If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
A 77-year-old man presented with a 5-day history of right eye pain, redness, and decreased vision. He was known for cataracts and Fuch's corneal dystrophy. He was also followed for a past Waldenström macroglobulinemia (WM) and rituximab-associated hypogammaglobulinemia.
His visual acuity (VA) was hand motion in the right eye and 20/50 in the left eye. Anterior segment examination of the right eye revealed endothelial guttata, diffuse corneal granulomatous keratic precipitates, a severe anterior chamber reaction with 4+ cells, and intense flare and fibrin. Snowballs were seen in the vitreous. The funduscopic examination was limited by the anterior segment and vitreous haze, but a peripheral superonasal infiltrate was suspected. Ultrasonography revealed no retinal detachment or choroidal lesions. Examination of the fellow eye was unremarkable except for the known Fuch's dystrophy. A presumptive diagnosis of right eye panuveitis with possible acute retinal necrosis (ARN) secondary to Herpesviridae or toxoplasmosis was made. The patient was started on oral valacyclovir, trimethoprim/sulfamethoxazole, and clindamycin in addition to topical corticosteroids and cycloplegia.
Laboratory testing showed lymphopenia (total, 0.40 × 109/L; CD4, 84 cells/mm3). The initial work-up for infectious etiologies was nonconclusive (Table 1). A diagnostic pars plana vitrectomy (PPV) was performed with gas endotamponade, phacoemulsification, and intraocular lens implantation, and an intravitreal injection of foscarnet. Diluted (200 mL) and nondiluted (0.2 mL) vitreous specimens were obtained and sent for viral PCR, aerobic, anaerobic, mycologic, and mycobacterial cultures. A week later, an inferior tractional retinal detachment was detected on ultrasonography (Fig. 1), and a second therapeutic PPV with silicone oil endotamponade was performed. The intraoperative findings supported the clinical suspicion of ARN (Fig. 2). Intravitreal injections of ceftazidime, amphotericin B, and vancomycin were carried out. Gram stain of the first vitreous specimens revealed no bacteria. PCRs for toxoplasmosis and Herpesviridae were negative. Twenty days after the initial PPV, mycobacterial cultures of the remaining diluted vitreous (13 mL) came back positive (Fig. 3A). To empirically treat both slowly and rapidly growing mycobacteria, intravitreal injections of amikacin (0.4 mg/0.1 mL) and moxifloxacin (0.1 mg/0.1 mL) were performed, pending the results of further identification and antibiotic susceptibility. PCR testing for Mycobacterium tuberculosis complex (MTBC) is performed twice a week in the hospital microbiology laboratory on all the positive mycobacterial culture broth. It confirmed the presence of MTBC 5 days later. The isolate was identified as Mycobacterium bovis using genomic deletion analysis at the public health laboratory, 15 days after the PCR results.
Table 1Summary of serologic and complementary tests
Not compatible with an active infection or sarcoidosis. CMV, cytomegalovirus; EIA, enzyme immunoassay; HIV, human immunodeficiency virus; HSV, herpes simplex virus; IFA, indirect immunofluorescence assays; VZV, varicella zoster virus.
Not compatible with an active infection or sarcoidosis.CMV, cytomegalovirus; EIA, enzyme immunoassay; HIV, human immunodeficiency virus; HSV, herpes simplex virus; IFA, indirect immunofluorescence assays; VZV, varicella zoster virus.
Fig. 2(A) Intraoperative right eye fundus color photograph obtained during the second pars plana vitrectomy showing a superonasal creamy infiltrate (red arrow). (B) The infiltrate extends inferonasally (blue arrows) and is surrounded by subretinal hemorrhage (white arrows), consistent with acute retinal necrosis (ARN).
Fig. 3(A) Ziehl-Neelsen smear showing aggregation of acid-fast bacilli (AFB) end-to-end and side-to-side to form serpentine cord-like structures, compatible with M. tuberculosis complex (MTBC). Mycobacterium bovis, also called zoonotic tuberculosis, is part of that complex. (B) Computed tomography (CT) scan of the chest showing a diffuse bilateral micronodular lung pattern consistent with miliary pulmonary tuberculosis. (C) Magnetic resonance imaging (MRI) of the brain obtained with gadolinium. In the T1/fat suppression sequence, bilateral tumefaction and enhancement of the prechiasmatic and intracanalicular portions of the optic nerves are seen, suggestive of bilateral optic neuropathy (ON). (D) Fifteen intraparenchymal circular lesions associated with vasogenic edema were noted. A 7 mm × 6 mm enhancing lesion in the right frontal lobe is shown on this image.
Cerebral imaging revealed scattered intraparenchymal lesions and bilateral prechiasmatic optic nerve enhancement (Fig. 3C, D). Although the chest x-ray was unremarkable apart from the presence of a calcified middle lobe granuloma, the chest computed tomography showed a micronodular pattern consistent with miliary pulmonary tuberculosis (Fig. 3B). Induced sputum cultures were obtained 4 days after the vitreous culture came back positive. Acid-fast bacilli were detected by direct auramine-rhodamine stain after 3 days, and PCR testing confirmed the presence of MTBC 4 days later. M. bovis was identified 2 months later. Mycobacterial culture and PCR of the cerebrospinal fluid were negative. Antitubercular therapy (ATT) was started, combining isoniazid (with pyridoxine), rifampin, and ethambutol, which was promptly replaced by moxifloxacin to decrease the risk of optic neuropathy (ON). High-dose oral corticotherapy (prednisone 60 mg daily) was added 10 days after the initiation of ATT in response to cerebral imaging to prevent paradoxical reaction of the lesions, and prophylactic trimethoprim/sulfamethoxazole and valacyclovir were continued. After 3 months of therapy, improvement was noted on brain and chest imaging, although the optic nerves enhancement remained unchanged. At the latest follow-up, phthisis bulbi was noted in the right eye with no light perception. Left eye examination was normal but VA deteriorated to hand motion.
The patient's presentation with panuveitis and a peripheral retinal infiltrate was concerning for ARN. Given the patient's history of WM and immunosuppression, lymphoma and other opportunistic etiologies were also considered. Intraocular tuberculosis has no pathognomonic eye findings and can occur without apparent signs of systemic involvement.
In our case, the tuberculin skin test was not suggestive of a prior exposure to tuberculosis, but this could be explained by anergy, secondary to immunosuppression or disseminated tuberculosis. This shows the limits of this test to diagnose an active tuberculosis infection. Additionally, the analysis of vitreous samples as supportive evidence may be of low yield with the sensitivity of PCR ranging from 33.3% to 46.9%.
Fortunately, our patient's vitreous samples revealed the tuberculous etiology although diagnosis was delayed due to slow growth of mycobacterial cultures. Only subsequently were pulmonary and cerebral tuberculosis uncovered. The patient's bilateral ON was presumed to be tuberculosis related. Tuberculosis-associated ON is thought to be either secondary to direct infection of the optic nerves or from an immune-related reaction.
In light of the follow-up brain imaging, a toxic etiology was then deemed more likely. Toxic neuropathies secondary to ethambutol and more rarely isoniazid are typically dose dependent and associated with prolonged use.
Giving the high mortality associated with a disseminated infection and the paucity of active antituberculous agents penetrating the central nervous system, isoniazid was not initially interrupted.
Tuberculosis is most commonly caused by M. tuberculosis and rarely by M. bovis (1.6% of cases in North America).
Endogenous endophthalmitis secondary to tuberculosis is rare, and to our knowledge, this is the first report of confirmed endogenous endophthalmitis secondary to the M. bovis isolate.
Interestingly, endogenous endophthalmitis secondary to M. bovis BCG, a live attenuated strain of M. bovis, has been previously reported in patients undergoing intravesical immunotherapy with BCG.
This strain is genomically distinct from M. bovis and was created at the beginning of the 20th century to be nonpathogenic. M. bovis isolation in the vitreous was reported a single time, co-existent with Mycobacterium fortuitum, in a patient with a large tubercular choroidal granuloma without endophthalmitis. Despite ATT, the infection lead to profound visual loss.
but relative success with intravitreal amikacin, combined with ATT, in treating tuberculous chorioretinitis has been reported, with regression of the chorioretinal lesion but without VA improvement.
The only case where VA improved (from light perception to 20/200) was reported by Hase et al; the patient was treated with prompt PPV and started on ATT and oral corticosteroids before culture results.
The decision to begin empiric antituberculous treatment in the absence of a laboratory-based diagnosis remains difficult given the known toxicity of those medications. Because ocular tuberculosis is associated with protean manifestations, mycobacterial cultures should be performed on every diluted vitreous specimen obtained in the context of an infection. PCR for MTBC should be considered in unusual cases and if the vitreal sample volume permits. Prompt consultation with an infectious disease specialist should be made as soon as a diagnosis of tuberculous endophthalmitis is raised in order to investigate for systemic findings and manage the treatment. Early detection and treatment of intraocular tuberculosis may be associated with better ocular and systemic outcomes.