Original Research

Culture-positive mould infections in patients with haemato-oncological diseases

Bonita van der Westhuizen, Stephanie Kennedy, Liska Budding, Samantha Potgieter

Received: 11 Feb. 2025; Accepted: 21 May 2025; Published: 31 July 2025

Copyright: © 2025. The Author(s). Licensee: AOSIS.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract

Background: Moulds are increasingly recognised as important pathogens carrying a high morbidity and mortality in critically ill and immune-compromised patients, including those with haematological neoplasms (HNs) and aplastic anaemia (AA). Our understanding of these diseases remains incomplete, largely due to the lack of surveillance data.

Aim: This study aimed to describe the distribution, patient characteristics, risk factors, therapy and treatment outcome in culture-positive mould infections in patients with haematological conditions.

Setting: The study was conducted at Universitas Academic Hospital (UAH), Bloemfontein, South Africa.

Methods: All mould isolates cultured from sterile and selected respiratory specimens were identified retrospectively from 01 July 2014 to 30 June 2017. Laboratory and clinical data were reviewed for patients meeting the inclusion criteria.

Results: Eleven patients with underlying HNs (n = 9/11; 81.8%) or AA (n = 2/11; 18.2%) were identified. Acute myeloid leukaemia was the most common HN. Aspergillus species were most frequently isolated, followed by mucoraceous moulds. Histological examination proved a useful adjunct in the diagnosis of these infections. Risk factors for invasive mould infections (IMIs) included chemotherapy and antimicrobial exposure. Most patients had prolonged severe neutropenia. Seven (63.6%) patients received antifungal therapy. Only two patients underwent surgery. The mortality rate was 45.5% (n = 5/11).

Conclusion: Invasive mould infections are a major cause of morbidity and mortality in patients with HNs or AA, especially in the setting of severe neutropenia.

Contribution: The findings complement current evidence on IMIs, especially in patients with HNs and AA in the province, and will stimulate future research in this field.

Keywords: invasive mould infection; treatment; outcome; haematological conditions; leukaemia; aplastic anaemia; neutropenia.

Introduction

Fungi are increasingly recognised as important pathogens associated with significant morbidity and mortality in critically ill and immune-compromised patients.^1,2,3,4 The risk of invasive mould infections (IMIs) is particularly high in patients with haematological neoplasms (HNs) and/or aplastic anaemia (AA).⁵

Data on mould infections in South African patients are limited, probably due to diagnostic challenges as symptoms are often non-specific and available diagnostic tests lack sensitivity and specificity. Epidemiological data to guide the treatment of IMIs in patients with haemato-oncological diseases at Universitas Academic Hospital (UAH), Bloemfontein, South Africa, are lacking. Local culture data are needed to optimise treatment and improve outcomes in these high-risk patients.

The aim of this study was to characterise the epidemiology, clinical and laboratory features, treatment, and outcomes of patients with haematological diseases and IMIs admitted to UAH for a 3-year period extending from July 2014 to June 2017.

Methods

Study design and setting

A retrospective, observational descriptive study was conducted, investigating patients with haematological diseases and fungal infections admitted to UAH between 01 July 2014 and 30 June 2017. All patients whose clinical specimens yielded positive fungal cultures were included. Patients with culture-positive yeast infections were excluded. Moulds isolated from non-sterile sites were excluded, except for selected bronchoalveolar lavage (BAL) fluid, endotracheal aspirates (ETA) and sputum specimens. Positive cultures from these non-sterile respiratory samples were considered significant if the same organism was cultured from a second specimen, and the patient had clinical and/or radiological features of an invasive fungal respiratory tract infection (RTI). Signs and symptoms that were considered supportive of a fungal RTI included tachypnoea, cough, fever, dyspnoea, haemoptysis or chest pain. Radiological features that were in keeping with a possible fungal RTI as reported by an experienced radiologist were taken as supportive evidence on imaging studies.

Sample analysis

Isolates were cultured at the Department of Medical Microbiology, National Health Laboratory Service (NHLS) Universitas Academic Laboratory. Samples were processed according to the laboratory’s standard protocols, being inoculated onto two Sabouraud dextrose agar plates and incubated at 25°C and 37°C for up to 14 days. Plates were inspected daily for fungal growth. Once growth was detected, the colony’s macroscopic characteristics were recorded, and a lactophenol cotton blue stain was used for microscopic identification. Isolates that were challenging to identify were sent to the mycology laboratory at the National Institute of Communicable Diseases (NICD) for broad-range fungal polymerase chain reaction (PCR) analysis.

Histology reports were evaluated where available. Pathologists at the Department of Anatomical Pathology, NHLS Universitas Academic Laboratory reported the histological features.

Data collection

Mycology culture results were retrieved from the NHLS Central Data Warehouse. All samples submitted for fungal culture at Universitas Academic Laboratory during the study timeframe were evaluated. Patients with mould-positive cultures were identified, and their clinical information was extracted from the electronic patient management system, Meditech. Where electronic records were incomplete, paper files were consulted.

Collected variables included demographic details (age, sex), type of specimen, mould species isolated, presenting symptoms, imaging findings, underlying diagnoses, additional risk factors, management, and patient outcomes. Outcomes were classified as improved, unchanged, in-hospital death, or unknown at 12 weeks post-sampling. Supplementary laboratory data – such as histology, neutrophil counts, C-reactive protein (CRP), procalcitonin (PCT), human immunodeficiency virus (HIV) status, CD4 count, and bacterial culture results – were obtained from the NHLS TrakCare Laboratory Information System. Patients lacking sufficient clinical data were excluded.

Data were captured on a Microsoft Excel (version 2016) spreadsheet (Microsoft Corporation; Redmond, WA, United States) and analysed by the Department of Biostatistics, University of the Free State (UFS).

Data analysis

Continuous variables were described using medians, minimums, maximums or percentiles, while categorical variables were reported as frequencies and percentages. Group differences were assessed using the chi-square or Fisher’s exact test for unpaired data. The analysis was performed by the Department of Biostatistics, UFS, using SAS version 9.4 (SAS Institute Inc.; Cary, NC, United States).

Ethical considerations

Permission to conduct the study was obtained from the NHLS business manager, the Head of the Department of Medical Microbiology and the Free State Province Department of Health. Ethical approval was granted by the Health Sciences Research Ethics Committee (HSREC) of the University of the Free State (reference number: UFS-HSD2017/1122). Due to the retrospective nature of the study and only using archived patient records for data collection, informed consent was not required. Confidentiality and anonymity were ensured by allocating a number to each patient’s record and excluding all personal information.

Results

During the 3-year study period, 988 specimens were received by the Medical Microbiology laboratory for mycology culture. Figure 1 shows that, of the 48 culture-positive moulds isolated from patients with clinical information available, 11 (22.9%) were from patients with underlying haematological conditions. These isolates represented 1.1% of the total number of mycology requests and 6.1% of all positive cultures over the 3 years. One patient had two separate mould infections diagnosed a week apart. All patients were admitted to the UAH clinical haematology ward and/or the multidisciplinary intensive care unit (MICU).

FIGURE 1: Moulds isolated from patients with haematological neoplasms (HN) and aplastic anaemia (AA) over the 3-year study period.

The baseline patient characteristics and results of all variables measured are summarised in Table 1. Table 2 quantifies the distribution of underlying haematological conditions, the mould species isolated and their significance, histology results, associated neutropenia, treatment and outcome data.

The median age of this cohort was 38 years and included five (45.5%) male and six (54.5%) female patients. Nine patients (81.8%) had a HN, while two (18.2%) had severe AA. Four patients had undergone imaging studies, of whom two (50%) had results suggestive of an invasive fungal infection (IFI).

Figure 2 and Figure 3 show the histological evidence of an IMI in one of the patients in this study.

FIGURE 2: Specimen from the lower lip of a 22-year-old male patient with acute myeloid leukaemia who developed a lip ulcer. Haematoxylin and eosin (H&E) stain showing numerous fungal hyphae with (a) acute angle branching (*) and (b) septations (arrows) in the lower lip. The presence of angio-invasion in this specimen (c) resulted in a histopathological differential diagnosis of angio-invasive mucormycosis. Aspergillosis was diagnosed on microbiological culture.

FIGURE 3: Specimen from the lower lip of a 22-year-old male patient with acute myeloid leukaemia who developed a lip ulcer. Periodic acid-Schiff (PAS) (a) and Grocott (b) histochemical stains highlighting fungal hyphae.

Discussion

Mould infections represent an important cause of morbidity and mortality in patients with underlying haemato-oncological diseases.^1,2,3,4,5 Numerous risk factors predispose this patient population to IMIs. These include disease- or chemotherapy-related neutropenia, exposure to broad-spectrum antibiotics, haematopoietic stem cell transplant (HSCT), and immune therapies.^1,2,3,4,5 Despite recent advances, IMIs remain difficult to diagnose and treat, especially in resource-constrained settings. However, increased awareness and early detection may improve outcomes in this high-risk patient population.⁶

This single-centre study is the first to describe the epidemiology of IMIs in the setting of HNs and AA in central South Africa.

Mould species isolated

The most common mould species isolated on mycology culture was Aspergillus species in 36.4% (n = 4/11) of patients, which was similar to findings reported in the literature.⁷ Consistent with other studies that have demonstrated an increase in infections with the Mucorales, Fusarium species and Bipolaris species,⁸ the mucoraceous moulds were the second most common fungal isolate in our setting. This information is important when a decision regarding empiric therapy must be made, as different antifungal agents cover different mould species.⁹

Underlying haematological conditions

The distribution of HNs in our cohort was in keeping with other studies, which demonstrated an increased risk of IMIs in acute myeloid leukaemia (AML) compared to lymphoma.⁵ All patients with IMIs caused by a true pathogen (n = 8/12; 66.7%) had either acute leukaemia (n = 5/11; 45.45%) or severe AA (n = 2/11; 18.2%). As shown in Table 1, two cultured moulds (Chaetomium and Cladosporium species) were regarded as contaminants and were both isolated from patients with lymphoma (n = 2/11; 18.2%). The pathogenicity of one isolate (n = 1/12; 8.3%) (Fusarium species) cultured from a patient with AML was doubtful, since histological evaluation was negative for fungal elements and confirmed myeloid sarcoma, while the patient improved without antifungal therapy. This case highlights the value of multidisciplinary investigations, in particular microbiology and histopathology, in the diagnosis of IMIs.¹⁰

Risk of invasive mould infections in patients with haematological neoplasms

Patients with AML and allogeneic HSCT recipients have a high risk of IMIs.⁵ Additional factors increase the risk of opportunistic mould infections, independent of the underlying disease, including neutropenia, relapse and/or refractory disease, previous history of IFIs, salvage therapy, and high dose corticosteroids.⁵

As part of a larger study at our centre, Van der Westhuizen and Potgieter¹¹ identified HNs as a common risk factor in patients with culture-positive mould infections, second only to HIV. However, mortality in the cohort with haematological diseases was three-fold the mortality rate observed in the HIV-positive group (45.1% vs 14.3%).¹¹

In this study, only two (18.2%) of the 11 patients with underlying haematological diseases were also HIV-positive, with CD4 counts of 484 cells/µL and 112 cells/µL, respectively. Although HIV infection with a low CD4 count is a known risk factor for IFIs,¹² no conclusions could be drawn from our small sample.

Neutropenia, which may be chemotherapy- and/or disease-related in patients with HNs, is a well-described risk factor for IFIs, especially if severe and long-lasting.^1,2,3,4,5,13 Patients with prolonged neutropenia (≥ 7 days), such as patients with AML or myelodysplastic syndrome (MDS) receiving remission induction chemotherapy, or severe AA, continue to represent the group with the highest risk of IFIs.⁵ In this study, seven (63.6%) patients were neutropenic, with a median neutrophil count of 0.13 × 10⁹/L. Six patients (54.5%) had severe neutropenia (< 0.5 × 10⁹/L), and one patient (9.1%) had moderate neutropenia (< 1.0 × 10⁹/L). Neutropenia was prolonged ≥ 7 days in all patients where the duration was recorded. Three patients (27.3%) had normal neutrophil counts (> 1.5 × 10⁹/L) at the time of diagnosing the fungal infection. In two of these patients (nr. 2 and nr. 9), no prior laboratory records were available to evaluate for recent neutropenic events. The fourth patient had a neutrophilic leucocytosis of 126.76 × 10⁹/L. In the latter patient, who had chronic myelomonocytic leukaemia (CMML), the clonal neutrophils likely had impaired function, which may have predisposed the patient to an IMI. In CMML and myelodysplastic neoplasms (MDS), the clonal myeloid cells harbour genetic aberrations that may lead to abnormal behaviour and physiology.¹⁴

Febrile neutropenia is characterised by an oral temperature of > 38.3ºC, or two consecutive readings of > 38.0ºC for 2 h, in a patient with an absolute neutrophil count of < 0.5 × 10⁹/L, or where it is expected to fall below 0.5 × 10⁹/L.¹⁵ Early in the course of neutropenia, IFIs are a notably less common cause of fever compared to bacterial infections, and when identified, Aspergillus species and Candida species are the most commonly isolated fungi.¹³ However, the risk for IMIs increases markedly when the neutropenia is prolonged, typically > 7–14 days.^5,13 Therefore, antifungal prophylaxis with mould-active azoles is recommended in haematology patients with long-lasting neutropenia (< 0.5 × 10⁹/L for a duration of ≥ 7 days), independent of the underlying disease.⁵ It is not recommended where the expected duration of neutropenia is < 7 days, as the risk of IFIs is not significantly increased.⁵ When implementing antifungal prophylaxis regimens, institutional epidemiology should be considered and adjustments made accordingly.¹⁶ Unfortunately, we could not report on the use of antifungal prophylaxis in this study due to missing data from the clinical notes.

Treatment with chemotherapy, immunosuppressive agents, and antibiotics were additional risk factors in this cohort. Eight (72.7%) patients received chemotherapy. The two patients with AA were treated with equine anti-thymocyte gamma-globulin (ATGAM) and steroids. Eight (72.7%) patients were receiving antibiotics at the time of sample collection. No patient was treated with HSCT, radiotherapy or novel immune therapies.

When considering the clinical diversity of patients in our cohort, the risk of IMIs may vary significantly. If underlying risk factors are not appropriately evaluated, the possibility of over- or insufficient treatment becomes likely.¹⁷

Outcome

The mortality in this haematology cohort was 45.1% (n = 5/11). Due to the inherently complicated clinical nature of patients with HNs, causality of death could not be assumed to be the mould infection. However, these infections were probably a contributory factor.

Three (27.3%) patients in whom the isolate was considered a true pathogen improved with appropriate antifungal therapy. All the patients in whom the mould was considered a contaminant (n = 3/11; 27.3%) or of uncertain significance (n = 1/11; 9.1%) improved without antifungal therapy.

Diagnostic and treatment challenges of invasive fungal infections in patients with haematological neoplasms

Despite recent advances in the diagnosis and treatment of IFIs, patients with HNs and AA remain a difficult group to treat.⁵ Challenges exist in terms of the availability of and high costs related to both antifungal prophylaxis and treatment.¹⁷ Furthermore, moulds are ubiquitous environmental organisms and are, therefore, a frequent contaminant of clinical specimens.^18,19,20 However, prompt diagnosis of IFIs remains imperative to reduce morbidity and mortality in high-risk patient populations.⁶

Diagnostic challenges

Fungal infections are usually difficult to diagnose due to their non-specific clinical features. The lack of dependable available diagnostic methods poses a challenge to determine the actual extent of diseases caused by fungi.²¹ Available diagnostic modalities in the laboratory diagnosis of IFIs include mycology culture, molecular identification, histology and serological biomarkers.^10,22 These tests are limited by availability, the requirement for adequate specimens, contamination by environmental organisms, and long turn-around times. A single laboratory investigation is not ideal when diagnosing fungal infections and a multidisciplinary approach, including at least both histopathological and microbiological examination, is recommended for the diagnosis of invasive fungal disease.^10,22 Furthermore, laboratory results should always be interpreted in conjunction with radiological and/or clinical findings.⁹

Histopathology

Although microbiological culture and molecular investigations provide sufficient evidence of an IFI in the appropriate clinicopathological context, histopathological examination provides valuable information. The reported sensitivity of histopathological methods for diagnosing IFIs is ~78%, compared to 8% – 60% for culture.^23,24

Histopathological examination can rapidly alert clinicians to the presence of fungal organisms that may require extended incubation periods or specific growth media.^10,25 Histopathological investigation also allows pathologists to determine whether a fungal infection has elicited an inflammatory response, thus distinguishing fungal contamination from true fungal infection.¹⁰ Not only is histopathological examination important for identifying fungal organisms, but histopathologists can also provide alternate diagnoses of non-infectious aetiologies, such as malignancies.²²

On histopathological examination, morphological features of fungi are similar among genera and species.⁸ Histochemical stains, including periodic acid-Schiff (PAS), Grocott-Gomori methenamine silver (GMS) and Masson Fontana stains, are beneficial ancillary tests that aid in the identification of fungi.^10,25 However, morphological examination is not sufficiently specific to allow accurate species identification.⁸ In the presence of certain inflammatory reaction patterns (e.g. granulomatous inflammation, angio-invasion or abscess formation), a fungal infection can be suspected, even when histochemistry is negative for fungal elements.¹⁰ Polymerase chain reaction-based techniques and/or microbiological cultures are required before a final, accurate diagnosis can be rendered.^8,10 Polymerase chain reaction-based techniques can also be employed on formalin-fixed paraffin embedded (FFPE) tissues.^10,22 However, this method yields more reliable results when performed on fresh tissue samples.

Radiology

The European Organisation for Research and Treatment of Cancer/Invasive Fungal Infections Cooperative Group and the National Institute of Allergy and Infectious Diseases Mycoses Study Group (EORTC/MSG) definitions for the diagnosis of IMIs include radiological criteria to assist with these diagnostically challenging infections. Radiological patterns are divided into angio-invasive or airway-invasive forms, and criteria for aspergillosis and other pulmonary mould diseases are also separated.⁹ In our study, of the four patients who had undergone imaging studies, two (50%) had results suggestive of an IFI. The first patient (nr. 4) had non-specific radiological findings, but fungal cultures positive on two different sample types. The second patient (nr. 5) had radiological features suggestive of an IFI together with positive culture and histological findings. Patient nr. 8, on the other hand, had non-specific radiological findings paired with positive culture and histological findings. Lastly, the fourth patient (nr. 9) had radiological findings suggestive of an IMI paired with positive culture findings. These findings highlight the fact, as stated previously, that a single diagnostic test is not sufficient for the diagnosis of IFIs.

Other biomarkers

The value of inflammatory biomarkers for the diagnosis of IFIs in patients with HNs has been investigated, but no clear roles have been defined. Some authors have shown increased CRP and low PCT levels to be indicative of IFIs in adult haematology patients with neutropenic fever.²⁶ In our study, nine (81.8%) patients had only CRPs done. The median CRP value at the time of specimen collection was 197.4 mg/L. Only two patients had PCTs done, with a median value of 8.35 µg/L. Most patients had additional positive cultures with bacterial growth and were treated with antimicrobial agents. Therefore, no clear conclusions could be drawn regarding the usefulness of these biomarkers for the diagnosis of mould infections in our setting.

Serum galactomannan (GM) and 1,3-β-d-glucan (BDG) are two additional serological tests that can assist in diagnosing IFIs. 1,3-β-d-glucan (BDG) testing was performed in only one (9.1%) patient in this study. 1,3-β-d-glucan is a cell wall component of multiple fungal species whereas GM is more specific for the diagnosis of invasive aspergillosis. Galactomannan can be performed on serum and BAL samples and is especially useful in the diagnosis of pulmonary aspergillosis. These tests may be used in addition to microbiological cultures and histological examination. Combining the different testing modalities may increase the sensitivity of diagnosing IFIs.²⁷

Treatment challenges

Voriconazole is the antifungal agent of choice for the treatment of aspergillosis, with isavuconazole or amphotericin B (AMB) as primary alternatives. Due to the invasive nature and rapid progression of mucormycosis, a combination of surgical debridement and AMB is advised. Treatment of other moulds is not well described, and management will depend on the availability of drugs, the site of infection and the clinical status of the patient.⁹

In this study, all patients that required treatment were started on intravenous AMB. However, antifungals were combined with surgical debridement in only two patients who presented with skin and/or mucus membrane manifestations, which resulted in a good outcome for both patients. Three patients with skin and/or mucus membrane manifestations received no surgical treatment, and despite appropriate antifungal therapy, still demised, emphasising the importance of source control in the management of mould infections.

Consulting an infectious diseases specialist has been shown to improve treatment with antifungal agents, proper laboratory investigations and follow-up of patients with fungal infections. Patients in these studies most commonly presented with candidemia.^28,29 Nevertheless, one would expect the same results in the context of IMIs.

Limitations

Limitations of this study include the small sample size, and the fact that only sterile and selected non-sterile respiratory specimen types were included in the analysis. The true burden of mould infections might, therefore, have been underestimated. Due to the retrospective nature of the study, we were only able to report most of the moulds to genus level, and were unable to collect certain clinical data, such as the use of antifungal prophylaxis. Additional limitations are that other biomarkers such as GM and BDG, as well as radiological investigations, were used infrequently to assist in the diagnosis of IMIs in this cohort.

Conclusion

Our findings suggest that patients with HNs (particularly AML) and severe AA in central South Africa have a high risk for IMIs and associated mortality (45.5%). Severe and prolonged neutropenia, chemotherapy, immunosuppressants and antibiotic exposure were significant predisposing factors in this cohort. Prompt diagnosis and treatment of IMIs may reduce morbidity and mortality in these patient populations. However, diagnosis and treatment of mould infections remain a challenge.

A multidisciplinary approach for the diagnosis and treatment of IMIs in patients with haemato-oncological diseases is recommended. In this study, the diagnostic value of histopathology in addition to mycology culture, and the addition of surgical intervention to appropriate antifungal therapy when required, have been emphasised. When treating patients diagnosed with HNs and AA, we suggest a high index of clinical suspicion for IMIs and recommend that clinical consultation with an infectious diseases specialist and microbiologist is considered for all patients.

As with bacterial infections, local epidemiology should be considered before deciding on antifungal prophylaxis or empirical antifungal therapy. However, prior to this report, there were no local data available on mould infections in patients with haemato-oncological diseases in central South Africa. Therefore, this study contributes to the growing knowledge on the distribution, patient characteristics, and outcomes of IMIs. It is the first local report on mould infections in patients with HNs and AA and lays the foundation for further research on IFIs in high-risk patient populations.

Acknowledgements

The authors express their gratitude to Dr. Daleen Struwig, medical writer and editor, Faculty of Health Sciences, University of the Free State, for technical and editorial preparation of the article. This article is partially based on the first author’s dissertation entitled ‘The distribution, patient characteristics, therapy and patient outcome in culture positive invasive mould infections in a tertiary hospital in the Free State province, South Africa’ towards the Master’s Degree MMed in the Department of Medical Microbiology at the University of the Free State, South Africa, 2019, with supervisor Prof. Yacoob Coovadia (who passed away shortly after the completion of this research), and co-supervisors Prof. S. Potgieter and Dr M.S. Abrahams. It is available here: http://hdl.handle.net/11660/10396.

The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.

B.v.d.W. and S.P. conceptualised the project and prepared the protocol. B.v.d.W. collected the data and drafted the article. L.B. contributed to the histology images. S.P., S.K. and L.B. contributed to editing of the article. All the authors, B.v.d.W., S.K., L.B. and S.P. approved the final article.

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

The data that support the findings of this study are available from the corresponding author, B.v.d.W., upon reasonable request.

The views and opinions expressed in this article are those of the authors and are the product of professional research. The article does not necessarily reflect the official policy or position of any affiliated institution, funder, agency or that of the publisher. The authors are responsible for this article’s results, findings and content.

References