Laryngeal candidiasis: Our experience from sixty biopsy specimens

Persistent throat symptoms, such as dysphonia, globus and throat pain, are highly prevalent and are a significant cause of morbidity. In a number of cases, a clear cause of these symptoms is not identified, and many patients are treated empirically with lifestyle advice and/or anti-reflux medication. There is an increasing frequency of respiratory diseases, such as asthma and chronic obstructive pulmonary disease (COPD), with associated increased use of inhaled corticosteroids (ICS). Oropharyngeal candidiasis is a well-recognised complication of ICS, particularly when inhaler technique and oropharyngeal hygiene are poor. There is, however, limited evidence on the prevalence of laryngeal candidiasis in those taking ICS. While clinical diagnosis is sometimes possible, it has been highlighted in oropharyngeal candidiasis that clinical findings do not always correlate with the presence of fungi. We hypothesised that laryngeal candidiasis may be an underrecognised cause of laryngeal inflammation and persistent throat symptoms in a subgroup of patients presenting to ENT clinics, particularly those taking ICS. We therefore aimed to retrospectively review the presence of candidiasis in a series of laryngeal biopsies.


| INTRODUCTION
Persistent throat symptoms, such as dysphonia, globus and throat pain, are highly prevalent and are a significant cause of morbidity. 1 In a number of cases, a clear cause of these symptoms is not identified, and many patients are treated empirically with lifestyle advice and/or anti-reflux medication. 2 There is an increasing frequency of respiratory diseases, such as asthma and chronic obstructive pulmonary disease (COPD), with associated increased use of inhaled corticosteroids (ICS). 3 Oropharyngeal candidiasis is a well-recognised complication of ICS, particularly when inhaler technique and oropharyngeal hygiene are poor. 4 There is, however, limited evidence on the prevalence of laryngeal candidiasis in those taking ICS. 5 While clinical diagnosis is sometimes possible, it has been highlighted in oropharyngeal candidiasis that clinical findings do not always correlate with the presence of fungi. 6,7 We hypothesised that laryngeal candidiasis may be an underrecognised cause of laryngeal inflammation and persistent throat symptoms in a subgroup of patients presenting to ENT clinics, particularly those taking ICS. We therefore aimed to retrospectively review the presence of candidiasis in a series of laryngeal biopsies.

| ME TH ODS
We retrospectively reviewed the laryngeal biopsy samples and case notes of patients presenting to the ENT outpatients department of the Newcastle upon Tyne Hospitals NHS Foundation Trust. We identified a series of patients who had previously undergone laryngeal biopsy for persistent throat symptoms and had suspicious findings on flexible endoscopy and had documented current ICS use. We also identified from the same database of biopsy samples an age and sex-matched group of patients without respiratory co-morbidities or documented ICS use.

| Statistical analysis
All clinical data were sorted on secure NHS computers, and analysis was performed on Microsoft Excel 14.6.5 (Redmond, Washington, USA).

| Strengths and weaknesses of the study
This is a small retrospective study and subjects to several limitations.
Firstly, the documentation of ICS may be unreliable, and it is conceivable that some of the non-ICS use group were actually ICS users.
DPAS is used for visualising the carbohydrate components of candida and as with any interpretive test will have limitations and the potential for reporter error. There are no guidelines on how many sections to stain when performing DPAS testing. DPAS is able to identify the glycoprotein content of fungi, although it may not be able to distinguish small early fungal stages due to their size or indeed specific candida species. The samples were all tested in a quality-assured CAP-endorsed pathology laboratory and reported by a specialist head and neck pathologist.
We present a relatively elderly cohort of patients reflecting the indications for laryngeal biopsies and need to exclude malignancy.
There is a possibility post-menopausal hormone changes may affect candida colonisation on laryngeal mucosa secondary to pH influences. This limits our conclusions with regard to the incidence of laryngeal candida in younger ICS users.
We were also unable to perform subgroup analysis of the EUA findings (such as inflammation or cysts) and the presence or the absence of laryngeal candidiasis due to the small study numbers. Similarly, we are not able to distinguish in this study whether throat symptoms are specifically due to candida infection or due to other benign pathologies such as Reinke's oedema or cysts. Furthermore, it would be ideal to compare the prevalence of candida in a control group of ICS users without throat symptoms. However, we would not routinely be taking laryngeal biopsies from normal larynx in Keypoints • Candidiasis was found in a number of laryngeal biopsy specimens tested. These patients had no reported visual evidence of candidiasis on laryngoscopy.
• Laryngeal candidiasis was more commonly identified in the pathology samples of those with documented use of inhaled corticosteroid therapy (20%) than those without documented use (10%).
• Local discussion between otolaryngologists and pathologists on the testing of laryngeal samples for candidiasis is required to ensure appropriate investigations are performed in a standardised manner.

(A) (B)
F I G U R E 1 Laryngeal vocal cord biopsy stained with Diastase Periodic Acid Schiff (DPAS) at 920 magnification. A, DPAS negative sample without fungal hyphae (920 magnification). B, DPAS positive sample with magenta-coloured fungal hyphae patients without such symptoms. Observational evidence from this pilot study will assist in planning studies that further consider asymptomatic patients, adequate power and causality rather than correlation.

| Comparison with other studies
The potential of laryngeal candidiasis to cause persistent throat symptoms is supported by Wong et al 8