The Epidemiological, Clinical, Mycological, and Pathological Features of Rhino-cerebral Mucormycosis: A Systematic Review

Cerebral mucormycosis (CM) is a life-threatening manifestation of mucormycosis, an angioinvasive fungal infection caused by Mucorales. We sought to systematically review all available case reports to describe epidemiologic features, clinical manifestations, predisposing factors, and diagnostic and treatment strategies of CM. A systematic search was conducted using a combination of the following keywords: "Mucor", "Zygomycetes", "mucormycosis", "cereb*", "brain", "central nervous system", and "intracranial", separately and in combination until December 31st 2018. Data sources included PubMed, Scopus, EMBASE, Web of Science, Science Direct, and Proquest without limiting the time of publication. We included 287 articles corresponding to 345 cases of CM. Out of the 345 cases, 206 (60%) were male with a median age of 44 years; 130 (38%) were reported from North America; 87 (25%) from Asia; and 84 (24%) from Europe. The median time from onset of symptoms to presentation was 3-7 days (65/345, 65%). The highest mortality was observed among patients with diabetes mellitus (P=0.003). Debridement of infected brain tissue was associated with improved survival in CM cases (OR 1.5; 95% CI 01.3-1.8; P<0.0001). The use of liposomal amphotericin B (L-AMB) was significantly associated with patients' recovery (OR 2.09; 95% CI 1.2-3.4; P=0.003). The combination of L-AMB and posaconazole (12.5%) was more effective than the monotherapy treatment of CM cases (P=0.009). Clinicians should consider DM as an important risk factor for CM. Moreover, surgical debridement and antifungal combination therapy could be an effective approach in the management of CM patients.


Introduction
Mucormycosis is a devastatingly angioinvasive fungal infection caused by filamentous fungi belonging to the order Mucorales (1). Mucormycosis is being sporadically reported; however, its aggressive nature, the increasing number of susceptible individuals, the challenge in diagnosis, and associated high mortality would be possible reasons of great concern to the epidemiologists and clinicians (2)(3)(4)(5). Some of the  (4,6,7). Mucor-mycosis invades various tissues and can present as rhino-orbito-cerebral, pulmonary, cutaneous, gastrointestinal, and disseminated infection (4). In contrast, rhino-orbito-cerebral mucormycosis (ROCM) is the most common presentation among diabetic patients; the HM patients and HSCT recipients usually present with pulmonary symptoms (4,8). Cerebral mucormycosis (CM), especially in patients with diabetic ketoacidosis (DKA) or profound neutropenia, can develop both via direct invasion from sino-orbital infection or hematogenously from an infected organ such as lungs (9,10). Currently, a paucity of information regarding clinical manifestations and predisposing factors affect the diagnosis and treatment of CM (9). We systematically reviewed previously reported cases of mucormycosis to provide a more clear perspective on epidemiologic features, clinical manifestations, radiographic features, predisposing factors, diagnosis, treatment modalities, and outcome of CM.

Material and Methods
In this study, we followed the preferred reporting items for systematic reviews and meta-analysis (PRISMA) guidelines (10). We designed the clinical question according to the PICO process and selected keywords to cover the question. Figure 1 depicts a flow diagram describing the different phases of this systematic review. To find related articles, a comprehensive literature review was conducted in global databases such as PubMed, Scopus, EMBASE, Web of Science (WOS), Science Direct, and Proquest without limiting the publication time until 31 st December 2018. The search process was accomplished using the following keywords: (((((mucor) OR mucor*) OR zygomycosis) OR zygomyc*)) AND (((((cerebral) OR brain) OR central nervous system) OR intracranial) OR cereb*) both separately and in combination. We also did a manual search in google using the mentioned keywords and selected any additional relevant studies.
Two researchers (M.S and S.K) independently searched the mentioned databases and screened articles. Only case reports or case series in the English literature were included. CM was defined as an invasive mold infection with histopathological or mycological evidence plus the presence of radiological features of brain involvement. Review articles, conference proceedings, and other studies that did not fulfill the inclusion criteria were not included.
The quality of the included articles was assessed using strengthening the reporting of observational studies in epidemiology (STROBE) checklist; accordingly, low quality and irrelevant papers (<10.5) were excluded (11). The quality assessment of the studies was performed independently by two reviewers (S.K and S.M), and disagreements were resolved by the third researcher (M.S).
The rticles by their titles were screened and the full text of relevant articles were reviewed. Afterward, the data containing the name of the first author, year of the publication, country, age, gender, comorbidities, background diseases, clinical symptoms, sources of infection, duration of symptoms, prescribed drugs, surgical procedures, follow up duration, samples types, diagnosis methods, CNS imaging results, antifungal treatment drugs, and outcome were extracted.
A meta-analysis using the STATA software ver. 11 (STATA Corp, College Station, TX, USA) was performed. Proportions were presented as percentages and continuous data as means with standard deviations (SD). A multiple logistic regression was done, and the results were reported as odds ratios (ORs) with a 95% confidence interval (CI). P-value≤0.05 was considered statistically significant.

Results
In the initial search, 1462 articles were retrieved; 1175 articles were excluded according to the exclusion criteria; 287 articles containing 345 cases of CM were included in the meta-review ( Figure 1).

Epidemiological Features
Two-hundred and six (60%) out of 345 reported CM cases were males with a median age of 44 years (range, 2 months-89 years). Cases were reported from 47 countries, with 130 (38%) from North America, 87 (25%) from Asia, and 84 (24%) from Europe ( Figure  2). Epidemiological features and underlying conditions plus results of the multivariable logistic regression in cases who survived or died following CM are shown in Table 1. Seventy-nine (23%) reported cases were immunocompetent and did not receive immunosuppressive agents. DM and HMs were reported in 176 (51%) and 93 (26.9%) of the reviewed cases, respectively (Table 1). Among HMs, acute lymphoblastic leukemia (ALL) was reported in 25 (7.2%) cases (P<0.001). We divided the published articles according to the publication year (before and after 2000); as such 56% and 47% had DM (P=0. 132), while 16% and 30% had HMs (P=0.04) before and after the year 2000, respectively. Hypertension, chronic kidney disease, and injection drug use (IDU) were reported in 34 (9.8%), 18 (7.5%), and 18 (5.2%) cases, respectively. IRANIAN JOURNAL OF PATHOLOGY

Laboratory Variables
From blood samples, the white blood cell (WBC) count was reported in 144 (42%) of 345 reported cases. WBC count was <3.5x10 3 microL in 31 (74%) and >12x10 3 microL in 72 (20%) of 144 cases. Information about cerebrospinal fluid (CSF) variables was not reported in all reported cases. WBC count in CSF was higher than 3 cells/microL in 37 (95%) out of 39 cases, low glucose (lower than 40 mg/dl) was reported in 5 (13%) out of 39 cases, and high CSF protein (higher than 50 mg/dL) was reported in 25 (57%) out of 44 reported cases. The antifungal therapy with amphotericin B-deoxycholate (AmB-d), liposomal amphotericin B (L-AMB), or posaconazole was documented in 186 (54%), 87 (25%), 37 (11%) patients, respectively. The survival OR was 2.09 in patients who received L-AMB (95% CI 1.2-3.4; P=0.003). Combination therapy with L-AMB and posaconazole was reported in 45 (13%) out of the 348 cases. There was a significant relation (P=0.009) between survival and L-AMB-posaconazole combination therapy (12.5% survivors vs. 5.3% dead among cases who had received the combined antifungal regimen; this relationship was not significant for other combined antifungal regimens). Table 3. Prescribed antifungals in cases of cerebral mucormycosis plus the results of the multivariable logistic regression in cases who survived or died following cerebral mucormycosis.

Discussion
Despite the clinical significance of cerebral mucormycosis (12), comprehensive data regarding its epidemiology and clinical manifestations are lacking. Accordingly, the present systematic review was conducted to fill this gap of knowledge. CM is globally distributed, and reports of this infection have been published from 47 countries, with the majority being from North America (130/345, 37.68%). This finding is similar to that of Kerezoudis et al. who reported a meta-analysis on isolated cerebral mucormycosis (13) but different from the results of a recent systematic review of mucormycosis case reports that found the majority of reports from Europe (290/851, 34%) followed by Asia (267/851, 31%) (14). This difference might have emanated from different inclusion criteria used in the studies or due to a higher incidence of CM in North America in comparison to other forms of mucormycosis. Furthermore, because the definitive diagnosis of mucormycosis is challenging and more than 90% of cases might have been confirmed by post-mortem examinations (15,16), in resource-limited and developing countries, where access to diagnostic methods and devices is limited, diagnosis and reporting of CM cases might be difficult. These limitations, in turn, result in underreporting of CM in some geographical regions.
CM can affect humans in all age groups. In the present study, patients' age spanned a wide range: from 2 months to 89 years with a median of 44.2 years. In another systematic review of isolated CM cases, the mean age of 28 years was reported (13). Moreover, similar to the results of other systematic reviews that reported 57% (13), 58% (17), and 63% (14) of patients being males, in the present study, males (57.9%) constituted a larger proportion of CM cases. Future studies need to be done to determine the exact role of the male gender as a potential risk factor for mucormycosis. Mucormycosis is generally known as an infection affecting patients with underlying conditions. Diabetes has been and is still the most IRANIAN JOURNAL OF PATHOLOGY common risk factor for mucormycosis in resourcelimited countries (16).
In the present study, 176 (51%) cases of CM were diabetic and 14 (23.8%) had diabetic ketoacidosis. In recent years, we have witnessed the rise in the number of patients with immunosuppressed conditions such as solid organ/hematopoietic stem cell transplant recipients and those under chemotherapy; these patients are known to be highly vulnerable to mucormycosis (16). In agreement with this statement, a large proportion (n=96, 27.8%) of the reported CM cases belong to these groups.
In addition to the above-mentioned underlying conditions, a divergent set of other diseases have also been reported among CM patients (Table 1). Surprisingly, in the current study, 22.9% of CM patients were immunocompetent and not receiving any immunosuppressant drugs. A similar finding was reported in another systematic review of mucormycosis cases (14). Accordingly, CM should be taken into account in all patient groups, even among healthy individuals with suspicious signs and symptoms. Orbital (46.4%) and neurological (41.2%) symptoms were the most common presentations in CM patients followed by headache (35.9%). In a review of isolated CM cases (13), altered mental status (54%) and headache (51%) were reported as the common signs and symptoms, which differ from our findings. However, the unspecific signs and symptoms similarly add to the complexity of the diagnosis of CM. Orbital involvement was the most common and interesting symptom reported previously (18), indicating the need to consider cerebral mucormycosis in patients with orbital symptoms.
As mentioned previously, neurological symptoms are often associated with other extra-cerebral symptoms (19). Although headache was a common clinical sign, fever was reported in just over one-fifth of cases. However, in some studies, up to 50% of patients had a fever (19). Overall, fever does not appear to be a reliable clinical sign to diagnose or rule out this invasive infection. In Kursun et al. study, cavernous sinus involvement was the most common sign of cerebral imaging in 27 patients followed by brain infarction, while in our study, abscess-like lesions and contrast-enhancing were the most common (18).
Rhizopus arrhizus was dominant Rhizopus spp. identified to cause CM in this study. Similarly, other reviews on different patient groups have reported Rhizopus spp. to be the leading etiologic agents of mucormycosis (13,14,17,20). Despite this agreement between various studies, results regarding the second and third-ranked causes of the disease are controversial. Although in the present study, Lichtemia spp. and Rhizomucor spp. were respectively the second and third common causes of CM, studies focused on all forms of mucormycosis (14,17,20) and even another review of isolated CM (13), have reported Mucor spp. to be the second common cause of the disease. Based on our results, obtained from a large number of cases, it could be hypothesized that Lichtemia spp. have a higher tendency for cerebral invasion in comparison to Mucor spp. These findings need to be confirmed by further investigations.

Conclusion
In conclusion, based on our findings, CM would be associated with global distribution and might present with a variety of orbital and neurological symptoms. Mucormycosis can invade any part of the brain but shows more predilection to the temporal and frontal lobes. DM and HMs are considered as common risk factors. Moreover, surgical debridement and antifungal combination treatment (L-AMB plus Posaconazole) could be an effective approach in management of the CM patients .