Journal of Neuro-Ophthalmology, June 1998, Volume 18, Issue 2

Annual Review of Systemic Disease

The intention of this series of semiannual reviews of systemic diseases is not to develop an annual encyclopedic compendium of every article published in the prior calendar year on each systemic illness but rather to select for each issue a few systemic diseases of import to the neuro-ophthalmologist, and to summarize the literature that has developed over the past few years. For each systemic disease selected, the review will cover new developments in its epidemiology; new reports of its neuro-ophthalmic, ocular, and neurologic manifestations; unique clinical presentations; radiologic and laboratory diagnostic tools; and new therapies available to the neuro-ophthalmologist. If appropriate, new information on the etiology of the illness will be reported. In this installment, progress in the clinical recognition and treatment of tuberculosis, sarcoidosis, and human immunodeficiency virus- (nonretinal) related illness is discussed.

Journal of Neuiv- Oplilluilinology IS( 2): 127- 142, 1998. © 1998 l. ippincotl- Raven Publishers, Philadelphia Annual Review of Systemic Disease- 1997- 11 Paul Lama, M. D., and Larry Frohman, M. D. The intention of this series of semiannual reviews of systemic diseases is not to develop an annual encyclopedic compendium of every article published in the prior calendar year on each systemic illness but rather to select for each issue a few sys­temic diseases of import to the nemo- ophthalmologist, and to summarize the literature that has developed over the past few years. For each systemic disease selected, the review will cover new developments in its epidemiology; new reports of its neuro- ophthalmic, ocular, and neurologic manifestations; unique clinical presentations; radiologic and laboratory diag­nostic tools; and new therapies available to the neuro-ophthalmologist. If appropriate, new information on the etiol­ogy of the illness will be reported. In this installment, progress in the clinical recognition and treatment of tuberculosis, sar­coidosis, and human immunodeficiency virus- ( nonretinal) re­lated illness is discussed. Kay Words: Tuberculosis- Sarcoidosis- HIV- Ocular dis­ease- Neuro- opthalmic disorders. TUBERCULOSIS Tuberculosis ( TB) remains a worldwide health prob­lem especially in association with the acquired immune deficiency syndrome ( AIDS). Of diseases resulting from a single pathogen, TB is the leading cause of infectious disease death in the world, and it is estimated that a third of the world's population is infected. Although TB rates have been decreasing in the United States over the past 3 years after a sustained rise since 1985, multidrug-resistant tuberculosis ( MDR- TB) remains an important medical and epidemiologic issue ( 1). In contrast, TB rates in developing countries are increasing, and in con­junction with AIDS there has been a rapid acceleration of MDR- TB. Statistics provided by the World Health Or­ganization and the International Union Against Tubercu­losis and Lung Disease clearly demonstrate the magni­tude of this problem. A review of 63 surveys of resis­tance to antituberculous drugs performed between 1985 and 1994 showed the staggering rates were as follows: Nepal ( 48.8%); Gujarat, India ( 33.8%); New York City Manuscript received November 17, 1997; accepted January 12, 1998. From the Departments of Ophthalmology ( P. L.) and Neurosciences ( P. L., L. F.), University of Medicine and Dentistry of New Jersey- New Jersey Medical School, Newark, New Jersey, U. S. A. Address correspondence and reprint requests to Larry Frohman, M. D., 90 Bergen Street, Sixth Floor, Newark, NJ 07103, U. S. A. Supported in part by an unrestricted grant from Research to Prevent Blindness, Inc. ( 30.1%); Bolivia ( 15.3%); and Korea ( 14.5%) ( 2). The rates in New York City, however, have declined dramati­cally from 442 cases in 1992 to 109 cases in 1995, which is currently 6.0% of all TB cases ( 3). The following section discusses various unusual TB manifestations af­fecting the eye and visual pathway as well as the central nervous system ( CNS) that have been reported over the past 2 years. New diagnostic and therapeutic modalities are also discussed. Clinical Manifestations That over the past 2 years a number of cases of ocular or CNS TB have appeared in the medical literature that have posed diagnostic dilemmas underscores the fact that TB should remain on the differential of inflammatory and mass lesions affecting the eye, visual pathways, and CNS. Chin et al. presented a case of a 6- year- old boy from a non- TB endemic region with orbital TB that ini­tially manifested as a malignancy ( 4). Treatment with multiple antituberculous agents resulted in prompt reso­lution of the infection. The preceding case illustrates two important points. First, extrapulmonary manifestations of TB are not unusual in individuals residing in nonendemic regions if there is heavy migration from regions with a high prevalence of infection. Second, TB can occur with­out pulmonary signs and symptoms in young immuno­competent individuals. Apropos of the preceding case was a report of a 4 '/ 2- year- old girl from Yugoslavia who presented with bilateral upper eyelid tumescence with extension to the forehead unilaterally associated with in­flammatory changes on the more involved side. Initial evaluation showed a fluctuant mobile mass and regional lymphadenopathy. The child underwent incision and drainage of the abscess on the left along with a prolonged course of antibiotics complicated by development of a draining fistula. It was not until an cxcisional biopsy of the tumor mass on the right side was performed revealing histopathologic features of TB ( caseating granulomas) that tuberculous infection was considered. Ziehl- Neelsen stains and culture in Lowenstein- Jensen medium pro­duced the typical yellow, cauliflowerlike colonies of My­cobacterium tuberculosis confirming the diagnosis of tu­berculous infection ( 5). Once again we have an exceed­ingly rare presentation of extrapulmonary TB without pulmonary symptoms occurring in an otherwise healthy young child. In contrast to the previous case, however, 127 128 P. LAMA AND L. FROHMAN this child had a calcified granuloma on a chest radio­graph as well as a highly reactive purified protein de­rivative ( PPD). Kadioglu et al. ( 6) reported a highly un­usual form of TB affecting the optic nerve and chiasm. To date there have been about 30 reported cases of op-tochiasmatic TB. The patient described was a 58- year-old, otherwise healthy woman who complained of sud­den initial followed by progressive bilateral visual loss. The patient was found to have a mass lesion encasing the optic nerves and chiasm. She underwent surgical explo­ration, and the diagnosis of TB was confirmed by poly­merase chain reaction ( PCR). Complete resolution of in­fection occurred after a prolonged course of corticoste­roids and anti- TB antibiotics ( 6). Recently, Berinstein et al. ( 7) reported a case of presumed primary choroidal tuberculoma without any evidence of systemic disease in a young, healthy, immunocompetent male. The only positive manifestation was a reactive PPD ( 25 mm at 48 hours). The patient was treated with isoniazid ( INH), rifampin ( RIF), ethambutol ( EMB), and pyridoxine for 16 weeks with complete resolution and restoration of vision to 20/ 20 ( 7). Although ocular TB infection is usually associated with visual loss, Reny et al. ( 8) described an unusual case of systemic TB infection in an immunocompetent 31- year- old man with persistent fevers and lymphadenopa-thy. During the course of the examination, he was found to have a retinal vasculitis characterized by periphlebitis, preretinal hemorrhage, and serous exudation. Behcet's syndrome, systemic vasculitis, and syphilis were clini­cally and serologically ruled out. Lymph node biopsy confirmed mycobacterial infection with M. tuberculosis. The patient was prescribed a four- drug regimen for 2 months followed by two drugs for 4 months. Despite development of moderately severe peripheral retinal ischemic changes requiring prophylactic photocoagula­tion, the fundus findings completely resolved after 7 months without loss of vision. Thus, albeit rare, TB reti­nal vasculitis may occur and be asymptomatic ( 8). In addition to unusual ocular manifestations of TB, there have been several reports of CNS TB. Berger et al. ( 9) described a 4- year- old girl who presented with com­plex febrile convulsions and was found to have brain lesions that could only be seen on computed tomography ( CT) scan after contrast enhancement. Results of lumbar puncture studies were entirely normal. She finally under­went gastric aspiration and liver biopsy that demon­strated mycobacterial infection. She was treated with INH and RIF for 12 months, pyrazinamide ( PZA) for 9 months, and EMB for 6 weeks, which resulted in reso­lution of the brain lesions. However, a macular scar de­veloped that resulted in a reduction of visual acuity ( 9). Tuberculosis coinfection in persons with human im­munodeficiency virus ( HIV) has resulted in a number of interesting clinical manifestations. There are several fun­damental differences between HIV- positive patients with TB and immunocompetent patients. Since HIV- positive persons have deficient cell- mediated immunity, the fol­lowing are more likely to occur: ( 1) development of re­activation disease from remote infection, ( 2) contraction of new infection, ( 3) development of disseminated TB, ( 4) unusual and varied pulmonary and extrapulmonary manifestations, ( 5) development of only extrapulmonary manifestations, and most important ( 6) contraction or development of MDR- TB. From the diagnostic and therapeutic standpoint, there was one particular case of ocular TB in an HIV- infected patient reported in the past 2 years worthy of mention. Muccioli and Belfort ( 10) described a 35- year- old intra­venous drug abuser in Sao Paulo, Brazil, with AIDS and CNS TB in whom a monocular yellow- white chorioret­inal infiltrate developed with indistinct borders and an associated mild vitritis. The patient was given INH, RIF, and EMB with complete resolution of the ocular lesion and improvement of visual acuity as well as the neuro­logic symptoms ( 10). This case is a prime example of a dilemma facing the ophthalmologist who discovers a choroidal lesion without any other supporting physical findings. A definitive diagnosis requires aspiration of vitreous with or without a choroidal biopsy, which may not be feasible and carries significant morbidity. As a result, empiric treatment is usually initiated, and subse­quent treatment is guided by the response to therapy. This patient was fortunate enough to be infected by a sensitive TB strain. As mentioned, the risk of MDR- TB is high in an HIV- infected patient, especially an intrave­nous drug abuser from an endemic region. Therefore, it would be extremely helpful if we could identify patients with ocular TB infection more effectively. Generally, such a diagnosis is difficult to establish even when clini­cal suspicion is high, not only because ocular TB is tra­ditionally considered unusual but also because our cur­rent diagnostic techniques have suboptimal sensitivity and specificity. If ocular criteria can be established in patients with culture- proven TB infection, these criteria may be used subsequently to identify those patients with ocular findings in whom culture specimens are not readily available. Bouza et al. ( 11) randomly selected 100 of 300 patients with culture- proven TB infection for systematic ophthalmologic evaluation using the follow­ing criteria for defining ocular TB infection: certainty ( isolation of M. tuberculosis from ocular specimens), probability ( patients with isolation of M. tuberculosis from extraocular samples, with ocular lesions not attrib­utable to other causes that responded to anti- TB treat­ment), and possibility ( same as probability but follow- up was impossible). They identified 18 patients with ocular TB, 10 of whom fulfilled the probability category and 8 of whom fulfilled the possibility category. None fulfilled the certainty qualification. Of note, 11 of 18 were HIV positive and 11 of 18 were asymptomatic at diagnosis. The most common ocular finding was choroiditis ( 17/ 18). Other findings included papillitis, retinitis, vitritis, dacryoadenitis, vasculitis, and scleritis. In addition, these researchers identified several independent risk factors for ocular involvement, including miliary disease and symp­toms of reduction in visual acuity. Miliary TB was a risk factor for development of ocular TB in both HIV- and non- HIV- infected patients. ./ Neum- Ophlhulmol, Vol. 18, No. 2, 1998 ANNUAL REVIEW OF SYSTEMIC DISEASE 129 New Diagnostic Techniques for Rapid Diagnosis Although the gold standard for TB infection remains a positive culture, more rapid diagnostic techniques are critical especially in the presence of an epidemic or in cases of suspected high pharmacologic resistance. Traditionally, Ziehl- Neelsen or acid- fast bacilli stain­ing of patient specimens and examination under oil im­mersion is the usual method of detecting a positive smear. However, especially in extrapulmonary isolates, the number of bacilli is often low and smears using this method are often negative. The development of Accu-probe ( Gen- Probe Inc., San Diego, CA), which is a nucleic acid probe, requires at least a 2- week incubation period. With the emergence of PCR techniques, the abil­ity to detect a positive sample despite low numbers of infecting bacilli has improved. However, there are limi­tations that reduce the sensitivity and specificity of PCR identification methods such as contamination and the presence of inhibitors in sputum and other isolates. Cur­rently, there are two newly developed, direct-identification kits available using PCR technology, the Gen- Probe Amplified M. tuberculosis Direct Test ( AMTD) ( Gen- Probe Inc., San Diego, CA) and the Roche Amplicor M. tuberculosis test ( Amplicor, Hoff-man- LaRoche Ltd., Basel, Switzerland). The first test is based on the isothermal amplification of rRNA by DNA intermediates, whereas the Amplicor method employs PCR to amplify a specific DNA sequence of the 16s rRNA gene. Results can be obtained in fewer than 24 hours. Recently, a comparative evaluation of these tech­niques was performed by Piersimoni et al. ( 12). In their study, 327 digested, decontaminated respiratory speci­mens collected from 236 patients clinically suspected to have TB pulmonary infection were studied. The " gold standard" for comparison was a positive culture or a clinical diagnosis. Statistical analysis of the results trans­lated to sensitivity, specificity, positive, and negative predictive values of 77, 100, 100, and 95 for staining; 87, 100, 100, and 97.4 for culture; 95.9, 98.9, 94, and 99.2 for AMTD; and 85.4, 99.6, 97.9, and 97.1 for Amplicor, respectively. Therefore, in this study AMTD was more sensitive than Amplicor, although both tests were found to be rapid and specific for detection of M. tuberculosis. Another important finding in this study was that both PCR techniques identified patients who were still posi­tive for M. tuberculosis during treatment, whereas stains and cultures were negative. Thus, these techniques may be more sensitive in identifying those patients who are at potential risk of relapse. Currently, from this report and additional scientific studies, the AMTD assay stands as the most sensitive direct identification test available for detection of M. tuberculosis. If the clinical manifestations of TB are such that there is relatively easy access to potentially infected fluid or tissue, then the above methods for detection of M. tuber­culosis infection are useful. However, in patients in whom the sole manifestation of extrapulmonary disease is ocular or CNS infection, obtaining specimens for labo­ratory study is difficult and often impractical. In many cases a diagnosis cannot be made unless the eye is lost or the patient dies and a postmortem examination is per­formed. Although skin testing for TB can identify pre­vious exposure, the test lacks specificity for M. tubercu­losis. Lack of specificity is secondary to previous sensi­tization with environmental mycobacteria or bacillus Calmette- Guerin vaccine. In addition, a substantial pro­portion of patients at risk for TB infection are HIV posi­tive and have cutaneous anergy. Obviously, a more sen­sitive and specific test than the PPD test would be useful in those situations in which obtaining samples for iden­tification is impossible or impractical. Wilkinson et al. ( 13) developed a 38- kd recombinant antigen that con­tains two M. tuberculosis- specific B- cell epitopes. The recombinant product was evaluated in 292 patients with presumed pulmonary TB infection and in 58 controls. Sensitivity and specificity were found to be 72.6% and 94.9%, respectively. The immunogen was also tested in guinea pigs previously sensitized with M. tuberculosis, bacillus Calmette- Guerin, and six other species of my­cobacteria. The recombinant was shown to be signifi­cantly more specific for TB than PPD ( 13). As an alternative to skin testing, Kawamura et al. ( 14) developed an enzyme immunoassay to detect specific anti- TB glycolipid antibodies in serum. When the test was used on healthy subjects and those with pulmonary TB, sensitivity and specificity were 90% and 98%, re­spectively. The incubation time for this assay was 2.5 hours ( 14). Thus, novel immunologic assays may be use­ful in situations in which there is a high suspicion of TB, but acquisition of specimen for smear and culture is not possible or sample sizes are inadequate. Therapy Treatment of TB has become significantly more com­plicated with the increase in the percentage of MDR- TB cases, which may occur in an individual through infec­tion from a patient with existing MDR- TB, defined as primary resistance, or may be acquired during a course of treatment for an initially drug- sensitive TB infection, termed secondary resistance. The most powerful predic­tor of MDR- TB is a history of treatment for TB. Inad­equate therapy is the most common means by which resistant organisms are acquired. In addition, patients with cavitary lesions have a higher frequency of resis­tance in that they are assumed to harbor greater numbers of organisms ( 15). Based on current epidemiologic shifts and changes in drug susceptibility patterns of the infect­ing mycobacteria, the Centers for Diseases Control, in conjunction with the American Thoracic Society, out­lined specific recommendations for the treatment of TB infection ( 16). In patients from communities with low prevalence of drug resistance, defined as INH resistance less than 4%, a three- drug regimen consisting of INH, RIF, and PZA should be administered for the first 2 months, followed by INH and RIF for 4 additional months. Drug susceptibility testing should always be per­formed regardless of initial assumptions regarding drug-resistance patterns, and therapy should be adjusted ap­propriately. In patients from communities with a high prevalence of drug- resistant TB, a three- drug regimen J Neum- Ophthalmol, Vol. 18. No. 2, 1998 130 P. LAMA AND L. FROHMAN may be inadequate. In addition to INH, RIF, and PZA, EMB or streptomycin should be added at the start of therapy until drug susceptibility data are available. Be­cause the prognosis of death increases dramatically in immunosuppresscd patients infected with resistant M. tu­berculosis, treatment should initially consist of at least five or six drugs in patients who reside in communities with a high prevalence of INH- resistant TB or MDR- TB. Similarly, in those who had previous antituberculous therapy, the likelihood of secondary resistance is high and treatment of relapsed TB should also be initiated with five or six or even seven drugs. The regimen should consist of at least three drugs to which the infecting mycobacteria is susceptible. The duration of therapy should be at least 9 months or 6 months after sputum smears and cultures convert to negative. If cultures are still positive after 3 months, drug susceptibility testing should be repeated. These recommendations are specifi­cally for isolates that are resistant to multiple agents including INH but excluding RIF. In isolates that are resistant to both INH and RIF, the duration of therapy should be extended for at least 18 to 24 months; if iso­lates are resistant to INH, RIF, PZA, EMB, or a combi­nation of the four, the duration of treatment should be 24 months after conversion to negative cultures with adju­vant surgical resection if possible ( 15). The advent of drug resistance and the expansion of drug regimens to five to seven drugs means that new agents with anti- TB activity will be useful. Quinolones have evolved as important second- line and third- line agents, especially in patients with MDR- TB. Ciprofloxa­cin has known activity against TB, including highly re­sistant species; furthermore, the high bactericidal con­centrations achieved allow for once- daily dosing ( 17). Ofloxacin and levofloxacin are two relatively new qui­nolones with potent anti- TB activity ( 18). Sparfloxacin is still under investigation but is believed to be the most potent and promising of all quinolones ( 19). Although there have been no well- designed, large, randomized tri­als currently evaluating quinolones for treatment or pro­phylaxis of TB, there arc two smaller trials reported in the literature, one comparing ofloxacin with EMB and the second comparing ciprofloxacin with RIF. The re­sults demonstrated that ofloxacin was as effective as EMB when combined with INH and RIF and was well tolerated with similar frequency of adverse reactions ( 20). Ciprofloxacin, however, was less effective than RIF ( 21). Thus, it appears that quinolones are well tolerated and appear to be useful antimycobacterial agents, par­ticularly in the setting of MDR- TB. Further prospective studies will define the role of quinolones and establish whether a new congener may demonstrate efficacy com­parable with current first- line agents. Interferon- gamma ( IFN- 7) is a cytokine produced mainly by CD4 lymphocytes that has been shown to activate macrophages important in providing host immu­nity against M. tuberculosis ( 22). Several studies have reported the beneficial immuno- enhancing effects of IFN- 7 in patients with disseminated, atypical mycobac­terial infection ( 23- 25). In addition, IFN- 7 has been used adjunctively in patients with MDR- TB that did not re­spond to medical therapy and has demonstrated clinical efficacy in pulmonary and extrapulmonary TB infection. There was a report of a patient with acute lymphocytic leukemia and CNS involvement with MDR- TB unre­sponsive to a six- drug regimen and steroids who had substantial radiologic and neurologic improvement after 5 months of IFN- 7 and granulocyte colon- stimulating factors. After 12 months of therapy, there was complete resolution of the brain and spinal cord lesions ( 26). More recently, Condos, Rom, and Schluger ( 27) used aero­solized IFN- 7 for 1 month in five patients with MDR- TB and persistent positive sputum cultures despite docu­mented adherence to therapy. It was well tolerated and all sputum smears became negative. The size of the cavi­tary lesions was also reduced in all patients 2 months after treatment ended ( 27). Although large randomized clinical trials using IFN- 7 in TB management are lacking, it nevertheless appears to be a promising addition to the agents that are effective against M. tuberculosis and may be an especially useful adjunct in patients with depressed cell- mediated immu­nity. Since the resurgence of TB, one of the major causes of drug resistance has been nonadherence to the medical regimen. Directly observed therapy was implemented as a measure for controlling the spread of drug- resistant TB. While patients are on DOT, their adherence to therapy is ensured, leading to a reduction in the spread of infection. Since DOT and other measures to control TB were implemented, the prevalence of TB, including drug-resistant TB, has declined dramatically. The Centers for Disease Control currently recommends that all patients considered high risk for medical noncompliance should be under DOT. Drug Toxicity There have been no new or unusual reports of visual system toxicity related to antituberculous therapy in the English language literature in the past 2 years. Summary In summary, we are currently experiencing a decline in the total number of TB cases in the United States. In 1995, the TB rate was 8.7 per 100,000, which was the lowest number of reported cases since the start of na­tional surveillance in 1953. The substantial decline in TB cases in the United States is attributable to many factors, including the following: more rapid diagnosis as a result of improved laboratory identification of M. tuberculosis, improved infection control measures reducing the spread of infection, increased preventive therapy in high- risk groups, increased resources from the state and federal agencies to implement and maintain TB control mea­sures, more widespread use of susceptibility testing so that therapy can be tailored appropriately, and wide­spread use of DOT in patients likely to be noncompliant with their treatment regimen ( 28). Despite the positive current trend, the number of TB cases affecting foreign-born individuals has continued to increase. In fact, since 1986, the proportion of cases affecting immigrants has ./ Nairo- Oplillwlnwl, Vol. 18, No. 2, I99H ANNUAL REVIEW OF SYSTEMIC DISEASE 131 increased to 63.3%. High rates of infection in the country of origin, low income and overcrowding in urban areas, and language and cultural barriers lead to poor access to health care and increased rates of disease. Moreover, these factors also lead to poor compliance even when medical care is available and hence to an increased risk of drug- resistant strains being acquired or developing. Therefore, it is imperative that we continue to seek new ways to diagnose TB quickly, persuade the local and federal authorities to maintain the funding necessary to implement infection control strategies, and develop new pharmacologic agents to combat drug- resistant TB. SARCOIDOSIS The standard figure quoted in the literature is that 22% of patients with sarcoidosis worldwide have ocular in­volvement ( 29). In a Russian series of 502 patients with sarcoidosis, Khomenko and Ozerova ( 31) found that some extrapulmonary manifestation of sarcoidosis devel­oped in 19% of cases. In another Russian series, Borod-ulina et al. ( 31) found that the incidence of ocular in­volvement was 36% of all those with pulmonary involve­ment. Similarly, Makarov et al. ( 32) defined the incidence of CNS involvement in those Russians with respiratory tract sarcoidosis to be 14%. Neurosarcoidosis is usually reported to occur in 5% of all cases of sarcoidosis. Recio et al. ( 33) reported a series of 520 patients with biopsy- proven sarcoidosis from a single institution. Thirty- two ( 6.3%) had neurologic symptoms; of these, 4 1% had their neurologic symptoms associated with systemic signs, the most common of which were weakness ( n = 14), facial paresis ( n = 7), headache ( n = 8), seizures ( n = 8), and visual loss ( n = 7). Of the 42, 38 underwent magnetic resonance imaging ( MRI) and 28 ( 74%) showed changes including both enhancing and nonenhancing parenchymal lesions, cra­nial nerve involvement, cranial nerve enhancement, meningeal enhancement, and cord lesions ( 33). The cause of sarcoidosis remains controversial, al­though infectious causes are likely. There is significant controversy as to whether there is a role of mycobacteria TB in the etiopathogenesis of sarcoidosis. Evidence has been cited in the past few years to support and refute the claim that mycobacteria TB is the causal agent in sar­coidosis. Other studies claim that evidence for M. tuber­culosis'$ presence in sarcoid patients is irrefutable, al­though they stop short of claiming a role for the organ­ism in the pathogenesis of the disease itself. Some recent articles reported evidence that both supported and re­futed the involvement of mycobacteria. Richter et al. ( 34) studied the presence of mycobacte­rial coding for 16s rRNA, which is present in all myco­bacterial species. They found amplification of this rRNA by PCR in lung tissue in only 1 ( 4%) of 24 patients with sarcoidosis and concluded that the granulomatous lesions in sarcoidosis might not be from mycobacterial infection ( 34). The cases that support the role of mycobacterial in­volvement in sarcoidosis are typified by the case re­ported by Baselga et al. ( 35). This patient, in whom a test for PPD was negative, had a cutaneous lesion with a histopathology that resembled sarcoidosis, but M. tuber­culosis complex DNA was demonstrated by PCR ampli­fication. The patient responded to anti- TB therapy ( 35). Almenoff et al. ( 36), supporting a role for mycobac­terial species, pointed out that acid- fast cell wall-deficient forms of bacteria can be grown from blood, aqueous humor, and bronchial washings of patients with sarcoidosis. In their study, cell wall deficient forms were grown in samples from 19 ( 95%) of 20 patients with sarcoidosis but in none from controls. All 19 of these stained with a monoclonal antibody directed against my­cobacteria TB whole cell antigen. This evidence supports the presence of mycobacterial TB or TB- like organisms in patients with sarcoidosis, although their role in sar­coidosis is not clear ( 36). el- Zaatari et al. ( 37) demonstrated amplification and hybridization with M. avium, M. paratuberculosis, or both in skin biopsies from two patients with sarcoidosis and in the CSF of a patient with sarcoidosis. Further­more, serum from 100% of seven patients with sarcoid­osis showed strong reactivity on immunoblot with a 36,000 molecular weight antigen ( p36) derived from my­cobacteria paratuberculosis, as opposed to only weak re­actions in 13 ( 34%) of 38 control subjects ( p < 0.002). When tested against another mycobacteria paratubercu-losis- derived antigen ( PTB65K), the sera from 5 ( 23%) of 22 control subjects, compared with 5 ( 100%) of 5 patients with sarcoidosis ( 100%, p < 0.001), demon­strated significant reactivity. The authors believe that these data support a mycobacterial cause for sarcoidosis ( 37). In an attempt to further establish that the cause of sarcoidosis is mycobacterial, el- Zaatari et al. ( 38) em­ployed PCR assays to study CSF from a patient with neurosarcoidosis. The authors showed positive hybrid­ization with probes derived from M. avium and M. paratuberculosis but not from M. tuberculosis or M. avium wood pidgeon strain. They believe that these find­ings prove that M. paratuberculosis or closely related M. avium species are associated with some cases of sarcoid­osis ( 38). Mycobacterial species are not the only organisms that have been implicated in sarcoidosis. Chlamydia pneumo­niae is a recently characterized chlamydial species that has been established as an important cause of respiratory disease, causing up to 10% of community- based cases of pneumonia. Although 40% to 60% of adults may harbor antibodies to C. pneumoniae, the presence of persistently elevated serum antibody titers, the presence of serum IgA antibodies, and the detection of immune complexes specific for the organism are said to be signs of chronic infection. Antibodies against a 60- kd chlamydial heat shock protein ( HSP60) have been incriminated in a de­layed hypersensitivity reaction in chronic chlamydial in­fection. Puolakkainen et al. ( 39), using a microimmun-ofluorescent technique, assayed for the presence of an­tibodies to C. pneumoniae in the serum of 24 patients with sarcoidosis. Five patients ( 21 %) had acute antibody, J Neiiro- Ophllmlmol, Vol. 18, No. 2, I99H 132 P. LAMA AND L. FROHMAN 15 ( 62%) had chronic antibody, and 4 ( 17%) did not exhibit antibodies. When an immunoblot technique was used, the sera of sarcoidosis patients showed recognition of C. pneumoniae proteins. A 98- kd protein was recog­nized by 55% of the sarcoid group ( controls = 3%), a 68- kd protein by 25% ( controls = 7%), and a 60- kd protein by 65% ( controls = 7%). The authors used these data to support the possible involvement of C. pneumo­niae in the pathogenesis of sarcoidosis ( 39). Studies have attempted to look at the etiopathogenesis of sarcoidosis on a histochemical level. Tsukada et al. ( 40) studied circulating immune complexes ( by a Raji cell assay) in the sera of patients with sarcoidosis. Pa­tients showed high levels of IgG binding to endothelial cells compared with healthy subjects ( p < 0.01). Patients with neurosarcoidosis had increased levels of IgG bind­ing to brain endothelial cells compared with controls and those with pulmonary sarcoidosis. The authors postu­lated that these autoantibodies to endothelial cells may play a role in the breakdown of the blood- brain barrier ( 40). Milburn et al. ( 41) investigated cytokine activity in 21 patients with acute untreated sarcoidosis. Bronchoalveo-lar lavage was performed on these patients and 11 healthy subjects. Of the 16 patients with sarcoidosis who were subsequently treated, 7 underwent repeat broncho-alveolar lavage. The ratios of IgG4 and IgE in the lavage fluid: serum were significantly reduced in the patients with sarcoidosis compared with the healthy subjects. Al­though extremely high levels of IgG were produced in the lungs of patients with sarcoidosis, the fraction of this that was IgG4 was greatly and significantly reduced compared with that in healthy subjects ( 1.6% for patients with sarcoidosis and 38.5% for healthy subjects). Inter-lcukin- 4 in the lavage fluid was also significantly re­duced in the patients with sarcoidosis, whereas IFN- 7 was significantly increased. Therapy with corticosteroids led to a significant increase in the ratios of IgG4 and IgE in the lavage fluid compared with the serum, a significant decrease in local pulmonary IgG production, a signifi­cant increase in the portion of IgG that was IgG4, a significant increase in the levels of interleukins 2 and 4, and a significant decrease in the level of IFN- 7 in the lavage fluid. The authors interpreted this as signifying that pulmonary sarcoidosis is accompanied by an imbal­ance in the activity of Th- 1 and Th- 2 cytokines, with the activity of Th- 1 cytokine increased and the activity of Th- 2 cytokines decreased. This imbalance is reversed by corticosteroid therapy ( 41). Clinical Manifestations Several reports of neuro- ophthalmic manifestations of sarcoidosis have been reported in the recent literature. A recurrent theme has been that neuro- ophthalmic, particu­larly optic nerve, involvement in sarcoidosis can occur both as a presenting sign of sarcoidosis and, on subse­quent systemic evaluation, in the absence of other sys­temic foci. Kosmorsky and Prayson ( 42) described a 38- year- old man in whom the initial manifestation of sar­coidosis was painless progressive visual loss. He was initially diagnosed as having neuroretinitis; neuroimag-ing showed an enhancing optic nerve mass that extended to the chiasm. A spinal tap showed 15 white cells, and systemic workup included normal results on tests for serum angiotensin- converting enzyme ( ACE) and lysozyme, chest radiograph, and anergy panel. In this patient, progressive visual loss developed despite admin­istration of high doses of prednisone. This led to a meningeal biopsy, results of which were normal, and the involved eye progressed to complete blindness. When symptoms developed in the contralateral optic nerve, a biopsy of the blind optic nerve was done, showing non-caseating granulomas consistent with sarcoidosis ( 42). Ing et al. ( 43) reported on a 22- year- old woman with progressive visual loss to complete blindness over 2 years, without uveitis. The MRI exhibited an enhancing, thickened apical optic nerve with a tram track sign. The preoperative diagnosis was optic nerve sheath meningi­oma, but the pathology from the biopsy specimen showed sarcoidosis. Systemic scrutiny did not find evi­dence of sarcoidosis elsewhere. The authors reported finding 17 other cases in the literature in which an optic nerve tumor was mimicked by biopsy- proven sarcoid­osis. Most of these were believed to be a meningioma before surgery. None of the patients had systemic sar­coidosis on examination. The authors concluded that op­tic nerve sarcoid is nearly impossible to diagnose without biopsy ( 43). el Zarka et al. ( 44) reiterated that direct optic nerve infiltration of the optic nerve is a rare event, and reported a case of systemic sarcoidosis with a sole symptom re­lated to an optic nerve granuloma. Ng et al. ( 45) reported on a 14- year- old girl who pre­sented with 2 years of visual loss OD and headache. A computed axial tomography scan showed an enhancing suprasellar mass, and MRI showed enlargement of the optic chiasm and posterior optic nerves with enhance­ment of the periphery of the chiasm and nerves. This was believed to be an optic nerve glioma. Rapid deterioration led to a follow- up MRI that exhibited marked soft tissue enhancement in the region of the basilar cisterns. This finding necessitated a biopsy, which established the di­agnosis of sarcoidosis ( 45). Kao and Rootman ( 46) reported on a case of solitary optic nerve sarcoid granuloma that underwent angiitic spread by the posterior ciliary vessels. The patient was a 65- year- old woman with bilateral primary open angle glaucoma. Disc pallor OS and a left relative afferent pupillary defect led to further testing. Computed tomog­raphy scan showed a mass of the anterior nasal optic nerve, and a systemic workup included normal findings on chest radiograph and gallium scan and normal results on tests for ACE, calcium, lysozyme, liver function, and 24- hour urine calcium. Progressive visual loss led to suc­cessful institution of prednisone therapy. Four years later, progressive visual loss again developed, this time accompanied by patchy choroidal infiltrates. A biopsy of the optic nerve sheath was done, which demonstrated choroidal granulomas. Infiltration along the perivascular space of the posterior ciliary vessels was present. ./ Neuro- Ophtlmlmol, Vol. 18. No. 2, 1998 ANNUAL REVIEW OF SYSTEMIC DISEASE 133 Kao and Rootman ( 46) also reported on a 41 - year- old man who presented with progressive proptosis. Com­puted tomography scan of the orbit showed bilateral dif­fuse enlargement of the extraocular muscles, and he was initially thought to have Graves' Disease, although re­sults of thyroid testing were normal. Further review of the images revealed bilateral sheets of periorbital infil­trate, with erosion of the greater wings of the sphenoid. There was infiltration of the sellar region, with normal lacrimal glands. Systemic evaluation included normal findings on chest radiograph and normal results on ACE, calcium, lysozyme, liver function tests, and 24- hour urine calcium. A biopsy of the medial rectus showed sarcoid granulomas. Prednisone controlled his symp­toms, although a maintenance dose of 10 mg every other day was required ( 46). Neuro- ophthalmic sarcoidosis may be seen in the pe­diatric population. Monfort- Gourard et al. reported on a 13- year- old boy with uveitis and orbital inflammation. Results of serum ACE were normal. Although cortico­steroids alleviated the symptoms, a bilateral orbital re­lapse occurred, with the serum ACE becoming abnormal. Ten months of oral corticosteroids led to resolution of symptoms ( 47). Leiba et al. ( 48) reported on a 13- year- old child who presented with bilateral panuveitis and a superior oblique palsy. Bilateral disc edema and choroidal infiltrates were seen. Retinal vasculitis and snow- banking were not seen. Evaluation included tests for ACE, calcium, and PPD and a chest radiography, all of which were negative. When a right superior oblique palsy developed, MRI scan was performed that showed an enhancing pon­tomesencephalic mass. There was no meningeal en­hancement or lacrimal or parotid gland enlargement. The mass was completely resected. Analysis of CSF was nor­mal for cells and protein, but the pathology identified a necrotizing granuloma. Tests for acid- fast bacilli and gene amplification studies for M. tuberculosis were negative, as was a test for CSF ACE. The child ulti­mately responded to prednisone therapy with resolution of the choroidal nodules ( 48). Thus, it is clear that the neuro- ophthalmologists may be the first physician in contact with the patient with undiagnosed systemic sarcoidosis; despite appropriate evaluation, diagnosis may be elusive, necessitating pathologic confirmation. We agree with Lee, who be­lieves that normal results on ACE and chest radiograph are not adequate to rule- out sarcoidosis in a patient in whom ocular or neurologic involvement is suspected. He has suggested that a gallium citrate 67 scan also be per­formed ( 49). Our initial evaluation includes this, as well as a PPD and anergy panel. We also find tests for serum calcium and 24- hour urine calcium useful. We agree that in cases of an isolated sarcoid optic nerve mass, it may be quite difficult to diagnose sarcoidosis. We consider sar­coidosis if the optic nerve is involved from the globe to the chiasm; if the contralateral optic nerve is involved, especially at a site radiologically discontinuous from the original optic nerve's involvement; and if the en­hancement pattern on MRI scan has a " nodular" appear­ance. In these cases, we have found that gallium scan and lumbar puncture are useful, with the gallium scan often guiding us to perform a biopsy of a site other than the optic nerve, thereby preserving vision. Several authors have published articles on ophthalmic manifestations of sarcoidosis. Rizzato et al. ( 50) re­viewed 1,156 white patients with biopsy- proven sarcoid­osis. In 9 of the 1,156 patients, uveitis led to the evalu­ation that ultimately showed systemic sarcoidosis. In an­other eight cases, the eye findings preceded the diagnosis of sarcoidosis by 1 to 11 years, although unattended sys­temic symptoms were present. They claimed that uveitis was the presenting sign in 15 ( 1.5%) of the 1,196 patients with sarcoidosis. These data must be interpreted in that this article was from a medical unit and not an ophthal­mology service ( 50). Braun and Jones ( 51) stated that although the overall incidence of ocular disease in sarcoidosis is 25%, band keratopathy is rare. Their patient was a 31- year- old man with no ocular signs except bilateral band keratopathy. The patient did not have active systemic signs at the time of the ocular presentation, but he had a history of sym­metric swelling of the large joints. On workup, this man had an elevated serum calcium, ACE, and bilateral nodu­lar changes in his lungs. A renal biopsy for nephrocal-cinosis confirmed sarcoidosis ( 51). DeRosa et al. ( 52) reported on a patient in whom the initial manifestation was a unilateral hemorrhagic reti­nopathy. Sarcoidosis was suspected, but the chest radio­graph and tests for ACE were repeatedly normal. Diag­nosis was only made by pathology on enucleation once the eye was blind, with a large noncaseating granuloma of the ciliary body present ( 52). Dodds et al. ( 53) described a patient with posterior scleritis and an annular ciliochoroidal detachment caus­ing angle closure glaucoma. That the cause was sarcoid­osis was proved by a parotid gland biopsy ( 53). Taches de bougie may resemble birdshot or multifocal choroiditis. Vrabec et al. ( 54) studied the pattern of ta­ches de bougie in 22 patients with sarcoidosis. Two pat­terns were seen. First, small discrete white spots in in­ferior or nasal periphery were seen in 16 ( 73%) of 22 patients and were indistinguishable from the lesions of multifocal choroiditis. In six patients ( 23%), the lesions were identical to those of birdshot chorioretinopathy, with larger posterior pale yellow- orange streaks seen. Findings on the chest radiograph were normal in 8 ( 36%) of 22 cases. Levels of ACE were normal in 5 ( 36%) of 14 patients tested. The authors concluded that sarcoidosis should be considered in patients with birdshot or multi­focal choroiditis. They believe that conjunctival biopsy and gallium scan should be considered in cases of bird-shot and multifocal choroiditis with normal findings on chest radiograph and elevated ACE, in HLA A29- negative birdshot, and in patients with multifocal cho­roiditis older than 50 years ( 54). Although neurologic involvement has been found to be present in 5% of cases of sarcoidosis, CNS mass J Neuro- Ophllmlmol, Vol. IH, No. 2, I99H 134 P. LAMA AND L. FROHMAN lesions are rare. Caparros- Lefebvre et al. ( 55) presented three such cases. The diagnosis in patient case was made by conjunctival biopsy, in the second by the presence of mediastinal lymph nodes and infiltrative lung lesions, and in the last by lymph node biopsy. Radiologic signs included diffuse subcortical high signal suggestive of pachymeningitis with vasogenic edema and multifocal granuloma. All patients improved clinically and radio-logically with corticosteroid therapy ( 55). Grand et al. ( 56) reported on a patient with altered cognitive functioning as the presenting sign of an en­hancing right frontal- parietal mass that was proved to be neurosarcoidosis on stereotactic biopsy. Spinal cord involvement may occur as well. Weissman et al. ( 57) reported on a patient with a cauda equina syndrome with an expanding mass in the lumbar epidural space extending from LI to SI. Preoperatively, there was no known systemic illness, and the mass was excised and found to be sarcoid granuloma. The patient was treated with 4 months of oral prednisone, and at 7 months post­operative follow- up, there was no recurrence. Waubant et al. ( 58) described a 68- year- old woman in whom sar­coidosis mimicked an intramedullary tumor. Magnetic resonance imaging showed an area of low signal inten­sity on T2- weighted spin echo at the core of the lesion, consistent with calcification. With therapy, both the clinical and radiologic findings resolved except for the MRI finding of calcification. Fitzpatrick et al. ( 59) re­ported on the first case of neurogenic lower urinary tract dysfunction from neurosarcoidosis with midthoracic in­tramedullary involvement. Other neurologic illness recently described in patients with sarcoidosis include a patient with hemidystonia ( 60), two patients with myasthenia gravis ( 61), and a young male with coexistent Guillain- Barre and pulmo­nary sarcoidosis ( 62). Several unusual or unique systemic manifestations of sarcoidosis have been recently presented. Cardiac sar­coidosis is often subclinical, with typical findings being bradyarrhythmias or tachyarrhythmias with or without congestive heart failure. Sekiguchi et al. ( 63) recom­mended electrocardiograms and echocardiography to de­tect the disease. They found that the incidence of elec­trocardiographic abnormalities in 963 patients with sar­coidosis was 22.1%, higher than for age- matched controls ( 17.9%, p < 0.025). These authors also deter­mined that the sensitivity of diagnosis by endomyocar­dial biopsy is low ( 20- 30%). Japanese patients with sar­coidosis may be more susceptible to sarcoid- induced car­diac death. Sekiguchi et al. ( 63) recommended early steroid therapy with small maintenance doses. Nelson et al. ( 64) found three cases of sarcoidosis in which the presenting sign was a cardiac conduction defi­cit. In two of three cases, the conduction deficits pre­ceded other signs by years. They concluded that sarcoid­osis must be considered in young patients with unex­plained conduction disturbances, arrhythmias, or cardiomyopathy ( 64). Vatti and Sharma ( 65) studied the clinical course of hepatic sarcoidosis. At their institution, 44 ( 35%) of 125 patients with sarcoidosis had evidence of hepatic in­volvement as evidenced by liver enzyme abnormalities. Twenty- five of these 44 patients received therapy for sarcoidosis; 12 ( 48%) of 25 improved. Of the 13 un­treated patients, 10 ( 77%) also improved, suggesting that hepatic involvement in sarcoidosis may be a self- limited process not requiring corticosteroid therapy ( 65). Pancreatic sarcoidosis is less commonly seen. Boru-chowicz et al. ( 66) reported on a case of pancreatic sar­coid, with simultaneous liver, colonic, renal, and medi­astinal involvement, and stated that there had been only 14 prior cases with pancreatic involvement in the litera­ture. Gynecologic presentations of sarcoidosis have been described. Pearce and Nolan ( 67) described a 53- year- old woman who presented with 5 days of vaginal bleeding. Biopsy showed many noncaseating granulomas. Results of the systemic evaluation were negative, and the symp­toms spontaneously resolved. Sherman et al. ( 68) re­ported on a 47- year- old patient who was being evaluated for bilateral recurrent uveitis and also had menometror-rhagia. Findings on chest radiograph and a test for ACE were negative. Result of gallium scan was positive, and endometrial biopsy made the diagnosis of sarcoidosis ( 68). De Bandt et al. ( 69) reported on five cases in which sarcoidosis was coincident with systemic sclerosis. They reported that 14 prior cases of these two illnesses coex­isting have been reported. Diagnostic Techniques Carmody et al. ( 70) reported the imaging findings in 15 cases of sarcoidosis of the orbit and anterior visual pathways. They found three cases of orbital sarcoidosis, and reported that the findings had signal characteristics resembling those of orbital pseudotumor. Five of the pa­tients had periventricular lesions resembling demyelinat-ing disease. They recommended that sarcoid be included in the differential of optic nerve or nerve sheath enhance­ment on MRI scanning, although they stated that isolated optic nerve sarcoidosis without systemic involvement is nearly impossible to diagnose definitively radiologically and requires a biopsy ( 70). Westlake et al. ( 71) reported on a 20- year- old African- Caribbean woman who had a 4- week history of visual loss OD and a 3- month history of amenorrhea as her presenting signs of neurosarcoid­osis. She had vitritis, an infiltrated optic disc, and pe­riphlebitis in the affected eye. Magnetic resonance im­aging scan showed what the authors believed were the typical findings of a thickened chiasm with infiltration of the hypothalamus, meninges, and in this instance the third ventricle and pituitary gland ( 71). Several authors reviewed the results of employing laboratory studies to establish a clinical diagnosis of sar­coidosis or to predict in which patients with sarcoidosis, ocular or neurosarcoidosis would develop. Results con­flicted as to whether the ACE and gallium scan tests are useful. In a Spanish study of 209 patients diagnosed over 14 ./ Neum- Ophllwlmol, Vol. 18. No. 2. 1998 ANNUAL REVIEW OF SYSTEMIC DISEASE 135 years as having sarcoidosis, Mana et al. ( 72) determined that in their patients, pulmonary function tests and ACE were useful to predict persistence of sarcoid activity. The gallium scan was not helpful. Agbogu et al. ( 73), in reporting their series of 28 pa­tients with neurosarcoidosis, found that no diagnostic test was useful for predicting in which patients with systemic sarcoidosis refractory neurosarcoidosis would develop. Power et al. ( 74) reported on 22 patients who were shown to have clinical ocular sarcoidosis with a positive biopsy. Of these, all had abnormal results on ACE test or gallium scan. In 16 ( 73%) of the 22, both tests were abnormal. The authors also reviewed their laboratory findings in 70 patients with nonsarcoid uveitis. In none of these 70 patients were both the ACE test and gallium scan abnormal. The sensitivity of an elevated ACE was 73% and its specificity was 83%. If both tests were re­quired to produce positive results, the sensitivity would be 73% and the specificity 100%. Therefore, the combi­nation of an elevated ACE level and abnormal gallium scan increases the specificity without lowering the sen­sitivity ( 74). Wolfensberger et al. ( 75) reported on the use of in-docyanine green angiography to recognize retinal and choroidal sarcoidosis. He found five patterns of abnor­mality in patients with sarcoidosis. These patterns were ( 1) multiple hypofluorescent lesions that were isofluo-rescent or hyperfluorescent in late stages, ( 2) multiple hypofluorescent lesions that remained hypofluorescent in late stages, ( 3) focal hyperfluorescence, ( 4) linear hyper-fluorescence ( so- called fuzzy choroidal vessels), and ( 5) diffuse zonal hypofluorescence. They suggested that the patterns typically seen in sarcoidosis can also be pro­duced by birdshot choroidopathy and Vogt- Koyanagi- Harada, with the exception of the first pattern of uneven zonal distribution with focal pinpoint hyperfluorescence, which they believed might be specific for sarcoidosis ( 75). Finally, Bacchella et al. ( 76) investigated a potentially new assay for sarcoidosis. Type I and type III are the most abundant collagens in lung tissue. Bacchella et al. examined levels of procollagen I and III peptides by radioimmunoassay and of ACE by colorimetric assay. They found that, compared with controls, newly diag­nosed patients with sarcoidosis have elevated serum pro­collagen 1 peptide ( p = 0.0014). They also reported that this is not a good marker for patients with the fibrotic stage of the disease ( 76). Therapy The clinical treatment of systemic, ocular, and neuro­sarcoidosis has been expanded with new routes of ad­ministration of standard agents, with application of com­bination therapies, and with novel new therapies. Sader et al. ( 77) reported on a case of neurosarcoidosis with biopsy- proven optic nerve involvement refractory to in­creasing doses of corticosteroids. The patient was treated with intrathecal corticosteroids, which were tolerated without systemic side effects. Lower and Baughman ( 78) treated 50 patients with chronic biopsy- proven sarcoidosis with once- weekly methotrexate for a minimum of 2 years each. Improve­ment was noted in 33 ( 66%) of 50 of patients. Six addi­tional patients were able to be removed from corticoste­roids. The major toxicities seen were hepatic ( 12%), leu­kopenia ( 2%), and cough ( 2%). Of 33 patients who underwent liver biopsy, 6 had changes that were believed to be secondary to methotrexate, causing discontinuation of the agent. They concluded that methotrexate is a well-tolerated drug for use in the treatment of chronic sar­coidosis ( 78). Gedalia et al. ( 79) used low- dose ( 10- 15 mg/ m2/ wk) oral methotrexate in seven children with biopsy- proven sarcoidosis in an open- ended, noncontrolled pilot study. This regimen allowed the mean daily dose of corticoste­roid to be tapered from 1.3 mg/ kg at the onset of the study to 0.2 mg/ kg ( an 85% reduction) after 6 months of therapy. Besides its steroid- sparing effects, institution of the methotrexate was correlated with a marked reduction in clinical severity, reduction in ACE level and erythro­cyte sedimentation rate, and improvement of hemoglo­bin. None of the children suffered significant side effects ( 79). Brechtel et al. ( 80) reported on a single patient in whom allopurinol was useful in treating a disseminated cutaneous sarcoid eruption ( 80). A promising new therapy was employed by Carlesimo et al. ( 81), who used low- dose ( 200 mg/ d) thalidomide for 2 weeks followed by 100 mg daily for 11 weeks in two patients with sar­coidosis who either did not respond to or suffered con­sequences from corticosteroid therapy. Both patients re­sponded to therapy, and a 100- mg every other day main­tenance dosage was employed that has prevented relapse. Thalidomide was felt to exert its action by inhibition of macrophage function ( 81). Agbogu et al. ( 73), in a multiinstitution retrospective analysis, reported on 26 patients with refractory neuro­sarcoidosis who required agents either to supplement prednisone or to reduce prednisone to a tolerable dosage. Agents used included azathioprine, cyclophosphamide, chlorambucil, methotrexate, and radiation therapy. They found that they were able to taper prednisone to 10 to 20 mg/ d in 10 ( 38%) of 26 patients without seeing disease progression or relapse. The therapy was not benign; four patients ( 15%) did not respond and died either of their illness or of complications of therapy. No patient in this study was treated with three- drug combination therapy. It was noted that even with alternative medication added, symptoms usually flared if prednisone was tapered to a level less than 10 mg/ d. They recommended that alter­native therapies be started as soon as signs of corticoste­roid complications develop or if frequent large increases in corticosteroids are required to control the neurologic symptoms. They believe, and we agree, that cyclosporine is a reasonable first agent to add to corticosteroids ( 73). We treated a patient with chiasmal sarcoidosis with triple immunosuppressant therapy. This patient required dosages of at least 60 mg of prednisone as well as 200 J Neiira- OplulKiliiml. Vol. IS, No. 2. 1998 136 P. LAMA AND L. FROHMAN mg cyclosporine daily to control the disease. This patient already had visual acuity of No Light Perception in one eye and a superotemporal field defect in the only seeing eye. Attempts to taper the prednisone to less than 60 mg led to progression of field loss on multiple occasions. Azathioprine ( Imuran, Burroughs Wellcome, Research Triangle Park, NC) was added, and despite initial suc­cess, the patient again experienced a flare- up as pred­nisone was tapered. The addition of 4 mg daily chloram­bucil ( Leukeran, Burroughs Wellcome) to the cyclospor­ine and prednisone allowed tapering of the prednisone to a dose tolerated by the patient ( 10- 20 mg daily) without further visual loss ( L. P. Frohman, unpublished data). It is anticipated that as new diagnostic tests become available, these new specific therapies will be introduced at an earlier point in the clinical course, thereby preserv­ing visual function. UPDATE ON HIV- RELATED OCULAR DISEASE The following discussion on HIV- related ocular mani­festations will not include a discussion on cytomegalo­virus ( CMV) or other viral retinitis as this is beyond the scope and the intent of this review. There have been only a few reports in the literature regarding new HIV- related ocular manifestations during the past 2 years. The vast majority concern the new developments in therapy for CMV retinitis. There have been no new reports regarding HIV optic neuropathy since 1995. Clinical Manifestations The following case illustrates the need for aggressive evaluation of HIV- positive patients presenting with vi­sual loss and disc swelling, even when unilateral, as the likelihood of discovery of a life- threatening problem is significantly greater than in HIV- negative patients given the frequency of CNS mass lesions that occur in the HIV- infected population. Wei et al. ( 82) reported on a patient who presented with precipitous vision loss asso­ciated with a papillitis. The patient was otherwise asymp­tomatic. Further workup revealed severe life- threatening intracranial lesions consistent with toxoplasmosis. Central retinal vein occlusion has rarely been reported in HIV- positive or AIDS patients in the absence of ret­initis. To date there have only been five cases reported in the literature. The most recent case report included a histopathologic study of a patient in whom there was loss of acuity to no light perception after neovascular glau­coma had developed ( 83). The findings were similar to those in nonimmunocompromised persons with ischemic central retinal vein occlusion. There was no evidence of HIV in vascular endothelial cells using electron micros­copy or in situ hybridization techniques. There were no structural abnormalities noted, indicating the cause in such cases appears to be related to rheologic abnormali­ties rather than virus infection- related damage to vascu­lar endothelial cells. Although anticardiolipin antibodies and circulating anticoagulants are associated with a pro-thrombotic tendency, this was not identified in this pa­tient or in two of four patients previously reported. In a recent autopsy study conducted by Glasgow ( 84), the integrity of the microvasculature was examined in HIV-infected patients and in those with immunosuppression of different causes as it compared with that in controls. Vascular breaches permeable to 200- nm microspheres were noted in the eyes of patients with AIDS. Of 14 patients with AIDS, 7 had ruptured microaneurysms at the center of retinal hemorrhages. However, five of eight patients with non- HIV immunosuppression were also found to have ruptured aneurysms at the center of retinal hemorrhages, but leakage around microaneurysms that occurred even in the absence of hemorrhage was more frequently found in the eyes of patients with AIDS ( 11/ 14) than in the other eyes ( 84). Thus, although there were no obvious structural differences noted in the eyes of patients with AIDS with a vein occlusion compared with controls with a vein occlusion, there were other structural abnormalities noted in those with AIDS that contributed to the retinal findings associated with HIV retinopathy. These structural abnormalities may provide an explana­tion for the development of ocular opportunistic infec­tions in that they allow increased transmission of virions and other organisms through the blood- retina barrier. When HIV infection occurs in the setting of another illness associated with immunodeficiency such as diabe­tes, the infectious manifestations may be more profound and wider in spectrum, especially if a different arm of the immune system is impaired. Lee et al. ( 85) reported on a patient in whom there was sudden loss of vision to no light perception OS. The patient had a history of type I diabetes in addition to AIDS and had been receiving foscarnet for CMV retinitis, which was in remission. Visual acuity was 20/ 20 in the affected eye 4 days before presentation. Beginning 3 days before presentation, the patient complained of paranasal sinus pressure and nasal congestion followed by sudden blindness OS. On exami­nation, there was no fever and visual acuity was 20/ 25 OD and no light perception OS, with a left relative af­ferent pupillary defect. Fundus examination was unre­markable, and the CMV lesions were quiescent. Twelve hours later, the patient reported neck soreness and a lum­bar puncture was performed. Cerebral spinal fluid analy­sis was remarkable for pleocytosis without a positive cryptococcal antigen. Within 36 hours total occlusion of the internal carotid artery developed, leading to a diffuse left cerebral infarction. Despite initiation of Amphoteri­cin B, the patient died 6 days later. Postmortem findings demonstrated Rhizopus species in the orbit and sinuses. There was a severe purulent meningoencephalitis that involved the base of the brain from the optic chiasm to the pontomedullary junction. The optic chiasm had un­dergone infarction, and the left optic nerve demonstrated focal necrosis associated with granulomata- containing organisms. This case is only the second case of rhinoor-bital mucormycosis in an AIDS patient and only the 10th reported case of mucormycosis affecting any organ sys­tem in an AIDS patient. The only other patient with rhinoorbital involvement also had diabetes ( 85). This im- ./ Neum- Ophlhulmol, Vol. 18. No. 2, 1998 ANNUAL REVIEW OF SYSTEMIC DISEASE 137 plies that although depressed immune status is necessary for the development of mucormycosis, cell- mediated im­mune dysfunction is not sufficient. Because persons with diabetes have alterations in neutrophil function and hu­moral immunity, it can be speculated that the reasons for such a fulminant infection in this patient were likely the combined immunosuppressive effects of both HIV and diabetes. Several case reports involving the anterior segment have been published including the following: a case of angle involvement with histoplasma in a patient with disseminated histoplasmosis and massive involvement of the choroidal vasculature including the choriocapillaris that presumably occurred by hematogenous dissemina­tion ( 86), Acanthamoeba endophthalmitis presenting as a granulomatous uveitis ( 87), Kaposi's sarcoma of the bul­bar conjunctiva presenting as the initial clinical manifes­tation of AIDS ( 88), bilateral endogenous fusarium en­dophthalmitis ( 89), bilateral uveal effusion syndrome leading to angle closure glaucoma ( 90), and CMV infec­tion of the cornea presenting as a branching nonulcer­ative keratopathy associated with stromal keratouveitis ( 91). Central Nervous System Manifestations Ischemic infarction and intracerebral hemorrhage ( ICH) are infrequently reported manifestations in AIDS patients. In several instances, however, stroke was the first manifestation of HIV infection ( 92). Recently the relationship between stroke and AIDS was reported on by Pinto who performed an extensive search of the lit­erature that spanned 18 years ( 93). Six clinical series and 11 autopsy series were identified in which clinical stroke or pathologic evidence of stroke were found. Stroke in the pediatric AIDS population was not included because the pathophysiologic mechanisms were believed to be different than in adults. Clinical or autopsy series in which cerebrovascular disease ( CVD) was not men­tioned were excluded, and subarachnoid hemorrhage was the only stroke subtype not included. A total of 1,885 patients infected with HIV at different stages of HIV infection were found. Although 40% had neurologic complications, only 1.3% had a stroke syndrome. Isch­emic infarctions were more common than ICH. Although the percentage of patients with neuropathologic autopsy findings of CVD was much higher than in the clinical series, the autopsy findings were mostly not associated with clinical stroke before death; therefore, their clinical significance is not entirely clear. However, these find­ings do indicate that stroke does occur more commonly than previously thought. The most common cause of cerebral infarction in both clinical and autopsy series was nonbacterial thrombotic endocarditis, a common manifestation of hypercoagula­bility in terminally ill patients ( cancer patients in par­ticular). Intracerebral hemorrhages were usually associ­ated with thrombocytopenia, primary CNS lymphoma, and metastatic Kaposi's sarcoma. In a few cases, cerebral vasculitis was related to thrombotic stroke. The vasculitis was due to herpes infection in some, whereas in others the cause was unknown and may have been related to HIV- induced angiopathy or secondary to deposition of immune complexes. In a letter written as a response to Pinto's article, Roquer et al. ( 92) maintained that throm­bocytopenia and toxoplasmosis were the most common causes associated with ICH rather than primary CNS lymphoma and Kaposi's sarcoma. Moreover, they iden­tified 7 of 720 patients with ICH and added that protein S deficiency has been identified in AIDS patients with cerebral infarction predisposing them to thromboembolic complications. Despite an attempt to identify an association between AIDS and CVD, there were limitations in the data pre­venting any such definitive conclusions. In addition, no conclusions could be made regarding the relationship between risk factors for HIV infection and stroke except that the intravenous drug- abusing subset of patients may be at higher risk due to foreign body embolism from intravenous injection, development of endocarditis, and formation of mycotic aneurysms ( 92,93). Zunker et al. ( 94) reported on two interesting cases of striatocapsular infarction associated with rapidly chang­ing transcranial ultrasound findings, fundus abnormali­ties consisting of cotton wool spots, and absence of CNS opportunistic infection. In both patients, results on elec­trocardiography and transthoracic and transesophageal echocardiography were normal, as well were those on duplex ultrasonography of the extracranial vasculature. Blood studies including rheumatoid factor, antithrombin III, protein S and C, anticardiolipin antibodies, and VDRL were negative. Given the above findings, the au­thors concluded that the only diagnoses possible in young patients at low risk for atherosclerotic disease with a rapidly changing transcranial blood flow study were either vasospasm or vasculitis. Because there was no clinical history of vasospasm elsewhere, HIV angi­opathy and an associated vasculitis was believed to be causal ( 94). In summary the accumulating information regarding CVD in AIDS suggests that CVD is more common than previously thought. With the increased survival of AIDS patients associated with improved treatment of opportu­nistic infection, we can expect an increase in the number of reported cases of CVD in AIDS patients. In light of this, prospective studies looking at the incidence of CVD in AIDS patients compared with patients with other ter­minal disorders will help identify the relationship be­tween HIV infection and stroke. It will be important to correlate clinical stroke findings in the presence and ab­sence of other CNS pathology and HIV infection risk factors that are possible variables contributing to the de­velopment of stroke. Intracranial mass lesions in AIDS patients are most often due to toxoplasmosis and CNS lymphoma. Al­though astrocytomas are rarely reported in AIDS pa­tients, their actual incidence compared with that in age-matched, non- HlV- infected individuals appears to be greater. Neal et al. ( 95) reported on a 35- year- old man who was treated empirically for toxoplasmosis without J Neiim- Oplillialmol, Vol. 18, No. 2, I99H 138 P. LAMA AND L. FROHMAN success leading to death within 2 months after stereotac­tic biopsy failed to identify the cause of the lesion. Neu-ropathologic findings at autopsy revealed an astrocy­toma. As of the date of publication, there were only nine other cases of astrocytoma among the HIV- infected population reported in the world literature. Because the incidence of astrocytoma in the general population in persons aged 25 to 35 years is 1 per 100,000 and because this case was the sixth case in Western Europe among approximately 14,665 AIDS patients over the interval studied, there may be a predisposition to development of astrocytomas in AIDS patients. The authors believed that immunosuppression secondary to HIV infection may have predisposed the patient to nonhematologic malig­nancies ( 95). The proximity of astrocytes to the blood-brain barrier and the presence of CD4 and complement receptors on the surface may result in infection and trans­formation of these infected astrocytes to malignant cells ( 96). Currently, space- occupying lesions of the CNS in HIV- infected patients are treated empirically for toxo­plasmosis. If no response is noted after 2 weeks, stereo­tactic biopsy is performed to determine the nature of the lesion that is usually primary CNS lymphoma. However, stereotactic biopsy may not provide a definitive diagno­sis and is associated with significant morbidity and in some cases death. A study was conducted to determine the effectiveness of stereotactic biopsy compared with no biopsy using a decision analysis. The results demon­strated an overall survival benefit for the biopsy group, 91 days versus 67 days, compared with the no biopsy group. The net survival benefit, however, was sensitive to the diagnostic success rate, operative mortality, like­lihood of lymphoma diagnosis, and life expectancy of patients being diagnosed and treated for lymphoma ( 97). Thus it would certainly be advantageous if noninvasive methods could accurately distinguish the cause of focal CNS mass lesions, thereby obviating the need for a ste­reotactic biopsy. Heald et al. ( 98) performed a prospec­tive trial evaluating the efficacy of positron emission scanning ( PET) using F- 18 fluorodeoxyglucose ( FDG) to distinguish between malignant and infectious lesions ( 98). Eighteen patients who were HIV positive and had focal mass lesions on CT and MRI scans were examined. They underwent FDG- PET scanning as a diagnostic ad­junctive test and the metabolic activity of each patient's lesion was graded using a qualitative visual score and a semiquantitative count ratio comparing the lesion with the contralateral brain. Patients with lymphoma were found to have statistically higher visual scores and count ratios than those with CNS lesions diagnosed as infec­tions. This test differentiated between CNS lymphoma and infection in 16 of 18 cases but could not differentiate lymphoma from progressive multifocal leukoencephal-opathy. If these results can be duplicated in future larger prospective trials, FDG- PET scanning may prove to be an extremely advantageous noninvasive rapid diagnostic adjunct reducing the need for invasive biopsy proce­dures. In addition to mass lesions, meningoencephalitis of cryptococcal origin is a major cause of CNS disease in the AIDS population. Some of the manifestations include multiple intracranial neuropathies, radiculitis, encephali­tis, and elevated intracranial pressure ( 99). Elevation of intracranial pressure is often due to obstruction of the arachnoid villi leading to impaired CSF outflow. There have been, however, instances of HIV- infected patients with elevated intracranial pressure and no evidence of infection or sagittal vein thrombosis. Prevett and Plant ( 100) described two cases of meningoradiculitis with signs and symptoms of raised intracranial pressure with­out evidence of CNS opportunistic infection, suggesting that the cause of obstruction of CSF outflow is due to HIV- related arachnoiditis. They concluded that such pa­tients may be managed by lumbar puncture alone. Hence, a pseudotumor cerebri- like syndrome may occur in the AIDS population as a diagnosis of exclusion. Therapy Pyrimethamine and sulfadiazine have conventionally been the agents used in the treatment of toxoplasmosis encephalitis. However, drug intolerance occurs fre­quently, in part due to bone marrow suppression. Clin­damycin is often used in lieu of sulfadiazine but intoler­ance secondary to diarrhea and enterocolitis related to Clostridium difficile toxin occurs in a significant fraction of patients, thus leaving the patient with limited options for long- term therapy. Atovaquone was investigated as a possible agent for long- term suppression against conven­tional therapy with pyrimethamine- sulfadiazine or clin­damycin. The study involved 65 patients with known intolerance to conventional therapy who, after resolution of acute toxoplasmic encephalitis, were administered atovaquone as maintenance therapy. Before atovaquone was used, there had been 129 episodes of intolerance to conventional antitoxoplasmic drugs. In 75% of patients, atovaquone was used as a single agent. The outcome revealed a 26% relapse rate, but the drug was extremely well tolerated and only two patients ( 3%) had to discon­tinue therapy. The duration of pyrimethamine-sulfadiazine treatment during the acute phase of toxo­plasmic encephalitis was the only factor significantly as­sociated with a decreased risk of relapse ( 101). HIV Diagnosis and Drug Therapy The past few years have brought exciting new devel­opments in HIV biology. Studies of viral kinetics have demonstrated that the turnover rate of plasma virions is continuous and highly productive leading to rapid de­struction of CD4 lymphocytes. Using a potent antiviral protease inhibitor, Ho et al. ( 102) demonstrated a turn­over rate of plasma virions equal to approximately 48 hours and a half- life of 6 hours. In addition, the rate constant of viral clearance did not vary according to disease stage, indicating that the extent of viremia de­pended on the rate of virus production ( 102). Similar data on viral kinetics were also noted by Wei et al. ( 103), and it is now known that viral load is directly correlated with disease progression ( 104). Traditionally, total CD4 lym- J Neuro- Ophthahnol, Vol. 18, No. 2, 1998 ANNUAL REVIEW OF SYSTEMIC DISEASE 139 phocyte count has been used as a marker of staging and susceptibility for development of different opportunistic infections ( 105). Although CD4 counts provide a rough estimation of immune status and prognosis, the absolute counts do not correlate with viral load. It follows then that because the amount of plasma HIV- 1 RNA ( measure of viral load) is an independent risk factor for disease progression, patients with similar CD4 counts but differ­ent plasma HIV- 1 RNA values are at unequal risk for progression. This is likely the reason for widely varying rates of progression in patients with similar initial CD4 counts. A recent prospective study confirmed these find­ings. The investigators noted that 50% of individuals with CD4 counts greater than or equal to 500 and greater than 10,900 HIV- 1 RNA molecules per milliliter died within 6 years of study entry, compared with only 5% of those with similar CD4 counts but with viral RNA levels less than 10,900 molecules per milliliter ( 106). However, there was no observable threshold with respect to viral load and disease progression to indicate that progression can occur at any level with varying degrees of risk. With this knowledge, measurement of plasma HIV- 1 RNA has been studied as a surrogate marker not only as a predictor of future progression but also as a monitor of drug effi­cacy. There appears to be a good correlation between reduction in levels of plasma HIV- 1 RNA during drug therapy and lowered risk of AIDS progression. The con­trary also appears to be true, that is, rising plasma HIV- 1 RNA during drug therapy implies acquisition of drug resistance and a greater risk of disease progression ( 107). Currently, there are three different methods of measuring plasma HIV- 1 RNA levels: reverse polymerase chain reaction, signal amplification, and nucleic acid se-quenced- based amplification. Because there may be widely varying differences in measured values between these different assay methods, the same assay should be used for each patient when monitoring disease status or drug efficacy. Due to the above findings and those of numerous stud­ies demonstrating the potency of multidrug regimens combining nucleoside reverse transcriptase inhibitors ( NRTI) and protease inhibitors, the International AIDS Society of the United States has significantly revised its recommendations for antiretroviral therapy since July 1996 ( 108). It is now recommended that all patients with HIV RNA levels greater than 5,000 to 10,000 copies per milliliter, regardless of CD4 cell count, should begin therapy, whereas previously it was suggested that pa­tients with CD4 counts less than 500 per microliter and those with plasma HIV- 1 RNA levels greater than 30,000 to 50,000 copies per milliliter should begin therapy. In addition, therapy should be considered for all those with detectable HIV- 1 RNA in plasma. Therapy for patients considered to be low risk, that is, those with low plasma HIV- 1 RNA and high CD4 counts who are not especially committed to complex multidrug regimens, may defer such treatment for 3 to 6 months, after which their con­dition should be reevaluated. Therapeutic options for ini­tial therapy include three- drug regimens consisting of two NRTI and a protease inhibitor or two NRTI and a nonnucleoside reverse transcriptase inhibitor such as nevirapine and delavirdine. If a patient who is on therapy is experiencing treatment failure as indicated by a rising plasma HIV- 1 RNA level, failure to achieve the desired reduction in viral load, a decline in CD4 cell count, or evidence of clinical disease progression, a change in therapy is warranted. Other indications for changing therapy include intolerable toxicity, medical noncompli­ance, or inappropriate medical treatment, that is, mono­therapy. Depending on the initial regimen, the panel lists a number of combinations of the currently available agents that may be substituted if a change is clinically warranted. For more details, we refer the reader to Table 1 of this article and Tables 4 and 5 in the article by Carpenter et al. ( 108). The changes in antiretroviral therapy have provided new hope for controlling HIV infection and staving off the inexorable decline in immune function. However, in spite of the development of new pharmacologic agents, currently we have only two classes available, and the development of viral resistance will eventually limit the efficacy of the treatment regimen. Research in antiretro- TABLE 1. Antiretroviral chemotherapy NRTI* Zidovudine, AZT ( Retrovir, Glaxo, Wellcome) Didansine, ddl ( Videx, Bristol- Meyers, Squibb) Lamivudine, 3TC ( Epivir, Glaxo, Wellcome) Stavudine, d4T ( Zerit, Bristol- Meyers, Squibb) Zalcitabine, ddC ( Hivid, Roche Pharmaceuticals) NNRTI** Delaviridine ( Rescriptor, Upjohn) Nevirapine ( Viramune, Roxane) Protease inhibitors Indinavir ( Crixivan, Merck) Saquinavir ( Invirase, Roche Pharmaceuticals) Ritonavir ( Norvir, Abbott) Nelfinavir ( Viracept, Agouron) Major toxicity Hematologic, myopathy Pancreatitis Pancreatitis Peripheral neuropathy Peripheral neuropathy, pancreatitis Rash, gastrointestinal, blood dyscrasia Severe rash, Stevens- Johnson syndrome Nephrolithiasis, potentially serious drug interactions Gastrointestinal- nausea, abdominal pain, diarrhea Potentially serious drug interactions similar to indinavir Diarrhea * Nucleoside reserve transcriptase inhibitors. ** Non- nucleoside reverse transcriptase inhibitors. ./ Neum- Ophlhalmol, Vol. 18, No. 2, 1998 140 P. LAMA AND L. FROHMAN viral pharmacology and the development of novel agents targeting different molecular processes necessary for vi­ral replication will be necessary to prevent emergence of viral resistance. It seems logical to develop treatment protocols that consist of an initial " remission induction" regimen followed by intensification and consolidation with different classes of agents to maintain plasma viral RNA levels at an undetectable level, mimicking cancer chemotherapeutic protocols. This also implies that more sensitive assays than currently are available will be nec­essary to monitor drug therapy. Other avenues of treatment appear to be on the hori­zon. Currently the hormone dehydroepiandrosterone ( DHEA) has been studied as a possible immune-enhancing agent because AIDS and other immunosup-pressed patients appear to have lower than normal levels of DHEA. Levels of DHEA are also lowest in healthy aging adults and young children ( 109). It appears that DHEA may function by enhancing the secretion of IL- 2 from activated T cells. Moreover, in that DHEA levels decrease with the progression of AIDS, there is potential to use serum DHEA levels as markers of disease activity. Although addition of DHEA to a treatment protocol may provide an additional boost to the immune system, pro­spective clinical trials will determine the role of DHEA in treating AIDS patients. Future Directions During the past 2 years, a number of major discoveries were made by researchers in HIV biology. In 1986, Levy and co- workers noted that a soluble factor resulted in noncytolytic suppression of HTV replication ( 110). It was not until 1995, however, that Cocchi discovered that there were multiple soluble factors involved in noncy­tolytic HIV suppression called beta chemokines ( 111). Chemokines are specific factors separate from lympho-kines that function as chemoattractants and cytokines. There are two main groups of chemokines, alpha and beta. Alpha chemokines primarily attract neutrophils, whereas beta chemokines attract and activate lympho­cytes, monocytes, basophils, and eosinophils. The semi­nal discovery of specific chemokine receptors on CD4 lymphocytes and on peripheral blood mononuclear cells essential for HIV entry clarified the previous observation that patients infected with HIV in whom beta chemo­kines were elevated had long- term survival and did not progress to AIDS. It also formed the basis of the impor­tant observation that homozygous deletion of a 32- base pair allele of the CCR- 5 chemokine receptor confers pro­tection against sexual and parenteral HIV transmission. Cocchi demonstrated that the gp 120 portion of the HIV molecule binds to CD4, exposing a variable loop domain ( V3 loop) that subsequently binds to either receptor CCR- 5 or CXCR- 4. Formation of this complex is essen­tial for membrane fusion and subsequent viral entry. Be­cause each subtype receptor is specific for a particular HIV isolate ( T- tropic vs. M- tropic), loss or blockage of one receptor would theoretically prevent infection by that specific isolate ( 112). Thus, this exciting new infor­mation theoretically opens the door for several potential therapeutic strategies, including pharmacologic blockade with beta chemokine- like ligands, genetic downregula-tion of CCR- 5 or CXCR- 4, and development of vaccines with portions of these receptor subtypes. REFERENCES 1. Barnes PF, Barrows SA. Tuberculosis in the 1990' s. Ann Intern Med 1993; 119: 400- 10. 2. Cohn DL, Bustreo F, Raviglione MC. Drug- resistant tuberculosis: review of the worldwide situation and the WHO/ 1UATLD global surveillance project. International Union Against Tuberculosis and Lung Disease. Clin Infect Dis I997; l( suppl 24): S121- 30. 3. Kamholz SL. Resurgence of tuberculosis: the perspective a dozen years later. ./ Assoc Acad Minor Phys 1996; 7( 3): 83- 6. 4. Chin PK, Jacobs MB, Hing SJ. Orbital tuberculoma masquerad­ing as an orbital malignancy. Aust N Z J Ophthalmol I997; 25( l): 67- 9. 5. Zoric LD, Zoric DL, Zoric DM. Bilateral tuberculous abscesses on the face ( eyelids) of a child. Am J Ophthalmol 1996; 121 ( 6): 717- 8. 6. Kadioglu HH, Gundogdu C, Dcniz O, et al. Optochiasmatic tu-berculoma- case report and review. Zentralbl Neurochir 1996; 57 ( l): 30- 6. 7. Berinstein DM, Gentile RC, McCormick SA, Walsh JB. Primary choroidal tuberculoma. Arch Ophthalmol 1997; 115( 3): 430- 1. 8. Reny JL, Challe G, Geisert P, ct al. Tuberculosis- related retinal vasculitis in an immunocompetent patient. Clin Infect Dis 1996; 22( 5): 873- 4. 9. Bcrger C, Braegger CP, Albisetti M, et al. Cerebral tuberculosis presenting as complex febrile convulsions. Neuropediatrics 1996; 27( 3): 161- 3. 10. Muccioli C, Bel fort R Jr. Presumed ocular and central nervous system tuberculosis in a patient with the acquired immunedefi-ciency syndrome. Am J Ophthalmol 1996; 121 ( 2): 217- 9. 11. Bouza E, Merino P, Munoz P, et al. Ocular tuberculosis: a pro­spective study in a general hospital. Medicine 1997; 76( 1): 53- 61. 12. Piersimoni C, Callegaro A, Nista D, et al. Comparative evaluation of two commercial amplification assays for direct detection of mycobacteria tuberculosis complex in respiratory specimens. J Clin Microbiol 1997; 35( 1): 193- 6. 13. Wilkinson RJ, Haslov K, Rappuoli R, et al. Evaluation of the recombinant 38- kiIodalton antigen of mycobactcrium tuberculo­sis as a potential immunodiagnostic reagent. ./ Clin Microbiol 1997; 35( 3): 553- 7. 14. Kawamura M, Sueshigc N, Imayoshi K, et al. Enzyme immuno­assay to detect antituberculous glycolipid antigen ( anti- TBGL antigen) antibodies in serum for diagnosis of tuberculosis. J Clin Lab Anal 1997; 1 1( 3): 140- 5. 15. Iscman M. Treatment of multidrug- resistant tuberculosis. A' Engl J Med 1993; 329( 11): 784- 91. 16. Centers for Disease Control. Initial therapy for tuberculosis in the era of multidrug- resistance: recommendations of the advisory council for the elimination of tuberculosis. MMWR CDC Surveill Summ 1993; 42( RR- 7): l- 8. 17. Peloquin CA. Pharmacology of the antimycobacterial drugs. Med Clin North Am 1993; 77: 1253- 62. 18. Peloquin CA. Quinolones and tuberculosis. Ann Pharmacother 1996; 30: 1034- 5. 19. Kaplan J, Kricff D. Quinolones for treatment and prophylaxis of TB. Ann Pharmacother 1996: 30: 1020- 2. 20. Kohno S, Koga H, Kaku M, et al. Prospective comparative study of ofloxacin or ethambutol for the treatment of pulmonary tuber­culosis. Chest 1993; 104: 1194- 8. 21. Mohanty KL, Dhamgaye TM. Controlled trial of ciprofloxacin short- term chemotherapy for pulmonary tuberculosis. Chest 1993: 104; l 194- 8. 22. Vilcek J, Oliveira I. Recent progress in the elucidation of inter-feron- gamma actions: molecular biology and biological func­tions, lnt Arch Allergy Immunol 1994; 104: 311- 6. 23. Holland SM, Eisenstein EM, Kuhus DB, et al. Treatment of re- J Neitro- Oplillialmol, Vol. 18, No. 2, I99K ANNUAL REVIEW OF SYSTEMIC DISEASE 141 fractory disseminated nontubcrculous mycobacterial infection with interferon gamma: a preliminary report. NEngl .1 Med 1994; 330: 1348- 55. 24. Squires KE, Brown ST, Armstrong DA, et al. Intcrferon- gamma treatment for mycobacterinm avium- intracellulare complex bacil-lemia in patients with AIDS ( letter)../ Infect Dis 1992; 166: 686- 7. 25. Murray HW. Intcrferon- gamma and host antimicrobial defense: current and future clinical applications. Am J Med I994; 97: 459- 67. 26. Raad I, Hachem R, Leeds N, et al. Use of adjunctive treatment with inlerferon- gamma in an immunocompromised patient who had refractory multidrug- resistant tuberculosis of the brain. Clin Infect Dis 1996; 22( 3): 572- 4. 27. Condos R, Rom WN, Schluger NW. Treatment of mullidrug-resistant pulmonary tuberculosis with inteiTeron- gamma aerosol. Lancet 1997; 349( 9064): 15 13- 5. 28. Centers for Disease Control. Tuberculosis morbidity: United Slates, 1995. MMWR CDC Suiveill Siimm 1996; 45( 18): 365- 70. 29. Jabs DA, Johns CJ. Ocular involvement in chronic sarcoidosis. Am J Ophthalmol 1986; 102: 297- 301. 30. Khomenko AG, Ozerova LV. Clinical aspects and characteristics of the course of most common non- pulmonary localizations of sarcoidosis. Vestn Ross Akad Med Nauk I995; 7: 21^ t. 31. Borodulina EA, Branchevskaia SI, Borodulin BE. Eye lesions in patients with pulmonary sarcoidosis. Vestn Oftalmol 1996; 112: 50- 1. 32. Makarov Al, Turicin VI, Kostina ZI. The diagnosis and treatment of ncurosarcoidosis. Zh Nevropatol Psikhiatr 1995; 95: 90- 3. 33. Recio V, Chistofondis G, Skarf B, Similar G. Clinical and radio­logical analysis of neurosarcoidosis. Presented at the 1997 North American Ncuro- Ophthalmology Society Meeting, Keystone, Colorado, February 1997. 34. Richter E, Greinerl U, Kirtsen D, et al. Assessment of mycobac­terial DNA in cells and tissue of mycobacteria and sarcoid le­sions. Am .1 Respir Crit Care Med 1996; 153: 375- 80. 35. Baselga E, Baradas MA, Margall N, de Moragas JM. Detection of tuberculosis complex DNA in a lesion resembling sarcoidosis. Clin Exp Dermatol 1996; 21: 235- 8. 36. Almenoff PL, Johnson A, Lesser M, Mattman LH. Growth of acid fast L forms from the blood of patients with sarcoidosis. Thorax 1996; 51: 530- 3. 37. el- Zaatari FA, Naser SA, Markesich DC, Kalter DC, Engsland L, Graham DY. Identification of mycobacterinm avium complex in sarcoidosis. J Clin Microbiol 1996; 34: 2240- 5. 38. cl- Zaatari FA, Graham DY, Samuelsson K, Engstrand L. Detec­tion on mycobacterial avium complex in cerebrospinal fluid of a sarcoid patient by specific polymerase reaction assays. Scand .1 Infect Dis 1997; 29: 202- 4. 39. Puolakkainen M, Campbell LA, Kuo C- C, Leinonen M, Gron-hagen- Riska C, Saikku P. Serological response to chlamydia pneumoniae in patients with sarcoidosis. J Infect 1996; 33: I99- 205. 40. Tsukada N, Yanigasawa N, Mochizuki I. Endothelial cell damage in sarcoidosis and ncurosarcoidosis: autoantibodies to endothelial cells. Eur Neurol 1995; 32: 108- 12. 41. Milburn HJ, Poulter LW, Milmec A, Cochrane GM, Kemeny DM. Corticosteroids restore the balance between locally pro­duced Thl and Th2 cytokines and immunoglobulin isotypes to normal in sarcoid lung. Clin Exp Immunol 1997; 108: 105- 13. 42. Kosmorsky GS, Prayson R. Primary optic pathway sarcoidosis in a 38- ycar- old while man. ./ Neuro- Ophthalmol 1996: 16: 188 90. 43. Ing EB, Garrity JA, Cross SA, Ebcrsold MJ. Sarcoid masquer­ading as optic nerve sheath meningioma. Mayo Clin Proc 1997; 72: 38- 43. 44. el Zarka A, Hille K, Ruprccht KW. Granulomatous process of the optic papilla as an initial sign of systemic sarcoidosis. Klin Monatsbl Augenheilkd 1995; 206: 134- 5. 45. Ng KL, McDermott N, Romanowski CAJ, Jackson A. Ncurosar­coidosis masquerading as glioma of the optic chiasm in a child. Postgrad Med .1 1995; 71: 265- 8. 46. Kao SC, Rootman J. Unusual orbital presentations of dural sar­coidosis. Can J Ophthalmol 1996; 31: 195- 200. 47. Monfort- Gouraud M, Chokrc R, Dubicz M, Ratsimiah T, Haniza A, Sauvageon G. Inflammatory pseudotumor of the orbit and suspected sarcoidosis. Arch Pediatr 1996; 3: 697- 700. 48. Leiba H, Siatkowski RM, Culbcrtson WW, Glaser JS. Ncurosar­coidosis presenting as an intracranial mass in childhood. J Neuro- Ophthalmol 1996; 16: 269- 73. 49. Lee Andrew G. Unusual ocular feature to the possible protean manifestations of sarcoidosis | Letter/ Comment |. Retina 1996; 16: 266- 7. 50. Rizzato G, Angi M, Fraioli P, Montemuuro L, Pilotto E, Tom-niasini A. Uveitis as a presenting feature of chronic sarcoidosis. Eur Respir J 1996; 9: 1201- 5. 51. Braun M, Jonas JB. Band- like corneal degeneration as the initial sign of sarcoidosis. Klin Monatsbl Augenheilkd 1006; 209: 328- 9. 52. DeRosa AJ, Margo CE, Orlick ME. Hemorrhagic retinopathy as the presenting manifestation of sarcoidosis. Retina 1995; 15: 422- 7. 53. Dodds EM, Lowder CY, Barnhorst DA, Lavcrtu P, Caravella LP, White DE. Posterior sclerilis with annular ciliochoroidal detach­ment. Am J Ophthalmol 1995; 120: 677- 9. 54. Vrabec TR, Augsburger J. I, Fischer DH, Belmont . IB. Tachcs dc bougie. Ophthalmology 1995; 102: 1712- 21. 55. Caparros- Lefebvre, Wallaert B, Girard- buttaz I, et al. MRI aspect and course of supra- tenlorial sarcoidosic lesions. Rev Neurol 1996; 153: 196- 201. 56. Grand S, Hoffman D, Bosl F, Francois- Jouberl A, Pasquicr B, Lc Bas JF. Case report: pscudotumoral brain lesion as the presenting feature of sarcoidosis. Br J Radiol 1996; 69: 272- 5. 57. Weissman MN, Lange R, Kellcy C, Belgca K, Abel L. Intraspinal epidural sarcoidosis: case report. Neurosurgery 1996; 39: 179- 81. 58. Waubanl E, Manelfe C, Bonal'e A, Berry 1,' cinel M. MRI of intramedullary sarcoidosis: follow- up of a case. Neuroradiology 1997; 39: 357- 60. 59. Fitzpatrick KJ, Chancellor MB, Rivas DA, Kumon H, Mandcl S, Manon- Espaillat R. Urologic manifestation of spinal cord sar­coidosis. ./ Spinal Cord Med 1996; 19: 201- 3. 60. Cavincss JN, Knox CA. Hemisdystonia occurring in a patient with sarcoidosis. Mov Disord 1996: 1 1: 340- 1. 61. De Bleeckcr J, van Akcn L, van Landegcm W, van den Abcele G, Cuvelicr C. Myasthenia gravis and sarcoidosis: report of 2 cases. Eur Neurol 1996; 36: 326- 7. 62. Jamrozik Z, Dabrowski A, Drac H, Kwiecinski H. Ncurosarcoid­osis or Guillain- Barre syndrome complicating sarcoidosis. Neurol Neurochir Pol 1996; 30: 481- 7. 63. Sekiguchi M, Yazaki Y, Isobe, Hiroe M. Cardiac sarcoidosis: diagnostic, prognostic, and therapeutic considerations. Cardio-vasc Drugs Ther 1996: 10: 495- 510. 64. Nelson JE, Kirschner PA, Teirstein AS. Sarcoidosis presenting as heart disease. Sarcoidosis Vase Diffuse Lung Dis 1996; 13: 178- 82. 65. Vatti R, Sharma OP. Course of asymptomatic liver involvement in sarcoidosis: role of therapy in selected cases. Sarcoidosis Vase Diffuse Lung Dis 1997; 14: 73- 6. 66. Boruchowicz A, Hachulla E, Cortot A, Paris JC, Colombel JF. Chronic pancreatitis: a complication of sarcoidosis? Eur .1 Gas­troenterol Hepatol I996; 8: l 125- 7. 67. Pearce KF, Nolan TE. Endometrial sarcoidosis as a cause of postmenopausal bleeding: a case report. ./ Reprod Med 1996; 4I: 878- 80. 68. Sherman MD, Pincc KJ, Farahmand SM. Sarcoidosis manifesting as uveitis and menomelrorrhagia. Am J Ophthalmol 1997; 12.3: 703 5. 69. De Bandt M, Perrol S, Masson C, Meyer O. Systemic sclerosis and sarcoidosis: a report of five cases. Br J Rheumatol I997; 36: 117- 9. 70. Carmody RF, Mafee MF, Goodwin JA, Small K, Hacry C. Orbital and optic pathway sarcoidosis: MR findings. AJNR Am .1 Neuro-radiol 1994; 15: 775- 83. 71. Westlakc WH, Heath JD, Spalton DJ. Sarcoidosis involving Ihe optic nerve and hypothalamus. Arch Ophthalmol 1995; 1 13: 669- 70. 72. Mana J, Salazar A, Pujol R, Manresa F. Are the pulmonary func­tion tests and the markers of activity helpful to establish the prognosis of sarcoidosis? Respiration 1996; 63: 298- 303. ./ Neiim- Oplulmlmot. Vol. 1H, No. 2, I99K 142 P. LAMA AND L. FROHMAN 73. Agbogu BN, Stern BJ, Sewell C, Yang G. Therapeutic consider­ations in patients with refractory neurosarcoidosis. Arch Neurol 1995; 52: 875- 9. 74. Power WJ, Neves RA, Rodriguez A, Pedroza- Scres M, Foster CS. The value of combined serum angiotensin- converting enzyme and gallium scan in diagnosing ocular sarcoidosis. Ophthalmol­ogy 1995; 102: 2007- 11. 75. Wolfensbcrger TJ, Guex- Crosier Y, Herbort CP. Indocyanine green angiography ( ICGA) features in ocular sarcoidosis. Poster presentation number 64 at the American Academy of Ophthal­mology, San Francisco, October 1997. 76. Bacchclla L, Tinelli C, Gile LS, et al. Serum Type I and Type III procollagen peptide levels in sarcoidosis. Eur Respir J 1996; 9: 1648- 51. 77. Sadcr A, Gramer E, Korner U. Sarcoidosis of the optic nerve. Klin Monatsblat Augenheilkd 1995; 206: 128- 33. 78. Lower EE, Baughman RP. Prolonged use of methotrexate for sarcoidosis. Arch Intern Med 1995; 155: 846- 51. 79. Gedalia A, Molina JF, Ellis GS Jr, Galen W, Moore C, Espinosa LR. Low dose methotrexate therapy for childhood sarcoidosis. J Pediatr 1997; 130: 25- 9. 80. Brechtel B, Haas N, Hen/. BM, Kokle G. Allopurinol: a thera­peutic alternative for disseminated cutaneous sarcoidosis. Br J Dermatol 1996; 135: 307- 9. 81. Carlesimo M, Giustini S, Rossi A, Bonaccorsi P. Treatment of cutaneous and pulmonary sarcoidosis with thalidomide. J Am Acad Dermatol 1995; 32( 5, part 2): 866- 9. 82. Wei ME, Campbell SH, Taylor C. Precipitous visual loss second­ary to optic nerve toxoplasmosis as an unusual presentation of AIDS. Aust NZJ Ophthalmol 1996; 24( l): 75- 7. 83. Friedman SM, Margo CE. Bilateral central retinal vein occlusions in a patient with acquired immune deficiency syndrome: clinico-pathologic correlation. Arch Ophthalmol 1995; 113( 9): 1184- 8. 84. Glasgow BJ. Evidence for breaches of the retinal vasculature in acquired immune deficiency syndrome angiopathy: a fluorescent microsphere study. Ophthalmology 1997; 104( 5): 753- 60. 85. Lee BL, Holland GN, Glasgow BJ. Chiasmal infarction and sud­den blindness caused by mucormycosis. Am J Ophthalmol 1996; 122( 6): 895- 6. 86. Font RL, Parsons MA, Keener MJ, et al. Involvement of anterior chamber angle structures in disseminated histoplasmosis: report of three cases. Ger J Ophthalmol 1995; 4( 2): 107- 15. 87. Heftier KF, Eckhardl TJ, Reboli AC, et al. Acantliamoeba en­dophthalmitis in acquired immune deficiency syndrome. Am J Ophthalmol 1996 122( 4): 584- 66. 88. Kurumetry UR, Lustbader JM. Kaposi's sarcoma of the bulbar conjunctiva as an initial clinical manifestation of acquired im­mune deficiency syndrome. Arch Ophthalmol 1995;! 13( 8): 978. 89. Glasgow BJ, Engstrom RE Jr, Holland GN, Kreiger AE, Wool MG. Bilateral endogenous Fusarium endophthalmitis associated with acquired immunodeficiency syndrome. Arch Ophthalmol 1996; 114( 7): 873- 7. 90. Krzystolik MG, Kuperwasser M, Low RM, Dreyer EB. Anterior segment ultrasound biomicroscopy in a patient with AIDS and bilateral angle- closure glaucoma secondary to uveal effusions [ Letter], Arch Ophthalmol 1996; 114( 7): 878- 90. 91. Wilhemus KR, Font RL, Lehmann RP, Cernoch PL. Cytomega­lovirus keratitis in acquired immunodeficiency syndrome [ Clini­cal Conference], Arch Ophthalmol 1996;! l4( 7): 869- 72. 92. Roquer J, Palomeras E, Pou A. AIDS and cerebrovascular disease [ Letter]. Stroke 1996; 27( 9): 1694. 93. Pinto AN. AIDS and cerebrovascular disease. Stroke 1996; 27( 3): 538- 43. 94. Zunker P, Nabavi DG, Allardt A, Husstedt IW. HIV- associated stroke: report of two unusual cases. Stroke 1996; 27( 9): 1694- 6. 95. Neal JW, Llewelyn MB, Morrison HL, et al. A malignant astro­cytoma in a patient with AIDS: a possible association between Astrocytomas and HIV Infection. J Infect 1996; 33: 159- 62. 96. Saito Y, Sharer LR, Epstein LG, et al. Over expression of NEF as a marker for restricted HIV- 1 infection of astrocytes in post­mortem pediatric central nervous system tissues. Neurology 1994; 44: 474- 81. 97. Holloway RG, Mushlin Al. Intracranial mass lesions in acquired immunodeficiency syndrome: using decision analysis to deter­mine the effectiveness of stereotactic brain biopsy. Neurology 1996; 46( 4): 1010- 5. 98. Heald AE, Hoffman JM, Bartlett JA, Waskin HA. Differentiation of central nervous system lesions in AIDS patients using positron emission tomography ( PET). Int J STD AIDS 1996; 7( 5): 337- 46. 99. Laverda AM, Ruga E, Pagliaro A, et al. Intracranial hypertension and cryptococcal meningitis in a girl with AIDS. Brain Dev 1996; 18: 330- 1. 100. Prevett MC, Plant GT. Intracranial hypertension and HIV asso­ciated meningoradiculitis. J Neurol Neurosurg Psychiatry 1997; 62( 4): 407- 9. 101. Katlama C, Mouthon B, Gourdon D, et al. Atovaquone as long-term suppressive therapy for toxoplasmic encephalitis in patients with AIDS and multiple drug intolerance: Atovaquone Expanded Access Group. AIDS 1996; 10( 10): 1107- 12. 102. Ho D, Neumann AU, Perelson AS, et al. Rapid turnover of plasma virions and CD4 lymphocytes in HIV- 1 infection. Nature 1995; 373: 123- 6. 103. Wei X, Ghosh SK, Taylor ME, et al. Viral dynamics in human immunodeficiency virus Type 1 infection. Nature 1995; 373: 117- 22. 104. Mellors JW, Kingsley LA, Rinald CR, et al. Quantitation of HIV- 1 RNA in plasma predicts outcome after seroconversion. Ann Intern Med 1995; 122: 573- 9. 105. Phillips AN, Lee CA, Elford J, et al. Serial CD4 counts and development of AIDS. Lancet 1991; 337: 389- 92. 106. Mellors JW, Rinaldo CR, Gupta P, et al. Prognosis in HIV- 1 infection predicted by the quantity of virus in plasma. Science 1996; 272: 1167- 70. 107. Coombs RW, Welles SL, Hooper C. Association of plasma hu­man immunodeficiency virus Type 1 RNA level with risk of clinical progression in patients with advanced infection. J Infect Dis 1996; 174: 704- 12. 108. Carpenter CCJ, Fischl MA, Hammer SM, et al. Antiretroviral therapy for HIV infection in 1996: recommendations of an inter­national panel. JAMA 1996; 276: 146- 54. 109. Centurelli MA, Abate MA. The role of dehydroepiandrosterone in AIDS. Ann Pharmacol 1997; 31( 5): 639- 42. 110. Walker CM, Levy JA. A diffusible lymphokine produced by CD8+ T lymphocytes suppresses HIV replication. Immunology 1989; 66: 628- 30. 111. Cocchi F, DeVico A, Garzino- Demo A, et al. Identification of RANTES, MIP- 1 alpha, and MIP- 1 beta as the major HIV sup­pressor factors produced by CD8+ T cells. Science 1995; 270: 1811- 5. 112. D'Souza MP, Harden VA. Chemokines and HIV- 1 second recep­tors. Nat Med 1996; 2( 12): 1293- 1300. ./ Neitm- Opluhahnol. Vol. 18, No. 2, 1998