Volume 101, Number 3, March 1994
due to previous trauma. Results of examination of the left eye were normal. He had positive CMV blood cultures and serum antibodies to varicella-zoster virus and herpes simplex virus. Nine months later, chronic graft-versus-host disease developed , and a retinal lesion in the periphery of the left eye was discovered. Visual acuit y was 20/20 in the left eye. A tentative diagnosis ofviral retinitis was made , and oral corticosteroid dosage was tapered to 12.5 mg twice daily. Visual field constriction was eventuall y noted, and increased opacification of the retinal lesion was found on examination. The patient was treated for presumed CMV retinopathy with intravenous ganciclovir (5 mg/ kg twice daily), which was subsequently changed to intravenous foscamet (60 rug/kg 3 times daily) because ofleukopenia. Despite this therapy, the retinal lesion extended posteriorly (Fig 6). There was an oval area of retinal necrosis with thickened, densely opaque borders, a central atrophic area, and scant hemorrhage adjacent to a vessel that coursed through the area of necrosis. He had fine keratic precipitates, with I+ anterior chamber cells and a I+ vitreous inflammatory reaction. Intravenous acyclovir (IO mg/kg 3 times daily) was started because of the possibility of varicella-zoster virus retinopathy, but the lesion continued to worsen. Three months after development of the retinopathy and approximately 3 weeks after acyclovir was started , an endoretinal biopsy was performed. Tox oplasma gondii cysts were identified by electron microscopy (Fig 7). No viral particles were seen, and viral cultures were negative. The patient was treated with oral pyrimethamine (100 mg daily) and oral clindamycin (400 mg 3 times daily). Sulfadiazine therapy was discontinued after 3 days because of rash. Two weeks later, mental status changes developed, and results of computed tomography (CT) of the head showed probable intracranial toxoplasmosis . The retinal lesion resolved completely after 8 weeks of treatment, but the patient's vision remained poor because of a dense posterior subcapsular cataract. Mental status remained altered, and a brain biopsy was performed, which confirmed intracranial toxoplasmosis. The patient died shortly thereafter. Case 4. Decreased vision and severe pain developed in the right eye of a 44-year-old homosexual man with AIDS. Retinal necrosis without hemorrhage was noted in the right macula. His CD4 lymphocyte count was 20/mm 3• The patient was given intravenous ganciclovir (5 rng/kg twice daily) for presumed CMV retinopathy. Six weeks later, pain, redness, and decreased vision still existed. There was no resolution of retinal opacification, and the lesion had enlarged (Fig 8). Visual acuity was counting fingers (eccentric fixation) in the right eye and 20/2 0 in the left. There was marked ciliary flush, 3+ cells in the anterior chamber, extensive posterior synechiae , and a 3+ vitreous inflammatory reaction. A diagnosis of presumed toxoplasmic retinochoro iditis was made, and antiparasitic therapy with oral pyrimethamine (50 mg daily after a 200-mg loading dose) and oral sulfadiazine (4 g daily) was begun. The sulfadiazine was replaced with oral c1indamycin (600 mg three times daily) because of intolerance. The redness and pain resolved rapidly, and the patient was given oral tetracycline (500 mg 3 times daily). Ten weeks later, visual acuity remained unchanged because of macular destruction. There were only I+ cells in the anterior chamber, and the retinal lesion was inactive (Fig 9). There was no clinical or radiologic evidence of intracranial toxoplasmosis. Case 5. A 35-year-old homosexual man with AIDS presented with floaters in his right eye. Visual acuity was 20/20 in the right eye and 20/20 in the left. There were two areas of retinal necrosis without hemorrhage, superior and inferior to the optic nerve head. His CD4 lymphocyte count was 61/mm 3•
502
Figure 7. C ase 3. Retinal biops y with T oxoplasma gondii cysto zoite with micronemes (arrow), rhoptries (arr owh eads), and polysaccharide granules (asterisk) (original magnification , X28,OOO).
In the left eye, there were cotton-wool spots only. Intravenous ganciclovir (5 mg/kg twice daily) was started for presumed CMV retinopathy. One month later, the patient complained of a red eye. Visual acuit y had decreased to 20/50 in the right eye. There were 2+ cells in the anterior chamber with keratic precipitates and a 3+ vitreous inflammatory reaction. The areas of retinal necrosis had enlarged . (Fig 10). Visual acuity decreased further to counting fingers at 2 feet over the next 2 weeks. A diagnosis of presumed toxoplasmic retinochoroiditis was made, and therapy was changed to oral pyrimethamine (50 mg daily). Two weeks later, visual acuity had improved to 20/40 in the right eye. Anterior chamber cells and vitreous inflammatory reaction had diminished , and the retinal lesions were lessopaque . After 6 weeks of antiparasitic therapy, visual acuity was 20/25 in the right eye. He was maintained on oral pyrimethamine (25 mg daily) (Fig II). There was no clinical or radiologic evidence of intracranial toxoplasmosis . Approximately I year later, the patient experienced blurred vision in the left eye. Visual acuit y was 20/ 25 in the right eye and 20/15 in the left. Results of examination showed quiet anterior chambers in both eyes. In the right eye, there was an area ofgranular retinal opacity adjacent to the inferior temporal border of the old, inacti ve toxoplasmosis scar (Fig 12). In the left eye, there was a similar lesion in the macula. The CD4 lymphocyte count had decreased to 14/mm 3 • Cytomegalovirus retinopathy was diagnosed, and intravenous ganciclovir therapy was started (5 mg/kg twice daily), with resolution of the retinal opacification in both eyes (Fig 13). Cytomegalovirus retinopathy reactivated three times but responded to increased doses of antiviral medications on each occasion . The toxoplasmosis scars rema ined inactive. Th e patient died of aspergillosis.
Results Patient characteristics and ocular findings are summarized in Table I. All five patients were immunosuppressed. One
Elkins et al . Ocular Toxoplasmosis
Top left, Figure l:l. Case 4, right eye. Untreated toxoplasmic retinochoroiditis in the macula. There is a marked vitreous intlammatory reaction overlying the lesion. The patient had the acquired immune deficiency syndrome. Top right, Figure 9. Case 4, right eye. The same area shown in Figure 8, photographed 3 months later, while the patient was receiving antiparasitic therapy. There is an inactive retinochoroidal scar with subretinal fibrosis. Center left, Figure 10. Case 5, right eye. Two peripapillary foci of untreated toxoplasmic retinochoroiditis. Center right, Figure 11. Case 5, right eye. The same area shown in Figure 10, 5 months later, while the patient was receiving maintenance antiparasitic therapy. The patient had the acquired immune deficiency syndrome. Bottom left, Figure 12. Case 5, right eye. Fundus montage shows a new focus of untreated cytomegalovirus retinopathy (arrow) adjacent to the inferotemporal border of the retinochoroidal scar shown in Figure 11. The scar is inactive. The photograph was taken 10 months after Figure 11 was taken. Bottom right, Figure 13. Case 5, right eye. The same area shown in Figure 12, photographed 1 year after Figure 11 was photographed. The patient was receiving intravenous ganciclovir. The area of cytomegalovirus retinopathy appears inactive.
503
Ophthalmology
Volume 101, Number 3, March 1994
.;:;; '"
o
-.~ E ~ c~
C
o
Z
C
o Z
..'" 0. 0 ><
bB u
.f:
o Z
+ ......
o
o Z
Z
+ ......
c o
0
o
0
C C
ZZ
ZZ
+
N
o
(J)
o
+ +
"'" "'" ~
0
:>-Z
+ +
"'" N
00
00
II
o
[f)
•
'" 0
~z
504
II
c
o
was being treated with prednisone for systemic lupus erythematosus and had received azathioprine, which had been discontinued because of bone marrow suppression before development of ocular disease. Two patients had graft-versus-host disease, which was diagnosed after allogeneic bone marrow transplantation for chronic myelogenous leukemia , and were being treated with immunosuppressive drugs. Two patients had AIDS, as defined by the Centers for Disease Control." Before presentation, none of the patients had a clinically apparent nonocular Toxoplasma gondii infection, although one patient was found to have intracranial toxoplasmosis on subsequent evaluation, which included CT scanning. None showed any evidence of a pre-existing chorioretinal scar. Toxoplasmic retinochoroiditis was unilateral in all five patients and was unifocal in four of five patients. One patient had two discrete lesions. All patients had anterior chamber cells ranging from trace to 3+, as defined by Hogan et al,? at the time toxoplasmic retinochoroiditis was diagnosed. Two of the five patients had posterior iris synechiae. All patients had vitreous cells, with four of the five having a 2 or 3+ inflammatory reaction , as defined by Nussenblatt et al." Inflammatory reactions were least severe in the patients with graft-versus-host disease; in one patient (case 3), the reactions in both the anterior chamber and vitreous were believed to be no more severe than those seen in patients with AIDS and CMV retinopathy. Toxoplasmic retinochoroiditis lesions were characterized by full-thickness necrosis, a thick and densely opaque yellow- white appearance , and well-demarcated, smooth borders (Figs l , 3, 4, 6, 8, lO). None of the borders had the dry granular appearance associated with CMV retinopathy. There was no retinal hemorrhage in four patients; in the fifth patient, there was a trace of hemorrhage in one spot only (Fig 6). All patients were diagnosed initially with CMV retinopathy . The time between introduction of ganciclovir therapy and the switch to antiparasitic therapy ranged from 3 weeks to 3 months (median, 6 weeks). Toxoplasma gondii infection was confirmed by biopsy in two patients. The others were presumed to have toxoplasmic retinochoroiditis on the basis of rapid clinical improvement with antiparasitic therapy. Response to antiparasitic therapy was good in all patients. Anterior chamber and vitreous inflammatory reactions decreased noticeably within 3 days to 3 weeks. Retinal lesions became less opaque over a similar time period, and evolved into inactive, atrophic retinochoroidal scars within 2 to 6 months in those patients who survived. Two patients had improvement of their visual acuities: from 20/70 to 20/30 in one case and from counting fingers to 20/25 in the other. The other three patients had no change in their visual acuities. One died of an unrelated infection soon after his retinal biopsy; one had a dense posterior subcapsular cataract that was not removed; and one had destruction of the macula from the infection. The patient described in case 5 had toxoplasmic retinochoroiditis and CMV retinopathy in the same eye, but
Elkins et al . Ocular Toxoplasmosis in different locations and at different times. The toxoplasmic lesions were similar to those described for other patients in this series and was associated with a marked inflammatory reaction (Fig 10). In contrast, the CMV infection was characterized by a dry, granular opacification of the retina that was much less dense than the previous toxoplasmic lesion and had little associated inflammatory reaction in the vitreous (Fig 12).
Discussion Cytomegalovirus retinopathy is the most common infection of the eye in patients with AIDS. It is believed to occur in 20% to 25% of patients," Although less common, toxoplasmic retinochoroiditis is another well-documented infection in patients with AIDS. I - 3 It probably accounts for only I% to 3% of retinal infections associated with AIDS, even in populations with a high prevalence of Toxoplasma gondii infection.? In patients who are immunosuppressed for reasons other than AIDS, the prevalence of CMV retinopathy is only 1% to 5%.10-12 The prevalence of toxoplasmic retinochoroiditis in patients who are immunosuppressed for reasons other than AIDS is not known, although there are numerous case reports in the literature.P:" In our experience, toxoplasmic retinochoroiditis may be as common as CMV retinopathy in immunosuppressed patients without AIDS. Because of emphasis in the recent medical literature on CMV as the leading cause of necrotizing retinopathy in patients with AIDS, ophthalmologists may fail to consider other causes of necrotizing retinopathy in immunocompromised patients. Several clinical factors distinguished these cases of toxoplasmic retinochoroiditis from the typical features of CMV retinopathy in immunocompromised patients. They included a prominent anterior chamber and vitreous inflammatory reaction, absence of retinal hemorrhage, and opacified retina that appeared thicker and more densely white-yellow. Borders were better defined and smoother in contour than the irregular, granular borders associated with CMV retinopathy. Prominent intraocular inflammatory reactions should suggest a diagnosis of ocular toxoplasmosis, since only minimal inflammation is seen in cases of CMV retinopathy. Conversely, there may be a spectrum to the amount of inflammation seen in immunosuppressed patients with ocular toxoplasmosis; therefore, a lack of inflammation does not rule out toxoplasmic retinochoroiditis, as in our case 3. Inflammatory signs were most prominent in the two patients with AIDS in this series, although other investigators have found little intraocular inflammation in patients with AIDS and ocular toxoplasmosis." Nevertheless, in our overall experience, patients with AIDS and toxoplasmic retinochoroiditis have much more severe inflammatory reactions associated with their retinal infections than do patients with CMV retinopathy. They can present with red, painful eyes and may develop pos-
terior synechiae, which rarely occurs in patients with CMV retinopathy. Cytomegalovirus retinopathy occurs only in patients with severe degrees of immunosuppression. Reported median CD4 lymphocyte counts in patients with AIDS and CMV end-organ disease have varied from 0 to 29 mm' with a range from 0 to 42 mm 3 •5, 17, 18 The median CD4 count associated with ocular toxoplasmosis in immunocompromised patients is not known, but in patients with AIDS and central nervous system toxoplasmosis, it has been reported to be 50 mrn' (range, 0 to 730 mm 3) . 19 Although these ranges overlap, it is likely that patients with AIDS and ocular toxoplasmosis will frequently have CD4 counts higher than those seen in patients with CMV retinopathy. Although conclusions cannot be drawn from the two patients in this study (CD4 counts of 20 and 61 mm:'), determination of CD4 counts may help to differentiate these two diseases in some cases; in particular, retinopathy in a patient with AIDS and a CD4 count higher than 50 mm' is statistically unlikely to be CMV retinopathy, and toxoplasmic retinochoroiditis should be considered. 17,18 In most cases, antibody titers are of little help in differentiating these infections. The prevalence of anti-CMV and anti-Toxoplasma gondii antibodies in healthy adults in the United States may be approximately 40% or higher for each. 2o-23 Anti-CMV antibodies are present in at least 95% of patients with AIDS,22,23 and anti-Toxoplasma gondii antibodies are present in up to 74% of patients with AIDS.22 Lack of detectable antibodies makes the diagnosis of either CMV retinopathy or toxoplasmic retinochoroiditis very unlikely, but both have been reported in patients with negative serologies.r':" Blood and urine cultures are also not helpful, since there is a poor correlation between CMV viremia, viuria, and active CMV disease in patients with AIDS. 25 Response to empiric therapy may help distinguish between these two infections. Lack of response to antiviral therapy points the clinician toward a diagnosis of ocular toxoplasmosis, since 80% to 100% of patients with CMV retinopathy are believed to have at least some response to ganciclovir or foscarnet therapy.V' In our experience, all patients with ocular toxoplasmosis respond to pyrimethamine and sulfadiazine or other antiparasitic drugs. The best medical therapy for ocular toxoplasmosis in immunosuppressed patients has not been determined; a variety of regimens appear to be effective.':" Retinal biopsy will confirm a diagnosis in lieu of a therapeutic drug trial. Biopsy specimens should come from the active border of lesions. It is important to distinguish between these two infections. Both can be successfullycontrolled in many patients with appropriate medications, and therefore early and accurate diagnosis may allow preservation of vision. Furthermore, the development of ocular infections by either Toxoplasma gondii or CMV have important implications for the overall health of the patient. Autopsy studies have shown that all patients with CMV retinopathy have tissueinvasive infections in other parts of the body, and treating
505
Ophthalmology
Volume 101, Number 3, March 1994
CMV retinopathy appears to control concurrent, nonocular CMV infections." Ocular toxoplasmosis may be the first manifestation of multifocal disease that includes intracranial infection. The reported prevalence of concurrent cerebral toxoplasmosis in patients with AIDS and ocular toxoplasmosis ranges from 29% to 50%.1 ,2 Therefore, CT scanning of the head should be considered when a diagnosis ofocular toxoplasmosis is made. In cases where a diagnosis of ocular toxoplasmosis is uncertain, a ringenhancing lesion on CT scan of the head may also help to support the diagnosis, although a negative CT scan does not rule it out. Scanning may also allow diagnosis of an intracranial lesion before it becomes clinically apparent. Toxoplasmosis and CMV infections can coexist in immunosuppressed patients, as occurred in case 5, and may in fact be present concurrently, at rates higher than would be expected from chance alone." lt has been hypothesized that Toxoplasma gondii may carry CMV into the eye, or there may be some synergism between the agents." Concurrent CMV retinopathy and toxoplasmic retinochoroiditis in patients with AIDS have been reported previously.' Whether these cases reflect some interaction between the agents or whether it simply reflects the high prevalence of CMV retinopathy in patients with AIDS has not been determined. It is interesting to note, however, that in case 5, CMV retinopathy arose at the border of the previous Toxoplasma gondii infection. Perhaps the CMV was simply better able to invade the retina at this previous site of injury. It is important to remember that the appearance of toxoplasmic retinochoroiditis in immunosuppressed patients is very different from that in immunocompetent patients, in whom the appearance is characterized by a discrete focus of intense retinal inflammation adjacent to a chorioretinal scar. Also, it should be remembered that not all cases oftoxoplasmic retinochoroiditis in immunocompromised patients will have the features described herein. There can be a spectrum offindings, which include single discrete lesions in one or both eyes, or multifocal discrete lesions, instead of diffuse areas of retinal necrosis. 1 The cases presented in this series fall into the category that can be most easily confused with CMV retinopathy, and therefore represent an important subset of patients. Acknowledgment Some aspects of patient evaluations were performed by Brian P. Conway, MD, Charlottesville, Virginia, Susan Lawrence, MD , Lancaster, California, and R. Scott Hitt, MD, Beverly Hills, California.
References I. Holland GN, Engstrom RE Jr, Glasgow BJ, et al. Ocular
toxoplasmosis in patients with the acquired immunodeficicncy syndrome. Am J Ophthalmol 1988;106:653-67. 2. Cochereau-Massin I, LeHoang P, Lautier-Frau M, et al. Ocular toxoplasmosis in human immunodeficiency virusinfected patients. Am J Ophthalmol 1992;114:130-5. 3. Gagliuso DJ, Teich SA, Friedman AH, Orellana J. Ocular toxoplasmosis in AIDS patients. Trans Am Ophthalmol Soc 1990;88:63-88.
506
4. Engstrom RE Jr, Holland GN, Nussenblatt RB, Jabs DA. Current practices in the management of ocular toxoplasmosis. Am J Ophthalmol 1991;111:601-10. 5. Studies of Ocular Complications of AIDS Research Group, in collaboration with the AIDS Clinical Trials Group. Mortality in patients with the acquired immunodeficiency syndrome treated with either foscarnet or ganciclovir for cytomegalovirus retinitis. N Eng) J Med 1992;326:213-20. 6. Centers for Disease Control and Prevention. 1993 revised classification system for HIV infection and expanded survcillance case definition for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep 1992;41:1-19. 7. Hogan MJ, Kimura SJ, Thygeson P. Signs and symptoms of uveitis. I. Anterior uveitis. Am J Ophthalmol I959;47(no.5, pt.2): 155-76 . 8. Nussenblatt RB, Palestine AC, Chan C-C, Roberge F. Standardization ofvitreal inflammatory activity in intermediate and posterior uveitis. Ophthalmology 1985;92:467-71. 9. Holland GN. Retinal and choroidal infections. In: Ryan SJ, Lewis H, eds. Medical and Surgical Retina: Advances, Controversies, and Management. St. Louis:Mosby Year Book [in press]. 10. Porter R, Crombie AL, Gardner PS, Uldall RP. Incidence of ocular complications in patients undergoing renal transplantation. Br Med J 1972;3:133-6. II. Astle IN, Ellis PP. Ocular complications in renal transplant patients. Ann Ophthalmol 1974;6:1269-74. 12. Oberman AE, Chatterjee SN. Ocular complications in renal transplant recipients. West J Med 1975;123:184-7. 13. Nicholson DH , Wolchok EB. Ocular toxoplasmosis in an adult receiving long-term corticosteroid therapy. Arch Ophthalmol 1976;94:248-54. 14. Hoerni B, VallatM, Durand M, Pesme D. Ocular toxoplasmosis and Hodgkin's disease. Arch Ophthalmol 1978;96:62-3. 15. Yeo JH, Jakobiec FA, Iwamoto T, et al. Opportunistic toxoplasmic retinochoroiditis following chemotherapy for systemic lymphoma. A light and electron microscopic study. Ophthalmology 1983;90:885-98. 16. Bottoni F, Gonnella P, Autelitano A, Orzalesi N. Diffuse necrotizing retinochoroiditis in a child with AIDS and toxoplasmic encephalitis. Graefcs Arch Clin Exp Ophthalmol 1990;228:36-9. 17. Pcrtel P, Hirschtick R, Phair J, et al. Risk of developing cytomegalovirus retinitis in persons infected with the human immunodeficiency virus. J Acquir Immune Defic Syndr 1992;5:1069-74. 18. Kuppermann BD, Petty JG, Richman DD, et al. Correlation between CD4+ counts and prevalence of cytomegalovirus retinitis and human immunodeficiency virus-related noninfectious retinal vasculopathy in patients with acquired immunodeficiency syndrome. Am J Ophthalmol 1993;115: 575-82. 19. Porter SB, Sande MA. Toxoplasmosis of the central nervous system in the acquired immunodeficiency syndrome. N Eng) J Med 1992;327:1643-8. 20. McCabe R, Remington JS. Toxoplasmosis: the time has come [editorial]. N Eng) J Mcd 1988;318:313-15. 21. Kean BH, Kimball AC, Christenson WN. An epidemic of acute toxoplasmosis. JAMA 1969;208:1002-4. 22. Quinn TC, Piot P, McCormick JB, et al. Serologic and immunologic studies in patients with AIDS in North America and Africa. The potential role of infectious agents as cofactors in human immunodeficiency virus infection . JAMA 1987;257:2617-21.
Elkins et al . Ocular Toxoplasmosis 23. Jordan Me. Cytomegalovirus infections. In: Hoeprich PD, Jordan MC, eds. Infectious Diseases: A Modern Treatise of Infectious Processes, 4th ed. Philadelphia: JB Lippincott, 1989; chap. 83. 24. Luft BJ, Brooks RG, Conley FK, et al. Toxoplasmic encephalitis in patients with acquired immune deficiency syndrome. JAMA 1984;252:913-17. 25. Neuwirth J, Gutman I, Hofeldt AJ, et al. Cytomegalovirus retinitis in a young homosexual male with acquired immunodeficiency. Ophthalmology 1982;89:805-8.
26. Zurlo JJ, O'Neill D, Polis MA, et al. Lack of clinical utility of cytomegalovirus blood and urine cultures in patients with HIV infection. Ann Intern Med 1993; 118: 12-17. 27. Morinelli EN, Dugel PD, Lee M, et al. Opportunistic intraocular infections in AIDS. Trans Am Ophthalmol Soc 1992;90:97-109. 28. Ryning FW, Mills J. Pneumocystis carinii, Toxoplasma gondii, Cytomegalovirus and the compromised host. West J Med 1979;130:18-34.
507