Outcomes in Previously Healthy Cryptococcal Meningoencephalitis Patients Treated With Pulse Taper Corticosteroids for Post-infectious Inflammatory Syndrome

Abstract

Background

Cryptococcal meningoencephalitis (CM) is a major cause of mortality in immunosuppressed patients and previously healthy individuals. In the latter, a post-infectious inflammatory response syndrome (PIIRS) is associated with poor clinical response despite antifungal therapy and negative cerebrospinal fluid (CSF) cultures. Data on effective treatment are limited.

Methods

Between March 2015 and March 2020, 15 consecutive previously healthy patients with CM and PIIRS were treated with adjunctive pulse corticosteroid taper therapy (PCT) consisting of intravenous methylprednisolone 1 gm daily for 1 week followed by oral prednisone 1 mg/kg/day, tapered based on clinical and radiological response plus oral fluconazole. Montreal cognitive assessments (MOCA), Karnofsky performance scores, magnetic resonance imaging (MRI) brain scanning, ophthalmic and audiologic exams, and CSF parameters including cellular and soluble immune responses were compared at PIIRS diagnosis and after methylprednisolone completion.

Results

The median time from antifungal treatment to steroid initiation was 6 weeks. The most common symptoms at PIIRS diagnosis were altered mental status and vision changes. All patients demonstrated significant improvements in MOCA and Karnofsky scores at 1 month (P < .0003), which was accompanied by improvements in CSF glucose, white blood cell (WBC) count, protein, cellular and soluble inflammatory markers 1 week after receiving corticosteroids (CS) (P < .003). All patients with papilledema and visual field deficits also exhibited improvement (P < .0005). Five out of 7 patients who underwent audiological testing demonstrated hearing improvement. Brain MRI showed significant improvement of radiological findings (P = .001). CSF cultures remained negative.

Conclusions

PCT in this small cohort of PIIRS was associated with improvements in CM-related complications with minimal toxicity in the acute setting.

This is a small case series describing functional outcomes in previously healthy cryptococcal meningitis patients treated with pulse corticosteroid taper therapy for postinfectious inflammatory syndrome (PIIRS).

In many regions of the developed world, control of human immunodeficiency virus/AIDS (HIV/AIDS)-related cryptococcal meningoencephalitis (CM) has been accompanied by a persistence in cases related to transplant conditioning and an increasing recognition of disease in the previously healthy. The control of bacterial meningitis by vaccines has resulted in CM becoming the principal cause of nonviral meningitis in the United States [1–3]. However, effective therapy remains elusive in all CM populations, with an estimated mortality of 30–50% despite treatment and significant disability in those who survive, principally related to cognitive, visual, and hearing deficits [2, 4–8]. Although historically, therapeutic efforts have focused on augmentation of pathogen killing [9, 10], recent consideration of host immune damage has been highlighted by modeling of a parabolic balance between microbial clearance and destructive immune responses [11]. The potential of a destructive immune response is particularly pertinent to central nervous system (CNS) infections where the close confines of the skull allow little expansion of edematous inflamed tissue, and its contribution has been well described in both bacterial [12] as well as cryptococcal infections. In case of the latter, an immune reconstitution inflammatory syndrome (IRIS) usually occurs post antiretroviral therapy (ART) initiation in AIDS [13].

Post-infectious inflammatory response syndrome (PIIRS) has been defined as a deterioration in neurological status in a previously healthy patient with CM after cerebrospinal fluid (CSF) fungal culture conversion to negative following optimal treatment [14, 15]. Unlike HIV-IRIS where immune reconstitution occurs after the initiation of combination antiretrovirals, PIIRS may be a response to released fungal antigens during therapy and reductions in immunomodulatory components including capsular fragments [16]. During PIIRS, an appropriate pro-inflammatory cytokine response including interferon (IFN)-γ and interleukin (IL)-6 stimulates T helper cells to cause immune-mediated host damage [17] but with an M2 or alternately activated macrophage subtype that is less effective at clearing nonviable fungal debris [13, 18, 19]. Levels of neurofilament light chain, a biomarker of axonal damage, have been found to be significantly elevated in PIIRS, consistent with a pathogenic immune response [18].

There is a lack of consensus regarding the management of inflammatory response syndromes associated with CM. IDSA guidelines recommend the use corticosteroids (CS) in those with severe CNS symptoms in HIV-IRIS, preferably once their CSF fungal cultures have become negative [4]. Apart from CS, nonsteroidal anti-inflammatory drugs (NSAIDS), IFN -γ, tumor necrosis factor (TNF) -α antagonists and thalidomide have been used with varying success [4, 20]. In those with PIIRS, anecdotal experience with successful CS use has been reported [18, 21]. Here we report the initial outcomes in a group of 15 consecutive previously healthy patients who developed PIIRS after conventional antifungal therapy and were treated with adjunctive pulse corticosteroid taper therapy (PCT).

METHODS

Ethics Statement

The National Institute of Allergy and Infectious Diseases (NIAID) Institutional Review Board (IRB) approved this study under NIAID Protocol number 93-I-0106. All subjects provided written informed consent directly or via their durable power of attorney.

Study Design and Participants

This is a prospective, observational study involving previously healthy patients who met criteria for PIIRS and were subsequently treated with a uniform regimen of corticosteroids. A diagnosis of CM was confirmed if patients had initial positive CSF fungal cultures for Cryptococcus.

Patients were excluded if they were <15 years old and had any underlying immune defects such as HIV or receipt of immunosuppressant medications including cancer chemotherapy or monoclonal antibodies.

Protocol

Fifteen patients consecutively admitted to the National Institutes of Health (NIH) clinical center between March 2015 and March 2020 were included in this study. All patients had already completed initial therapy with amphotericin B and flucytosine at another institution prior to their NIH transfer. At the NIH, patients were diagnosed with PIIRS (if they met 1 or more main and supporting criteria defined in Table 1), received pulse methylprednisolone 1 gm parenteral daily for 7 days followed by 1 mg/kg/day prednisone for 1 month and were then tapered by 5 mg every month based on their clinical response and magnetic resonance imaging (MRI) brain findings. Lumbar punctures were performed on admission, after pulse methylprednisolone and as clinically required based on opening pressures and symptoms. Spinal fluid was sent for routine analysis, fungal cultures, cryptococcal antigen titers, immunophenotyping, and soluble cytokine analysis as previously described [18, 22].

Table 1.

Post-infectious Inflammatory Response Syndrome (PIIRS) Definition

Post-infectious Inflammatory Response Syndromea
Main criteria:b
• Unchanged or declining mental status/cognitive skills
• Visual deficit not refractive in nature
• Change in hearing
In a previously healthy patient with CSF culture conversion to negative after initial therapy with amphotericin-containing regimens
Supportive criteria:
• Increased CSF WBC, protein, decreased CSF glucose
• Increased CSF inflammatory markers: IL-6 and soluble CD25 levels
• Elevation in CSF activated immune cells: HLADR+ CD4 cells, HLADR+ CD8 cells, NK cells, monocytes
• Abnormal brain MRI findings on post-contrast FLAIR (including but not limited to leptomeningeal enhancement, choroiditis, ependymitis, parenchymal lesions, hydrocephalus)

Abbreviations: CSF, cerebrospinal fluid; IL, interleukin; MRI, magnetic resonance imaging; NK, natural killer; WBC, white blood cells.

^aPatients that met 1 or more of the main and supportive criteria were diagnosed as PIIRS.

^bSymptoms not accounted for any additional active infections/malignancy/drugs.

MOCA and Karnofsky performance scores at the time of PIIRS diagnosis (baseline) and after the administration of pulse methylprednisolone were recorded to assess mental status. The MOCA tests 8 major neurocognitive domains and has been used in patients with neurocognitive conditions previously [23]. The maximum achievable score is 30. A MOCA score of < 22 was selected as a defining criteria for PIIRS as patients with this score were considered to be at higher risk for poor long-term cognitive outcomes based on a recent published multicenter CINCH study depicting functional outcomes in HIV negative adults with CM [8]. The Karnofsky performance score was utilized to assess functional ability and has been validated in other neurodegenerative infectious conditions such as progressive multifocal leukoencephalopathy [24]. A score of 0 indicates death, whereas a score of 100 implies the ability to work independently with no complaints.

Results of ophthalmological, audiological examinations, MRI brain scans were compared between baseline and post-pulse CS administration. Papilledema was defined as optical coherence tomography retinal nerve fiber layer thickness (RNFL)  > 115 microns [25]. Automated Humphrey visual fields were measured using mean deviation in decibels (dB) and considered abnormal at values <−2 decibels. To assess audiological outcomes, pure tone hearing thresholds were obtained from 250 to 8000 Hz, and pure-tone averages were determined for the low frequencies (0.25/0.5 kHz), high frequencies (4/8 kHz), and mid frequencies (0.5/1/2/4 kHz) as described [26]. Thresholds >25 dB HL (decibels in hearing level) were considered abnormal.

Radiological findings were scored based on the presence of 9 criteria; details of scoring can be found in the Supplementary Material (Appendix A). Aggregate scores were calculated for each patient ranging from a minimum of 0 to a maximum of 12. A decrease in aggregate scores on post-pulse MRI compared to baseline indicated radiological improvement.

Once discharged from the hospital, patients had regular monthly follow-up visits for a year and were then seen annually or more frequently if their clinical condition warranted.

Statistical Analysis

Wilcoxon matched pairs signed rank tests were conducted using GraphPad Prism 7.0 to analyze CSF parameters, neurological scores, ophthalmologic data, and radiological scores. The Spearman rank test was used to identify correlations between CSF parameters, biomarkers, and outcome measures. To assess whether neurofilament light chain (NFL) values changed over time, a generalized estimating equation (GEE) model was fitted using the statistical software R.

RESULTS

Demographic data are shown in Table 2. All 15 patients were referrals from outside facilities; Table 3 highlights the differences in symptomatology at diagnosis and at the time of referral to the NIH. In all cases, the reason for referral was a lack of improvement despite antifungal therapy and negative CSF fungal cultures. The median time from CM symptom onset to CM diagnosis was 47 days (interquartile range [IQR] 19–116 days). The median time from start of antifungal therapy administration to corticosteroid initiation was 6 weeks (IQR 3–9weeks). Eleven (73%) patients presented with headache, and 4 (27%) presented with fever upon initial CM diagnosis, but fewer had characteristic meningitis symptoms at the time of transfer. Symptoms more common at transfer were altered mental status (7 patients), vision change (7 patients), and imbalance (5 patients).

Table 2.

Demographic Data of Patients Treated With Pulse Corticosteroid Taper Therapy (PCT) for Cryptococcal Post-infectious Inflammatory Response Syndrome (PIIRS)

Demographic Data	N = 15
Gender, no. of males (%)	12 (80)
Median age in years (IQR)	51 (39–62)
Median time from CM symptom onset to CM diagnosis in days (IQR)	47 (19–116)
Median time from start of antifungal treatment to steroid initiation in weeks (IQR)	6 (3–9)
Median duration of amphotericin therapy in weeks	5 (4–6)
Race, no. (%)
Caucasian	13 (87)
African American	1 (7)
Hispanic	1(7)
Species, no. (%)
neoformans	10 (67)
gattii	1 (7)
unknown	4 (26)
Underlying comorbidities, no. (%)
Hypertension	6 (40)
Diabetes	2 (13)
Multiple sclerosis	1 (7)
Asthma	1 (7)
Factor V Leiden deficiency	1 (7)
Cardiac disease	4 (26)
Patient with ventriculoperitoneal shunts, no. (%)	7a (47)

Abbreviations: CM, cryptococcal meningoencephalitis; IQR, interquartile range; NIH, National Institutes of Health.

^a Three patients were shunted before NIH transfer.

Table 3.

Signs and Symptoms of Patients at the Time of Initial Cryptococcal Meningoencephalitis (CM) Diagnosis and at Post-infectious Inflammatory Response Syndrome (PIIRS) Diagnosis, n = 15 (the Majority of Patients Had >1 Symptom at Initial CM Presentation and at National Institutes of Health [NIH] Transfer)

Signs and Symptoms	At Initial Presentation, no. (%)	At PIIRS Diagnosis, no. (%)
Headache	11 (73)	4 (27)
Altered mental status	2 (13)	7 (47)
Vision change	4 (27)	7 (47)
Hearing loss	3 (20)	2 (13)
Fever	4 (27)	1 (7)
Imbalance	3 (20)	5 (33)
Speech deficits	0	2 (13)
Back pain	0	1 (7)
Seizures	1 (7)	0
Facial nerve palsy	2 (13)	1 (7)
Motor deficits	0	1 (7)

Mental Status Outcomes

One patient did not have MOCA scores recorded at baseline. Nine patients had MOCA scores of < 22 at baseline. There was a median improvement of 9 points in MOCA scores after receiving pulse CS (median 21 days after pulse; IQR 1–36, Figure 1A). All 15 patients had Karnofsky scores recorded at baseline and after pulse CS (median 14 days; IQR 5–23); 10 patients had a score of 50 or below at baseline, indicating that they required frequent medical care due to their condition. Similar to the MOCA score, there was a statistically significant improvement in Karnofsky scores after pulse CS (median improvement of 20 points; Figure 1B).

Figure 1.

PCT is associated with improved neurologic dysfunction in cryptococcal post-infectious inflammatory response syndrome. (a) MOCA (n = 14) and (b) Karnofsky (n = 15) scores were obtained at baseline and post methylprednisolone pulse. Three patients underwent ventricular peritoneal shunt prior to NIH transfer and 1 at the NIH clinical center by the time their post CS MOCA was measured. One patient had disparate scores (MOCA of 23 and a Karnofsky of 20) after therapy because the MOCA score was determined 10 days after the Karnofsky score post-pulse and the patient was continuing to exhibit clinical improvement. Patients unable to complete MOCA questions due to poor mental status were assigned a score of <8 to indicate an inability to complete 1 point of the 8 constituent MOCA categories. Scores were rated for each patient by a certified infectious disease specialist based on physical therapist and inpatient physician documentation. Abbreviations: MOCA, Montreal cognitive assessments; NIH, National Institutes of Health; PCT, pulse corticosteroid taper therapy.

Karnofsky scores repeated for 14 patients at a median of 2 years following pulse CS and showed a median 20-point improvement in 13 patients and a decline by 20 points in 1 patient who was diagnosed with depression a year after CM diagnosis. Supplemental videos 1 and 2 show 2 patients on admission and after initial therapy with PCT.

Ocular Outcomes

Eleven patients had comprehensive ocular exams performed at PIIRS diagnosis and at follow-up; patients who were too stuporous to tolerate a detailed ocular exam on initial presentation were excluded. Eight patients (73%) had abnormal RNFL measurements, and all 11 patients had abnormal visual field measurements in one or both eyes upon NIH transfer. Average RNFL thickness in micrometers (µm) and visual field measurements in dB are shown at baseline (within 1 week of pulse initiation) and after pulse completion (median 7 weeks; IQR 2–20) in Figure 2A, B. Improvements in both these parameters were significant post methylprednisolone. Four patients receiving a comprehensive ocular exam (4/11) required adjunctive shunt placement to control papilledema.

Figure 2.

Pulse corticosteroid therapy is associated with improved RNFL thickness (a) and visual fields (b) in PIIRS. Patients (n = 11) were assessed at baseline and post methylprednisolone pulse. Normal RNFL thickness typically ranges between 75 and 115 µm. Visual fields were measured using mean deviation and normal values range from 0 to –2 decibels (dB). (c) Median reduction of 5 cmCSF was observed in opening pressures recorded at the end of the pulse. Red dashed lines indicate the upper limit of normal (n = 12, ns, not significant). For RNFL thickness and visual fields, the average for both eyes was utilized for the analysis. When unilateral data were missing, contralateral values were used. Abbreviations: PIIRS, post-infectious inflammatory response syndrome; RNFL, retinal nerve fiber layer.

The median opening pressure recorded at baseline was 27 cmCSF (n = 15, IQR 16–35). Post-pulse opening pressures were recorded for 12 patients and a median reduction of 5 cmCSF was observed (Figure 2C). Four of 6 patients with elevated baseline opening pressures (>25 cmCSF) eventually developed intracranial hypertension and required shunt placement at a later date.

Audiologic and Neurotologic Outcomes

Of the 15 patients, 7 reported hearing abnormalities and had audiological assessments performed at baseline and follow-up. At baseline, 5 patients had abnormal hearing independent of pure-tone average (PTA) classification (low-frequency, 4-freqency, or high-frequency PTA), 1 patient had abnormal hearing according to both the 4-freqency and high-frequency PTA classification for both ears, and 1 patient had abnormal hearing according to only the high-frequency PTA in a single ear with normal hearing in the other ear. Pure-tone hearing thresholds for each ear are depicted for all 7 patients in Figure 3 for baseline (within 11 days of methylprednisolone initiation) and after methylprednisolone pulse completion (median 3 weeks; IQR 3–9). Five of 7 patients met criteria for hearing improvement in at least 1 ear as defined by the American Speech-Language Hearing Association (ASHA; >10 dB change at 2 consecutive frequencies): 3 bilateral and 2 unilateral.

Figure 3.

PCT is associated with pure-tone improvements in hearing. Bilateral hearing levels in dB HL for patients tested at baseline and post methylprednisolone pulse completion based on LF (a), 4F (b), and HF (c) PTA (n = 7). Average thresholds above 25 dB HL are considered abnormal. Dashed line indicates ≥10 dB improvement. Abbreviations: 4F, 4 frequency; dB HL, decibel hearing loss; HF, high frequency; LF, low frequency; PCT, pulse corticosteroid taper therapy; PTA, pure tone average.

Improvement in CSF Parameters During Initial CS Pulse

All 15 patients had lumbar punctures at PIIRS diagnosis and after pulse methylprednisolone, of which 9 had no additional lumbar punctures and 6 had at least one lumbar puncture in the interim. Initial CSF examination revealed an inflammatory milieu typical of PIIRS in the setting of negative CSF fungal cultures (Table 4). CSF/serum glucose ratios, previously associated with elevated CSF NFL, a marker of axonal damage [27], were positively associated with baseline MOCA scores within this population(Figure 4A, Spearman r = 0.53). After pulse CS completion, patients exhibited significant improvement in hypoglycorrhachia (Figure 4B), CSF pleocytosis (Figure 4C), and elevated protein (Figure 4D). Immunophenotyping of spinal fluid showed a decline in inflammatory HLADR⁺ CD4 cells (Figure 4E, F), HLADR + CD8 cells (Figure 4E, G) post CS pulse with additional significant reductions in monocytes (Figure 4H, I) and inflammatory CD56⁺NK cells (Figure 4J, K). Soluble CSF cytokine analysis showed a decrease in IL-6 levels (Figure 5A) and soluble CD25 levels (Figure 5B). The levels of NFL did not show reductions in the immediate post-pulse period (Figure 5C) but did show reductions over a prolonged period during the PCT, typical of the prolonged half-life of this marker(Figure 5D).

Table 4.

Cerebrospinal Fluid (CSF) Immune Parameters of Patients Treated With Pulse Corticosteroid Taper Therapy (PCT) for Cryptococcal Post-infectious Inflammatory Response Syndrome (PIIRS)

CSF Parameters at PIIRS Diagnosis n = 14a	Median (IQR)	Normal Range
WBC (/mm3)	77 (47–146)	(0–5)
Total protein (mg/dL)	121 (67–223)	(15–60)
CSF: serum glucose ratio	0.3 (0.17–0.41)	(50–80)
HLADR4+ CD4 cells/mL	9932 (4398–24 535)	(30–107)
HLADR4+ CD8 cells/mL	6912 (2351–16 419)	(24–85)
CD56+ NK cells/mL	787 (278–1405)	(10–33)
Monocytes/mL	1420 (729–3534)	(84–237)
IL 6 (pg/mL)	96 (37–675)	(≤5)
Soluble CD25 (pg/mL)	582(152–2131)	(≥1033)
NFL (pg/mL)	29 595 (7155–45 374)	(<1068)

Abbreviations: IL, interleukin; IQR, interquartile range; NFL, neurofilament light chain; NK, natural killer; NIH, National Institutes of Health; WBC, white blood cells.

^aOne patient had obstructive arachnoiditis which prevented adequate CSF sampling beyond that for a fungal culture prior to PCT. One of the patients in this study was treated with 60 mg oral prednisone for 5 days while awaiting transfer to the NIH but did meet criteria for PIIRS on admission.

Figure 4.

Improvements in CSF pro-inflammatory parameters are noted in the immediate post-pulse period. (a) Correlation between CSF/serum glucose ratios vs MOCA obtained at time of PIIRS diagnosis. (b) CSF/serum glucose ratios, (c) CSF WBC, and (d) total protein at baseline and post-pulse (n = 14). Representative flow cytometry of CSF of a 51-y/o patient and summary data of CSF HLADR+ CD4 T cells (red puncta; e, f), HLADR+ CD8 T cells (blue puncta; e, g), monocytes (h, i) and CD56+NK cells (j, k) obtained prior to PCT (baseline) and post-pulse (n = 12). Abbreviations: CSF, cerebrospinal fluid; MOCA, Montreal cognitive assessments; PCT, pulse corticosteroid taper therapy; PIIRS, post-infectious inflammatory response syndrome; WBC, white blood cells; y/o, year old.

Figure 5.

Improvements in soluble markers of inflammation and axonal damage after PCT. CSF IL-6 (a, n = 12), sCD25 (b, n = 12), and NFL (c, age adjusted, n = 10) were obtained at baseline and post-pulse CS. (d) Time course of CSF NFL during initial and taper period of PCT (n = 10). GEE estimated slope of reduction = –54.14. Abbreviations: CSF, cerebrospinal fluid; GEE, generalized estimating equation; IL, interleukin; NFL, neurofilament light chain; ns, not significant; PCT, pulse corticosteroid taper therapy.

Radiological Outcomes

Brain MRI with contrast was performed on all 15 patients at baseline (median 1 day from solumedrol initiation; IQR −3 to 5) and approximately 1 month after pulse completion (median 55 days, IQR 43–69). Examples of 2 patients with PIIRS are shown in Figure 6 with marked decrease in parenchymal and meningeal enhancement post-pulse. Results of aggregate MRI scores are shown in Figure 6C. In 11 of 14 patients, a significant reduction in scores was noted post-pulse and in the remainder, scores were unchanged. The most common radiological findings reported at baseline were meningeal enhancement (13 patients), hydrocephalus or shunted ventricular system (10 patients), and parenchymal enhancement (9 patients). In patients with hydrocephalus, 3 already had VP shunts in place at the time the baseline MRI was performed.

Figure 6.

MRI brain changes post methylprednisolone pulse. (a) 49 y/o patient: baseline enhanced T1-weighted and FLAIR images at PIIRS diagnosis demonstrating extensive meningeal enhancement (white arrows), especially noted along the Sylvian fissures and bilateral cerebral sulci, and parenchymal enhancement (open arrows) mainly in the basal ganglia. Findings of bilateral ependymitis (arrowheads) also seen. Post-pulse CS images obtained on hospital day 60 showing almost complete resolution of leptomeningeal/parenchymal enhancement and ependymitis. (b) 45 y/o patient: post contrast T1-weighted and FLAIR images demonstrating marked meningeal enhancement along the cerebellar folia and along the convexity sulci (white arrows). Post-pulse CS images obtained on hospital day 60 showing significantly improved leptomeningeal enhancement. (c) Aggregate MRI scores obtained prior to PCT (baseline) and during the post-CS pulse period (n = 15). Abbreviations: CS, corticosteroids; MRI, magnetic resonance imaging; PCT, pulse corticosteroid taper therapy; PIIRS, post-infectious inflammatory response syndrome; y/o, year old.

Clinical Course

All patients initiated on CS had baseline liver functions, QuantiFERON-TB gold test and DEXA scans performed. They were started on Bactrim DS thrice weekly for PJP (Pneumocystis jirovecii pneumonia) prophylaxis and daily calcium and vitamin D supplementation. Steroid-related complications are described in Table 5: 2 (13%) patients demonstrated osteopenia on DEXA scan and were started on alendronate, 2 (13%) patients developed elevated intraocular pressure, and 4 (27%) developed posterior subcapsular (PSC) cataracts, a known complication of prolonged steroid use. No infections and no recurrence of cryptococcal growth in cultures were noted.

Table 5.

Adverse Outcomes of Patients Treated With Pulse Corticosteroid Taper Therapy (PCT) for Cryptococcal Post-infectious Inflammatory Response Syndrome (PIIRS)

Complications	N = 15 (%)
Infections	0
Osteopenia requiring bisphosphonates	2(13)
Fractures	0
Posterior subcapsular cataracts	4 (27)
Refractory adrenal insufficiency	0
Aseptic necrosis	0
Ocular hypertension	2 (13)
Steroid-induced hyperglycemia	3a (20)
Weight gain > kg 6 months post-pulse	7 (47)
Neuropsychiatric symptoms	1b(7)

^a Patient 1 had an underlying history of diabetes mellitus; in the other 2 patients, it subsided with no requirement for long-term insulin use.

^b Resolved.

At the time of writing, 9 patients have completed the prednisone taper within a median duration of 22 months (IQR 20–28) post-pulse initiation, 5 patients are still in the taper, and 1 patient died of a pulmonary embolism.

DISCUSSION

Currently, there are scarce guidelines available on the management of previously healthy hosts who develop CM-PIIRS [4]. Herein we have described favorable short-term outcomes in a consecutive group of CM-PIIRS patients in whom a uniform regimen of PCT was administered after poor clinical responses over a median of 6 weeks in the setting of negative CSF fungal culture. The decision to use 1 gm of methylprednisolone daily for 1 week followed by a prolonged taper was based on past moderate responses and inflammatory rebounds with rapidly tapered regimens of oral corticosteroids as well as the severity of the patients’ condition [18, 27]. The use of high doses of methylprednisolone is similar to the 3–7 day initial pulse regimens used in neuro-inflammatory disorders such as multiple sclerosis [28] and resembles the recommended steroid regimens for AIDS-related c-IRIS [29, 30], except for the requirement for an initial pulse and a longer taper period in PIIRS, which may be due to the increased responsiveness of the immune system in the previously healthy compared to those with HIV/AIDS.

Despite this, 3 patients exhibited immunological flairs during their tapers defined as (1) clinical deterioration despite negative CSF cultures and or no change in serum cryptococcal antigens, and (2) worsening CSF findings or a return of lesions on brain MRI images in an area of previous inflammation. All flairs were treated successfully with a 3-day increase in oral prednisone followed by a reversion to the dose used prior to the flare. It is also important to note the difference between the success of adjunctive CS in PIIRS where the dominant pathology may be neuroinflammatory, versus its failure during the initial presentation of CM in HIV/AIDS where the neuropathology is due to the fungal pathogen in the absence of effective immunity [31].

This report also helps to emphasize the implications of analyzing biomarkers in CM. For example, we observed remarkable decreases in the immediate post-pulse period of CSF cellular and soluble inflammatory markers. CSF HLADR + CD4 cells are a well-known marker of intrathecal inflammation, shown previously to be elevated in PIIRS and other neuroinflammatory diseases [18, 22]. IL-6 is an important pro-inflammatory cytokine involved in T-cell activation produced by a variety of immune cells including CNS astrocytes [32]. The sCD25 is released by activated lymphocytes and is a biomarker of other inflammatory conditions including macrophage-activation syndromes [33]. CSF NFL levels also declined over time in response to the PCT regimen, suggesting that the latter may prevent ongoing axonal injury and resultant permanent neurological damage. NFL was first used as a biomarker in the treatment of multiple sclerosis (MS) where it showed consistent reductions over several months after therapy, similar to findings in the present study [34]. In the future, these markers of inflammation may be utilized to develop targeted steroid-sparing regimens that could replace prolonged oral CS tapers.

Finally, the fact that CSF fungal cultures in all patients remained negative stresses the safety of PCT in CM-PIIRS patients provided they are continued on azoles and prophylactic antibiotics throughout the CS course. One death was noted from a pulmonary embolism 3 months after discharge to a rehabilitation facility in a patient with persistent negative CSF fungal cultures. These results thus provide reassurance to physicians who may be understandably apprehensive regarding CS use due to their known risk of increasing susceptibility to CM infections in the absence of antifungal therapy [35]. A few of our patients did develop known steroid-related complications such as cataracts, which is treatable with surgery, and osteopenia [36], emphasizing the need to monitor for adverse effects while on PCT.

A limitation of this report is the small number of participants and a lack of randomization involved despite this being the largest study describing steroid-related outcomes in CM PIIRS patients. Although definitive evidence of improved outcomes would only come from a randomized trial, in this study where patients were chosen for PCT based on progression of severe disease after exhaustion of all other treatment options, we found [37] it difficult to ethically randomize them to a no-alternative arm. A similar scenario is encountered when administering “salvage” therapy for refractory aspergillosis, in which treatments have been accepted without randomization [38]. Additionally, we were not able to factor in the effect of shunting on outcomes. However, including those with severe complications of CM comprising PIIRS does make comparisons of PCT described here favorable, compared to overall historical mortalities of 15–30% [2, 8]. Whether the same intensity of CS therapy should be considered in those with less severe manifestations of CM is a matter of future study and may require randomized trials to demonstrate benefit. Therefore, the results of this report should be interpreted with caution when treating patients who are improving clinically without neurologic sequelae.

The findings of this case series of CM-PIIRS patients suggest that there may be clinical benefit of PCT use in severe cases presenting clinical deterioration despite effective microbiological control. In the future, the results of this report may pave the way for larger, well-designed studies aiming to investigate corticosteroid regimens in CM patients.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Disclaimer. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government.

Financial support. This research was supported in part by the Intramural Research Program of the National Institutes of Health (NIH) (grants AI001123 and AI001124 to P. R. W. and DC000064 to C. C. B.) and an extramural grant (UO1 AI109657 to K. A. M.). This project has also been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under contract number 75N91019D00024, task order number 75N91019F00130. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the US Government. O. D. reports personal fees from NIH Medical Research Scholars Program, a public-private partnership supported jointly by the NIH and contributions to the Foundation for the NIH from the Doris Duke Charitable Foundation, Genentech, the American Association for Dental Research, the Colgate-Palmolive Company, and other private donors.

Potential conflicts of interest. All authors: No reported conflicts of interest. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.