Lung nodules due to Candida parapsilosis in a person with cystic fibrosis

Abstract

We present the first reported case of Candida parapsilosis pulmonary infection presenting as lung nodules. The patient is a 31-year-old man with cystic fibrosis (CF) colonised with multidrug-resistant Escherichia coli and increased frequency of pulmonary exacerbations in the preceding months. While on intravenous antibiotics for a pulmonary exacerbation, he developed bilateral pulmonary nodules. Bronchoalveolar lavage cultures grew C. parapsilosis. He was initially treated with dual antifungal therapy, voriconazole and micafungin. Discontinuation of voriconazole due to transaminitis resulted in the development of new nodules, and isavuconazonium was added. Repeat imaging revealed no progression of disease. Micafungin was eventually discontinued. Monotherapy with isavuconazonium is planned for 1 year post lung transplant. In the CF population, C. parapsilosis may be an opportunistic pathogen. The case highlights that frequent CF exacerbations and antibiotic exposure increase the risk for opportunistic infections including Candida species and the implications for lung transplantation in this setting.

Background

Candida is a known coloniser of the human oral cavity and upper airways. Candida rarely causes infection in individuals with cystic fibrosis (CF).¹ In people with CF who develop infection with Candida, Candida albicans is the cause in 95% of cases.¹ C. parapsilosis, a normal commensal found frequently on human skin, is a known but uncommon respiratory coloniser of individuals with CF.² We present the first reported case of C. parapsilosis pulmonary infection presenting as multiple lung nodules in an individual with CF.

Global C. parapsilosis prevalence data varies. The ARTEMIS DISK Global Anti-fungal Surveillance Programme collected and tested 205 329 yeasts isolated at 134 sites between 1997 and 2005. The data showed there was an increased rate of isolation of C. parapsilosis (4.8%–6.6%) between the time periods from 1997 to 2000 and 2001 to 2005.³ However, the SENTRY Antifungal Surveillance Programme, which evaluated Candida species prevalence among clinical isolates from 1997 to 2016, has shown prevalence to be much higher at 15.9%, and does not show any consistent upward or downward trend prevalence over time.⁴

Data on colonisation and infection by C. parapsilosis in people with cystic fibrosis are limited and variable. An international prospective study demonstrated that C. parapsilosis grew in 4.8% of all sputum samples collected from individuals with CF.⁵ A separate study found C. parapsilosis in 12.5% of culture samples in a Turkish cohort of individuals with CF.⁶ In a study by Jha et al, C. parapsilosis was isolated in only 3% (1 individual) with lower respiratory tract infection.⁷ These studies suggest that C. parapsilosis is an uncommon coloniser and pathogen in the CF population. A more recent study published in March 2021 used DNA sequencing—next generation sequencing of the ITS2 region—from sputum to assess the fungal microbiome in CF.⁸ The most common operational taxonomic unit in patients with CF was C. parapsilosis, up to 20.4% of 172 patient isolates compared with only four samples that showed any fungal culture growth.⁸ This data suggest that C. parapsilosis is possibly more prevalent than suspected and not identified due to testing limitations.⁸

In the population at large, risk factors for C. parapsilosis infection include use of antibiotics, glucocorticoids, immunosuppressive agents and total parenteral nutrition.⁹ C. parapsilosis infections are associated with hyperalimentation solutions, prosthetic devices, indwelling catheters and nosocomial transmission through the hands of healthcare workers.^{2 10} Factors related to disease pathogenesis are the secretion of hydrolytic enzymes, adhesion to prosthetics and biofilm formation.² C. parapsilosis is unique and different from C. albicans in that invasive disease can occur without prior colonisation and is frequently transmitted horizontally via external sources.² The ability of C. parapsilosis to form biofilms varies with different surfaces, which may explain why certain areas of infection are more common, and others are virtually non-existent.²

In the general population, pulmonary infection (pneumonia (PNA), pulmonary nodules) related to C. parapsilosis is so uncommon that it is not mentioned in a subset of review articles about C. parapsilosis. C. parapsilosis has been associated with pulmonary infection in patients with cutaneous T cell lymphomas.¹¹ In an extensive literature search, we found no case reports of pulmonary nodules due to C. parapsilosis. The most common presentation of C. parapsilosis is fungemia. Other presentations of C. parapsilosis infection are less common but include endocarditis, meningitis, peritonitis, nail, ear, eye infections, urinary tract infections and vulvovaginal infections.^{2 10}

Case presentation

A 31-year-old man with CF presented with worsening dyspnoea. His medical history includes CF (delta F508/3876delA) with severe airway obstruction with baseline forced expiratory volume in 1 second 30%–39% of predicted, pancreatic exocrine insufficiency, CF-related diabetes (CFRD) and persistent airway colonisation with multidrug resistant (MDR) Escherichia coli, Pseudomonas aeruginosa and methicillin-susceptible Staphylococcus aureus (MSSA). His pulmonary exacerbations had increased in frequency from twice a year for the past 2 years to every other month at the time of admission. He received intermittent low-dose corticosteroid courses during this time. Transplant listing was deferred due to inadequate abstinence from alcohol use.

One month prior to admission, he was hospitalised for acute hypoxemic respiratory failure requiring high flow nasal cannula. He was found to have cavitary PNA with MSSA and MDR E. coli. He was treated with and discharged on intravenous meropenem and tobramycin. He failed to improve on this regimen despite over 4 weeks of treatment and presented for medical evaluation. He denied history of recent travel. He denied sick contacts or other new or unusual exposures. He had mediport in place.

On admission examination, he was found to be tachypnoeic to 24 breaths per minute with an oxygen saturation of 98% on room air, tachycardic to 120 beats per minute and afebrile. He was noted to have bilateral expiratory wheezing and crackles, chronic digital clubbing, but otherwise an unremarkable examination.

Investigations

Initial laboratory workup was significant for white blood cell count of 12.13×10⁹/L, 79% of which were neutrophils. Chemistries were normal other than mild elevation in alkaline phosphatase to 179 U/L (reference range: 39–106 U/L). Initial CT of the chest showed resolution of the previously seen cavitary lung lesion, but new bilateral pulmonary nodules, measuring up to 2.5 cm (figure 1, which demonstrates multiple CT cuts from the same time point).

Figure 1.

CT scan revealing multiple, bilateral pulmonary nodules.

Differential diagnosis

The broad differential for multiple pulmonary nodules includes infectious, oncologic, immune-mediated and other miscellaneous causes.¹² As the patient developed pulmonary nodules on CT within a month, infection was suspected. Highest on the differential for CF are fungi and non-tuberculosis mycobacteria (NTM), both which can cause pulmonary nodules.

Our suspicion for bacterial infections was low given that the nodules developed while the patient was on a 4-week course of meropenem. Initial consideration was given to organisms prevalent in CF airway infections including P. aeruginosa, S. aureus, Haemophilus influenzae, Stenotrophomonas maltophilia, Achromobacter xylosoxidans and Burkholderia cepacia.¹³ While the patient’s colonisers, E. coli, Staphylococcus and Pseudomonas, may have developed resistance to meropenem and presented as infection in the form of nodules, this scenario would be unlikely. The other bacteria, with the exception Stenotrophomonas, are usually susceptible to meropenem and are unlikely to develop while on meropenem. Pulmonary nodules due to Stenotrophomonas are uncommon. Another bacterium which can cause pulmonary nodules is Nocardia, but this bacterium is also typically meropenem susceptible and primarily affects immunocompromised patients.

The patient’s exposure to frequent broad-spectrum antibiotics certainly increased the risk for fungal infection. High on the differential were Aspergillus species, which have been isolated from more than 25% of respiratory cultures from patients with CF.¹³ The patient had prior cavitary PNA, and Aspergillus infection often occurs in pre-existing cavities. Candida species are the most frequent colonisers in the CF airway, isolated from almost 50%–75% of patients with CF.¹³ Candida species, though, are not typically associated with pulmonary nodules. Other fungi isolated from the respiratory tract of patients with CF, though less likely in our case, include Scedosporium species, Wangiella dermatitidis and Penicillium emersonii. Other fungal disorders which can cause pulmonary nodules but are not consistent with the patient’s clinical history or clinical picture include histoplasmosis, coccidiomycosis and blastomycosis.

Another potential cause of his pulmonary nodules was NTM, which has been increasingly reported from respiratory secretions in people with CF. High on the differential were Mycobacterium avium complex and Mycobacterium abscessus, the most common isolated NTM species in a prevalence study conducted at 21 CF centres across the USA.¹³

Non-infectious differential diagnoses were very low in the differential. We considered mucus impaction appearing as nodules and scattered nodular densities related to progression of chronic bronchiectasis, but the clinical picture was not consistent. The pulmonary nodules were not cavitary, so septic emboli were also less likely. Given the development of nodules in a rather short interval, we were much less concerned about aetiologies in the oncologic, immune-mediated or miscellaneous categories such as primary malignancy, metastases, rheumatoid arthritis, granulomatosis with polyangiitis, sarcoidosis, organising PNA, lymphoid granulomatosis or amyloidosis.

To determine the diagnosis, a bronchioalveolar lavage (BAL) was performed. Both sputum and BAL cultures grew C. parapsilosis without bacterial or mycobacterial growth. Sequence testing at a reference lab confirmed C. parapsilosis, susceptible to all triazoles, echinocandins, and amphotericin B. The patient’s blood cultures were negative for microbial growth and a transthoracic echocardiogram did not suggest findings of endocarditis.

Treatment

He was hospitalised for 2 weeks and improved on a regimen of oral voriconazole, intravenous micafungin and inhaled amphotericin B. He was discharged on intravenous micafungin and oral voriconazole with regimen and duration to be determined by clinical and radiographic progress.

Outcome and follow-up

Following 6 weeks of antifungal therapy, CT scan revealed improvement in bilateral pulmonary nodules. Due to elevation of liver function tests (LFTs), three times greater than baseline voriconazole was discontinued. The patient remained on micafungin monotherapy for an additional 3 months. Repeat CT scan was performed after 3 months of micafungin which revealed resolution of nodules in bilateral lower lobes but new small nodules in the left upper lobe. His LFTs had normalised, and isavuconazonium was started with a plan of transitioning to monotherapy after a period of overlap. One month after dual antifungal therapy, repeat CT showed no new nodules. Micafungin was discontinued after a 7-month total course, and the patient remains on isavuconazonium with plan to continue for 1 year post lung transplant. Eight months after the patient’s acute infection and diagnosis of candidal PNA, he was started on highly effective cystic fibrosis transmembrane conductance modulator (CFTR) modulator therapy, elexacaftor/tezacaftor/ivacaftor.

Discussion

There is a section in the Infectious Diseases Society of America Practice Guidelines for the Treatment of Candidiasis on candidal PNA which provides treatment options of intravenous amphotericin B or fluconazole with intended outcome of clearance of local sites of infection along with any associated sites of systemic infection.¹⁴ Other than these general candidal guidelines, there is minimal data regarding treatment of invasive pulmonary infection due to C. parapsilosis. Much of the data are from treatment of blood stream infections. In a case report of two patients with C. parapsilosis prosthetic valve endocarditis, treatment regimens were micafungin and fluconazole and micafungin, flucytosine and fluconazole. Both patients were not candidates for valve surgery (ie, mechanical extirpation of infection source) and were treated successfully with these regimens. Both patients remained on suppressive fluconazole.¹⁵

We modelled this treatment plan for our patient. We decided to give dual antifungal therapy because of:

1. The invasive nature of disease (multiple bilateral pulmonary nodules).

2. Time sensitivity: the inability to list for lung transplant in the setting of active fungal infection in the context of the anticipated need for transplant (advanced CF disease with recurrent exacerbations).

3. The extent of underlying structural lung disease which rendered eradication of this fungal pathogen (ie, source control) unlikely.

In multiple surveillance studies, C. parapsilosis has been found to be largely suscepticle to all antifungals. Per SENTRY International Surveillance Data, in the period from 2006 to 2016, only 3.9% of 2433 C. parapsilosis clinical isolates showed resistance to fluconazole.⁴ Despite this data, there are multiple reports of resistance which have developed, sometimes in the setting of prolonged use of fluconazole and other times in the setting of outbreaks.^{2 16} In a study which occurred in South Africa, only 37% of the 531 C. parapsilosis (bloodstream) isolates showed susceptibility to fluconazole and voriconazole.¹⁶ In another study investigating an intensive care unit outbreak of candidemia related to fluconazole-resistant C. parapsilosis (FRCP), 75% of candidemia cases were fluconazole resistant.¹⁷ This study also identified diabetes as an independent risk factor for FRCP,¹⁷ and our patient does have CFRD. Due to concerns of resistance, we opted to use a broad-spectrum triazole antifungal agent, voriconazole, in our patient.

Due to LFT elevation with voriconazole and failure of echinocandin monotherapy, we switched the triazole to isavuconazonium. There is a paucity of in vivo data regarding isavuconazonium. A study of 197 clinical isolates of C. parapsilosis found the organism to be largely susceptible to isavuconazonium in vitro.¹⁸ In vitro data also shows synergistic effect of micafungin and isavuconazonium in the treatment of invasive fungal infections, although not C. parapsilosis specifically.¹⁹ The patient’s candidiasis has remained suppressed on the isavuconazonium and the plan is to continue the isavuconazonium for 1 year post lung transplant (while he is on maximal immunosuppression).

Regarding treatment with echinocandins, the 2433 international isolates of C. parapsilosis evaluated as part of the SENTRY Surveillance showed 0% resistance to micafungin or caspofungin.⁴ However, there is not always a correlation between susceptibility testing and clinical outcomes, with some evidence of new C. parapsilosis infection in individuals already on echinocandin therapy and evidence of failed therapy despite low minimum inhibitory concentration (MIC) values.⁴ Of note, our patient developed new nodules while on a 3-month course of micafungin monotherapy. The addition of isavuconazonium halted disease progression within 1 month.

This case highlights, among other points, that opportunistic infections are difficult to culture, and the potential role of sequencing techniques for early identification and treatment of elusive pathogens in patients who are decompensating or at risk of decompensating, such as those with CF. Although Candida is a known coloniser and uncommon pathogen in those with CF, it can become an aggressive pathogen in the presence of certain risk factors, as demonstrated in this patient. In patients who develop pathogenic candida infection, early identification is important to obtain adequate treatment and prevent decompensation. Identification may require more invasive diagnostics such as BAL with sequencing. Although sequencing can be a strong and transformative tool for clinicians caring for patients who develop pathogenic infections, it also may identify colonisers and could result in inappropriate treatment of colonisers and subsequent development of resistance. Therefore, it is a tool that must be used in those who are most at risk, and in collaboration with infectious disease (ID) or CF specialists.

Of interest, the patient had mediport in place during the time he presented and was treated for candida PNA. As described above, C. parapsilosis infection is associated with indwelling devices, such as mediport. However, our case represents a case of pulmonary disease with pulmonary symptoms and radiology. Bacteraemia was not present to suggest line infection.

This case is important in that it describes a new presentation (bilateral pulmonary nodules) of C. parapsilosis as well as in that it describes this unique presentation in the context of CF. The case presents a successful outcome for this infection with combination triazole and echinocandin therapy followed by suppressive triazole therapy until and through source removal via bilateral lung transplantation. C. parapsilosis pulmonary infections are important to consider in CF as these individuals are at risk for less common, opportunistic pulmonary infections, especially following exposure to prolonged antibiotics. Fungal respiratory cultures should be evaluated and C. parapsilosis should be considered in cases of unresolving PNA or failure to respond to broad, appropriate antibiotics. The case highlights that frequent CF exacerbations and antibiotic exposure increase the risk for opportunistic infections and the implications of lung transplantation in this setting.

Patient’s perspective.

When I first began to feel sick I couldn’t breathe, I felt every time I would take a breathe razors were cutting through my lungs. I couldn’t do my vest chest PT, it was too overwhelming, I can tell when I’m about to get sick, I feel it ‘creeping up’ and know my body, but this was a feeling that I have never felt before, sputum was different, my breathing was different, and felt constantly as though I had 200 lbs of weight were on top of my chest, I managed to survive, thanks to my lord and savior…and the whole CF team for coming up with a great game plan, and making me feel at ease and confident that we would get through my difficult time when I believe I was on my deathbed. And again I like to thank my Lord and Savior Jesus Christ for putting the right people in my life who were there to help me at my weakest and toughest moments.

Learning points.

• We present a case of Candida parapsilosis pulmonary infection presenting as multiple lung nodules.

• In the cystic fibrosis population, C. parapsilosis may be an opportunistic pathogen causing pulmonary infection.

• DNA sequencing may be necessary to identify C. parapsilosis.

• In the absence of guidelines, multiple factors related to our patient’s unique situation led us to use combination triazole and echinocandin therapy, with successful outcome.

• Frequent cystic fibrosis exacerbations and antibiotic exposure increase the risk for opportunistic infections.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s)