- Email: [email protected]
Intramedullary Abscess Resulting from Disseminated Cryptococcosis Despite Immune Restoration in a Patient with AIDS
Journal of Infection (2002) 44, 185±188 doi:10.1053/jinf.2001.0955, available online at http://www.idealibrary.com on
Intramedullary Abscess Resulting from Disseminated Cryptococcosis Despite Immune Restoration in a Patient with AIDS J. Rambeloarisoaa, D. Batissea, J.-B. Thiebautb, J. Mikolc, S. Mrejena, M. Karmochkinea, M. D. Kazatchkinea, L. Weissa and C. Pikettya,* a
Service d'Immunologie Clinique, HoÃpital EuropeÂen Georges Pompidou, Paris; bService de Neurochirurgie et unite de traitement de la douleur, Fondation Rothschild, Paris; cService d'anatomopathologie, HoÃpital LariboisieÁre, Paris, France We report on a case of cryptococcal intramedullary abscess, which occurred three years after a disseminated cryptococcosis and two years after a lymph node cryptococcal recurrence in a HIV-infected patient who exhibited a long-standing immune restoration. At the time of diagnosis, CD4 lymphocyte-count was 640 106/l and HIV viral load was undetectable. Spinal involvement is rare during cryptococcosis of the central nervous system. As far as we are aware, there is only one case of proven intramedullary cryptococcal abscess reported in the literature and this case is then the second one. The significant and sustained increase in CD4 count following effective antiretroviral therapy was probably associated with only a partial immune restitution that did not allow to avoid the occurrence of the cryptococcal medullar abscess. Finally, this case raises the question of when to stop secondary prophylaxis of cryptococcal disease after increase in CD4 cell count under antiretroviral therapy. # 2002 The British Infection Society
We report on a case of cryptococcal intramedullary abscess, which occurred three years after a disseminated cryptococcosis and two years after a lymph node cryptococcal recurrence in a HIV-infected patient who exhibited a long-standing immune restoration. Human cryptoccocal infection is rare except in patients with disorders of cell-mediated immunity such as AIDS or immunosuppression following corticosteroid therapy or organ transplantation. Neurocryptococcosis is the third most common infection of the central nervous system (CNS) in the immunocompromised patients with AIDS after the encephalopathy due to HIV and encephalitis caused by Toxoplasma gondii [1,2]. Cryptococcal meningitis usually occurred in HIV-infected patients with CD4 cell counts below 100 106/l. Meningoencephalitis with or without intracranial mass lesion is the most common feature in neurocryptococcosis and spinal involvement including cauda equina is rare [3±6]. Intramedullary involvement in spinal cryptococcosis is exceptional [6±9]. * Please address all correspondence to: Dr. C. Piketty, Service d'Immunologie Clinique, HoÃpital EuropeÂen Georges Pompidou, 20 rue Leblanc, 75015 Paris, France. Tel.: 33 1 56 09 27 01; Fax: 33 1 56 09 28 59; E-mail address: [email protected] (C. Piketty). 0163±4453/02/$35.00
A 38-year-old male presented with sudden lower back pain radiating into both lower limbs. He reported no underlying traumatism or effort and no fever. He complained of permanent pain originating from the buttocks and radiating first into the right posterior thigh, the right posterior and lateral calf and the posterior and lateral part of the right knee, then later also into the left posterior thigh. Three years previously, he was found to have HIV infection, when a disseminated cryptococcosis with meningitis was diagnosed. At that time, CD4 cell count was 3 106/l. Sputum, blood and cerebrospinal fluid (CSF) cultures grew Cryptococcus neoformans serotype A. Cryptococcal antigen latex agglutination titre was 1 : 1000 in the CSF and 1 : 32000 in the serum. Brain CT scan and MRI revealed the presence of 4 lesions of cryptococcomas. Intravenous treatment with amphotericin B and flucytosin was started and switched after 3 weeks to oral fluconazole at high dosage (400 mg). Lumbar puncture performed after 5 weeks of treatment collected a CSF that was still India ink stain positive but did not grow cryptococcus any more. Sputum test became negative. He was discharged at that time and oral fluconazole was continued and decreased to 200 mg per day after 6 months. # 2002 The British Infection Society
186
J. Rambeloarisoa et al.
Two years prior to admission, after 10 months of highly active antiretroviral therapy (HAART) including stavudine, lamivudine and indinavir, he presented with an enlarged right supraclavicular lymph node with a diameter of 3 cm which was punctured. The recovered sample was Gomori Grocott stain positive and contained empty non-budding cryptococcus but cultures for fungi were negative. Thorax CT scan disclosed multiple enlarged lymph nodes of the mediastinum, one of which had 24 mm diameter, the right supraclavicular one and a centimetric excavated nodular lesion of the apex of the right lung. There was no evidence for tuberculosis or other infections. At that time, HIV viral load was undetectable since 8 months and CD4 lymphocyte-count was 175 106/l. The right supraclavicular enlarged lymph node shrank and disappeared after the reinforcement of the dosage of fluconazole from 200 mg per day to 400 mg per day. On the last admission in June 2000, the general physical examination was not remarkable apart from a non-inflammatory enlarged right supraclavicular lymph node of about 5 cm diameter. Neurologic examination revealed no abnormality of mental status or of the cranial nerves. Reflexes in the upper limbs as well as in the lower limbs were normal. No muscle weakness or sensory deficit were demonstrable. He had no sign of meningeal irritation. There was neither tenderness over the lower lumbar spine nor contraction of paravertebral muscles. Straight-leg raising was negative bilaterally. Plantar responses were flexor bilaterally. Position sense was normal in the toes. There was no cauda equina syndrome. Course in hospital. The right supraclavicular enlarged lymph node spontaneously and rapidly shrank within 48 h to 3 cm then vanished. Chest radiograph did not show evidence of mediastinal lymph nodes or pulmonary disease. Radiographs of the dorsal and lumbar spine were negative. One week after admission while the pain was soothed by morphine hydrochlorid, he stated that his legs were weak. Examination at that time revealed weakness of the proximal musculature. Both knee jerks were depressed without Babinski sign. Ankle and upper limbs jerks were normal. An hypesthesia of the lateral and anterior parts of the left calf and of the lateral and dorsal parts of the right foot was present. No cauda equina syndrome was found. A spinal MRI was accordingly performed that disclosed a round ring enhancing with gadolinium in a intramedullary lesion at the level of D11±D12, crushing the adjacent structures and reminiscent of abscess (Fig. 1). Cerebral CT scan without contrast injection was normal.
Figure 1. Gadolinium T1 weighted spinal MRI showing a ring enhancing intramedullary lesion at D11±D12 level. Sagittal section (a) Axial section (b).
Laboratory findings. The CSF contained 1 cell and 3 erythrocytes per mm3, normal glucose and 0.79 g/l protein; Gram and Ziehl-Neelsen stains were negative as well as cultures for bacteria and mycobacteria; India ink stain was positive whereas the cultures for cryptococcus were negative; DNA-polymerase chain reactions (PCR) for Herpes-viruses family, and JC papovavirus were negative. Cryptococcal antigen latex agglutination titre was 1 : 10 in the CSF and 1 : 100 in the serum. Blood cultures on fungal media were negative. CD4 lymphocyte-count was 640 106/l and HIV viral load was undetectable. Surgery. An emergency neurosurgical operation was performed. Spinous processes of D11 and D12 were removed then laminectomy and median incision followed that disclosed an abscess with minimal fair pus. The abscess and its hard capsule were removed and no drainage was necessary. Gram stain and culture failed to establish the presence of pyogenic organisms. India ink stain revealed encapsulated yeasts whereas the cultures
Disseminated Cryptococcosis in a Patient with AIDS for fungi were negative. DNA-PCR for Toxoplasma gondii was negative as well as the culture in diploid cells for cytomegalovirus. Histological examination showed an abscess wall rich in fibroblasts, myofibroblasts and inflammatory cells mixed with larged cells of the motoneuron type. The necrotic material inside of it was constituted by polymorphonuclear debris and inflammatory cells of lymphocytes and plasma cells type. By careful examination, there was no pathogenic organism apart from small round structures of slight different size, light brown by Brown and Brenn stain, hard to identify (Periodic Acid-Schiff, Alcian blue and Gomori Grocott stains were different from the control cryptococcal material). Mucicarmine stain was negative. Evolution. The day after the operation, an antituberculous treatment with an association of ethambutol, isoniazid and rifabutin was started as well as an anticryptococcal treatment with liposomal amphotericin B at the dose of 3 mg/kg per d. As the information about the presence of encapsulated yeasts in the pus was given 8 days later, the antituberculous treatment was discontinued and the anticryptococcal treatment was reinforced with flucytosin 100 mg/kg per d. The neurologic deficits of lower limbs slowly improved, in the left side more quickly than in the right one, where an amyotrophia was noted. Physical therapy was initiated. After 2 months of liposomal amphotericin B, the treatment was switched to fluconazole 400 mg/d. Spinal MRI performed 4 months after the operation disclosed a residual cyst limited to D11±D12 level and found no progressive signs. He can walk again and he actually does not need crutches any more. HAART was never interrupted and CD4 lymphocyte-count was 613 mmÿ3 at the last control. Fluconazole was continued at the same dose. Discussion. Spinal involvement is rare during cryptococcosis of the CNS [3±6]. Spinal cryptococcosis splits into 2 groups according to the location of the lesion i.e. extradural [5] and intradural. The cryptococcal infection involving the intradural part of the spinal cord divides up to extramedullary lesions, essentially arachnoiditis with or without mass lesions [3,4,10±17], and intramedullary lesions, namely granulomas and abscesses [6±9]. As far as we are aware, there is only one case of proven intramedullary cryptococcal abscess reported in the literature. It was described as a purulent material escaping from beneath the mass [7]. Our case is then the second one. Furthermore, the necrotic material was found to be constituted by polymorphonuclear debris and inflammatory cells at histological examination. Although the culture of the pus sample was negative, it is well known that India ink stain is effective to detect the presence of cryptococcus in fresh material [18] and it was indeed
187
positive in the present case. The lack of growth could be explained either by the too little available material, or by the fact that the fungus was biologically no more active in the capsule circumscribed pus collection [19]. Even in the absence of positive mucicarmine stain, another fungal aetiology like Histoplasma capsulatum or Penicillium marneffii could be excluded because the patient had never left Paris, France during his life and blood cultures on fungal media were negative. The symptoms are therefore due to inflammatory reaction by an underlying immunity restoration. Such situations were already described and reported for different infections [20±26]. Regarding the enlarged lymph node, our case somewhat fits the case reported by Blanche et al. because symptoms started long after the initiation of HAART [21]: 8 and 15 months for their two patients and 10 months for ours. In the 3 cases, the cultures for cryptococcus were negative. Yet, the similarity ends there because: (1) the enlarged lymph node in our patient relapsed 2 years later and spontaneously vanished within 5 days, moreover no anti-inflammatory drugs was prescribed in our case; (2) there was another cryptococcal location that was intramedullary and that became symptomatic by the time of the relapse. Like the sample from the supraclavicular lymph node punctured 2 years previously, the culture of the pus from the intramedullary abscess was negative for cryptococcus despite the positivity of India ink stain. Van Dellen et al. also reported on a case of cryptococcal infection diagnosed only on the basis of staining and histology [13]. Surgery with resection was successfully performed in our patient with the removal of the encapsulated pus, the postoperative course was uneventful. Grosse et al. are against such an operative process because of the risk of irreversible spinal cord damage, although they mentioned the successful case of Ramamurthi et al. [6,9]. On the cerebral level concerning granulomas, surgery is usually performed only when the absence of an associated diagnosis is doubtful according to Hospenthal et al. [27]. Considering that the enlarged lymph node relapsed for it probably holds residual cryptococcal antigens inside of it, the risk of relapse at the medullary level is conceivable if surgery was avoided. Anyway, its compressive effect on the adjacent structures, which was clinically and radiologically (MRI) evident, was an argument for surgery even that would be restricted to decompression. Although at present, the better surrogate marker for clinical protection is CD4 cell count, this case raises the question of the efficiency of immune restoration to protect against a relapse of cryptococcal infection. Unlike functional tests as T-cell proliferation assays to recall antigens such as cytomegalovirus and tuberculin, tests
188
J. Rambeloarisoa et al.
assessing T-cell reactivity against cryptococcus are not available. However, the correlation between results of such functional tests and clinical protection remains uncertain. The significant and sustained increase in CD4 count following effective antiretroviral therapy was probably associated with only a partial immune restitution that did not allow to avoid the occurrence of the cryptococcal medullar abscess. Finally, this case raises the question of when to stop secondary prophylaxis of cryptococcal disease after increase in CD4 cell count under antiretroviral therapy.
References 1 Powderly WG. Cryptococcal meningitis and AIDS. Clin Infect Dis 1993; 17: 837±842. 2 Gabuzda DH, Hirsch MS. Neurologic manifestations of infection with human immunodeficiency virus. Clinical features and pathogenesis. Ann Intern Med 1989; 107: 383±391. 3 Carton CA, Mount LA. Neurosurgical aspects of cryptococcosis. J Neurosurg 1951; 8: 143±156. 4 Davidson S. Cryptococcal spinal arachnoiditis. J Neurol Neurosurg Psychiatry 1968; 31: 76±80. 5 Litvinoff J, Nelson M. Extradural lumbar cryptococcosis. J Neurosurg 1978; 49: 921±923. 6 Grosse P, Tintelnot K, SoÈllner O et al. Encephalomyelitis due to Cryptococcus neoformans var. gattii presenting as spinal tumor: case report and review of the literature. J Neurol Neurosurg Psychiatry 2001; 70: 113±116. 7 Skultety FM. Cryptococcic granuloma of the dorsal spinal cord. Neurology 1961; 11: 1066±1070. 8 Su MC, Ho WL, Chen JH. Intramedullary cryptococcal granuloma of the spinal cord: a case report. Chung Hua J Hsueh Tsa Chih (Taipei) 1994; 53: 58±61. 9 Ramamurthi B, Anguli VC. Intramedullary cryptococcic granuloma of the spinal cord. J Neurosurg 1954; 11: 622±624. 10 Alajouanine TH, Houdart R, Drouhet E. Les formes chirurgicales spinales de la torulose. Torulome de la queue de cheval. Rev Neurol 1953; 88: 153±163. 11 HeÂnin D, Smith TW, De Girolami U et al. Neuropathology of the spinal cord in the acquired immunodeficiency syndrome. Hum Pathol 1992; 23: 1106±1114.
12 Woodall III WC, Bertorini TE, Bakhtian BJ et al. Spinal Arachnoiditis with Cryptococcus neoformans in a nonimmunocompromised child. Pediatr Neurol 1990; 6: 206±208. 13 Van Dellen JR, Buchanan N. Intrathecal cryptococcal lesion of the cauda equina successfully treated with intrathecal Amphotericin B. A Case Report. S Afr Med J 1980; 58: 137±138. 14 Niizuma H, Higuchi H, Tajima T. Cryptococcal granulomatous arachnitis of the spinal cordÐA case report. No Shinkei Geka 1979; 7: 805±808. 15 Ley A, Jacas R, Oliveras C. Torula granuloma of the cervical spinal cord. J Neurosurg 1951; 8: 327±335. 16 Rao SB, Rao KS, Dinakar I. Spinal extradural cryptococcal granuloma: a case report. Neurol India 1970; 18: 192±193. 17 Dromer F, Moulignier A, Dupont B et al. Myeloradiculitis due to Cryptococcus curvatus in AIDS. AIDS 1995; 9: 395±408. 18 Weinke T, RoÈgler G, Sixt C et al. Cryptococcosis in AIDS patients: observations concerning CNS involvement. J Neurol 1989; 236: 38± 42. 19 Kamezawa T, Shimozuru T, Niiro M et al. MRI of cerebral cryptococcal granuloma. Neuroradiology 2000; 42: 441±443. 20 Woods ML, MacGinley R, Eisen DP. HIV combination therapy: partial immune restitution unmasking latent cryptococcal infection. AIDS 1998; 12: 1491±1494. 21 Blanche P, Gombert B, Ginsburg C, et al. HIV combination therapy: immune restitution causing cryptococcal lymphadenitis dramatically improved by anti-inflammatory therapy. Scand J Infect Dis 1998; 30: 615±616. 22 Michelet C, Arvieux C, FrancËois C et al. Opportunistic infections occurring during highly active antiretroviral treatment. AIDS 1998; 12: 1815±1822. 23 DeSimone JA, Pomerantz RJ, Babinchak TJ. Inflammatory reactions in HIV-1-infected persons after initiation of highly active antiretroviral therapy. Ann Intern Med 2000; 133: 447±454. 24 Kunimoto DY, Chui L, Norbert E, Houston S. Immune mediated ``HAART'' attack during treatment for tuberculosis. Int J Tuberc Lung Dis 1999; 3: 944±947. 25 Race EM, Adelson-Mitty J, Kriegel GR et al. Focal mycobacterial lymphadenitis following initiation of protease-inhibitor therapy in patients with advanced HIV-1 disease. Lancet 1998; 351: 252±255. 26 Gilquin J, Piketty C, Thomas V et al. Acute cytomegalovirus infection in AIDS patients with CD4 counts above 100 106 cells/l following combination antiretroviral therapy including protease inhibitors. AIDS 1997; 11: 1659±1660. 27 Hospenthal DR, Bennett JE. Persistence of Cryptococcomas on Neuroimaging. Clin Infect Dis 2000; 31: 1303±1306.
Intramedullary Abscess Resulting from Disseminated Cryptococcosis Despite Immune Restoration in a Patient with AIDS J. Rambeloarisoaa, D. Batissea, J.-B. Thiebautb, J. Mikolc, S. Mrejena, M. Karmochkinea, M. D. Kazatchkinea, L. Weissa and C. Pikettya,* a
Service d'Immunologie Clinique, HoÃpital EuropeÂen Georges Pompidou, Paris; bService de Neurochirurgie et unite de traitement de la douleur, Fondation Rothschild, Paris; cService d'anatomopathologie, HoÃpital LariboisieÁre, Paris, France We report on a case of cryptococcal intramedullary abscess, which occurred three years after a disseminated cryptococcosis and two years after a lymph node cryptococcal recurrence in a HIV-infected patient who exhibited a long-standing immune restoration. At the time of diagnosis, CD4 lymphocyte-count was 640 106/l and HIV viral load was undetectable. Spinal involvement is rare during cryptococcosis of the central nervous system. As far as we are aware, there is only one case of proven intramedullary cryptococcal abscess reported in the literature and this case is then the second one. The significant and sustained increase in CD4 count following effective antiretroviral therapy was probably associated with only a partial immune restitution that did not allow to avoid the occurrence of the cryptococcal medullar abscess. Finally, this case raises the question of when to stop secondary prophylaxis of cryptococcal disease after increase in CD4 cell count under antiretroviral therapy. # 2002 The British Infection Society
We report on a case of cryptococcal intramedullary abscess, which occurred three years after a disseminated cryptococcosis and two years after a lymph node cryptococcal recurrence in a HIV-infected patient who exhibited a long-standing immune restoration. Human cryptoccocal infection is rare except in patients with disorders of cell-mediated immunity such as AIDS or immunosuppression following corticosteroid therapy or organ transplantation. Neurocryptococcosis is the third most common infection of the central nervous system (CNS) in the immunocompromised patients with AIDS after the encephalopathy due to HIV and encephalitis caused by Toxoplasma gondii [1,2]. Cryptococcal meningitis usually occurred in HIV-infected patients with CD4 cell counts below 100 106/l. Meningoencephalitis with or without intracranial mass lesion is the most common feature in neurocryptococcosis and spinal involvement including cauda equina is rare [3±6]. Intramedullary involvement in spinal cryptococcosis is exceptional [6±9]. * Please address all correspondence to: Dr. C. Piketty, Service d'Immunologie Clinique, HoÃpital EuropeÂen Georges Pompidou, 20 rue Leblanc, 75015 Paris, France. Tel.: 33 1 56 09 27 01; Fax: 33 1 56 09 28 59; E-mail address: [email protected] (C. Piketty). 0163±4453/02/$35.00
A 38-year-old male presented with sudden lower back pain radiating into both lower limbs. He reported no underlying traumatism or effort and no fever. He complained of permanent pain originating from the buttocks and radiating first into the right posterior thigh, the right posterior and lateral calf and the posterior and lateral part of the right knee, then later also into the left posterior thigh. Three years previously, he was found to have HIV infection, when a disseminated cryptococcosis with meningitis was diagnosed. At that time, CD4 cell count was 3 106/l. Sputum, blood and cerebrospinal fluid (CSF) cultures grew Cryptococcus neoformans serotype A. Cryptococcal antigen latex agglutination titre was 1 : 1000 in the CSF and 1 : 32000 in the serum. Brain CT scan and MRI revealed the presence of 4 lesions of cryptococcomas. Intravenous treatment with amphotericin B and flucytosin was started and switched after 3 weeks to oral fluconazole at high dosage (400 mg). Lumbar puncture performed after 5 weeks of treatment collected a CSF that was still India ink stain positive but did not grow cryptococcus any more. Sputum test became negative. He was discharged at that time and oral fluconazole was continued and decreased to 200 mg per day after 6 months. # 2002 The British Infection Society
186
J. Rambeloarisoa et al.
Two years prior to admission, after 10 months of highly active antiretroviral therapy (HAART) including stavudine, lamivudine and indinavir, he presented with an enlarged right supraclavicular lymph node with a diameter of 3 cm which was punctured. The recovered sample was Gomori Grocott stain positive and contained empty non-budding cryptococcus but cultures for fungi were negative. Thorax CT scan disclosed multiple enlarged lymph nodes of the mediastinum, one of which had 24 mm diameter, the right supraclavicular one and a centimetric excavated nodular lesion of the apex of the right lung. There was no evidence for tuberculosis or other infections. At that time, HIV viral load was undetectable since 8 months and CD4 lymphocyte-count was 175 106/l. The right supraclavicular enlarged lymph node shrank and disappeared after the reinforcement of the dosage of fluconazole from 200 mg per day to 400 mg per day. On the last admission in June 2000, the general physical examination was not remarkable apart from a non-inflammatory enlarged right supraclavicular lymph node of about 5 cm diameter. Neurologic examination revealed no abnormality of mental status or of the cranial nerves. Reflexes in the upper limbs as well as in the lower limbs were normal. No muscle weakness or sensory deficit were demonstrable. He had no sign of meningeal irritation. There was neither tenderness over the lower lumbar spine nor contraction of paravertebral muscles. Straight-leg raising was negative bilaterally. Plantar responses were flexor bilaterally. Position sense was normal in the toes. There was no cauda equina syndrome. Course in hospital. The right supraclavicular enlarged lymph node spontaneously and rapidly shrank within 48 h to 3 cm then vanished. Chest radiograph did not show evidence of mediastinal lymph nodes or pulmonary disease. Radiographs of the dorsal and lumbar spine were negative. One week after admission while the pain was soothed by morphine hydrochlorid, he stated that his legs were weak. Examination at that time revealed weakness of the proximal musculature. Both knee jerks were depressed without Babinski sign. Ankle and upper limbs jerks were normal. An hypesthesia of the lateral and anterior parts of the left calf and of the lateral and dorsal parts of the right foot was present. No cauda equina syndrome was found. A spinal MRI was accordingly performed that disclosed a round ring enhancing with gadolinium in a intramedullary lesion at the level of D11±D12, crushing the adjacent structures and reminiscent of abscess (Fig. 1). Cerebral CT scan without contrast injection was normal.
Figure 1. Gadolinium T1 weighted spinal MRI showing a ring enhancing intramedullary lesion at D11±D12 level. Sagittal section (a) Axial section (b).
Laboratory findings. The CSF contained 1 cell and 3 erythrocytes per mm3, normal glucose and 0.79 g/l protein; Gram and Ziehl-Neelsen stains were negative as well as cultures for bacteria and mycobacteria; India ink stain was positive whereas the cultures for cryptococcus were negative; DNA-polymerase chain reactions (PCR) for Herpes-viruses family, and JC papovavirus were negative. Cryptococcal antigen latex agglutination titre was 1 : 10 in the CSF and 1 : 100 in the serum. Blood cultures on fungal media were negative. CD4 lymphocyte-count was 640 106/l and HIV viral load was undetectable. Surgery. An emergency neurosurgical operation was performed. Spinous processes of D11 and D12 were removed then laminectomy and median incision followed that disclosed an abscess with minimal fair pus. The abscess and its hard capsule were removed and no drainage was necessary. Gram stain and culture failed to establish the presence of pyogenic organisms. India ink stain revealed encapsulated yeasts whereas the cultures
Disseminated Cryptococcosis in a Patient with AIDS for fungi were negative. DNA-PCR for Toxoplasma gondii was negative as well as the culture in diploid cells for cytomegalovirus. Histological examination showed an abscess wall rich in fibroblasts, myofibroblasts and inflammatory cells mixed with larged cells of the motoneuron type. The necrotic material inside of it was constituted by polymorphonuclear debris and inflammatory cells of lymphocytes and plasma cells type. By careful examination, there was no pathogenic organism apart from small round structures of slight different size, light brown by Brown and Brenn stain, hard to identify (Periodic Acid-Schiff, Alcian blue and Gomori Grocott stains were different from the control cryptococcal material). Mucicarmine stain was negative. Evolution. The day after the operation, an antituberculous treatment with an association of ethambutol, isoniazid and rifabutin was started as well as an anticryptococcal treatment with liposomal amphotericin B at the dose of 3 mg/kg per d. As the information about the presence of encapsulated yeasts in the pus was given 8 days later, the antituberculous treatment was discontinued and the anticryptococcal treatment was reinforced with flucytosin 100 mg/kg per d. The neurologic deficits of lower limbs slowly improved, in the left side more quickly than in the right one, where an amyotrophia was noted. Physical therapy was initiated. After 2 months of liposomal amphotericin B, the treatment was switched to fluconazole 400 mg/d. Spinal MRI performed 4 months after the operation disclosed a residual cyst limited to D11±D12 level and found no progressive signs. He can walk again and he actually does not need crutches any more. HAART was never interrupted and CD4 lymphocyte-count was 613 mmÿ3 at the last control. Fluconazole was continued at the same dose. Discussion. Spinal involvement is rare during cryptococcosis of the CNS [3±6]. Spinal cryptococcosis splits into 2 groups according to the location of the lesion i.e. extradural [5] and intradural. The cryptococcal infection involving the intradural part of the spinal cord divides up to extramedullary lesions, essentially arachnoiditis with or without mass lesions [3,4,10±17], and intramedullary lesions, namely granulomas and abscesses [6±9]. As far as we are aware, there is only one case of proven intramedullary cryptococcal abscess reported in the literature. It was described as a purulent material escaping from beneath the mass [7]. Our case is then the second one. Furthermore, the necrotic material was found to be constituted by polymorphonuclear debris and inflammatory cells at histological examination. Although the culture of the pus sample was negative, it is well known that India ink stain is effective to detect the presence of cryptococcus in fresh material [18] and it was indeed
187
positive in the present case. The lack of growth could be explained either by the too little available material, or by the fact that the fungus was biologically no more active in the capsule circumscribed pus collection [19]. Even in the absence of positive mucicarmine stain, another fungal aetiology like Histoplasma capsulatum or Penicillium marneffii could be excluded because the patient had never left Paris, France during his life and blood cultures on fungal media were negative. The symptoms are therefore due to inflammatory reaction by an underlying immunity restoration. Such situations were already described and reported for different infections [20±26]. Regarding the enlarged lymph node, our case somewhat fits the case reported by Blanche et al. because symptoms started long after the initiation of HAART [21]: 8 and 15 months for their two patients and 10 months for ours. In the 3 cases, the cultures for cryptococcus were negative. Yet, the similarity ends there because: (1) the enlarged lymph node in our patient relapsed 2 years later and spontaneously vanished within 5 days, moreover no anti-inflammatory drugs was prescribed in our case; (2) there was another cryptococcal location that was intramedullary and that became symptomatic by the time of the relapse. Like the sample from the supraclavicular lymph node punctured 2 years previously, the culture of the pus from the intramedullary abscess was negative for cryptococcus despite the positivity of India ink stain. Van Dellen et al. also reported on a case of cryptococcal infection diagnosed only on the basis of staining and histology [13]. Surgery with resection was successfully performed in our patient with the removal of the encapsulated pus, the postoperative course was uneventful. Grosse et al. are against such an operative process because of the risk of irreversible spinal cord damage, although they mentioned the successful case of Ramamurthi et al. [6,9]. On the cerebral level concerning granulomas, surgery is usually performed only when the absence of an associated diagnosis is doubtful according to Hospenthal et al. [27]. Considering that the enlarged lymph node relapsed for it probably holds residual cryptococcal antigens inside of it, the risk of relapse at the medullary level is conceivable if surgery was avoided. Anyway, its compressive effect on the adjacent structures, which was clinically and radiologically (MRI) evident, was an argument for surgery even that would be restricted to decompression. Although at present, the better surrogate marker for clinical protection is CD4 cell count, this case raises the question of the efficiency of immune restoration to protect against a relapse of cryptococcal infection. Unlike functional tests as T-cell proliferation assays to recall antigens such as cytomegalovirus and tuberculin, tests
188
J. Rambeloarisoa et al.
assessing T-cell reactivity against cryptococcus are not available. However, the correlation between results of such functional tests and clinical protection remains uncertain. The significant and sustained increase in CD4 count following effective antiretroviral therapy was probably associated with only a partial immune restitution that did not allow to avoid the occurrence of the cryptococcal medullar abscess. Finally, this case raises the question of when to stop secondary prophylaxis of cryptococcal disease after increase in CD4 cell count under antiretroviral therapy.
References 1 Powderly WG. Cryptococcal meningitis and AIDS. Clin Infect Dis 1993; 17: 837±842. 2 Gabuzda DH, Hirsch MS. Neurologic manifestations of infection with human immunodeficiency virus. Clinical features and pathogenesis. Ann Intern Med 1989; 107: 383±391. 3 Carton CA, Mount LA. Neurosurgical aspects of cryptococcosis. J Neurosurg 1951; 8: 143±156. 4 Davidson S. Cryptococcal spinal arachnoiditis. J Neurol Neurosurg Psychiatry 1968; 31: 76±80. 5 Litvinoff J, Nelson M. Extradural lumbar cryptococcosis. J Neurosurg 1978; 49: 921±923. 6 Grosse P, Tintelnot K, SoÈllner O et al. Encephalomyelitis due to Cryptococcus neoformans var. gattii presenting as spinal tumor: case report and review of the literature. J Neurol Neurosurg Psychiatry 2001; 70: 113±116. 7 Skultety FM. Cryptococcic granuloma of the dorsal spinal cord. Neurology 1961; 11: 1066±1070. 8 Su MC, Ho WL, Chen JH. Intramedullary cryptococcal granuloma of the spinal cord: a case report. Chung Hua J Hsueh Tsa Chih (Taipei) 1994; 53: 58±61. 9 Ramamurthi B, Anguli VC. Intramedullary cryptococcic granuloma of the spinal cord. J Neurosurg 1954; 11: 622±624. 10 Alajouanine TH, Houdart R, Drouhet E. Les formes chirurgicales spinales de la torulose. Torulome de la queue de cheval. Rev Neurol 1953; 88: 153±163. 11 HeÂnin D, Smith TW, De Girolami U et al. Neuropathology of the spinal cord in the acquired immunodeficiency syndrome. Hum Pathol 1992; 23: 1106±1114.
12 Woodall III WC, Bertorini TE, Bakhtian BJ et al. Spinal Arachnoiditis with Cryptococcus neoformans in a nonimmunocompromised child. Pediatr Neurol 1990; 6: 206±208. 13 Van Dellen JR, Buchanan N. Intrathecal cryptococcal lesion of the cauda equina successfully treated with intrathecal Amphotericin B. A Case Report. S Afr Med J 1980; 58: 137±138. 14 Niizuma H, Higuchi H, Tajima T. Cryptococcal granulomatous arachnitis of the spinal cordÐA case report. No Shinkei Geka 1979; 7: 805±808. 15 Ley A, Jacas R, Oliveras C. Torula granuloma of the cervical spinal cord. J Neurosurg 1951; 8: 327±335. 16 Rao SB, Rao KS, Dinakar I. Spinal extradural cryptococcal granuloma: a case report. Neurol India 1970; 18: 192±193. 17 Dromer F, Moulignier A, Dupont B et al. Myeloradiculitis due to Cryptococcus curvatus in AIDS. AIDS 1995; 9: 395±408. 18 Weinke T, RoÈgler G, Sixt C et al. Cryptococcosis in AIDS patients: observations concerning CNS involvement. J Neurol 1989; 236: 38± 42. 19 Kamezawa T, Shimozuru T, Niiro M et al. MRI of cerebral cryptococcal granuloma. Neuroradiology 2000; 42: 441±443. 20 Woods ML, MacGinley R, Eisen DP. HIV combination therapy: partial immune restitution unmasking latent cryptococcal infection. AIDS 1998; 12: 1491±1494. 21 Blanche P, Gombert B, Ginsburg C, et al. HIV combination therapy: immune restitution causing cryptococcal lymphadenitis dramatically improved by anti-inflammatory therapy. Scand J Infect Dis 1998; 30: 615±616. 22 Michelet C, Arvieux C, FrancËois C et al. Opportunistic infections occurring during highly active antiretroviral treatment. AIDS 1998; 12: 1815±1822. 23 DeSimone JA, Pomerantz RJ, Babinchak TJ. Inflammatory reactions in HIV-1-infected persons after initiation of highly active antiretroviral therapy. Ann Intern Med 2000; 133: 447±454. 24 Kunimoto DY, Chui L, Norbert E, Houston S. Immune mediated ``HAART'' attack during treatment for tuberculosis. Int J Tuberc Lung Dis 1999; 3: 944±947. 25 Race EM, Adelson-Mitty J, Kriegel GR et al. Focal mycobacterial lymphadenitis following initiation of protease-inhibitor therapy in patients with advanced HIV-1 disease. Lancet 1998; 351: 252±255. 26 Gilquin J, Piketty C, Thomas V et al. Acute cytomegalovirus infection in AIDS patients with CD4 counts above 100 106 cells/l following combination antiretroviral therapy including protease inhibitors. AIDS 1997; 11: 1659±1660. 27 Hospenthal DR, Bennett JE. Persistence of Cryptococcomas on Neuroimaging. Clin Infect Dis 2000; 31: 1303±1306.