Cerebral Cryptococcomas: A Systematic Scoping Review of Available Evidence to Facilitate Diagnosis and Treatment

Background: Recommendations for managing patients with cerebral cryptococcomas are scarce across multiple clinical guidelines. Due to the deficiency of high-quality data coupled with an increasing number of at-risk patients, the purpose of this review is to describe the demographic characteristics, causative pathogen, intracranial imaging, surgical and/or pharmacological interventions, as well as outcomes of patients with cerebral cryptococcomas to improve recognition and management. Methods: We conducted a scoping review in accordance with the PRISMA guidelines using PubMed and Web of Science. Reports were included if the following details were presented: (1) site of infection; (2) treatment details which at least include the specific antifungal therapy administered, if applicable; and (3) patient outcome. Results: A total of 40 records representing 47 individual patients were included, of which the median age was 48.5 years, 75% were male, and 60% reported a significant past medical, surgical, or social history. C. neoformans was isolated more often than C. gattii (74% vs. 26%, respectively). Patients most often presented with headache, altered mental status and/or confusion, and vomiting occurring over a median of 30 days; though few were noted to have significant findings on physical examination. More than 50% of patients had a single cerebral cryptococcoma lesion, whereas perilesional edema was present in 73% of cases. Surgical intervention occurred in 49% of patients. An amphotericin B-based formulation was administered as “induction” therapy to 91% of patients, but combined with flucytosine or fluconazole in only 58%, for an overall median of 42 days. Fifty two percent of patients received “maintenance” therapy for a median of 126 days, in which fluconazole was most often used. Corticosteroids were administered to approximately 30% of patients for a median of 31.5 days. Overall, mortality was 34%. Conclusion: Based on our findings, management should include antifungal therapy for a minimum of 6 months with considerations for concomitant corticosteroids in the setting of perilesional edema, as well as surgical intervention. Emphasis should be placed on providing well-documented treatment details in future case reports and series to allow for the development of more concise evidence-based recommendations.


Background
Despite a decreasing incidence among persons living with HIV due to antiretroviral therapy (ART), rates of cryptococcosis have been increasing in patients who have undergone solid organ transplant or hematopoietic cell transplant [1]. In addition, cryptococcosis is

Results
A total of 269 records were identified. After removing duplicates and exclusions, 74 full-text records were screened for eligibility, of which 49 were excluded (Supplementary Figure S1) [16]. An additional 15 records, some of which were published prior to 2010, were identified from references within articles of interest for a total of 40 records [9, representing 47 individual patients included in the systematic scoping review (Table 1).

Demographic Characteristics
Patient ages ranged from 19 to 75 years (median 48.5 years), and 75% (n = 35) were male. Though 40% (n = 19) of the patients did not report a past medical, surgical, or social history, of the 60% (n = 28) who did, 36% (n = 10) were HIV-infected, 21% (n = 6) had hypertension, and 18% (n = 5) had diabetes mellitus. One patient was in her second trimester of pregnancy with no other significant history [39], one patient had a history of polycythemia vera and monoclonal gammopathy of unknown significance [24], whereas two patients reported close contact with pigeons [41,46]. Of those with HIV, absolute CD4 count ranged from 0 to 157 cells/µL [9,28,36,42,43,50], whereas only three patients were noted to be on ART [36,42]. Six patients, all of whom were HIV-infected, were previously treated for CM which occurred from 2 to 22 months prior to their diagnosis of cerebral cryptococcoma [9,36,42,53].

Treatment Details
Treatment details were provided for 96% (n = 45) of patients, as two patients died prior to surgical or pharmacological intervention [36,39]. Forty-nine percent (n = 22) of patients underwent surgical resection of one or more lesions, in which one patient underwent complete surgical resection of the lesion and antifungal therapy was not administered postoperatively [20]. Notably, surgical intervention occurred more often in those with lesions measuring 3 or more centimeters. Alternatively, surgical intervention was performed less often in patients with more than one lesion (38%) despite high rates of perilesional edema (71%). Management of intracranial pressure (ICP) was described in only a few cases, but of those, ventricular shunts were placed in three patients [18,22,43], whereas another two patients underwent external ventricular drain (EVD) implantation [42,52] as a result of failure to control ICP with serial lumbar punctures (LPs) [26], clinical deterioration [30], or development of hydrocephalus [50].
Of the 44 patients who received antifungal therapy, "induction" regimens most often consisted of an amphotericin B-based formulation (91%, n = 40) combined with flucytosine or fluconazole in 45% (n = 18) or 13% (n = 5), respectively. The duration of "induction" antifungal therapy varied widely from 7 to 180 days (median 42 days). The specific amphotericin B formulation utilized was included in 70% (n = 28) of cases, whereas specific doses, including mg/kg/day or total mg/dose, were reported in 53% (n = 21). Among patients who received an amphotericin B-based formulation, the median duration of therapy was 38 days (range 7-180 days), but details were limited as to why some treatment durations were abbreviated or prolonged. However, in one patient, amphotericin B was discontinued after the development of hyperkalemia, hypomagnesemia, and a pruritic rash, all of which improved after the initiation of fluconazole [33]. Fluconazole was exclusively administered to patients who did not receive an amphotericin B-based formulation as "induction" therapy for a median duration of 56 days (range 32 to 84 days).
Corticosteroids were administered to 27% (n = 12) of the 44 patients who received "induction" antifungal therapy for a median of 31.5 days (range 1-60 days). The specific corticosteroid and dose utilized was often not mentioned, but where described, dexamethasone 12 to 28 mg/day in three to four divided doses were most frequently administered [18,42,50]. The rationale for starting corticosteroids were provided in very few cases, and included initiation during the early phases of antifungal therapy, as well as later on in the treatment course in response to complications, such as increased ICP despite serial LPs [42], presence of hydrocephalus on repeat imaging [42], spinal cord edema, and paradoxical immune reconstitution inflammatory syndrome (IRIS) during week 8 of antifungal therapy [18]. Adjunctive medications, besides corticosteroids, were not widely used. One patient was treated with interferon (INF)-γ three times per week after 3 weeks of antifungal therapy for refractory disease [18], whereas another was started on subcutaneous adalimumab every 2 weeks due to clinical deterioration despite antifungal therapy [9].
Patients who died were more likely to report a significant past medical or surgical history (64% vs. 40%), be infected with C. neoformans (69% vs. 45%), present with altered mental status or confusion (53% vs. 30%), nuchal rigidity (20% vs. 0%), as well as upper or lower extremity weakness (27% and 27% vs. 3% and 10%, respectively) compared to survivors. The presence of multiple lesions and perilesional edema were similar between groups. However, fewer patients that died underwent surgery (40% vs. 53%). Though an amphotericin B-based formulation was commonly administered as "induction" therapy in both groups (80% vs. 90%), those who died received a shorter duration of "induction" therapy (median, 21 days (range, 10-42 days) vs. 42 days (range, 7-180 days)). In addition, the duration of "maintenance" therapy was much shorter in those who died (median, 60 days (range, 6-76 days) vs. 180 days (range, 60-730 days)), despite 90% of patients in each group receiving fluconazole or voriconazole. Approximately 40% of patients in each group received corticosteroids, but the median duration of therapy was 4.5 days (range, 1-8 days) in patients who died compared to 49 days (range, 21 to 60 days) in those who survived.
Of the 31 patients who survived, 94% (n = 29) were described as having symptomatic, clinical, and/or radiographic improvement during follow-up. Median time to followup was 255 days (range, 7 to 4380 days). Though most patients were asymptomatic at follow-up, one described residual neuromotor symptoms on day 1460 [25], whereas another patient experienced improvement in symptoms initially, but later required readmission due to worsening neurologic symptoms [48].
Few cases described recurrent disease during the follow-up period [22,[28][29][30]41,42,48], with even fewer detailing subsequent surgical or pharmacological intervention. An HIVinfected patient was asymptomatic with near complete radiographic resolution after 7 months of antifungal therapy and 5 months of ART, but then experienced a recurrence with worsening basal ganglia lesions on MRI, prompting initiation of combination antifungal therapy and corticosteroids for 6 weeks, followed by long-term fluconazole [28]. Radiographic improvement was noted after 6 months of treatment, but the patient experienced another recurrence 4 months later despite remaining on fluconazole and ART. The patient was treated with combination antifungal therapy for 6 weeks, then transitioned to long-term voriconazole, instead of fluconazole, without the development of new symptoms after 10 months. In another case, a patient who underwent surgical resection of left frontal lobe and left temporal lobe cryptococcomas followed by antifungal therapy, presented after 6 months due to non-adherence with new onset headaches and upper extremity weakness with the identification of two right parietal lobe lesions [29,30]. The patient underwent surgical resection, and was then started on oral fluconazole with reported symptom improvement after 8 weeks.

Discussion
In this systematic scoping review, we present a thorough analysis of the data, describing the demographic characteristics, diagnostic findings, treatment modalities, as well as outcomes for 47 individual patients with cerebral cryptococcomas. Though data presented across all 40 records were quite heterogeneous, as all but three were case reports, this review provides valuable insight into an understudied, and perhaps under recognized, area with limited guideline recommendations due to low quality evidence [13][14][15] despite an increasing frequency of at-risk patients [2][3][4][5][6][7][8]. As a result, findings from this systematic scoping review may lead to improvements in patient care and better clinical outcomes.
Cerebral cryptococcomas are challenging to diagnose, and are often missed on initial evaluation due to their highly variable clinical presentation in patients with and without comorbidities or risk factor for cryptococcosis. Similar to epidemiologic findings in patients with CM [58], males comprised 75% of the cases included in our analyses. Almost 80% were at least 30 years of age, in which 14% of patients were aged 30 to 39 years, 29% were 40 to 49, 29% were 50 to 59, 14% were 60 to 69, and 14% were over 70 years. Overall, most patients reported a past medical or surgical history including known risk factors for cryptococcosis, such as HIV-infection, diabetes, or idiopathic CD4 lymphocytopenia [59]. Patients with C. gattii were less likely to report a significant past medical or surgical history, whereas those with C. neoformans were more likely to be HIV-infected, which is similar to findings from previous data [58,60]. A few patients reported social histories with established ecologic risk factors, including living in or traveling to Australia in the case of C. gattii [18,49], and exposure to birds and bird droppings in the case of C. neoformans [41,46]. Patients with cerebral cryptococcomas presented with a wide range of often non-specific symptoms and physical exam findings, from headache and altered mental status to vomiting, giddiness, word-finding difficulties, and cerebellar signs, which were more common among patients 50 years and older. Clinical manifestations differed between patients with C. neoformans and those with C. gattii ( Table 2).
All 47 individual patients included in our review underwent neuroimaging, of which MRI was utilized more often than CT (66% vs. 34%, respectively). Though CTs are often preferentially performed due to availability and timeliness [61], previous data suggest increased sensitivity for detecting cryptococcoma lesions using gadolinium-enhanced MRI compared to standard MRI or CT [10,54,61]. MRI findings vary, ranging from hyper-intense on T2-weighted images, to non-enhancing on postcontrast T1-weighted images ( Figure 1). Due to the rarity of cerebral cryptococcomas and challenges associated with diagnosis, it is common for them to be mistaken for neoplasms, as well as pyogenic or tubercular abscesses, prior to surgery [52]. Less often, cerebral cryptococcomas have been misdiagnosed as neurocysticercosis [39], or even a tumefactive demyelinating lesion, a rare focal demyelinating disease [62]. Cryptococcomas are sometimes confused in neuroimaging with dilated perivascular spaces (Virchow-Robin spaces) that coalesce to form gelatinous pseudocysts [63]. However, cryptococcomas resulting from invasion of Cryptococcus spp. into the brain parenchyma may develop in a variety of locations throughout the brain [10,11,52]. The frontal and parietal lobes (21% and 19%, respectively), as well as the basal ganglia (21%) were most often involved, whereas the thalamus (5%) or the pons (2%) were rarely involved among the 47 individual patients included in our analysis. The median number of lesions identified was one, but ranged from one to three among reports that provided specified details. Forty-seven percent of patients were noted to have more than one or multiple lesions throughout the brain parenchyma, which was more common amongst patients with C. gattii ( Table 2). Characteristics of the lesions were not universally reported, but the size of the cryptococcomas varied substantially from less than 1 cm to 5 or 6 cm. C. neoformans was identified more often among cases where measurements were provided [23,24,34,44,50]. Perilesional edema and hydrocephalus were slightly more common among patients with C. neoformans than those with C. gattii (Table 2). Intracranial imaging is a valuable tool to determine the extent of disease severity in patients with disseminated cryptococcosis, but is insufficient to establish a diagnosis of cerebral cryptococcoma.
Amphotericin B and flucytosine followed by long-term fluconazole remains the mainstay of treatment in CM [13][14][15]. Treatment strategies for patients with CM have largely been extrapolated to patients with cerebral cryptococcomas, as no prospective studies have been performed. Though guidelines suggest initial therapy for patients with cerebral cryptococcoma should include lipid-associated formulation of amphotericin B, in lieu of AmB-d when available, in combination with oral flucytosine for at least 42 days followed by oral fluconazole 400 mg to 800 mg per day for 180 to 540 days [14,15], regimen selection and duration were inconsistent across all most cases. More than 90% of patients received an amphotericin B-based formulation, of which 58% received combination therapy, whereas only 52% received "maintenance" therapy. Details about the specific formulation and/or dose utilized and associated rationale for "induction" and "maintenance" antifungal therapies were rarely provided.
reported, but the size of the cryptococcomas varied substantially from less than 1 cm to 5 or 6 cm. C. neoformans was identified more often among cases where measurements were provided [23,24,34,44,50]. Perilesional edema and hydrocephalus were slightly more common among patients with C. neoformans than those with C. gattii (Table 2). Intracranial imaging is a valuable tool to determine the extent of disease severity in patients with disseminated cryptococcosis, but is insufficient to establish a diagnosis of cerebral cryptococcoma. Amphotericin B and flucytosine followed by long-term fluconazole remains the mainstay of treatment in CM [13][14][15]. Treatment strategies for patients with CM have largely been extrapolated to patients with cerebral cryptococcomas, as no prospective studies have been performed. Though guidelines suggest initial therapy for patients with cerebral cryptococcoma should include lipid-associated formulation of amphotericin B, in lieu of AmB-d when available, in combination with oral flucytosine for at least 42 days followed by oral fluconazole 400 mg to 800 mg per day for 180 to 540 days [14,15], regimen selection and duration were inconsistent across all most cases. More than 90% of patients Approximately 25% of all patients received corticosteroids, of which only 50% had perilesional edema. Though limited details regarding the specific corticosteroid and/or dose administered were reported, the median duration of corticosteroid therapy was 21 days (range, 1-60 days) among patients with perilesional edema. In addition, details regarding corticosteroid tapers were also not provided. Although guidelines mention corticosteroids in the setting of cerebral cryptococcomas with surrounding edema, no recommendations for which specific corticosteroid and/or dose are available [14,15].
Of the 47 individual patients included in this analysis, 34% died prior to undergoing surgical or pharmacological intervention, discharge, or during the follow-up period, similar to that observed with CM [1]. Fewer patients who underwent surgical intervention died, which may necessitate greater emphasis on the role of surgery as a component of the overall management for cerebral cryptococcomas in the guidelines [13][14][15]. In addition, prolonged durations of "induction" and "maintenance" therapy were reported more often in patients who survived, and should continue to be considered the standard of care. However, the guidelines recommend a "gradual reduction" in corticosteroids [14,15], but the vague recommendations should be modified to emphasize the importance that the tapering should occur over at least a 3-to-6-week period.

Limitations
Limitations of this systematic scoping review must be acknowledged. First, we chose to complete a scoping systematic review due to the lack of an available comprehensive review about cerebral cryptococcomas. Second, our analysis only included 47 individual patients derived from 40 reports. In order to abstract consistent data each report, we had to exclude publications that did not provide information on the site of infection; treatment regimen at least including the specific antifungal therapy administered, if applicable; and patient outcome. Third, specific details about the pharmacological treatments administered, such as the route, dose, and/or duration, were often not reported, which limited our abilities to provide clear treatment recommendations, but may be due to the paucity of recommendations provided in the guidelines [13][14][15]. Additionally, few reports clearly delineated "induction" therapy versus "maintenance" therapy. Fourth, the duration of "induction", "maintenance", and corticosteroid therapy was longer in patients who survived than those who died, which could represent survival-related selection bias. Lastly, due to the inherent difficulties of diagnosing cerebral cryptococcomas, many additional cases may be unrecognized and, therefore, not reported.

Conclusions
Despite guideline recommendations, diagnosis and treatment of cerebral cryptococcomas is very heterogeneous due to the lack of high-quality data. In our analysis, C. neoformans was identified more often than C. gattii as the causative pathogen in cases where an organism was isolated and speciated, which may be due to difficulties differentiating C. neoformans from C. gattii, and underreporting. Furthermore, cerebral cryptococcomas are most often identified as a single lesion with surrounding edema on gadolinium-enhanced MRI of the brain, and occur more often in male patients over 30 years of age who frequently report a significant past medical, surgical, or social history, and present with non-specific symptoms and physical exam findings. Though the most efficacious treatment remains undefined, a multipronged approach should be utilized that includes corticosteroids in the setting of perilesional edema tapered over 3 to 6 weeks; considerations for surgical intervention, if feasible; as well as a prolonged duration of antifungal therapy. Emphasis should be placed on providing species identification and well-documented treatment details in future reports to allow for the development of more concise evidence-based recommendations.

Funding:
The authors received no financial support for the research, authorship, and/or publication of this article.