Special Feature: Systemic Fungal Infections: Epidemiology and Pathogenesis
By Stephen W. Crawford, MD
Fungal infections are increasingly common problems in the ICU. Fungi are now the fourth most commonly isolated pathogens in nosocomial bloodstream infections and represent almost 8% of all hospital-acquired bloodstream infections (see Table 1, below).1 Candida species account for more than 85% of all fungal bloodstream infections. The incidence of candidemia has increased dramatically.2-8 Between 1980 and 1989, the rate increased by 75% in small, nonteaching hospitals and by 487% in large teaching hospitals. While candidal species account for most of the blood-borne infections, molds are also increasingly problematic. By some estimates, aspergillus-species infections have grown by more than 14-fold in the last 2 decades.
|
Nosocomial Bloodstream Infections in 49 US Hospitals: The SCOPE Program (1995-1998)* |
|||
| Rank | Pathogen | % | Mortality (%) |
| 1 | Coag-negative staphylococci | 31.9 | 21 |
| 2 | Staphylococcus aureus | 15.7 | 25 |
| 3 | Enterococci | 11.1 | 32 |
| 4 | Candida species | 7.6 | 40 |
| Adapted from: Pfaller MA, et
al. Diagn Microbiol Infect Dis. 1998;30:121-129. |
|||
The reasons for the increasing problem with fungal infections is, in part, related to increased prevalence in the ICU of patients with significant risk factors. The patients with the greatest risks include those with HIV or AIDS, and those receiving immunosuppressive drugs or chemotherapy for cancer or organ transplantation. The numbers of these patients in the ICU are increasing. In addition, the risk of invasive candidal infections increases with the use of intravascular devices, which are increasingly present in the ICU. Profound neutropenia is sentinel risk factor for invasive fungal infections. This is evidenced by the fact that 40% of bone marrow transplant (BMT) patients develop such infections if neutropenia is present for more than 3 weeks. Among BMT recipients overall, the incidence of invasive fungal infections is between 15% and 25%.
Candidemia
Several studies have illustrated an often under-appreciated high mortality rate associated with fungal bloodstream infection.9-11 As noted in Table 1, the overall death rate associated with candidemia is 40%. Two studies within the last decade demonstrated that up to 23% of patients die within 2 to 7 days of the positive culture.8,9 Anaissie et al11 examined the death rate as a function of neutropenia and time after culture. The mortality rate among neutropenic patients was high, 24% in 7 days and 63% in 3 months. (See Table 2, below.) This may be a reflection of the severity of the underlying immunological process causing the neutropenia. However, the mortality among non-neutropenic patients was also alarmingly high, 10% in 7 days and 24% in 3 months. While these data do not prove that the fungemia is a direct contributor to the deaths, I believe many of us have not appreciated the grave prognostic implication of fungemia.
Death Rates at 3 Intervals After Index Positive Blood Culture for Candida Species
Neutropenic Non-neutropenic Within 7 d: 24% 10% Within 14 d: 34% 16% Within 90 d: 63% 43% Adapted from: Anaissie EJ, et al. Am J Med.1998;104:238-245.
Aspergillosis
The incidence of invasive infections with Aspergillus species has increased over the past 2 decades. Patients with neutropenia or neutrophil or monocyte/macrophage dysfunction, and those treated with corticosteroids, are especially susceptible to these infections. Current estimates are that 10% to 20% of patients with leukemia and 5% to 25% of heart or lung transplant recipients develop aspergillosis. Infections with Aspergillus species occur in about 5% to 13% of marrow transplant recipients. As with past experiences, the sino-pulmonary tree is the most common site of colonization. Unfortunately, only 34% of patients respond to currently available antifungal therapy and relapses often occur. This dismal response rate underscores the need to pursue new treatment options.
The spectrum of invasive aspergillosis and its outcomes were recently reviewed in a large multicenter study.12 This review of several hundred patients found that complete response rates with conventional treatments (ie, either amphotericin B or itraconazole) was only 27% overall. Patients with marrow transplantation fared worst (10-12% response) while those with solid organ transplantation did best (48%). While the results of this study are important, its use is limited. Patterson and associates collected "the most recent cases treated for invasive aspergillosis" from clinicians around the world. Thus, this paper provides data regarding those treated for aspergillosis, but does not represent the full spectrum of the disease, much of which goes unrecognized.
Kaiser et al published a small but incredibly useful review of the spectrum of invasive aspergillosis.13 They identified all the cases of aspergillosis within their institution by combing through clinical, pathology and microbiology records. 35 cases were identified and reviewed. Surprisingly, the diagnosis not suspected in 40% of cases of aspergillosis. As in other reports, the lungs were almost always involved (94% of cases), and in 74% of patients the infection was limited to the lungs. However, this means that there was metastasis in one fourth of the cases. Other sites involved, in descending order were: heart, CNS, liver, spleen, and skin. Pathologists have noted the finding of cardiac involvement in the past; however, I fear that most clinicians are unaware of the prevalence. I believe this is because of the great difficulty in establishing the diagnosis before death. In my experience, cardiac arrhythmia in patients with aspergillosis is a grim harbinger.
The review by Kaiser et al13 revealed interesting information regarding risk factors. Whereas most authors have concluded that prolonged neutropenia is a major risk for invasive aspergillosis, only 40% of cases in the Kaiser review were neutropenic. However, corticosteroid use occurred in 91% before diagnosis. This fact is consistent with recent reports of invasive aspergillosis occurring in patients with COPD after exposure to large numbers of fungal spores in an ICU environment.14
Culture of both sputum and bronchoalveolar lavage contributed to the diagnostic sensitivity (75% and 50% yield, respectively). In my experience, methenamine silver staining of respiratory secretions adds significantly to the ability to detect fungal hyphae. I believe the yield is much higher than with KOH smear analysis, even with the addition of calcofluor agent. Interpretation of a positive culture is difficult, as it may merely represent colonization without invasion. However, the presence of fungal hyphae on cytological examination strongly suggests growth of the organism within the respiratory tree. In a patient with appropriate risk factors and a clinical presentation consistent with invasive aspergillosis, I believe that the finding of either positive cytology or fungal culture should be assumed to signify true infection.
Among the important findings of the Kaiser study was the discovery that concurrent infection with other nonfungal organisms occurred in 83% of cases. Thus, secondary infection is the rule rather than the exception. These infections may indicate the severity of the immune deficit in these patients. However, these other infections may mask the evidence of the fungal process, misdirect our antimicrobial therapy and may increase the mortality in these patients. In this study, the mortality rate was 94%. This is exactly the same as the one-year mortality for invasive aspergillus among marrow transplant recipients.15
The pathogenesis of invasive Aspergillus infection usually begins when patients inhale airborne spores. The primary mechanisms of host defense against growth of the organism are phagocytic cells. If the defenses are inadequate or are overcome, fungal growth occurs in the airways. At this stage, the process is radiologically and clinically silent. However, I have seen cases with small (1 cm diameter) radiographic rounded opacities due to obstruction of a bronchiole with a matted mass of intra-airway fungus. Hyphae were evident on histological examination beginning to penetrate the bronchiolar epithelium in a radial fashion. Thus it is possible that the earliest radiographic lesion may be a small area of obstructive pneumonia.
From this position in the lung, invasion of the bronchiole (or bronchus) may lead to angioinvasion of the adjoining artery. Occlusion of the vessel causes lung infarction. Radiologically, clinicians recognize a sudden increase in the size of a mass lesion. Since most of the "mass" is infarcted lung, the size of the lesion does not change appreciably over the next days. I think that this lack of radiographic progression lulls us into believing that whatever therapy we begin is effective. In the meantime, the fungus slowly invades the infarcted tissue. Obtaining a diagnostic sample of the lesion is complicated by the large amount of infarcted tissue and relatively small amount of fungus. Often, early in the process, unless the proximal aspect of the early infarct is sampled, the fungus will be missed.
Unlike bacterial pneumonia where there is a large number of organism throughout the infected tissue, there are relatively few and very localized number of aspergillus organism responsible for a sizable amount of lung infarction. In my experience, aspergillus infection was not initially detected in approximately 20% of pathological analyses of lung resection specimens due to sampling errors. If the pathologist who is handed the entire lesion en bloc cannot regularly find the fungus that is present, it is little wonder we clinicians have such a poor diagnostic yield with bronchoscopes and needle aspirations. This pathology also explains why the bronchoscopic yield from transbronchial biopsy is very low, while airway wash specimens has a higher yield.
Mortality with invasive aspergillosis is very high and approaches 94% in some populations. I believe that probably only 50% of infected patients die due to the fungus. Most likely die either from the concomitant infection (which is very common), or from the underlying disease that predisposed to the aspergillus infection. Death due to aspergillus infection is either due to bleeding after erosion into large blood vessels or to metastasis to heart or brain. In this respect, the infection behaves similarly to bronchogenic carcinoma. There are similarities in the attempts to treat these infections.
A subsequent Critical Care Alert Special Feature will discuss treatment approaches and new antifungal agents in managing this group of important ICU infections.
[Note: The views and opinions expressed are not necessarily those of the US Navy or US Government. Dr. Crawford is a member of the Ortho-Biotech, Inc speaker’s bureau.]
References
1. Pfaller MA, et al. Diagn Microbiol Infect Dis. 1998;30: 121-129.
2. Barchiesi F, et al. Eur J Clin Microbiol Infect Dis. 1999;18:184-187.
3. Fortun J, et al. Eur J Clin Microbiol Dis. 1997;16: 314-318.
4. Gubbins PO, et al. Pharmacotherapy. 1998;18:549-564.
5. Kauffman CA, et al. Clin Infect Dis. 2000;30:14-18.
6. Klepser ME, et al. Ann Pharmacother. 1998;32: 1353-1361.
7. Milan EP, et al. Diagn Microbiol Infect Dis. 1998;32: 211-216.
8. Sobel JD, et al. Clin Infect Dis. 2000;30:19-24.
9. Fraser VJ, et al. Clin Infect Dis. 1992;15:414-421.
10. Taylor GD, et al. Mycoses. 1994;37:187-190.
11. Anaissie EJ, et al. Am J Med. 1998;104:238-245.
12. Patterson TF, et al. Medicine. 2000;79:250-260.
13. Kaiser L, et al. Medicine. 1998;77:188-194.
14. Rello J, et al. Clin Infect Dis. 1998;26(6):1473-1475.
15. Wald A, et al. J Infect Dis. 1997;175(6):1459-1466.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content