Aspergillosis and other invasive fungal infections

For a fungus, this thing attracts a lot of attention from the college. Question 9  from the second paper of 2014 asked for a substantial amount of detail regarding the diagnosis of aspergillosis, and Question 21 from the second paper of 2022 had required some discussion of the risk factors.

As with many of these, the Intensive Care Manual offers little help. Instead, one should turn to LITFL's CCC entry on this subject, which is precise and brief. However if one is not into brevity and precision, one can enjoy several hours submerged in the reference swamp listed below. Selected reading could be limited to the following exceptional articles:

• Sherif et al (2010)
• Meersseman et al (2007)

Invasive fungal infections in critically ill patients

Question 21 from the second paper of 2022 asked for the "risk factors for developing an invasive fungal infection in a critically ill patient", which then pivots into a discussion of Aspergillus. The impression this might give is that the college wanted the candidates to mainly discuss this specific pathogen, but the wording is deliberately more broad. This begs the question: what other invasive fungi did the college want us to know about?

Well.  The excellent 2011 article by Muskett et al identifies Candida species as by far the most common pathogen. There are obviously several others, and the following list was compiled using a table from Shoham & Marwaha (2010):

• Candida (albicans and non-albicans)
• Aspergillus 
• Zygomyces (eg. Mucor)
• Cryptococcus
• Histoplasmosis
• Trichosporon

The risk factors identified by Muskett et al (2011) were:

• surgery
• total parenteral nutrition
• fungal colonisation (eg. Candida in the urine)
• renal replacement therapy
• infection and sepsis
• mechanical ventilation
• diabetes
• high APACHE score (i.e. severe illness)

However these were the risk factors identified from the meta-analysis of a whole host of historical studies, and so represents a sort of pure distilled product from which you could manufacture a risk stratification tool or scoring system (as an example). CICM exams probably need something a bit more broad, and ideally with structure and categories (to aid memory deposition as well as to answer the examiner's comments from Question 21. A good example of such a list, with some distinctive categories, was found in an article on voriconazole by Zabalza et al (2013), and modified to suit the CICM answer:

• Patient factors (comorbidities)
  • Extremes of age (<1 and >70 years)
  • Prolonged neutropenia (more than 14 days)
  • Malnutrition
  • Renal failure
  • Diabetes
• Disease factors (features of the critical illness)
  • Bone marrow transplantation
    • GVHD
    • Delayed engraftment
  • Haematological malignancy with sepsis
  • Complicated abdominal surgery
• Treatment factors
  • ​​​​​​​Total body irradiation
  • Steroid use (1mg/kg prednisolone for more than 2 weeks)
  • Broad spectrum antibiotics
  • Indwelling (urinary and venous) catheters
  • Mechanical ventilation
  • Parenteral nutrition
  • Chemotherapy (esp. ATG, high dose cytarabine and etoposide)

Risk factors for aspergillus infection

One of the articles referenced by LITFL contains within it Table 2 (p.207) which lists the risk categories for invasive aspergillosis among ICU patients. In a contorted form, those data are displayed below:

Low risk	Intermediate risk	High risk
Heart, kidney, liver transplant Burns Stay in ICU > 21 days Malnutrition Cardiac surgery Short term steroids	Bone marrow transplant (auto) COPD Cirrhosis Solid malignancy HIV Lung transplant Chronic steroids Chronic immunesuppression	Bone marrow transplant (allo) Neutropenia Haem malignancy

It should be noted (from the same article) that merely being intubated places one at an increased risk of forming Aspergillus colonies in one's respiratory tract. This has the effect of causing positive test results, but no actual invasive disease.

Diagnosis of aspergillosis

An old 1977 article (from before PCRs and whatnot) and a new 2002 article describe the clinical picture. Their suggestions can be compiled into a table:

Clinical Features and Associated Findings in Aspergillosis

Pulmonary: An asthma-like clinical picture Haemoptysis Chronic cough Extrapulmonary: immunocompromised host endopthalmitis endocarditis

Eosinophilia Elevated serum IgE Spherical lesions on CXR Incidental lesions on CT

Difficulty of identifying aspergillosis in the ICU

This answers Question 9 from the second paper of 2014.

Why the diagnosis is not straightforward:

• ICU patients will have nonspecific signs, buried under other signs, and often no history.
• Radiological diagnosis may be obscured by othr pathology, or may be logistically difficult
• Biopsy is frequently impossible
• Immunocompetent individuals are not suspected, yet may still develop the disease
• Clinically insignificant colonisation is common, but will yield the same positive diagnostic results as active invasive disease.
• Diagnostic test accuracy may be confounded by many factors, eg. concurrent β-lactam therapy.

Laboratory tests

The references which this author thought were the most authoritative were Hites et al (2016) for galactomannan and Marty & Koo (2009) for β-glucan.

Aspergillus galactomannan: blood

• Galactomannan is a polysaccharide composed of galactose side groups nested on a mannose backbone. It is a ubiquitous organic molecule, and can be found everywhere, including in ice cream (as a texture stabiliser). It is also a fungal cell wall component.
• Sensitivity is 71% and specificity is 89% according to a 2006 meta-analysis.
• Causes of confused results include:
  • Beta-lactams (false positives) - particularly Augmentin and Tazocin
  • Other fungi (false positives) - especially Candida species
  • Antifungal therapy (decreased sensitivity)
  • GI tract mucositis (false positives)
  • Plasmalyte (false positives) - perhaps just that one batch was contaminated?
  • Immunocompetent patients (your normal neutrophils clear this molecule from the blood)

Aspergillus galactomannan: bronchoalveolar lavage

• Sensitivity is 61% and specificity is 98% with a 1.0 cutoff
• Though "specific" for the presence of Aspergillus, this test cannot discriminate among those who have invasive aspergillosis and those who are merely colonised.

Aspergillus PCR on bronchoalveolar lavage specimens

• Old techniques had limited usefulness: poor sensitivity and specificity according to some studies.
• In more recent studies, BAL PCR had higher sensitivity but lower specificity than BAl galactomannan.

Aspergillus hyphae identified

• The gold standard
• Again, as all of these tests, it cannot discriminate among those who have invasive aspergillosis and those who are merely colonised.

β-glucan antigen

• Otherwise known as β-(1,3)-d-glucan
• More sensitive than blood galactomannan in detecting invasive aspergillosis; improves confidence when both are positive
• Present in the cell wall of many other fungi (eg. Pneumocystis)
• Can be elevated in the absence of invasive fungal disease, eg. haemodialysis, immunoglobulin therapy, albumin, gauze packing, IV amoxycillin/clavulanate

Radiological identification

LITFL link to an excellent Radiopedia page which has some lovely fungus balls. A definitive journal reference would have to be this 2001 article.

In brief, the radiological features are:

• Pulmonary nodule on CXR
• A "halo" of ground-glass opacity surrounds the nodules on CT
• Wedge-like haemorrhagic pulmonary infarcts
• Air crescents following resolution (necrotic lung separating from the rest of the parenchyma)

Tissue biopsy

Open lung biopsy is not to be taken lightly. Furthermore, it is not always helpful. In one 2001 study the authors claim that they could histopathologically confirm invasive fungal infections only in 53.1% by open lung biopsy. Generally speaking these days people rely on laboratory data and radiology. The open biopsy remains at the hopeless end of the diagnostic algorithm flowcharts. Its only benefit is to spare the patient from a course of amphotericin therapy.

Management of Aspergillosis

The Sanford Guide recommends the following:

• Voriconazole (loading dose of 6mg/kg, followed by 4mg/kg)

Alternatives include

• Amphotericin B
• Posaconazole
• Caspofungin
• Itraconazole