Recurrent Cryptococcal Meningitis in a Late Presenter of HIV: A Rare Case Report and Review of Literature

Background

Cryptococcus neoformans is a saprophytic yeast with a worldwide distribution, associated with a wide range of clinical conditions from harmless colonization to disseminated disease [1]. Cryptococcal meningitis in immunosuppressed patients such as people living with HIV, and eventually in late presenters, is one of the major opportunistic infections causing morbidity and mortality [2]. Late presenters are defined as patients presenting for care with a CD4 cell count below 350 cells/mm3 or with an AIDS-defining event, regardless of the CD4 cell count43]. Most HIV-infected patients diagnosed with C. neoformans infection have a CD4+ T cell count below 100 cells/µl [4]. The highly active antiretroviral treatment and the primary prophylaxis in newly diagnosed people living with HIV (PLWHIV) with a low CD4 cell count has significantly reduced the incidence rate of cryptococcal meningitis [5].

The secondary antifungal prophylaxis following diagnosis and treatment of cryptococcal meningitis is also vital to prevent relapse of the infection [6]. Studies have shown that the annual relapse rates for cryptococcal disease in developed countries is <5% during maintenance fluconazole treatment [7]. There is little information on CM relapse in the era of HAART and widely available fluconazole [7–9]. Relapse of CM due to C. neoformans non-resistant to fluconazole isolates is rare [7]. Initial induction treatment with fluconazole monotherapy, development of fluconazole resistance, immune reconstitution inflammatory syndrome (IRIS), and discontinuation of maintenance therapy while CD4+ T cell count remains <100 cells/ mL are possible factors associated with a relapse [8,9]. Since 2000, when fluconazole became widely available in the public sector through the Diflucan donation program, we have noted symptomatic relapses of CM due to fluconazole-resistant strains of C. neoformans. Furthermore, surveillance data from South Africa have revealed an increase in the percentage of C. neoformans isolates with fluconazole resistance (MIC, ≥64 µg/mL), from 0% in 1999 (when 233 isolates were evaluated) to 12.7% in 2004 (when 63 isolates were evaluated).

The aim of this case presentation is to report the diagnostic approach, clinical progress, and treatment of relapse of cryptococcal meningitis in a late presenter PLHIV after the increase of CD4 cell count, achievement of undetectable viral load, and disruption of secondary antifungal prophylaxis. This case presentation highlights the importance of timely diagnosis of HIV infection to reduce the frequency of AIDS-defining conditions that increase the risk of unfavourable clinical outcome. The management of PLHIV with advanced HIV disease is often associated with difficult medical treatment decisions, and the risk of complications is increased.

Case Report

FIRST EPISODE OF CRYPTOCOCCAL MENINGITIS:

A man in his 40s was hospitalized in the HIV unit due to fever of 3 weeks, persistent headache and vomiting for 15 days, and disorientation to time and place for 24 hours. Neck rigidity and Kernig’s and Brudzinski’s sign were present. No episodes of seizures were mentioned. In examination of cranial nerves, diplopia was detected. During Romberg test, the patient had difficulty maintaining his balance with closed eyes. No motor weakness, cerebellar ataxia, or sensory impairment were noted. Tendon reflexes were normal and symmetrical, without any pathological reflexes. The patient had a weight loss of 10 kilograms during last 6 months. The clinical examination revealed fungal stomatitis. His lymph nodes, liver, and spleen were not palpable. There was no relevant past medical history. He smoked cigarettes (30 pack years) and consumed alcohol and non-intravenous recreational drugs.

He was diagnosed with human immunodeficiency virus infection, with a CD4 cell count of 22 cells/mm3. Blood count revealed normocytic normochromic anemia, thrombocytopenia, and white blood cell (WBC) count of 4100 cells/mm3 (2020 neutrophils/mm3 and 1020 lymphocytes/mm3). A chest X-ray was normal. The results of brain contrast-enhanced computed tomography and MRI were normal. No mass, hemorrhage, major vessel vascular territory infarct, intra- or extra-axial collection, or hydrocephalus were found. Basal ganglia and posterior fossa structures were normal. Lumbar puncture was conducted. In cerebrospinal fluid (CSF), 30 mononuclear cells were observed, the glucose level was 79 mg/dL (serum glucose level 98 mg/dL), and the protein level (94 mg/dL) was elevated. CSF cultures were negative. Herpes simplex virus, varicella-zoster virus, Epstein-Barr virus, and cytomegalovirus DNA were negative. Cryptococcal antigen was detected in CSF and serum. Cryptococcus neoformans was isolated in CSF with India ink stain (Figure 1).

Firstly, he received empiric antibiotic treatment with ceftriaxone, vancomycin, ampicillin, and acyclovir. After the results were obtained from CSF examination, antifungal treatment was initiated. Induction therapy consisted of liposomal amphotericin B (3 mg/kg/24 hours intravenously) and flucytosine (25 mg/kg/24 hours intravenously) for 14 days. Consolidation treatment was continued with fluconazole for 8 weeks. His clinical condition and laboratory exams were improved in 1 week. Highly active antiretroviral treatment (tenofovir disoproxil fumarate, emtricitabine, and dolutegravir) began 4 weeks after the initiation of antifungal treatment. The antiretroviral regimen with high genetic barrier and without interactions with antifungal treatment was chosen. Primary prophylaxis with azithromycin for mycobacterium avium complex (MAC) and cotrimoxazole for Pneumocystis jirovecii (PJ) and secondary prophylaxis with fluconazole after completed consolidation treatment were initiated. The patient did not suffer from any neurological sequels after the first CM episode.

RELAPSE OF CRYPTOCOCCAL MENINGITIS:

One year later, the patient was in good clinical condition without findings in clinical examination and laboratory exams. He had excellent compliance with his treatment. The CD4 cell count was increased (365 cells/mm3), and HIV RNA was below level of detection; therefore, primary prophylaxis for MAC and PJ and secondary prophylaxis for Cryptococcus neoformans were interrupted. However, 3 months later, he had 1 episode of partial seizure with secondary generalization, followed by a short-term memory loss. He had persistent headache without fever, neck rigidity, or other neurological symptoms. No diplopia was observed. Evaluation of cranial nerves, tendon reflexes, cerebellar examinations, and sensory and motor system was normal. No weight loss was documented. Peripheral lymph nodes, spleen, and liver were not palpable and no exanthema or en-anthem was found. Brain MRI showed a focal region in the right frontal-parietal region with increased intensity in T2 and FLAIR images, which was attributed to meningoencephalitis (Figure 2). No surrounding edema or choroid plexitis was reported. Perivascular spaces were normal. A new lumbar puncture was conducted and Cryptococcus neoformans non-resistant to fluconazole was isolated. In CSF, 10 mononuclear cells were observed, the glucose level was normal, and the protein level (68 mg/dL) was elevated. Antiepileptic treatment with levetiracetam and similar antifungal treatment with the previous was initiated. He received induction treatment with liposomal amphotericin and fluconazole, consolidation treatment with fluconazole, and secondary prophylaxis with fluconazole. He had another episode of partial seizure lasting 5 minutes without memory loss 4 days after the relapse. Lacosamide was added to the antiepileptic treatment. During the second week of antifungal treatment, an electroencephalogram was performed, showing slow activities of theta and delta waves, mainly in the right hemisphere and occipital electrodes (Figure 3). During the last 12 months, he has not had any episodes of seizure or other neurological symptoms. He continues receiving secondary prophylaxis with fluconazole, while there are insufficient data in the literature on CM relapses.

Discussion

Studies have reported that the annual relapse rates of cryptococcal disease in developed countries with amphotericin-based regimens as initial therapy is <5% during maintenance treatment with fluconazole [7]. However, a South African study reported a significant proportion (23%) of cases with relapse of cryptococcal meningitis [10]. Almost half (43%) of these cases occurred in patients who did not receive secondary prophylaxis with fluconazole [10]. The clinical outcome was poor for many patients and the mortality rate was approximately 30% [10]. Forty-five percent of relapse cases were due to IRIS at a median of 61 days after ART initiation, and 12% in patients prior to ART despite treatment with fluconazole [10]. Patients presenting with IRIS in this study had less severe clinical disease, lower fungal burden, and a lower mortality (13%) [10]. In another retrospective study of Chinese patients with HIV-associated CM, 25% experienced relapse of CM [11]. Among ART-experienced patients with CM relapse, more than 50% had poor compliance with antiretroviral and antifungal treatment. In our case, the patient had 100% compliance with his treatment, the HIV viral load was undetectable, and the CD4 cell count was significantly increased (>200 cells/mm3).

Several studies have reported potential risk factors that could lead to relapse of CM among people living with HIV. Fluconazole monotherapy as induction therapy compared with amphotericin-based regimens and no initial treatment with flucytosine have been associated with higher relapse rate and mortality risk [12,13]. Treatment with itraconazole has been also documented as a risk factor for relapse [13]. Inadequate antifungal maintenance treatment, poor compliance with secondary prophylaxis, inappropriate ART initiation timing, and low baseline CD4+ cell count are the main reasons for CM relapse [14]. A systematic review including 4995 clinical isolates of Cryptococcus species from 3210 patients, with 248 (5%) of the isolates from relapsing episodes, showed 24.1% fluconazole resistance in relapse cases [15]. Not all cryptococcal meningoencephalitis cases have positive imaging findings, and normal brain imaging does not exclude the diagnosis [16,17]. Approximately 21-27% of cases have typical cryptococcal meningoencephalitis findings on MRI, which include the dilated Virchow-Robin spaces, pseudocysts, cryptococcoma, or leptomeningeal- or parenchymal-enhancing lesions and hazy brain base [16,17].

Some studies separated CM relapse cases into IRIS and non-IRIS cases [12]. IRIS cases are defined when there is a micro-biologically confirmed first episode of CM with resolution of symptoms before starting ART, excellent adherence to antifungal and antiretroviral treatment, recurrence of CM symptoms after initiation of ART, and no other possible diagnosis such as fluconazole resistance [12–14]. Diagnosis of IRIS-associated relapse of CM cannot be excluded in our case report. However, our patient’s first episode improved with antifungal treatment before ART initiation, and relapse occurred 1 year after, with 100% compliance, undetectable viral load, and CD4 cell count over 300 cells/mm3. Clinical manifestations of IRIS symptoms typically present a few days to 6 months after ART initiation; therefore, an IRIS diagnosis may be unlikely [16]. No resistance to fluconazole was reported, and CM symptoms occurred 3 months after discontinuation of fluconazole, highlighting the need for evaluation of duration of secondary prophylaxis.

A prospective observational study conducted in Cape Town, South Africa indicated that CM relapse occurred in 23.0% of all cases of HIV-associated CM [17]. A randomized controlled trial showed that relapse of CM is associated with initial timing of ART among HIV-infected patients, while the relapse rate for HIV-associated CM was 2% (95% CI, 1–7%) in those initiating ART at 2–4 weeks after initiating antifungals, and 9% (95% CI, 4–16%) in those initiating ART at 4–6 weeks after initiating antifungals, P=0.06 [18]. A systematic review comparing the outcomes of early and delayed initiation of ART in HIV-infected patients with CM showed that early ART initiation can reduce CM relapse (RR with 95% CI: 0.27, 0.07–1.04; low-certainty evidence) [19]. Our patient initiated HAART 4 weeks after initiation of antifungal treatment, which is associated with low rates of relapse. A randomized controlled trial suggested that for patients with CSF white-cell counts <5 cells/mm3 at baseline, the 26-week mortality rate in the early ART initiation group (1–2 weeks after diagnosis) was significantly higher than in the delayed ART initiation group (5 weeks after diagnosis) (hazard ratio with 95% CI, 1.73, 1.06–2.82, P=0.03) [20]. Based on the guidelines of the European AIDS Clinical Society (EACS), initiation of ART should be delayed for at least 4 weeks, while the WHO recommends a delay of 4–6 weeks, and some specialists recommend a delay of 6–10 weeks in severe cryptococcal meningitis [21].

Secondary prophylaxis with fluconazole should be maintained at least 12 months and stopped if CD4 count >100 cells/μL and the HIV viral load is undetectable for 3 months [21]. These guidelines were followed in our case. In a prospective, multi-center, randomized study with PLHIV successfully treated for acute cryptococcal meningitis and receiving secondary prophylaxis with fluconazole, patients were randomized to continue or discontinue secondary prophylaxis when the CD4 cell count had increased to >100 cells/μL and an undetectable HIV RNA level had been sustained for 3 months [22]. At a median of 48 weeks after randomization, there were no episodes of cryptococcal meningitis in either group [22].

Shelburne et al found that PLHIV with IRIS (n=18) were more likely to have initiated HAART (100%) with significant virological response and increase of CD4 cell count (median increase: 93 cells/μL) compared with PLHIV who had relapse (n=12, 33%; 4 cells/μL) [23]. An observational prospective study that enrolled 724 PLHIV Ugandans with cryptococcal meningitis concluded that those with poor immune reconstitution were more often in relapse than those with IRIS with lower CSF white blood cells (WBC) and blood CD4 counts [24]. Early mortality among those with IRIS (14.3%) and relapse (12.3%) was similar at day 14 [24]. However, by 18 weeks, IRIS mortality remained stable and mortality in those with relapse had increased to 30.9% [24]. Li et al analyzed a cohort of 348 PLHIV in which 25% of patients with HIV-associated cryptococcal meningitis experienced relapse [13]. Multivariate logistic regression found that CD4 counts of less than 20 cells/μL, previous ART experience, and fewer than 4 weeks from symptom onset to presentation were significantly associated with the occurrence of relapse [13]. Studies have shown that median CSF WBC count was higher in those with IRIS compared to those with relapse [24,25]. Although an elevated CSF WBC count should raise suspicion for IRIS, it is not confirmatory [25]. In our case, during the relapse episode the patient had initiated HAART and achieved virological suppression and CD4 cell count above 300 cells/mm3, while the CSF WBC count was low. Thus, IRIS cannot be excluded.

Conclusions

Relapse of previously treated cryptococcal meningitis is a serious clinical condition in people living with HIV, leading to increased risk of morbidity and mortality. Relapse may occur in cases with excellent adherence to antiretroviral and antifungal treatment, choice of appropriate antifungal drugs, without resistance in fluconazole, and more than 6 months after ART initiation. Further studies are needed for the evaluation of parameters such as duration of secondary prophylaxis and options for induction and consolidation treatment.