Introduction

Since the advent of lung transplantation (LT) in 1963, the annual rate has been increasing with 4600 LTs performed worldwide in 2019, of which about 50% were performed in the USA, with a 5-year survival rate between 62 and 75% [1,2,3,4]. The benefits of transplantation come with the challenge of balancing immunosuppression (IS) with infection risks. Despite antifungal therapies and prophylactic strategies, lung transplant recipients (LTRs) still have a high risk for developing invasive fungal infections (IFIs) [5•, 6••], which can increase the post-transplant mortality rate by as much as threefold [7•, 8]. According to the transplant associated infection surveillance network (TRANSNET), 8.6% of LTRs will develop IFIs in the first 3 to 12 months post-LT [9]. The successful management of fungal infections is a great clinical challenge. Antifungals require intensive therapeutic drug monitoring (TDM) [10] as adequate levels are crucial for successful treatment and prevention of drug-related toxicities [11]. This review will highlight fundamental issues in managing fungal infections in LTRs including risk factors, diagnosis, treatment, antifungal prophylaxis, and recommendations in drug monitoring.

Epidemiology

The incidence of IFIs is lower than fungal colonization after LT, with the rate of 3–14% compared to 20–50% respectively [12••]. Invasive pulmonary aspergillosis (IPA) is the most common IFI post-LT with mortality rates of 23–82%, whereas invasive candidiasis (IC) follows with a mortality rate as high as 40% [7•]. Risk factors include IS regimen, impairment of mucociliary clearance, such as underlying cystic fibrosis, airway injury, altered alveolar macrophage function, underlying pulmonary architectural distortion, and mucosal defects, and geography and environmental exposure [8, 12••, 13••]. The use of tacrolimus or sirolimus was also demonstrated to be an independent risk factor for developing IFIs [14]. Other risk factors recognized are chronic rejection, cytomegalovirus (CMV) infection, and hypogammaglobulinemia [12••]. The lower respiratory tract is the most common site of mold diseases, especially for the multi-drug-resistant (MDR) infections [15].

Diagnosis

The diagnosis of IFI is divided into categories ranging from possible, probable, or proven depending on the presence of symptoms, radiologic changes, and a positive culture (sputum, bronchial washings, or urine). Proven IFIs have histologic findings supporting fungal elements [14]. Several molecular tests have been used for diagnosis which we will briefly discuss in this section.

The role of serum galactomannan (GM), an enzyme-linked immunosorbent assay that detects polysaccharides present in the fungal cell wall, has been controversial since its sensitivity in non-neutropenic cardiothoracic recipients is about 30% [6••, 16, 17]. When a GM is obtained from a bronchoalveolar lavage (BAL) sample, the sensitivity rises to 82–86% and specificity 89–92% with the positivity cutoff of 0.5–1.5 [18,19,20,21]. In comparison, the use of BAL Aspergillus polymerase chain reaction (PCR) showed a median pooled sensitivity of 79% compared to serum PCR which had a sensitivity ranging 75–88% [22]. Respiratory PCR testing is considered more sensitive than fungal culture and can also help in antifungal resistance testing; unfortunately, it cannot distinguish between colonization versus invasive infection, nor can it discern between Aspergillus subspecies [12••, 18].

In contrast, when looking at (1→3) beta-D-glucan (BDG), a component of the fungal cell wall released into circulation during IFI, the sensitivity is 76–80% and specificity is 82–85% (not specific for any particular mold, or yeast) [23]. Notably, iatrogenic contamination with blood fractionation products (IVIg and albumin), invasive use of surgical materials, and cellulosic dialysis membranes are associated with falsely elevated BDG levels [12••, 24]. New diagnostic tools are under development, such as urinary antigen for Aspergillus detection, lateral flow devices using monoclonal antibodies, and other non-specific biomarkers like Pentraxin-related protein and cytokines [6••].

Radiologic criteria include a “halo sign” observed in 56% and 8% of neutropenic and solid organ transplantation (SOT) patients, respectively [6••]. Other diagnostic signs include macronodules, less commonly peribronchial consolidations, and ground glass opacities. Tree-in-bud nodules/bronchial wall thickening were also reported [12••]. Different diagnostic strategies will be discussed under each IFI.

Special Clinical Presentation

Most common fungal infections

Aspergillosis

Aspergillus spp. is an important cause of life-threatening infection and is the most common IFI in LTRs from either colonization or inhalation of spores [25]. Aspergillosis occurs typically within 1 year but can affect patients up to 3 years after transplant [26]. The overall incidence of IPA in LTRs ranges from 4 to 23% [27••]. The mortality of IPA in LTRs varies according to the clinical presentation, ranging 23–29% in patients with tracheobronchitis to 67–82% in patients with invasive pulmonary disease [27••, 28]. The most common infecting species is Aspergillus fumigatus, with A. niger, A. terreus, and A. flavus being less common.

IPA in LTRs can originate from latent infection, colonization of trachea or retained lung, or donor-derived [25]. Aspergillus tracheobronchitis is seen only in LTRs, and requires systemic therapy with voriconazole in addition to nebulized amphotericin B for at least 3 months [26]. In LTRs, the risk factors for infection include single-LT, early airway ischemia, colonization of airway, CMV infection, and increased IS [27••, 29]. Cystic fibrosis increases the risk of pre-transplant airway colonization with Aspergillus spp.

Voriconazole is the treatment of choice [27••, 30], although some studies have demonstrated non-inferiority of posaconazole and isavuconazole in the treatment of IPA [31•, 32, 33•]. Decreasing IS also plays an important role. Since LTRs are at high risk of IPA, the published guidelines currently recommend prophylaxis for patients with Aspergillus colonization pre-transplant and within the first year after transplantation, cystic fibrosis patients with positive intraoperative Aspergillus culture, and single LT. Targeted prophylaxis can be considered in those with two or more of the following risk factors: early airway ischemia, induction with alemtuzumab or thymoglobulin, CMV infection, augmented IS due to rejection, and hypogammaglobulinemia [25, 27••]. However, in the systematic review by Bhaskaran et al., no reduction in IPA was found when comparing prophylaxis versus no prophylaxis [34]. There have been studies supporting the use of voriconazole [35] and posaconazole for IPA prophylaxis [36]. However, a multicenter randomized controlled trial to determine the most appropriate prophylactic regimen is still needed [29].

Candidiasis

Candida spp. can be found in the pharynx oralis; therefore, it is difficult to distinguish between colonization and IC. When Candida is detected in sputum culture, it is important to note it rarely causes pulmonary infections [37]. In the first month following LT, candidiasis usually presents as candidemia, and is associated with high mortality (54.5%) [6••]. Factors associated with candidemia are high-dose steroids, immunomodulators, long-term catheters [38], as well as open chest and ECMO support post-transplant. Other manifestations of infection include pleural space and local anastomotic site infections [29, 39, 40]. The treatment for IC for LTR is an echinocandin as empiric therapy then transitioning to an azole once the organism’s susceptibilities are available [6••].

Endemic fungi

These mycoses are a group of organisms with similar characteristics. They are dimorphic in nature and are found in different geographic areas (Fig. 1) [41]. It is important to counsel patients on the risk of exposures in these endemic areas post-transplant.

Fig. 1
figure 1

Endemic mycoses—map of the USA with the distribution of endemic mycoses

Cryptococcosis

Largely caused by Cryptococcus neoformans, though in the past few years Cryptococcus gatti has been prevalent in the Pacific northwest region [25]. Cryptococcosis tends to occur between 2 and 5 years post SOT [25], however could be sooner in the case of donor-derived, especially in LTR [42, 43]. In SOT patients, the rate of cryptococcosis is 6–7% [25] with reactivation of quiescent infection being the most common cause [44]. Serum cryptococcal antigen (CRAG) may be a useful tool, though they are frequently negative when the organism burden is low such as those limited to the lung or with single nodule. According to current literature, there is insufficient data to determine the role of CRAG monitoring during treatment for pulmonary cryptococcosis [45]. Central nervous system (CNS) involvement is observed in nearly 50% of transplant recipients with pulmonary cryptococcosis; therefore, a lumbar puncture should be performed in all LTRs to rule this out [25, 44]. Notably, as per Husain et al., transplant recipients on calcineurin inhibitors have shown lower prevalence of CNS involvement suggesting a potential degree of anticryptococcal activity [46, 47]. Whenever possible, gradual reduction in IS during cryptococcosis treatment is advised. However, this could be difficult to do if the patient is receiving T-cell-depleting agents such as alemtuzumab or thymoglobulin; in which case, the rapid reduction in IS may cause adverse acute organ rejection or immune reconstitution inflammatory syndrome (IRIS) [44]. IRIS may cause ventricular obstruction with increased intracranial pressure and hydrocephalus [29].

The treatment for cryptococcosis in LTRs is the same as in other patients, including a lipid formulation amphotericin B plus 5-flucytosine (5-FC) as induction therapy for CNS disease, disseminated disease, and moderate to severe pulmonary disease. This should be followed by fluconazole consolidation and maintenance regimen. However, in mild or asymptomatic disease, initial treatment with fluconazole is the preferred therapy. Dexamethasone does not seem to be effective for cryptococcal meningitis treatment [44].

Coccidiomycosis

Endemic mycosis caused by organism Coccidioides immitis and Coccidioides posadasii, prevalent in the desert soil of the north of Mexico, southwest USA, and California’s central valley [48, 49]. The most common exposure is inhalation of spores (or arthroconidia). Transmission of coccidiomycosis via organ transplantation is common in LT with a rate of 1.4–6.9% in endemic regions. Most cases occur within the first year post-transplantation with a mortality rate up to 30% [23, 49]. Clinical infection is uncommon and can be prevented or mitigated in patients receiving preemptive therapy [49, 50]. Manifestations range from asymptomatic infection to severe pneumonia or disseminated disease, with the latter being more common in the immunocompromised host. This can then progress to acute respiratory distress syndrome (ARDS) and respiratory failure.

Radiologic findings include mass-like lesions, lobar consolidations, pulmonary nodules, cavities, or interstitial infiltrates [49]. Peripheral eosinophilia, though not diagnostic, is present in a third to a half of patients with coccidiomycosis [23]. Diagnosis of coccidiomycosis includes histopathologic findings of spherules containing endospores; Coccidioides species also grow well in most mycologic and bacteriologic media within 5 to 7 days [23, 49]. Immunologic assays have been largely utilizing immunoglobulin detection with tube precipitin (TP) and complement fixation (CF). TP turns positive within weeks of infection, whereas CF took 2 to 3 months to turn positive, demonstrating that TP corresponded to immunoglobulin M (IgM) and CF to immunoglobulin G (IgG). CF tends to uptrend when the infection is poorly controlled [48]. Similar to TP, latex particle agglutination assay (LPA) also detects IgM. Currently, a serological ELISA method based on detection of IgM and IgG is typically used for initial screen, with a sensitivity of 95.5% and specificity of 98.5% [48, 49]. The enzyme immunoassay (EIA) IgM test is the least compelling diagnostic evidence and can produce false-positive results due to interference from other fungal infections, medications, or technical issues [51]. It is recommended to repeat testing for anticoccidioidal antibodies over subsequent weeks to help resolve these discrepancies and improve the certainty of a diagnosis [49, 52]. Other methods of diagnosis are antigen enzyme immunoassay (available for urine, serum, BAL, and cerebrospinal fluid (CSF)) and molecular assays based on DNA hybridization and PCR/qPCR methods [49].

The treatment for coccidiomycosis depends on the severity of the disease, ranging from 3 to 12 months to lifelong treatment, as in the case of CNS infections. The drug of choice for treatment is fluconazole [48]. However, in severe or disseminated coccidiomycosis, lipid formulation amphotericin B is preferred until patient is stabilized and then can transition to fluconazole. There have been reports of infection relapse of coccidioidal meningitis after discontinuation of azole; therefore, treatment is recommended indefinitely or until withdrawal of IS [23, 49, 50].

Pre-transplant evaluation should include history of exposure or residence in an endemic area, as well as current or past symptoms of infection, radiologic evaluation, and serologic testing [23]. Lifelong fungal prophylaxis with an azole is recommended in endemic areas [49], in the setting of positive serological screening, and active infection of the donor [53]. Currently, there are no concrete guidelines on either universal or targeted screening for donor-derived infection [50].

Histoplasmosis

Histoplasma capsulatum is endemic to the Mississippi and Ohio River Valleys [54, 55]. Exposure to the spores is from soil disruption around construction and agricultural sites with large concentrations of bird droppings [25, 56]. In immunocompromised hosts with impaired cell immunity, such as LTRs, the organism remains viable within macrophages, which poses a risk for disseminated disease [49, 57]. Fortunately, histoplasmosis is rare in SOT recipients with an estimated incidence of less than 1% in endemic areas.

Histoplasmosis can be acquired most commonly via inhalation or reactivation of prior disease while on IS, as well as in rare cases (1:10,000 transplants) through donor-derived allograft transmission [49]. Unexpected histoplasmosis was found in 18 of 1000 LTR in endemic areas in a case series [58•].

Most infections are reported within the first 2 years of transplantation. It can present in an occult manner in the transplant population but most commonly (81% of transplants) presents as disseminated infection with subacute febrile illness, progressing to hepatosplenomegaly, pneumonia, GI involvement, and weight loss [59]. Mucocutaneous histoplasmosis presents in 25% of transplant recipients, and CNS involvement is also described in this population [59]. The use of mycophenolate and fungemia are risk factors for severe disease.

Histopathologic visualization of yeast forms (with or without granulomas) confirms the diagnosis as culture can take up to 4 weeks. In SOT recipients, urine Histoplasma antigen EIA demonstrates the highest sensitivity at 92%, with a lower sensitivity in pulmonary disease versus disseminated disease. This is also true about the slightly less sensitive serum Histoplasma antigen (86%), which can be followed to evaluate therapeutic response. However, the specificity of the test is compromised as there is a 90% cross-reactivity with other endemic fungi such as Blastomyces and in lower proportion with Coccidioides [49, 60]. Serologic testing is not recommended for diagnosis in immunosuppressed host.

The treatment of histoplasmosis depends on the severity of illness. Itraconazole is often used in mild to moderate illness. For moderate to severe infections, amphotericin B is utilized as the initial treatment for 2 weeks followed by itraconazole for 12 months [60]. Second-line therapy is fluconazole, though voriconazole, posaconazole, and isavuconazole have also been reported in case reports as successful treatments and may be the preferred choice for non-HIV infected immunosuppressed patients given that fluconazole has high relapse rates in this subpopulation [49]. The concomitant reduction of IS, especially calcineurin inhibitors, is recommended when feasible to decrease relapse risk.

Pre-transplant screening for histoplasmosis is not recommended even in endemic areas because of the poor serologic predictive value of current tests. Secondary prophylaxis as well as antigen monitoring may be considered with recent infection within the past 2 years. Primary prophylaxis might be considered in LT with evidence of donor-derived allograft infection [49, 58•].

Blastomycosis

Blastomyces dermatitidis is endemic to the Mississippi and Ohio River Valleys, the Great Lakes region, and the St. Lawrence Seaway. Infection with B. dermatitidis is through inhalation of spores and less commonly direct cutaneous inoculation. Blastomycosis in immunocompromised individuals is associated with disseminated disease [61] and can increase the risk of allograft loss and overall mortality [62]. However, blastomycosis remains very rare in post-transplant recipients even in endemic areas. There is no significant amount of data reporting the rate of blastomycosis in LTRs, with studies underlying transmission in other SOT but no evidence of transmission in lungs [62]. The clinical presentation includes pneumonia with or without extrapulmonary dissemination. The spectrum of infection ranges from subclinical pulmonary disease to acute or chronic pneumonia, with a subset of patients developing fulminant multilobar pneumonia and ARDS.

The definitive diagnosis of blastomycosis is made from culture isolation of the organism. However, due to the 2 to 4 weeks growth period, histopathologic visualization of yeast forms is the most commonly used method of diagnosis. EIA is also available to detect antigens in body fluids (urine, serum, BAL, or CSF) with a sensitivity of 62–83% but has a low specificity given the cross-reactivity with H. capulatum [63].

Treatment of blastomycosis, particularly in severe pulmonary cases, should start with lipid formulation of amphotericin B for 1 to 2 weeks or until clinical improvement. It should be followed by itraconazole for 12 months or longer if symptoms have not resolved. The exact duration of therapy has not been determined [64]. In the setting of CNS infection, amphotericin B should be extended for 4 to 6 weeks, followed by voriconazole instead of itraconazole given the lower CSF penetration from the latter (< 1%) [49, 64]. Itraconazole monotherapy can be considered initial therapy in mild to moderate cases with close monitoring.

Pre-transplant screening can be done in candidates with prior history of exposure. There is no recommendation on primary prophylaxis against blastomycosis given the lack of supporting studies [49].

MDR mold infections

Non-Aspergillus spp. mold infections have posed an increased challenge in LTRs [29] given the difficulty to discern them from Aspergillus spp. and each other, their intrinsic resistance to antifungals, and their aggressive characteristics of disease [13••, 15]. Exposure to these emerging molds could be from cutaneous contact or spore inhalation from the environment. It has also been noted that exposure to fungal prophylaxis such as voriconazole or inhaled amphotericin can predispose LTRs to these emerging molds [13••].

Scedosporium and Lomentospora prolificans (formerly Scedosporium prolificans)

They are soil saprophytes that are commonly found in temperate climates. LTRs are at higher risk than other organ transplants during the first 12 months post-transplantation [15]. A recent survey found that 48% of a total of 45 LT centers had positive cultures [65]. Pre-transplant colonization plays an important role in Scedosporiosis, which becomes a contraindication for many LT centers [66•]. Infection can occur within a month after transplantation in those previously colonized, but develops 6 months or after in those not previously colonized [13••]. Some of the risk factors for scedosporiosis are underlying cystic fibrosis, prior use of amphotericin, and enhanced IS [13••]. The treatment response depends on the site of infection, the extent of dissemination, and the host’s degree of IS. Outcomes are better with localized disease to either the skin or lungs compared to disseminated disease. In vitro, voriconazole has the most potent activity against Scedosporium. Surgical debridement is the preferred treatment against Lomentospora since it is virtually resistant to all antifungals available, and with reduced susceptibility to echinocandins, especially caspofungin and anidulafungin [67]. Some reports suggested voriconazole [26] or a combination of voriconazole and terbinafine [68].

Mucormycoses

Invasive mucormycosis is a devastating disease with an overall mortality rate of 40–50% [69], and even reported up to 90% [26]. Mucormycetes are ubiquitous in the air but are associated mostly with natural composts and soils of potted plants [13••]. LTRs have the highest incidence of pulmonary mucormycosis in the first year after transplant [13••, 15], with 78% of infections occurring within the first year and 40% within the first month [70]. Given its angiotropic nature, mucormycetes tend to cause tissue infarction and necrosis [70]. Rhino-orbital-cerebral infection is one of the most common presentations as the fungal spores get inhaled through the sinuses, which is more common in patients with uncontrolled diabetes mellitus but also found in one-third of SOT patients [71].

Surgical excision and debridement is the standard of care for all non-pulmonary infectious processes, with amphotericin being the treatment of choice for induction therapy [13••, 26], in addition to reduction of IS. Isavuconazole is the newest triazole approved for treatment of invasive mucormycosis and IPA [13••, 72]. However, in the absence of prospective studies of mucormycosis in LTRs, the management is mainly based on case reports and retrospective studies [70].

Fusariosis

Pulmonary disease is common with Fusarium spp. in LTRs; however, their larger conidia (compared for instance with Aspergillus) can get trapped in the upper airway and sinuses causing upper airway disease. In other severely immunosuppressed individuals, cutaneous manifestations tend to be more common. Voriconazole is the first line of treatment, though surgical excision alone of localized cutaneous disease can effectively treat the infection, in addition to reduction of IS [26].

Prophylaxis

There are several prophylactic strategies described for LTRs. Universal prophylaxis is defined as antifungal agent(s) administered to all patients during the immediate post-transplant period [6••]. Preemptive treatment is the administration of antifungal agents for mold isolated during the surveillance post-transplant bronchoscopy without evidence of invasive disease. A third strategy, “targeted prophylaxis,” refers to an antifungal medication started in the post-transplantation period prior to isolating any fungal pathogen in patients who are deemed high risk for infection, such as in cystic fibrosis or prior fungal colonization [6••]. No randomized trials have been performed comparing these prophylactic strategies [12••]. Though a recent meta-analysis concluded that anti-Aspergillus prophylaxis did not result in significant reduction in IPA or Aspergillus colonization [34]; another meta-analysis from 2016 concluded that universal prophylaxis reduced the incidence of IA in LTRs compared to no or targeted prophylaxis [26, 73]. Universal prophylaxis has several disadvantages especially adverse events associated with azole use: hepatotoxicity, neurotoxicity, QT interval prolongation, and drug interactions. The exposure to universal prophylaxis has also increased the emergent resistance of other fungal infections [6••]. The difficulty determining the appropriate approach highlights the need for a multicenter randomized trial in LTRs [34, 74].

Therapeutic Drug Monitoring

As discussed previously, TDM of azoles is crucial in ensuring treatment success and minimizing drug toxicity, as detailed in Table 1. All azoles can cause hepatotoxicity at any point during therapy. Liver function test abnormalities were reported in up to 60% of LTRs receiving voriconazole whereas less than 10% of patients developed hepatotoxicity on posaconazole and isavuconazole [33•, 35, 75,76,77,78•]. In a meta-analysis, fluconazole was found to have better hepatic safety profiles than other antifungal agents [79, 80]. Liver enzyme abnormality is reversible upon azole discontinuation or by switching to an alternative azole therapy.

Table 1 Therapeutic drug monitoring (TDM)

Amphotericin B is associated with high incidence of infusion-related reactions and nephrotoxicity. The lipid formulations have less nephrotoxicity compared to conventional amphotericin B deoxycholate [80, 91, 92]. Hypokalemia and hypomagnesemia are common side effects so close monitoring of renal function and electrolytes is recommended [93].

The triazole antifungals are inhibitors of the cytochrome P450 system, which results in significant drug interactions [12••]. The coadministration of mTOR inhibitors and voriconazole or posaconazole is contraindicated per manufacturers’ recommendations. However, the use of these combinations seemed to be safe as demonstrated in retrospective studies and case reports, as well as in our center’s experience [94,95,96,97,98,99,100,101]. Recommendations for IS dose adjustment when starting triazoles are detailed in Table 2. However, because of the significant interpatient variability, providers should weigh the risk of drug toxicity and rejection risk in deciding dose modification for a given patient.

Table 2 Immunosuppression drug interactions

Clinical Trials and Future Studies

There are several ongoing clinical trials investigating new therapies and novel approaches to IFI. IA-DUET from The Netherlands investigates the combination of azole-echinocandin for IPA in neutropenic stem cell transplants patients [122]. There is also a phase IIb clinical trial studying F901318, FORMULA-OLS for the treatment of IPA and MDR fungal infections such as Scedosporium and Lamentospora [123]. Ibrexafungerp—a glucan synthase inhibitor—is being evaluated for the treatment of several IFIs including refractory endemic mycoses [124]. Lastly, AEGIS is a phase II clinical trial studying the efficacy and safety of fosmanogepix (APX001), a novel antifungal targeting the Gwt1 enzyme required for localization of glycosylphosphatidylinositol-anchored mannoproteins in fungi [125]. These trials highlight the interest and need for novel therapies for the treatment of fungal infections.

Conclusion

This review describes different fungal organisms that have the potential to cause invasive infections in LTRs. We discussed their epidemiology, clinical presentation, diagnosis, treatment, and prevention of disease. We also delved into TDM and drug interactions in the setting of immunosuppressive agents, which are important factors in the treatment of these IFIs.