Central nervous system infections in the compromised host: a diagnostic approach

Abstract

The diagnostic approach to the compromised host with CNS infection depends on an analysis of the patient's clinical manifestations of CNS disease, the acuteness or subacuteness of the clinical presentation, and an analysis of the type of immune defect compromising the patient's host defenses. Most patients with CNS infections may be grouped into those with meningeal signs, or those with mass lesions. Other common manifestations of CNS infection include encephalopathy, seizures, or a stroke-like presentation. Most pathogens have a predictable clinical presentation that differs from that of the normal host. CNS Aspergillus infections present either as mass lesions (e.g., brain abscess), or as cerebral infarcts, but rarely as meningitis. Cryptococcus neoformans, in contrast, usually presents as a meningitis but not as a cerebral mass lesion even when cryptococcal elements are present. Aspergillus and Cryptococcus CNS infections are manifestations of impaired host defenses, and rarely occur in immunocompetent hosts. In contrast, the clinical presentation of Nocardia infections in the CNS is the same in normal and compromised hosts, although more frequent in compromised hosts. The acuteness of the clinical presentation coupled with the CNS symptomatology further adds to limit differential diagnostic possibilities. Excluding stroke-like presentations, CNS mass lesions tend to present subacutely or chronically. Meningitis and encephalitis tend to present more acutely, which is of some assistance in limiting differential diagnostic possibilities. The analysis of the type of immune defect predicts the range of possible pathogens likely to be responsible for the patient's CNS signs and symptoms. Patients with diseases and disorders that decrease B-lymphocyte function are particularly susceptible to meningitis caused by encapsulated bacterial pathogens. The presentation of bacterial meningitis is essentially the same in normal and compromised hosts with impaired B-lymphocyte immunity. Compromised hosts with impaired T-lymphocyte or macrophage function are prone to develop CNS infections caused by intracellular pathogens. The most common intracellular pathogens are the fungi, particularly Aspergillus, other bacteria (e.g., Nocardia), viruses (i.e., HSV, JC, CMV, HHV-6), and parasites (e.g., T. gondii). The clinical syndromic approach is most accurate when combining the rapidity of clinical presentation and the expression of CNS infection with the defect in host defenses. The presence of extra-CNS sites of involvement also may be helpful in the diagnosis. A patient with impaired cellular immunity with mass lesions in the lungs and brain that have appeared subacutely or chronically should suggest Nocardia or Aspergillus rather than cryptococcosis or toxoplasmosis. Patients with T-lymphocyte defects presenting with meningitis generally have meningitis caused by Listeria or Cryptococcus rather than toxoplasmosis or CMV infection. The disorders that impair host defenses, and the therapeutic modalities used to treat these disorders, may have CNS manifestations that mimic infections of the CNS clinically. Clinicians must be ever vigilant to rule out the mimics of CNS infections caused by noninfectious etiologies. Although the syndromic approach is useful in limiting diagnostic possibilities, a specific diagnosis still is essential in compromised hosts in order to describe effective therapy. Bacterial meningitis, cryptococcal meningitis, and tuberculosis easily are diagnosed accurately from stain, culture, or serology of the CSF. In contrast, patients with CNS mass lesions usually require a tissue biopsy to arrive at a specific etiologic diagnosis. In a compromised host with impaired cellular immunity in which the differential diagnosis of a CNS mass lesion is between TB, lymphoma, and toxoplasmosis, a trial of empiric therapy is warranted. Antitoxoplasmosis therapy may be initiated empirically and usually results in clinical improvement after 2 to 3 weeks of therapy. The nonresponse to antitoxoplasmosis therapy in such a patient would warrant an empiric trial of antituberculous therapy. Lack of response to anti-Toxoplasma and antituberculous therapy should suggest a noninfectious etiology (e.g., CNS lymphoma). Fortunately, most infections in compromised hosts are similar in their clinical presentation to those in the normal host, particularly in the case of meningitis. The compromised host is different than the normal host in the distribution of pathogens, which is determined by the nature of the host defense defect. In compromised hosts, differential diagnostic possibilities are more extensive and the likelihood of noninfectious explanations for CNS symptomatology is greater. (ABSTRACT TRUNCATED)