Fungal colonization and infection during critical illness

Key points

• ■

  Fungal infections are a leading cause of nosocomial infections in patients, in particular those residing in the intensive care unit (ICU).

• ■

  Risk factors for developing fungal infections include diabetes mellitus, immunosuppression, previous antibiotic use, prolonged ICU length of stay, malignancy, parenteral nutrition, neutropenia, and prolonged central line catheterization.

• ■

  Although Candida albicans is the most common fungus infection seen, other mycoses including other Candida subspecies such as Candida glabrata , Candida krusei , and other mycoses such as Aspergillus create challenges as to identification and early appropriate treatment.

• ■

  There are a number of classes of antifungal agents available. Being familiar with their sensitivities and resistance is crucial to selecting the proper agent(s).

History and prevalence

Invasive mycoses have emerged as a major cause of morbidity and mortality in hospitalized surgical patients. It is estimated that the incidence of nosocomial candidemia, the most common form of invasive fungal infection in the United States, is about 9 per 100,000 population, with an estimated 25,000 cases per year. In the United States, candidemia increases hospital length of stay by 3 to 13 days and health care costs per patient by $6000 to $29,000. In the United States, Candida is a common cause of catheter-related bloodstream infection. Candida is one of the most common types of bloodstream infections reported in a recent U.S. point-prevalence survey. The incidence of fungal infection in the ICU appears to have remained relatively constant, as reported by the three Extended Prevalence of Infection in Intensive Care point prevalence studies. In the 1995 EPIC (Extended Prevalence of Infection in Intensive Care) study of 17 Western European countries, fungal infection accounted for 17.1% of organisms in 4501 infected patients.

In the EPIC II study of 7087 infected ICU patients in 75 countries in 2007, fungus was the third most commonly isolated organism (19%), preceded by gram-negative organisms (70%) and gram-positive organism (47%). Specifically, Candida accounted for 17% of all isolated organisms. In the follow-up 2017 EPIC III study in 88 countries, of 8135 patients with suspected or proven infection, 5259 had at least one positive microbiologic culture, with fungus remaining the third most common organism at 16%. Gram-negative (67%) and gram-positive organisms (37%) were more frequent. In both of these reports, patients may have had cultures positive for more than one organism. In Canada, there has been a rise in the number of Candida isolates since 1991. Reasons attributed to the rise of fungal nosocomial infections include increased number of patients with long-term venous access catheters, many in place for longer periods of time and patient immunosuppression.

Fungemia is the fourth most common type of septicemia in the United States, and every year about 20 new opportunistic pathogenic species of fungi are discovered. Outside the United States, several studies have reported a rise in candidemia and other forms of Candida infections. U.S. rates of candidemia declined between 2008 and 2013 and then stabilized between 2013 and 2017 In a meta-analysis of randomized, placebo-controlled trials with fluconazole prophylaxis the incidence of fungal infections was significantly reduced, but there was no survival advantage, raising the issue of the value of prophylaxis.

The first clinical description of Candida infection can be traced to Hippocrates with Parrot recognizing a link to severe illness, Lagenbeck implicating fungus as a source of infection, and Berg establishing causality between this organism and thrush by inoculating healthy babies with aphthous “membrane material.” The first description of a deep infection by C. albicans was made by Zenker in 1861, even though it was not named until 1923 by Berkout. On the other hand, the genus Aspergillus was first described in 1729 by Michaeli, and the first human cases of aspergillosis were described in the mid-1800s. With the introduction of antibiotics and the subsequent appearance of ICUs, new instances of opportunistic fungal infections began to emerge. The use of immunosuppression, organ transplantation, implantable devices, and the human immunodeficiency virus (HIV) have also radically changed the spectrum of fungal pathogenicity. To this end, while Candida infections can occur in nonimmunocompromised individuals, invasive infections with molds such as Aspergillus primarily occur in immunocompromised patients.

Fungi are ubiquitous heterotrophic eukaryotes, quite resilient to environmental stress and able to thrive in the most unusual places. They may belong to either kingdom Chromista (has some fungal-like organisms) or kingdom Eumycota (where most fungi reside). For identification purposes, the separation of taxa is based on the method of spore production, assisted by molecular biology techniques (ribosomal RNA and ribosomal DNA) that further refine fungi phylogeny and establish new relationships between groups. The most important human pathogens are the yeasts and the molds (from the Norse “mowlde,” fuzzy). The dual modality of fungal propagation (sexual/teleomorph and asexual/anamorph states) has meant that since the last century there has been a dual nomenclature.

Predictors for fungal infections

The National Healthcare Safety Network began collecting voluntary reported data in 2005. Its first report showed Candida spp. are the fourth most common infection in overall pathogenic isolates in all cases of health care–associated infections (HAI). They are a common cause of central line–associated bloodstream infection. Of note, an analysis of Centers for Disease Control NHSN central line–associated bloodstream infection (CLABSI) data at U.S. acute care facilities from 2011 to 2017 indicated that Candida sp./yeasts were the most common pathogen in adult ICUs, comprising 27% of CLABSI. Of note, Candida sp. and yeast stopped being included in the CDC definition for catheter-associated UTI in 2015; this definition change has resulted in decreased CAUTI rates. In addition, several conditions (both patient-dependent and disease-specific conditions) have been recognized as independent predictors for invasive fungal complications during critical illness. ICU length of stay was associated with Candida infection as were the degree of morbidity, alterations of immune response, and the number of medical devices involved. Neutropenia, diabetes mellitus, new-onset hemodialysis, total parenteral nutrition (TPN) use, broad-spectrum antibiotic administration, bladder catheterization, azotemia, diarrhea, use of corticosteroids, cytotoxic drugs, H ₂ blockers, acquired immunodeficiency syndrome (AIDS), radiation therapy, previous bacteremia, abdominal surgery, hemodialysis, extremes of age, recurrent mucocutaneous candidiasis, and prolonged duration of cardiopulmonary bypass have been associated with invasive candida infections. ICU and invasive mechanical ventilation duration are related: epidemiologic observation correlating the length of mechanical ventilation and the amount of intensive care required correlates with the occurrence of both systemic fungal infections and fungal colonization. A recent prospective multicenter case control study of 192 patients with candidemia and 411 matched controls, in and out of the ICU, confirmed these (central venous line, TPN, previous septic shock, renal replacement therapy, and exposure to certain antibiotics) associations.