Oesophageal mucormycosis managed with antifungal therapy alone with PET/CT for monitoring treatment response: a case report and a review of the literature

Abstract

Oesophageal mucormycosis is an uncommon life-threatening opportunistic fungal infection usually managed with antifungal therapy and aggressive surgical debridement. We describe a 54-year-old male who developed oesophageal mucormycosis 3 months post-allograft for multiply-relapsed leukaemia. Surgery was not performed due to prohibitive perioperative risk. He received 6 weeks of intravenous Liposomal Amphotericin B followed by long-term isavuconazole. Serial endoscopy and PET/CT imaging demonstrated resolution of infection to 7 months post-transplant where he died from relapsed leukaemia. We demonstrate that antifungal therapy alone may be efficacious where surgery is contraindicated and highlight the utility of PET/CT imaging for staging and monitoring this disease.

1. Introduction

Mucormycosis is a life-threatening opportunistic invasive fungal infection characterized by angioinvasion, thrombosis, tissue infarction and haematogenous dissemination. Common sites of involvement include rhino-orbito-cerebral, pulmonary, cutaneous, and gastrointestinal regions, with gastrointestinal involvement comprising only 5-10% of cases [1]. Within the gastrointestinal tract, the stomach, colon and ileum are most frequently affected, whereas oesophageal involvement is rare and has only been described in case reports.

Standard management of mucormycosis involves aggressive surgical debridement of involved tissue, systemic antifungal therapy and correction of predisposing factors including concomitant immunosuppression [1]. However, in cases of oesophageal mucormycosis, surgical intervention may not be feasible due to prohibitive perioperative risks associated with oesophagectomy in critically ill patients [2].

We describe a patient successfully managed with antifungal therapy alone, with sustained remission confirmed through serial endoscopic evaluations and positron emission tomography (PET) imaging; the use of the latter as a marker of therapeutic response to our knowledge has not been previously described in this context.

2. Case presentation

A 54-year-old male of Portuguese descent underwent a matched-unrelated-donor allogeneic hematopoetic stem cell transplantation for multiply-relapsed B cell acute lymphoblastic leukaemia in March 2022. Prior treatment included multiple chemotherapy regimens for marrow and central nervous system disease, and chimeric antigen receptor T-cell therapy in June 2021. Comorbidities included acquired hypogammaglobulinaemia on monthly subcutaneous immunoglobulin and C282Y homozygous hemochromatosis with iron overload (pre-allograft ferritin 2296ug/L, transferrin saturation 87%) managed with venesection and deferasirox, and a 40 pack-year smoking history having stopped smoking in 2019.

Pre-transplant computed tomography (CT) imaging of the sinuses, chest, abdomen, and pelvis did not identify an abnormality consistent with invasive fungal infection, though there was the presence of multiple spiculated peribronchovascular bilateral upper lobe lesions associated with emphysematous changes in the parenchyma. For further investigation of these upper lobe lung lesions, a bronchoscopy was performed, showing macroscopically normal airways, and cultures of bronchoalveolar lavage washings for bacteria, fungi and acid-fast bacilli were all negative. Additionally, galactomannan antigen testing and PCR assays on bronchoalveolar lavage washings for Pneumocystis jirovecii, Mycobacterium tuberculosis and cytomegalovirus (CMV) were negative. Donor and recipient pre-transplant CMV and Epstein-Barr virus were IgG positive and the patient had latent tuberculosis planned for eradication post-transplant.

Conditioning consisted of fludarabine (25mg/m²) and melphalan (140mg/m²). Graft-versus-host (GVHD) prophylaxis consisted of thymoglobulin (4.5mg/kg), cyclosporine and methotrexate. Early complications included grade 4 mucositis requiring prolonged total parenteral nutrition and culture-negative febrile neutropenia treated with broad spectrum antibiotics until neutrophil count recovery (neutrophils >0.5x10⁹/L) on day 18 post-transplant. Clostridioides difficile colitis was diagnosed on day 6 and responded to a 14-day course of oral vancomycin and i ntravenous (IV) metronidazole. Oral posaconazole was commenced on day −1 (pre-transplant) and switched to IV formulation on day 6 due to mucositis; trough levels remained subtherapeutic until day 17 (0.97mg/L).

Cross-sectional CT imaging at day 20 for new fevers after neutrophil recovery demonstrated marked thickening of the thoracic oesophageal wall extending to the gastro-oesophageal junction. The remaining GI tract appeared normal with no signs of typhlitis. The oesophageal thickening was presumed mucositis-related and not further evaluated. He was discharged on day 26 on oral posaconazole prophylaxis at 300mg daily. CMV DNAemia (peak viral load 1376 IU/mL) without evidence of CMV disease was noted on day 25. This was treated with induction valganciclovir from day 27 to day 53 and the patient then resumed CMV pre-emptive therapy monitoring.

At day 82 he reported two weeks of progressive dysphagia to solids and liquids, regurgitation, and post-prandial vomiting. CT revealed improved but persisting circumferential oesophageal thickening, more localized compared to day 20 imaging. Gastroscopy at day 84 identified an ulcerated exophytic 3cm lesion partially obstructing the mid-oesophagus (Fig. 1A). The stomach and duodenum appeared macroscopically normal, and a paediatric nasogastric tube was inserted under endoscopic guidance for nutrition due to 15% weight loss since allograft.

Fig. 1.

(A) Gastroscopy image of an ulcerated exophytic 3cm mid-oesophageal fungating lesion with raised edges and superficial slough causing partial obstruction to the oesophagus. The adjacent mucosa was congested, erythematous and friable with contact bleeding. (B) Thick ribbon-like pauci-septate fungal hyphae with right angle branching of Mucorales species (arrows), co-existing with clumps of filamentous delicate Actinomyces-like organisms (arrowheads). H&E x400.

The lesion was biopsied and histopathology showed necrotic tissue and fungal hyphae morphologically consistent with Mucormycetes species (Fig. 1B). Gram-positive Actinomyces-like organisms were also identified on histology but not isolated on culture of fresh tissue. Given the uncertainty of whether this represented a co-infection or colonization, a decision was made to treat empirically with one month of IV ampicillin followed by five months of oral suspension amoxicillin. CMV and Herpes Simplex virus immunohistochemistry were negative. As tissue samples were not sent for culture, another gastroscopy was performed three days later (day 87) for repeat biopsy. Fungal culture of oesophageal biopsy tissue was negative despite 28 days of incubation on Sabouraud dextrose agar plates and a brain-heart-infusion broth slope. However, PCR and DNA sequencing of oesophageal biopsy tissue detected Rhizopus microsporus DNA, confirming aninvasive mould infection.

Risk factors for mucormycosis included significant soil exposure as a gardener, and iron overload in the setting of homozygous C282Y hemochromatosis as well as multiple red cell transfusions in the setting of pancytopenia with a pre-allograft ferritin of 2296ug/L, and a day 42 ferritin of 1791ug/L. His albumin level was normal at 37g/L on day 84 when he presented with symptomatic infection, though this was noted to be lower in the first four weeks of transplant with a nadir of 22g/L on day 18. He was not diabetic, had no recent steroid use and he had not experienced neutropenia below 1 x 10^9/L since engraftment or GVHD requiring additional immunosuppression; cyclosporine had been progressively tapered over the preceding months to 25mg BD at day 82. Posaconazole was continued and therapeutic drug monitoring demonstrated subtherapeutic levels at day 60 (0.68mg/L) and day 74 (0.59mg/L), though a therapeutic level (1.5mg/L) was documented on day 82 (hospital admission).

IV liposomal amphotericin B (LAMB) (5mg/kg daily) was commenced and cyclosporine ceased with subsequent lymphocyte recovery (from 0.4x10⁹/L at time of diagnosis to >1.0x10⁹/L one week later). CT chest, brain and sinuses showed no evidence of dissemination. PET/CT showed localised fluoro-deoxyglucose (FDG) uptake at the mid-oesophageal lesion.

Endoscopic resection was contraindicated due to lesion depth and size, while surgical resection was contraindicated due to prohibitive perioperative risks. After two weeks of LAMB, repeat gastroscopy identified a severe 3mm diameter oesophageal stenosis, appearing benign and not biopsied. He underwent fortnightly oesophageal dilatations over two months; the oesophageal lesion continued to appear benign macroscopically. PET/CT scans after 3 and 6 weeks of LAMB showed resolution of the metabolic activity. After 6 weeks of LAMB, the patient was transitioned to isavuconazole due to concerns about acquired posaconazole resistance and the lack of a cultured isolate for antifungal susceptibility testing. Isavuconazole capsules were administered enterally via nasogastric tube for 2 weeks with a loading dose of 200mg three times a day for 2 days, followed by maintenance dosing at 200mg daily.

After 8 weeks of nasogastric enteral feeding, he could tolerate a pureed diet and swallow tablets, and was transitioned to oral isavuconazole with an intention to continue for one year. He remained clinically stable with three-monthly interval PET/CT scans and endoscopic monitoring while on isavuconazole. However, PET/CT at 6 months post-transplant showed relapsed leukaemia which was confirmed on bone marrow biopsy, with no signs of recurrent Mucormycosis. He died 7 months post-transplant from relapsed leukaemia.

3. Discussion

Risk factors for mucormycosis in this case included an allograft for a haematological malignancy, prolonged severe neutropenia post-transplant (<0.2 x10^9/L for 15 days), T-cell depletion with cyclosporine, iron overload from haemochromatosis and recurrent transfusions for pancytopenia as well as possibly, subtherapeutic posaconazole levels. The iron chelator deferasirox was notably withheld during the peri-transplant period which undoubtedly contributed to the iron overload state. Mucorales are environmental fungi found in soil and decaying organic matter, acquired typically from spore inhalation or ingestion [1]. The patient was an avid gardener, suggesting possible pre-transplant colonization, with severe mucositis providing a portal of entry. In retrospect, the oesophageal wall thickening seen on day 20 CT may have represented early mucormycosis. Recognition of the clinical significance of subsequent progressive dysphagia may have been delayed by communication difficulties with the patient who was non-English speaking. Hypoalbuminaemia could have been another risk factor for Mucormycosis infection in our case, with a mechanistic study suggesting albumin-bound free fatty acids suppress protein synthesis in Mucorales species [3]. Although albumin levels were normal at the time of symptomatic disease, they were low during the first four weeks post-transplant when early infection was suspected.

The most important observation of this case is that antifungal therapy, in conjunction with immune reconstitution, may be effective in oesophageal mucormycosis when surgical resection is contraindicated. While surgical debridement in mucormycosis is associated with improved survival, especially for non-oesophageal gastrointestinal, renal and rhino-orbito-cerebral involvement [4], the rarity of adult oesophageal mucormycosis makes the role of surgery difficult to delineate. To our knowledge only nine such cases have been reported which we have summarized (Table 1) [[5], [6], [7], [8], [9], [10], [11], [12], [13]]. Two were managed with surgical resection, including one patient who underwent oesophagagectomy and subsequently died from multiorgan failure [7], and one patient who underwent partial gastrectomy and survived out to 3 months [9]. Of the seven patients managed without surgical resection, four (57.1%) survived. Factors contributing to recovery in this case include the use of IV LAMB, a first line antifungal agent for the treatment of invasive mucormycosis, the early commencement of which is associated with reduced mortality [1] and immune recovery in the context of cessation of cyclosporine and resolution of lymphopenia, noting that CD4 T cell immune recovery is delayed after thymoglobulin GVHD prophylaxis [14]. Current guidelines recommend dosing of LAMB of 5-10mg/kg/day with higher doses of 10mg/kg/day considered in cases of central nervous system involvement or solid organ transplant recipients [15]. In our case, given relative clinical stability and to balance risks of nephrotoxicity and electrolyte disturbances associated with higher doses of IV LAMB, a 5mg/kg/day dosing strategy was selected. After an induction period of 4-6 weeks with IV LAMB, therapy is typically changed to either oral posaconazole or isavuconazole, with the choice of agent determined by antifungal susceptibility, the site of disease, drug-drug interactions and cost [4]. Total antifungal treatment duration is often individualized based on clinical, biochemical, radiological and endoscopic assessment.

Table 1.

Previous cases of oesophageal mucormycosis and our case highlighted below.

Authors/Year	Age/sex	Predisposing conditions	Additional sites	Presentation	Antifungal prophylaxis	Imaging	Endoscopic findings	Diagnosis	Medical Therapy	Surgery	Outcome
Mezhir et al., 2009 (4)	50M	ESRF on HDx, polysubstance and steroid use	Stomach and liver	Abdo pain and haematemesis	Nil	Nil of oesophagus CT: liver abscess	Necrotic ulcers	Histology and fungal culture	LAMB > maintenance posaconazole for 1 year	No	Survived: Resolution of liver abscess on serial CT
Boatright et al., 2014 (5)	63M	APML, T2DM	Nil	Dysphagia	Nil	CT: oesophageal thickening	Necrotic tissue, wall thickening	Histology	LAMB + echinocandins	No	Died 24 days after diagnosis (unclear cause)
Kraft et al., 2014 (6)	55M	AML	Nil	Fever, dysphagia, neutropenia	Posaconazole	CT: oesophageal thickening and mediastinitis	Ulceration	Histology and fungal culture	LAMB	Yes	Died day 5 post surgery due to multiorgan failure
Nandwani et al., 2015 (7)	45M	Renal Transplant	Stomach	Epigastric pain, vomiting	Nil	Nil	Haemorrhagic ulcers in oesophagus	Histology	LAMB	No	Died day 2 from GI perforation
Raviraj et al., 2015 (8)	19F	HUS	Stomach	Fever, haematemesis and melaena	Nil	CT: gastric ulcer with perforation	Bleeding ulcers in oesophagus and stomach	Histology and fungal culture	LAMB + posaconazole	Yes	Survived: Resolution at gastroscopy at 3 months
Gani et al., 2019 (9)	79M	Renal transplant, T2DM	Stomach	Dysphagia and odynophagia	Nil	Nil	Oesophagitis	Histology and fungal culture	Isavuconazole	No	Survived: Partial resolution on gastroscopy at day 20
Ringer et al., 2020 (10)	60M	AML	Stomach	Fever, odynophagia, neutropenia	Voriconazole	CT: oesophageal thickening	Severe circumferential oesophagitis	Histology and fungal PCR	Anidulafungin > LAMB.	No (only GI stenting)	Survived: Resolution on gastroscopy at day 27
Rajarajen et al., 2022 (11)	56M	Renal transplant, T2DM	Nil	Chest pain, progressive dysphagia, productive cough	Nil	Nil	Necrotic oesophageal mucosa	Histology	LAMB	No	Died at 5 days post diagnosis (unclear cause)
Harikrishnan et al., 2023 (12)	41M	T2DM, DKA	Nil	Malena, haemorrhagic shock	Nil	CT: oesophageal and GOJ thickening	Oesophageal ulcer with “unhealthy” appearing mucosa	Histology	Oral isavuconazole - > posaconazole	No	Survived: Partial resolution on gastroscopy at 1 month
Our case	54M	ALL, Iron overload, mucositis, neutropenia, cyclosporine	Nil	Dysphagia, vomiting	Posaconazole	CT: oesophageal thickening PET/CT avidity	Oesophageal obstructing mass, friable mucosa	Histology and fungal PCR	LAMB - > isavuconazole	No	Survived: Resolution on PET and endoscopy at 6 weeks

Abbreviations for table: ALL = Acute lymphoblastic leukaemia, APML = Acute promyelocytic myeloid leukaemia, CT = Computed tomography, CMV = Cytomegalovirus, DKA = Diabetic ketoacidosis, ESRF = End stage renal failure, GI = Gastrointestinal, GOJ = Gastro-oesophageal junction, HDx = Haemodialysis, HUS = Haemolytic uraemic syndrome, LAMB = Liposomal Amphotericin B, PCR = Polymerase chain reaction, PEG = Percutaneous endoscopic gastrostomy, PET = Positron Emission Tomography, T2DM = Type 2 diabetes mellitus.

Regarding the use of PET/CT, it served as a useful companion to endoscopy for staging and monitoring treatment response in our case. It helped confirm alongside routine CT scanning the absence of disseminated disease. After 6 weeks of intravenous LAMB, a complete metabolic response was seen and guided the transition to oral isavuconazole, enabling early cessation of potentially nephrotoxic therapy. Compared to conventional CT or MRI, PET/CT provides combined anatomical and functional assessment, enabling detection of metabolically active lesions. While PET/CT has been used for diagnosis and staging in gastrointestinal mucormycosis [16,17] its role in monitoring treatment response has not previously been described. In other invasive fungal infections (IFI), including Aspergillosis and Candidiasis, PET/CT is superior to CT for staging and monitoring response to therapy. In staging, it can detect 30% more disseminated disease than conventional CT at diagnosis, particularly in hepatosplenic disease, and performed well even in patients with severe neutropenia [18]. In monitoring, 61% of PET/CT showed normalization of FDG activity despite a residual lesion on conventional CT [18], thereby guiding de-escalation of therapy, suggesting that PET avidity may be more specific than CT in detecting resolution of infection. PET/CT was found to alter management in 50% of patients, resulting in cessation or switch to a different antifungal agent in an observational study on IFI [19].

Regarding microbiological diagnosis, species-specific PCR on fresh tissue was crucial in confirming a proven invasive mould infection in the setting of negative fungal cultures, supported by histopathology identifying fungal hyphae accompanied by evidence of associated tissue damage [20]. While histopathology can help differentiate invasion from colonization, accurately distinguishing Mucorales from other moulds such as Aspergillus species requires experienced personnel. Compared with culture, which yields positive results in only half of Mucorales species due to hyphae friability [2], PCR offers a sensitivity between 73 and 97%, depending on specimen type tested [21]. In culture-negative cases where histopathology is suggestive of mould infection, PCR may allow for species-level identification which can be critical in diagnosis and guiding empirical antifungal therapy. Limitations of fungal PCR include limited access to specialist laboratories required to run the test, prolonged turnaround times, and the inability to provide information on antifungal susceptibility. Moreover, the performance of this assay varies depending on the clinical specimen submitted, with a systematic review reporting the highest sensitivity in bronchoalveolar lavage fluid (97.5%), followed by tissue (86.4%), blood (81.6%) and formalin-fixed paraffin-embedded tissue specimens (73.0%) [22].

Regarding the development of this infection on posaconazole prophylaxis, one consideration was invasion with a posaconazole-resistant Mucorales species, which was deemed unlikely. A retrospective review of 854 clinical Mucorales isolates from a North American reference laboratory between 2015 and 2020 identified non-wild type posaconazole susceptibility in only 3% of isolates [23]. We suspect that subtherapeutic levels may have led to impaired resolution of initial infection and/or emergence of acquired resistance resulting in delayed presentation of symptomatic disease. Interestingly amongst reported oesophageal mucormycosis cases (Table 1), two of nine patients were on antifungal prophylaxis (posaconazole) at the time of diagnosis, although drug serum levels were not documented.

In summary, we describe the successful treatment of localised oesophageal mucormycosis infection in a stem cell allograft recipient managed with antifungal therapy and withdrawal of immunosuppression without surgery, and the utility of PET/CT imaging for staging and monitoring this disease. We also summarize the literature on oesophageal mucormycosis and highlight potential areas of research including the need for validation of PET/CT imaging use in Mucormycosis.

CRediT authorship contribution statement

Ethan X. Tan: Writing – review & editing, Writing – original draft, Project administration, Data curation. Brennan Collis: Writing – review & editing, Data curation. Natasha Holmes: Writing – review & editing, Supervision. Julie Lokan: Resources. Andrew Grigg: Writing – review & editing, Supervision, Conceptualization.

Conflict of interest

The authors have no conflicts of interest to disclose.

There was no funding received for the conduct of this study.

Informed consent was obtained for publication of this case report.