Primary Cryptococcal Cellulitis With High Antigenemia in an Immunocompromised Host: A Case Report and Laboratory Investigation of the Causative Cryptococcus neoformans Strain Authors: Hannah M Hamad, Charles A Specht, Diana Carlson, Vikas Yadav, Rajendra Upadhya, Joseph Heitman, Colleen Gabel, Patrick O’Donnell, Farha A P Karlath, Fozia Qamar, Michael Newstein, Stuart M Levitz Published: 2026 ABSTRACT AbstractCutaneous cryptococcosis, although rare, may arise as primary inoculation or as part of disseminated cryptococcal infection, particularly in immunocompromised hosts. We present the case of an 80-year-old man with type 2 diabetes mellitus receiving corticosteroid therapy for cryptogenic organizing pneumonia who developed cryptococcal cellulitis with a markedly elevated serum cryptococcal antigen (CrAg) titer but no discernible central nervous system (CNS) involvement. As high serum CrAg titers are strongly associated with CNS disease, we compared the patient isolate (designated MA-1) with the Cryptococcus neoformans reference strain KN99, using in vitro growth curves, a murine pulmonary infection model, and whole-genome sequencing. MA-1 grew more slowly than KN99 in tissue culture medium. Mice infected with MA-1 survived longer than those challenged with KN99, although brain dissemination of MA-1 still occurred by day 14, indicating that the strain was neurotropic. Genomic analysis revealed overall similarity to the C neoformans H99 strain but with distinctive structural features, including duplication of chromosome 12, a large deletion on chromosome 5, and multiple single-nucleotide polymorphisms. These alterations may have contributed to altered dissemination dynamics. This case underscores the importance of considering cryptococcosis in refractory cellulitis in immunocompromised patients. High CrAg titers, even without CNS findings, should prompt lumbar puncture and consideration of induction therapy per current guidelines. This integrated clinical and laboratory investigation suggests that pathogen-specific factors, as shown in vitro and in a murine model, may influence the tempo of dissemination in human disease. Virtually all cases of cryptococcosis are caused by the Cryptococcus neoformans and Cryptococcus gattii species complexes [1, 2]. Infection typically follows inhalation of environmental propagules, with alveolar granulomatous responses that are either contained or progress when immunity is impaired. Meningitis is the most common and clinically significant disease manifestation, reflecting the strong neurotropism of Cryptococcus. In persons with human immunodeficiency virus (HIV-1)/AIDS, cryptococcosis is a common opportunistic infection, accounting for an estimated 112 000 annual deaths [3]. In many resource-rich settings, due to decreased numbers of persons with advanced HIV infection, most cases occur in immunocompromised persons without HIV [4, 5]. Morphologically, Cryptococcus is a narrow-budding encapsulated yeast. Capsule shed during infection can circulate in the blood and cerebrospinal fluid (CSF) and be detected by the cryptococcal antigen (CrAg) test. Studies of environmental and clinical cryptococcal isolates have revealed substantial genotypic and phenotypic variations, some of which have been linked to virulence and clinical outcome [6–12]. Primary cutaneous cryptococcosis is rare relative to secondary cutaneous disease and can remain localized or progress to dissemination depending on host and strain factors [13]. Against this backdrop, we report a case of biopsy-proven primary cryptococcal cellulitis with a markedly elevated serum CrAg titer but no CNS involvement—an atypical outcome given correlations between high titers and meningitis [14–17]. We hypothesized that disease was captured before dissemination to the CNS, potentially influenced by slower growth kinetics and/or genomic alterations. To interrogate this, we combined clinical observation with in vitro growth curves, a murine pulmonary infection model, and whole-genome sequencing of the clinical isolate (which was named MA-1). CASE PRESENTATION An 80-year-old male resident of central Massachusetts presented with right upper-extremity cellulitis; he had with a history of poorly differentiated squamous cell carcinoma of the lung after left upper lobe resection in 2021, tobacco use disorder (quit smoking in 2018), poorly controlled type 2 diabetes mellitus (hemoglobin A1c, 10%), chronic kidney disease, coronary artery disease after coronary artery bypass grafting, atrial fibrillation, chronic obstructive pulmonary disease (COPD), and peripheral artery disease. In April 2024, the patient experienced cough, wheezing, and dyspnea. Following treatment for presumed COPD exacerbation in May 2024, his symptoms persisted despite antibiotics and corticosteroids, prompting hospitalization for multifocal pneumonia (25–27 May). Pulmonary imaging demonstrated bilateral infiltrates unresponsive to antibiotic therapy. The patient was subsequently readmitted (8–13 June) for worsening respiratory symptoms and was suspected to have cryptogenic organizing pneumonia. He received empiric high-dose prednisone (100 mg daily), which was tapered over approximately 2½ months. On 15 August 2024, the patient experienced right lower-extremity cellulitis, which was treated with cephalexin. Shortly thereafter, he sustained a skin tear on his right upper extremity after traumatic contact with a door hinge in a public bathroom and 1 week later with a wooden nightstand. This was managed locally with wound care. However, worsening swelling, redness, and pain in his right wrist and forearm ensued, and the patient was admitted to a community hospital on 27 August. The patient received daptomycin 27–28 August 2024 and then again on 30 August to 3 September, as well as ceftriaxone on 29 August, with minimal improvement (Figure 1). No debridement was performed, and bacterial cultures were negative. Wound culture from the right upper-extremity ulcer obtained on 30 August revealed C neoformans, and the serum CrAg test result (IMMY) was positive at a titer of 1:2560 on 3 September. The patient was started on fluconazole (200 mg daily) on 4 September, and the dose was reduced based on renal function, for an estimated creatinine clearance of 13 mL/min. His wounds initially improved after starting fluconazole, then worsened again with new areas of necrosis. On 6 September 2024, daptomycin was deescalated to cefadroxil, and the patient completed 14 days of therapy. Clinical images of the right hand and forearm cryptococcal cellulitis showing temporal changes with antifungal therapy. Due to the patient’s markedly elevated serum CrAg titer and suspicion for disseminated disease, a lumbar puncture was performed on 18 September and revealed an opening pressure of 23 cm H2O (normal <18 cm). The CSF had a white and red blood cell counts of 5/µL and 2710/µL, respectively, glucose and protein levels of 114 and 98 mg/dL, and an undetectable CrAg titer (IMMY). The CSF Gram stain and culture were negative, and blood cultures were negative for growth of Cryptococcus. A chest computed tomographic scan showed a persistent small left pleural effusion with consolidation and patchy opacities, which was read as improving compared with prior imaging. The patient's cutaneous lesions continued to worsen with purulent drainage (see Figure 1), and he was transferred to UMass Memorial Medical Center on 20 September 2024 for further management. The patient was started on induction therapy with liposomal amphotericin B (4 mg/kg daily) and flucytosine (25 mg/kg twice daily, dose adjusted for impaired renal function) on 20 September 2024. On this date his serum CrAg titer was noted to have decreased to 1:128 (Meridian Bioscience). A skin biopsy specimen from the right forearm lesion obtained on 21 September revealed necrotizing granulomatous inflammation with abundant yeast forms consistent with Cryptococcus species (see Figure 2). An acute kidney injury thought to be due to amphotericin B developed on 26 September, and the patient was switched to consolidation therapy with fluconazole (800 mg loading dose followed by 200 mg daily, dose adjusted for an estimated creatinine clearance of 28 mL/min) and flucytosine, per guidelines [18]. Images of representative histology and special stains from skin punch biopsy. A, Transected necrotizing granulomatous inflammation on low power (hematoxylin-eosin [HE]; original magnification ×4.). B, Transected necrotizing granulomatous and neutrophilic inflammation (HE; original magnification ×10). C, Round yeast forms in the deep dermis, 5–10 µm in diameter, with surrounding necrosis and neutrophilic karyorrhectic debris. D, Grocott methenamine silver stain highlights abundant yeast form in the deep dermis, with focal narrow-based budding (original magnification ×20). E, Periodic acid–Schiff stain highlights abundant yeast forms in the deep dermis with narrow-based budding (original magnification ×20). F, Mucicarmine stain highlights mucinous capsule surrounding yeast fungal forms (original magnification ×20). G, Alcian blue stain (pH 2.5) highlights mucinous capsule surrounding yeast fungal forms (original magnification ×20). H, Fontana-Masson highlights melanin within the fungal cell wall (original magnification ×20). Arrows in C–H point to yeast cells. The patient completed a 4-week course of fluconazole and flucytosine, and the cryptococcal isolate obtained on 20 September was confirmed to be sensitive to both fluconazole (minimal inhibitory concentration, 2 µg/mL) and amphotericin B (minimal inhibitory concentration, 0.5 µg/mL). The patient's right arm wounds improved significantly following antifungal treatment. By discharge, his cellulitis had resolved, and his renal function had stabilized. He received 8 months of maintenance therapy with fluconazole (200 mg daily). Figure 3 shows a timeline with key clinical events and treatment depicted. Timeline of clinical events (figure created in BioRender [19]). Abbreviations: Ampho, liposomal amphotericin B; Azithro, azithromycin; COPD, chronic obstructive pulmonary disease; CrAg, cryptococcal antigen; CSF, cerebrospinal fluid; CTX, ceftriaxone.′ The patient graciously provided verbal and written permission to publish this case report. METHODS Fungal Strains The C neoformans isolate MA-1 was recovered from a right forearm culture obtained on 20 September 2024. For phenotypic comparisons, KN99, a laboratory C neoformans reference strain, derived from the clinical strain H99, was used as a control. Fungal strains were aliquoted and stored in 25% glycerol at −80°C. Fungal Growth Curves Strains MA-1 and KN99 were grown in 4 mL of yeast peptone dextrose at 30°C for 24 hours with shaking. Yeast cells were counted on a TC20 automated cell counter (Bio-Rad) and washed in phosphate-buffered saline (PBS), and 104 cells per well were added to 24-well plates containing Roswell Park Memorial Institute 1640 medium supplemented with 10% fetal calf serum. Plates were then incubated at 37°C in humidified air containing 5% carbon dioxide. At designated time points, the plates were shaken at 500 rpm for 1 minute to suspend the cells. Cell counts were then determined using the TC20 cell counter. Two independent experiments were performed, each containing triplicate wells per time point. Cell counts on each of the triplicate wells were performed in duplicate, and the results were averaged. Murine Infection Model Mice were infected as in prior studies [6, 20, 21]. Briefly, BALB/c mice (purchased from Jackson Laboratory) were housed under specific pathogen-free conditions in the animal facilities at the University of Massachusetts Chan Medical School. Experiments were conducted under a protocol approved by Chan Medical School’s Institutional Animal Care and Use Committee. Male and female mice, aged 6 and 10 weeks, were used in approximately equal numbers. Strains MA-1 and KN99 were grown in yeast peptone dextrose and washed in PBS, as described above. Yeast cells were then suspended in PBS at 4 × 105/mL, and mice were anesthetized with isoflurane and given an orotracheal challenge with 50 µL. Mice were then followed up daily for survival, with euthanasia provided to mice that reached protocol-specified humane end points. For fungal burden experiments, mice were euthanized at designated time points, after which lungs, spleens, and brains were harvested. Organs were homogenized, and colony-forming unit (CFU) counts were determined by plating dilutions on Sabouraud dextrose agar. Genomic Analysis Whole-genome sequencing was performed on MA-1, and reads were mapped to the C neoformans H99 reference genome, which is about 19.5 Mb in size and is divided across 14 chromosomes. The DNA extraction was done from an overnight culture using the Monarch genomic DNA purification kit (no. T3010; New England Biolabs). The purified DNA was checked for quality by means of gel electrophoresis and NanoDrop and quantified by means of Qubit DNA measurement. The sample was submitted for whole-genome DNA sequencing at the Duke Sequencing and Genomic Technologies core facility. The DNA was subjected to library preparation using a KAPA HyperPrep kit (Roche no. 07962363001) and sequenced on the Illumina Novoseq platform. The sequencing data were quality checked and trimmed to remove adapters from the reads using Trim Galore (Babraham Bioinformatics) [22, 23]. The trimmed reads were aligned to the reference H99 genome using Bowtie2 software with default parameters [24]. The average coverage depth of the alignment was >200× with >82% of total reads aligning exactly 1 time and an overall alignment rate of >97%. The resulting BAM file was used to generate a bigWig (.bw) coverage file, which was then used for a genome coverage plot shown in Figure 4. The sequencing reads have been deposited at the National Center for Biotechnology Information (BioProject accession no. PRJNA1312108; ID 1312108). In vitro growth, mouse survival, and fungal burdens following pulmonary challenge with MA-1. A, Strains KN99 and MA-1 were grown in yeast peptone dextrose (YPD) medium at 30°C for 24 hours and were used to inoculate 1-mL Roswell Park Memorial Institute 1640 medium cultures at 104 cells per milliliter. Each strain was cultured at 37°C and 5% carbon dioxide. Data are from independent duplicate experiments with 3 determinations per time point. B, BALB/c mice were inoculated with 2 × 104 colony-forming units (CFUs) orotracheally with MA-1 and KN99. Mice were monitored over a period of 84 days after infection for survival assessment. C–E, Data are from independent duplicate experiments, each with 5 mice per group. Mice were euthanized at 7, 14, 21, 28, or 35 days after infection to assess fungal burdens in the lungs, brain, and spleen, respectively. There are no data (ND) for KN99 for days 28 and 35 as mice succumb to infection before those time points are reached. Bars represent geometric means. Data are from independent duplicate experiments, each with 3–5 mice per group. For A and C–E, the first experiment is shown with filled circles, and the second with open circles. Statistical comparisons are in relation to KN99. For A, unpaired t tests were used to compare significance at measured time points; for B, Kaplan-Meier curves were compared using the Mantel-Cox log-rank test; and for C–E, Mann-Whitney tests were used to determine significance. Only comparisons with significant differences are noted. **P < .01; ***P < .001. RESULTS In Vitro Growth Growth of strains MA-1 and KN99 were compared in tissue culture medium at 37°C to mimic in vivo growth conditions. MA-1 grew more slowly than KN99, with lower cell counts at all time points examined (Figure 4A). Despite this slower growth, MA-1 still replicated at host temperature and ultimately reached a similar peak cell concentration as KN99 once the stationary phase was reached. Mouse Infection In a pulmonary infection model using BALB/c mice, MA-1 demonstrated significantly delayed mortality compared with KN99 (P < .001; Figure 4B). Nevertheless, all mice died by day 83, demonstrating the virulence of MA-1. We next compared lung, brain, and spleen CFUs in mice challenged with MA-1 versus KN99. Mice were euthanized, and organs were plated for CFU counts on days 7, 14, 21, and 28. The 28-day time point was not studied in KN99 mice because most of the mice die before that time point. Progressive growth in the lungs was observed for both fungal strains, although CFU counts were significantly greater in lungs infected with KN99 compared with MA-1 (Figure 4C), a finding that has been noted with other cryptococcal isolates [7, 9]. Interestingly, for the MA-1 strain CFU counts started to plateau over time, whereas for KN99 growth increased logarithmically such that by day 21, lungs averaged nearly a billion CFUs. Dissemination to the brain occurred following infection with both fungal strains, although as with the lungs, greater increases in brain CFUs were observed following KN99 infection compared with MA-1. Spleen fungal burdens were undetectable at all time points in MA-1–infected mice but increased over time in the mice infected with KN99. Genomic Findings Whole-genome sequencing of the MA-1 isolate revealed >99% identity to the reference H99 genome with an average of 1 single-nucleotide polymorphism (SNP) per 400 base pairs. While the majority of the H99 genome sequence was uniformly covered, most subtelomeric regions exhibited low or no coverage, suggesting high level of sequence divergence in these regions (Figure 5). Importantly, a large subtelomeric deletion (approximately 80 kb) identified on chromosome 5 includes 31 genes involved in various cellular processes (highlighted in Figure 5). In addition, several genes were found to be missing or truncated in MA-1 based on the loss of genomic coverage (Supplementary Table 1). Genome sequencing of the MA-1 isolate, presented as sequencing coverage of reference type strain H99 using the MA-1 whole-genome sequencing reads. Each chromosome is plotted as a bar with color coding according to genomic average coverage as 100–base pair bins. Blue lines on top of each bar represent centromere locations. The approximately 80-kb region missing in MA-1 at the end of chromosome 5 is highlighted. Chromosome 12 shows approximately 2× coverage, consistent with aneuploidy. While most of the lost genes encode hypothetical proteins, several interesting genes were present in this pool. For example, 2 copies of the drug transporter Pdr5 were lost in MA-1, which might be consequential for drug resistance. Furthermore, 3 myo-inositol transporter genes were lost, among which Itr1a has been directly implicated in the virulence of C neoformans [25, 26]. Also absent in MA-1 are several transcription factors, such as Znf2, Fzc20, and Ecm2201, a phosphatase (PPZ1), and mitochondrial adenylate kinase. In addition to gene losses, genome sequencing revealed that MA-1 harbors >1 copy of chromosome 12, as indicated in Figure 5. Overall, whole-genome sequencing revealed that MA-1 has undergone substantial genomic changes, including the loss of genes implicated in C neoformans pathogenesis which could explain the attenuated virulence for this strain. The mitochondrial genome exhibited a few small deletions and insertions, 1 of which is in an intron of CNAG_09000 (nicotinamide adenine dinucleotide dehydrogenase), but no indication of differential divergence between mitochondrial and nuclear genomes was found. DISCUSSION Cryptococcal cellulitis is an uncommon manifestation of C neoformans, and in immunocompromised hosts it more often reflects hematogenous dissemination rather than primary skin inoculation. Primary cutaneous cryptococcosis is rarer still, accounting for only a small minority of cryptococcosis cases. In the French National Cryptococcosis Registry, primary cutaneous cryptococcosis accounted for 28 of 1974 cryptococcosis cases (1.4%), compared with 80 cases of secondary cutaneous cryptococcosis attributed to hematogenous dissemination [27]. The current case is unusual in that the patient had biopsy-proven cryptococcal cellulitis with a markedly elevated serum CrAg titer, yet CSF studies showed no evidence of CNS involvement. This outcome was unexpected because high serum CrAg titers are predictors of meningitis: in non-HIV, nontransplant populations, a titer ≥1:64 predicts CNS disease (with a sensitivity of 83% and a specificity of 65%) [14], with similar correlations in solid organ transplant recipients [15] and persons infected with HIV [16]. Studies examining screening of advanced HIV cohorts found that roughly 1 in 10 patients overall were serum CrAg positive, and among those with CD4 cell counts ≤200/µL, approximately 40% were CrAg positive and nearly half of those had positive CSF CrAg results—indicating meningitis or imminent progression [28]. Longitudinal data show that most patients with antigenemia develop CNS disease if untreated [16]. Within this context, our patient's course and very high serum CrAg titer without meningitis is atypical. Clinically, our patient's presentation was most consistent with primary cutaneous cryptococcosis arising from direct inoculation at a skin-trauma site. The high serum CrAg titer does raise a concern that extracutaneous infection may have occurred even if the CNS was not involved. Pulmonary involvement was not documented but could not be entirely ruled out in the setting of abnormalities in chest imaging due to COPD, left upper lobe lung resection, and cryptogenic organizing pneumonia. Current guidelines recommend induction therapy with liposomal amphotericin B plus flucytosine for patients with high serum CrAg titers and extensive cutaneous involvement, even without proven meningitis [18]. This should then be followed by consolidation and maintenance therapy with fluconazole. Our patient was originally treated with fluconazole, but his condition failed to improve. He was then switched to induction therapy with liposomal amphotericin B plus flucytosine. However, after 6 days of amphotericin B, the patient experienced worsening renal function. He was switched to fluconazole plus flucytosine. In AIDS-associated cryptococcal meningitis, transitioning from amphotericin B to fluconazole after a short course (while continuing the flucytosine) is effective induction therapy [29, 30]. However, such short courses of amphotericin B have not been well studied in patients with cryptococcosis not associated with AIDS. Our case report suggests that shorter courses of amphotericin B should be trialed in HIV-negative individuals with cryptococcosis, especially those at high risk for toxicity. Clinical and environmental C neoformans isolates can differ markedly when tested in mouse virulence studies, even if closely related genetically, with differences observed even at the level of SNPs [6–8, 11]. Genotype variations have also been linked to clinical presentation and mortality rates in human cryptococcosis [8, 10, 12, 31]. The ability to grow at 37°C is critical for the virulence of the fungus and is controlled by a complex network of signaling pathways [32]. Some fungal skin pathogens grow poorly at 37°C but well at the cooler temperatures of the skin. MA-1 grew at 37°C in tissue culture medium but reached its peak cell density more slowly than KN99 (Figure 4A), a factor that could have contributed to the lack of dissemination from the skin to the CNS in our patient. In vivo, mice infected with MA-1 via a pulmonary challenge had significantly prolonged survival, with lower organ CFU counts compared with KN99 (Figure 4B–4E). Notably, however, MA-1 was neurotropic, as brain CFUs were recovered from most mice, even as early as 1 week postinfection, albeit at lower levels compared to KN99-infected mice. The spleens of MA-1–infected mice had undetectable CFUs even on day 35 after infection, which we speculate is due to the mice having minimal systemic dissemination and the absence of prolonged fungemia. At the genomic level, MA-1 is closely related to the reference strain H99 with alpha mating type within the VNI lineage of C neoformans [33], but it demonstrates notable structural differences, including loss of several genomic regions, most prominently an approximately 80-kb deletion on chromosome 5 and duplication of chromosome 12. These alterations represent substantial genomic remodeling and microevolution relative to H99/KN99. Interestingly, repeated laboratory passage of H99 and KN99 can result in genomic changes leading to either hypervirulence or hypovirulence [34, 35]. Moreover, some of the SNPs and genome variations in MA-1 are in the same genomic regions found in strains isolated in HIV-infected patients from sub-Saharan Africa [7, 8]. Subtelomeric regions, where these changes localize, are known hot spots for transposon activity and gene clustering that can influence stress responses, nutrient acquisition, and virulence traits [36, 37]. In our analysis, we also found that several break points that resulted in loss of subtelomeric regions are associated with the presence of transposable elements, including C. neoformans LINE-1 (Cnl1), Tcn1-long terminal repeat (LTR), and Tcn10-LTR. Among these, the Cnl1 elements have been shown to drive significant changes in the subtelomeric regions via their amplification [37, 38]. In other fungi, high recombination rates between subtelomeric regions have been shown to result in gene amplification and rapid evolution of these regions, affecting their persistence in the host [39, 40]. Finally, variations in chromosome copy number are well described in C neoformans clinical isolates, may evolve under the antifungal selection pressure, and have been associated with expression of virulence factors and antifungal resistance [41–44]. Specifically, chromosome 12 duplication, which is present in MA-1, has been identified as one of the common duplication events among clinical isolates in several studies [44–47]. While the exact phenotype resulting from this duplication is not clear, it has been implicated in fluconazole resistance and harbors loci that are implicated in thermotolerance. In addition, chromosome 12 harbors 324 genes, which include genes involved in galactose metabolism, target of rapamycin signaling, and transcriptional responses that could be directly relevant to virulence. Our brain CFU data show that although both MA-1 and KN99 ultimately reach the CNS, dissemination by MA-1 is slower and remains markedly lower at comparable time points. In murine models, CNS involvement typically occurs by day 14 after pulmonary inoculation; thus, the infection was plausibly captured before dissemination, a pattern aligned with impaired trafficking across the blood–brain barrier. The myo-inositol transporter deletions present in MA-1 may further contribute to this phenotype by disrupting genes required for efficient blood brain barrier traversal [25, 26]. Our data add to the growing literature demonstrating that patient isolates of C neoformans, even ones that are closely related genetically, can differ markedly in virulence [8–12, 31]. The phenotypic and genotypic factors responsible are only starting to be unraveled. The Chr12 duplication may be adaptive, compensating for gene losses elsewhere in the genome, but this hypothesis requires targeted functional validation. The mitochondrial genome appeared largely conserved, differing by only minor insertions and deletions, and overall sequence divergence was minimal (approximately 0.25%), underscoring MA-1's close phylogenetic relationship to H99 despite these chromosomal changes. Because the coverage plot does not resolve telomeric repeats, telomere integrity cannot be inferred. Functional studies—such as gene restoration or targeted correction of aneuploidy—will be needed to directly link genotype to phenotype. The coverage plot (Figure 5) and gene-loss table (Supplementary Table 1) summarize MA-1's structural architecture. In summary, this case highlights an unusual presentation of apparent primary cutaneous cryptococcosis in an older, corticosteroid-exposed patient with a markedly elevated serum CrAg titer but no demonstrable CNS involvement. The clinical course raised the possibility that disease was identified before CNS dissemination became detectable; however, the accompanying laboratory findings suggest that pathogen-specific factors may also have shaped this presentation. The causative strain, MA-1, remained neurotropic in mice, but compared with the reference strain, KN99 it had slower in vitro growth, delayed death, lower pulmonary and brain fungal burdens, and no detectable splenic dissemination. Genomic analysis further demonstrated that MA-1 had substantial remodeling, including subtelomeric gene losses and a chromosome 12 duplication. Taken together, these findings suggest that the patient's atypical course may have reflected both host and pathogen factors, and they underscore the value of studying clinical C neoformans isolates directly when clinical behavior diverges from expected patterns. Supplementary Material