A Careful Neurologic Examination Should Precede Neuroimaging Studies in HIV-Infected Patients with Headache

Headache is a common symptom in patients with HIV infection. It is the most common form of pain, more common than painful peripheral neuropathy, and has been reported as the presenting manifestation of AIDS in 12.5% to more than 55% of patients (1). Headaches in HIV-infected patients may be caused by HIV-associated aseptic meningitis, antiretroviral and other therapies, or opportunistic infections and neoplasms arising in the face of impaired cellular immunity. Causes of the latter are legion and include opportunistic infections such as CNS toxoplasmosis, cryptococcal meningitis, primary central nervous system lymphoma, and lymphomatous meningitis. Sinusitis is common with advanced HIV immunosuppression and is also a potential cause of headaches. The frequency with which significant disease is determined to be the cause of headache in this population is highly dependent on the nature of the population studied. For instance, Goldstein found serious underlying disease in 40 (82%) of 49 patients in a study of hospitalized patients (2), whereas Singer found that intracranial opportunistic infection or tumor was present in only 4% of HIV-infected patients presenting as outpatients (3). Nevertheless, the vast majority of headaches in HIV-infected patients are no different from headaches in a control HIV-uninfected population (l). These headaches bear no relation to CD4 T-lymphocyte counts, cerebrospinal fluid parameters, cranial MR imaging abnormalities, the presence of sinusitis, or the use of zidovudine (1). They have been associated with polypharmacy, depression, anxiety, and insomnia, and are often relatively unresponsive to conventional headache therapies (4). Multiple causes of headaches may be observed in a single patient.

When confronted by an HIV-infected patient complaining of headache, the treating physician's chief concern is the presence of a life-threatening, treatable illness. Does the patient have an intracranial mass lesion or meningitis? Unfortunately, no consensus has evolved regarding the best diagnostic approach for the evaluation of the HIV-infected person presenting with headache. The absence of fever, neck stiffness, and altered mental status has been suggested as effectively eliminating the possibility of meningitis, but whether this remains true in the HIV-infected population is open to question. Neurologists with experience in treating patients with AIDS recognize the importance of CD4 cell counts in their approach to management of the HIV-infected patient with headache. Generally, in the absence of significant immunosuppression, CT or MR imaging is not suggested unless focal findings are present on neurologic examination.

The findings in the study by Graham and colleagues (page 451) are not unexpected; their population was screened in advance. Allegedly, none had altered mental status, meningeal signs, neurologic findings, or “the worst headache of their life.” In their HIV-seropositive population without significant immunosuppression, defined as having a CD4 cell count >200 cells/cu mm, no abnormal findings on unenhanced and contrast-enhanced cranial CT scans were noted. Even in the group with CD4 cell counts <200 cells/cu mm, only 14 (15.7%) of 89 had mass lesions and another 4 (4.5%) had white matter lesions. This is not unlike observations found with CSF analysis in which CD4 cell counts >200 cells/cu mm militate against the likelihood of finding opportunistic infection. Although it is clear that the risk of opportunistic CNS infections greatly increases when CD4 cell counts drop below 200 cells/cu mm, these infections certainly occur in patients with higher cell counts. Approximately 10% of AIDS patients with cerebral toxoplasmosis or progressive multifocal leukoencephalopathy (PML) have CD4 cell counts >200 cells/cu mm.

Importantly, Graham and colleagues do not address the quality and thoroughness of the neurologic examinations performed. Surely there is a difference between the hurried house officer's examination in an emergency room and the refined examination performed by an experienced neurologist for probing the characteristics of a headache and eliciting subtle focal neurologic abnormalities. This is an essential point. How many of the 14 patients with intracranial mass lesions would have exhibited absent spontaneous venous pulsations on funduscopy, a subtle upper extremity drift, limb incoordination, increased tone, or sensory loss had a neurologist performed the examination? A careful history is equally important. A convincing history of migraine or cluster headaches, even in the face of severe immunosuppression, militates strongly against opportunistic infection or space-occupying lesions (4). Similarly, chronic daily headaches, defined as daily and near-daily headaches lasting for more than 4 hours per day for more than 15 days per month for more than 1 month, argues against underlying identifiable intracranial disease, particularly in the absence of abnormalities on neurologic examination (4). A neurologic consultation is more cost-effective than either a CT scan or an MR examination. Additionally, other neurologic complications of HIV infection, such as early cognitive impairment, peripheral neuropathy, or myelopathy, may be recognized, whereas they easily may be overlooked without careful neurologic examination. While obtaining a CD4 cell count is certainly worthwhile, as the authors suggest, for patients on antiretroviral therapies, this result is generally available from outpatient studies proximate to the time of the patient's presentation with headache. CD4 cell counts are not inexpensive nor is a test performed in an urgent fashion—facts providing another argument for a formal neurologic consultation when evaluating the HIV-infected patient with headache.

The presence of mass lesions or meningeal enhancement (not observed among the patients in this study) is the major concern. Cerebral atrophy and white matter lesions as described by the authors are not particularly relevant when addressing the issue of headaches. Although both may be features of HIV dementia, HIV dementia does not appear to be a cause of headache. The authors present insufficient data to determine whether the white matter lesions seen were the consequence of PML, which may on occasion be associated with headache. Headache in isolation, ie, without focal neurologic abnormalities, however, would be highly unusual for PML.

Graham and colleagues have demonstrated a relationship between significant immunosuppression (CD4 cell count <200 cells/cu mm) and the presence of intracranial abnormalities at the time of CT scan. They demonstrate that even in the immunosuppressed population, clinically relevant lesions remain relatively rare. Many questions remain regarding the best approach to the evaluation of these patients; however, the value of a thorough and precise neurologic examination in the evaluation of these patients cannot be underestimated.