Invasive fungal infections

Fungal infections have dramatically emerged as important nosocomial infections in the last decade. Between 1986 and 1990, Candida was the fourth most commonly cultured organism from all sites in the ICU setting in the United States. The increase in Candida bloodstream infections has been explosive, particularly in teaching hospitals. Fungal infections now account for nearly 8 per cent of nosocomial infections hospital-wide, and Candida is responsible for 80 per cent of these. Invasive fungal infections must be included in the differential diagnosis of hospital-acquired infection in every patient in the ICU.

Candida species are resident flora of the gastrointestinal tract and oropharynx in 25 per cent of normal individuals. The gallbladder is also a common site of colonization. The colonization rate doubles in hospital in-patients. Virtually all the previously mentioned factors that impair host defenses (e.g. extremes of age, diabetes, disease- or therapy-related immunosuppression) are factors in the colonization of the host by Candida (TabJe,3). Broad-spectrum antibiotic therapy is the most important risk factor owing to suppression of normal intestinal flora.

Table 3 Risk factors for disseminated fungal infection*

Patients colonized with Candida may suffer life-threatening disease if the infection becomes disseminated. Dissemination occurs only after mucosal or skin barriers have been breached, and either bloodstream infection or locally invasive disease is the result. Fungal infection is less common than either bacterial or viral infections after solid organ transplantation, although liver and bone marrow transplant patients are at substantial risk. Most critically ill patients have more than one major risk factor (Table. 3).

The majority of infections in humans are caused by Candida albicans, although a few other species are increasingly isolated as the overall incidence of serious fungal infections continues to increase. Candida tropicalis is the second most common isolate, and accounts for as many as 25 per cent of cases. Candida krusei is common in patients with malignant hematological disease, whereas Candida parapsilosis is common in non-tumor-bearing patients and may primarily infect patients with lesser degrees of immunosuppression, such as those who develop catheter-related infections. An uncommon but dangerous pathogen is Candida glabrata (formerly Torulopsis glabrata), which also infects less immunosuppressed patients but carries a high mortality when disseminated.

The mechanism by which Candida organisms express virulence is unknown. There is some evidence to suggest that Candida antigen may promote host immunosuppression by downregulating T-cell function, thereby initiating the process of tissue invasion; cell wall proteins may promote some of the serious manifestations of Candida-related disease.

Life-threatening Candida infection may present as either generalized sepsis or localized organ involvement. In the former circumstance, fungal sepsis can be indistinguishable from bacterial infection. Since fungal superinfection is commonplace during or after broad-spectrum antibiotic therapy, an indolent fungal infection may be mistaken for persistence of the bacterial infection or even for bacterial superinfection. Disseminated candidiasis may produce focal disease or microabscesses in many organs. Endophthalmitis may be present in as many as 10 per cent of disseminated infections, particularly when C. albicans is the pathogen. When present, the 'cotton-wool' exudates visible by fundoscopy are diagnostic and are strongly suggestive of systemic disease. Careful fundoscopy is mandatory when disseminated disease is suspected.

Hematogenous dissemination to skin can occur in critically ill patients. Lesions present as maculopapular eruptions 0.5 to 1.0 cm in diameter. Candida arthritis, myositis, and osteomyelitis are unusual but should not be discounted. Hepatosplenic candidiasis is common, particularly in patients with hematological tumors and neutropenia. This manifestation tends to be indolent, with fever, abdominal pain, and non-specific biochemical abnormalities of liver function. Although indolent in its presentation and relatively uncommon, candidal involvement of liver and spleen can be very difficult to eradicate because of the often profound immunosuppression typical in these patients.

Candida may affect the heart in many ways. Fungal pericarditis has been reported after burns and cardiac surgery. Myocardial involvement manifests with electrocardiographic abnormalities or arrhythmias, and may occur in up to 60 per cent of cases of disseminated disease. Fungal endocarditis may complicate cardiac surgery or intravenous drug abuse. Candida is an important pathogen in prosthetic valve endocarditis.

Contamination of the peritoneal cavity with fungi may follow perforation of a hollow viscus or dehiscence of an anastomosis, or perforation of an acutely inflamed gallbladder. Fungal peritonitis is also seen in patients undergoing chronic ambulatory peritoneal dialysis. Peritonitis rivals only catheter-associated infection as a cause of disseminated fungal infection in critically ill general surgery patients ( Dean and Burchard 1996).

Since catheter-related infections frequently occur in critically ill patients, the presence of an indwelling central venous catheter should be suspected as the primary focus of infection in any patient with fungemia. However, Candida is also recognized as a cause of suppurative phlebitis of peripheral veins, a dangerous infection often characterized by a fulminant course after the onset of marked hyperpyrexia, a palpable cord with overlying cellulitis or induration, and the presence of intravascular pus.

Definitive diagnosis of invasive fungal infection is difficult, and is often made relatively late in the course of disease. Accurate timely diagnosis is of great importance because invasive infection accounts for substantial morbidity and mortality, and lack of antifungal therapy has been correlated with increased mortality. Therefore the decision to treat a patient with systemic antifungal therapy must often be made before the diagnosis has been established. In particular, a lack of positive blood cultures should not be counted as evidence against the presence of invasive candidiasis.

Other than the rare signs of cutaneous or ocular manifestations of invasive fungal infection, the clinician is reliant on the laboratory for the diagnosis. Unfortunately, the laboratory diagnosis of invasive fungal infection is often difficult, and sometimes the interpretation of even positive results from the laboratory is problematic. One of the major difficulties is that blood cultures are negative even in autopsy-proven disease up to 50 per cent of the time. False-negative blood cultures are particularly common in neutropenic patients, but can also occur in less immunocompromised patients. Moreover, a positive culture for Candida does not imply invasive infection in every circumstance, although the prevalence of the disease makes it prudent to consider every positive blood culture of importance.

Biopsy of infected tissue may have the highest diagnostic yield, but culturing of tissue samples may be as disappointing as blood cultures. However, the identification of fungal elements in tissue sections is extremely accurate. There are obvious difficulties in obtaining tissue for histology in critically ill patients, but skin lesions and percutaneous liver biopsies (particularly in cancer patients) are often diagnostic.

The difficulty with establishing the diagnosis of invasive fungal infection compounds the difficulty of the decision to treat the patient. Because it is hard to establish the presence of infection and mortality is high despite appropriate therapy, there are no definitive guidelines available to support the decision to treat. Currently, most authorities believe that most, if not all, patients with Candida fungemia should receive at least a short course of therapy even if fungemia is transient (e.g. a single positive culture with clinical improvement after removal of a colonized venous catheter), as untreated transient fungemia can lead to an invasive infection. Any positive blood culture requires a full course of therapy in a patient who is immunosuppressed or neutropenic, in whom eye lesions are present or Candida has been identified in tissue, or where there is a strong clinical suspicion of invasive infection. Anyone with a positive blood culture for Candida who is not treated requires very careful follow-up, including immediate removal and culturing of all central venous catheters, careful sequential eye examinations, and reculturing of multiple sites in 72 h.

In the surgical ICU, a decision to treat a patient with negative blood cultures must often be made; clinical suspicion is often the key factor in the decision-making process (Dean.ian.d...Burchard 1996). Clinical suspicion of systemic disease with positive serology, tissue isolation, eye lesions, or Candida isolated from two sites (e.g.

urine and sputum) justifies therapy; the last factor is commonplace in surgical illness. Because Candida peritonitis is being diagnosed with increasing frequency in critically ill surgical patients, an isolated finding of Candida in peritoneal fluid may sometimes warrant treatment.

Amphotericin B remains the treatment of choice for invasive fungal infections in critically ill patients. The drug acts by binding to sterols in the fungal cell membrane and increasing membrane permeability, and thus it can be fungicidal in high concentrations. After intravenous administration, the highest drug levels are achieved in lung, kidney, and spleen. Amphotericin B penetrates poorly into body cavities such as the subdural space or peritoneal cavity; thus direct intracavitary administration is sometimes necessary. Irrigation of the urinary bladder (50 mg in 1.0 liters of 5 per cent glucose at 50 ml/h) may also be employed for fungal cystitis. Dosage regimens for amphotericin B have been largely empirical, but accumulating evidence suggests that shorter courses of therapy may be appropriate, even for some serious Candida infections. A test dose (1.0 mg given over 4 h) is no longer considered mandatory, although anaphylaxis and ventricular fibrillation can occur. The usual dose is about 0.7 mg/kg/day. Some investigators suggest that total daily doses of up to 1.5 mg/kg/day may be given with safety, but it is not known whether these higher dosages confer a therapeutic advantage. A common regimen for amphotericin B administration is to treat patients other than those with endocarditis or invasive infections with 500 mg over 1 week. The decision to treat for another week is made on re-evaluation of the patient for signs of both efficacy and toxicity. Success with this type of regimen has been reported in treating surgical and trauma patients, and also for empirical therapy in immunosuppressed patients. However, a total dose of 1 to 2 g is still recommended for endocarditis, fungemia with dissemination or tissue invasion, and fungi other than Candida.

In addition to ototoxicity, amphotericin B may lead to nephrotoxicity in as many as 80 per cent of patients. Irreversible nephrotoxicity is unusual, particularly with smaller total doses and normal baseline renal function, but renal function may not return to baseline for months after a course of therapy. Renal damage may be attenuated by sodium loading or alternate-day therapy. Other important side-effects include renal tubular acidosis, hypokalemia, hypomagnesemia, cardiomyopathy, anemia, and thrombophlebitis. Liposomal amphotericin B has been approved in the United States. The total dose of active medication is probably the same, although toxicity is reduced. The cost of therapy may approach $500/day for a typical regimen of 5 mg/kg/day, although 1 g of liposomal amphotericin B can be administered to a 70-kg patient in only 3 days.

Fluconazole acts by inhibiting ergosterol biosynthesis and is relatively non-toxic. Adverse reactions may occur in 20 per cent of patients; they most commonly include nausea, headache, and skin rash. The recommended dose of fluconazole for serious infections is 400 to 800 mg on the first day followed by 200 to 400 mg/day for a minimum of 4 weeks, or 2 weeks after the resolution of symptoms. Randomized prospective studies of candidemia indicate that the effectiveness of fluconazole is comparable to that of amphotericin B (cure rates are about 80 per cent). The drug may be given orally or intravenously in the same dosage; the oral form is much less expensive. Infections of lesser magnitude (e.g. esophagitis) can be treated with 200 mg initially followed by 100 mg/day for 2 to 3 weeks ( De§.D,..§D.d...,§yichard 1996).

Was this article helpful?

0 0
Natural Arthritis Pain Remedies

Natural Arthritis Pain Remedies

It's time for a change. Finally A Way to Get Pain Relief for Your Arthritis Without Possibly Risking Your Health in the Process. You may not be aware of this, but taking prescription drugs to get relief for your Arthritis Pain is not the only solution. There are alternative pain relief treatments available.

Get My Free Ebook

Post a comment