Cryptococcal meningitis: a review for emergency clinicians—comment

Abstract

Reading the article by Fischer et al. entitled “Cryptococcal meningitis: a review for emergency clinicians”, we found several inaccuracies that deserve to be clarified so as not to give misleading information to clinicians [1]. First, although traditionally distinguished in two species (i.e. Cryptococcus neoformans and C. gattii) as reported in the paper, it is now recognized that the Cryptococcus genus responsible for human infections is more complex and heterogeneous than previously thought. Indeed, within the C. neoformans and C. gattii species complex twelve molecular lineages and eight species (C. neoformans, C. deneoformans, C. gattii, C. deuterogattii, C. bacillosporus, C. decagattii, C. tetragattii, unnamed new species) have been now recognized. These species have different geographical distribution and hit different hosts [2]. Second, the correct global average prevalence of cryptococcal antigenemia among HIV-positive patients with a lymphocyte CD4 + cell count of less than 100/μL should be 6% and not 60% as reported in the article; however, it ranges from 2 to 12.2% according to different studies. Third, the sentence “cryptococcal meningitis (CM) may be the initial AIDS-defining illness in up to 84% of patients in low-income settings and 60–70% in resource-rich environments” is quite misleading and inaccurate. It is misleading because it refers to the percentage of patients diagnosed as affected by CM in whom the disease is the AIDS-defining illness. However, the true incidence of CM as and AIDSdefining illness has been reported to be 0.05 per 100-person years in the United States and Canada and 1.1 to 1.6% in resource-limited settings [3]. Moreover, in the paper by Mirza et al. cited by Fischer in their review, CM was the AIDS-defining illness in 39% of subjects in resource-rich environments and not, as indicated, 60–70%. The description of clinical presentation of CM is oversimplistic and cannot be based on a single report from China dealing with HIV-negative patients. As shown in Table 1, the frequency of several clinical features appears either overestimated (fever, ocular involvement, seizure) or underestimated (neck stiffness). In addition, the incidence of concomitant opportunistic infections (up to 52%) reported by the authors seems to be an overestimate also in the population of HIVinfected subjects. Regarding the microbiological diagnosis of CM, we do not agree with the affirmation that India ink stain sensitivity on cerebrospinal fluid is “largely dependent on the microbiologist’s expertise”. The India ink is a simple staining method that can be employed and read also by inexperienced clinicians. As a matter of fact, sensitivity is mostly influenced by the fungal burden being, therefore, higher among HIV-positive subjects (ranging from 72 to 93.8%) and lower among, for instance, transplant patients (50–77%) [4]. Moreover, the percentage of positive blood cultures depends on the population under study ranging from 47% in HIV-positive patients outside Africa to 70.8% in HIVpositive patients in Africa but being only 26.8% among HIVnegative subjects [4]. As a matter of fact, CSF CrAg works also better when compared with the FilmArray PCR meningitis/encephalitis (ME) panel (Biofire, Salt Lake City, UT), a multiplex PCR assay that detects 14-meningitis causing pathogens including Cryptococcus spp. that is adopted in many laboratories in high-income countries for the rapid diagnosis of meningitis/ encephalitis in the Emergency Department. In a recent study by Van et al., the ME panel was able to detect correctly 84.2% cryptococcal-positive specimens vs 73.7% by culture and 97.4% by CrAg [5]. Not surprisingly, as reported for India ink staining, the ability to detect Cryptococcus by ME panel depends largely on fungal burden as negative tested samples showed a median CrAg titre of 1:128. * Spinello Antinori [email protected]