Multiple cerebral cryptococcomas associated with immune reconstitution in HIV-1 infection

Probably as a result of restored immunity, HIV-1-infected patients receiving highly active antiretroviral therapy (HAART) may experience atypical clinical manifestations of cryptococcosis, as well as other opportunistic disorders. We describe here a new presentation of multiple cerebral cryptococcomas occurring in two patients after the introduction of HAART. Case 1 In September 2001, a 32-year-old heterosexual woman had been contemporarily diagnosed as having cryptococcal meningitis and HIV infection. Her CD4 cell count was 17 cells/μl and her HIV-RNA level was 375 000 copies/ml. A brain magnetic resonance imaging (MRI) scan was negative. She was treated with a 21-day course of liposomal amphotericin B (3 mg/kg a day); thereafter, she started both fluconazole secondary prophylaxis against cryptococcal infection and antiretroviral therapy with HAART (nelfinavir, zidovudine and lamivudine). The cryptococcal antigen titre dropped to 1 : 1 both in blood and cerebrospinal fluid (CSF) from an initial value of 1 : 4096 and 1 : 1024 in blood and CSF, respectively. Seven months later (CD4 cell count 220 cells/μl; HIV-RNA level < 40 copies/ml) she complained of fever, vomiting, and headache. On admission, brain MRI disclosed multiple cortical masses and a diffuse meningeal enhancement (Fig. 1a). CSF analysis showed 50 cells/μl, a protein content of 85 mg/dl, and a glucose level of 45 mg/dl.Fig. 1. Gadolinium-enhanced T1-weighted magnetic resonance images of the brain.: (a) Multiple hypertense nodules with surrounding oedema, leptomeningeal enhancement and ventricular enlargement. (b) Six months after the onset of the disease: a reduction in the number and size of the pre-existing nodules, normal ventricular system and no leptomeningeal enhancement.Cultures from CSF failed to yield Cryptococcus spp. and other pathogens; additional laboratory tests excluded toxoplasmosis, tuberculosis, bacterial cerebral abscesses, syphilitic cerebral gummas, neuroborreliosis, and lymphoma. Three consecutive serum and CSF cryptococcal antigen titres were 1 : 16. The patient was thus retreated with a 15-day course of liposomal amphotericin B, and continued to receive previous antiretroviral therapy, whereas corticosteroids were not administered. A 6-month follow-up MRI showed a significant regression of the known cerebral lesions (Fig. 1b). Case 2 A 43-year-old homosexual man with HIV-1 infection developed cryptococcal meningitis in December 2001. His CD4 cell count was 27 cells/μl with a viral load of 108 574 copies/ml. Brain MRI did not show any focal lesion. Conventional amphotericin B and HAART (zidovudine, lamivudine, and nelfinavir) were initiated. After 4 weeks the patient was discharged and amphotericin B (cumulative dose of 1.6 g) was switched to fluconazole as secondary prophylaxis. The cryptococcal antigen titre dropped to 1 : 1 in serum and CSF from initial values of 1 : 2048 and 1 : 512 in blood and CSF, respectively. Six months later, the patient complained of fever and headache; the CD4 cell count was 205 cells/μl and the HIV-RNA level was less than 40 copies/ml. A new brain MRI revealed multiple cortical and subcortical lesions. Examination of CSF disclosed 20 cells/μl and normal glucose and protein levels. Both blood and CSF cultures were negative as well as polymerase chain reaction analysis for herpes simplex virus, Epstein–Barr virus, cytomegalovirus, JC virus, varicella zoster virus, and mycobacteria. Cryptococcal antigen titres were 1 : 8 in CSF and 1 : 32 in serum. The patient received a 15-day course of conventional amphotericin B whereas the antiretroviral regimen was left unchanged. A brain MRI after 6 months revealed a marked regression of the cerebral lesions. Immune reconstitution after HAART may cause undesirable effects, the so-called paradoxical reactions, or better the immune reconstitution syndromes that manifest as focal lymphadenitis in patients treated for cryptococcosis, a worsening of pulmonary infiltrates and intracranial tuberculomas in patients treated for tuberculosis, vitritis in patients treated for cytomegalovirus retinitis, or herpes zoster and acute hepatitis in patients who had previous varicella zoster virus and hepatitis B or C virus infection [1–9]. Discussion concerning the pathogenesis of these entities is still controversial, and two complementary scenarios have actually been proposed [3,10]. The first is that immune reconstitution syndrome (IRS) represents the emergence of a previously quiescent or incubating infection, precipitated by HAART-induced immunological changes. Generally, it occurs soon after the institution of HAART (within 1–4 weeks), and the pathogens are often recovered and cultured from the site of infection. The second is that IRS is the consequence of increased levels of circulating immune cells and the accessibility of immune cells to sites of infection where low amounts of pathogen antigens are still present. In this case, IRS appears after a longer time (> 2–4 months) of ongoing antiretroviral therapy, and the opportunistic agent can not be identified by cultures, but only revealed on special histological specimens, suggesting that microrganisms could replicate at a very low level or even be dead. Our two cases seem to belong to this second pathogenetic mechanism. To date, definitive diagnostic criteria and therapeutic guidelines for IRS have still not been achieved. In our patients, the diagnosis was sustained by the significant immune recovery associated with the slightly increased level of cryptococcal antigen, the negative results of all other studies, and the good clinical outcome without relapses during the follow-up period. A brain biopsy was not obtained because of the patient's refusal; however, in a previous reported case [11] stereotactic biopsy cultures failed to yield any pathogens, and Cryptococcus was only evidenced in the histological specimen. Similar results were described for other cryptococcal-related immune reconstitution syndromes [2,12]. With regard to therapy, it is difficult to postulate whether specific antimicrobial therapy rather than the continuation of HAART alone should be the treatment of choice in such cases. Given the severity of the disease, we decided to initiate a full course of antifungal therapy. As reported by some authors, it is possible that corticosteroid therapy alone may be worthwhile and effective [12,13].