Vacuolar myelopathy is the most common chronic myelopathy associated with HIV infection. HIV-associated vacuolar myelopathy occurs during the late stages of HIV infection, when CD4+ lymphocyte counts are very low, often in conjunction with AIDS dementia complex, peripheral neuropathies, and opportunistic infections or malignancies of the central or peripheral nervous system (eg, cytomegalovirus, progressive multifocal leukoencephalopathy, lymphoma).
Several hypotheses have been proposed to explain the development of this common complication of HIV-1 infection. One hypothesis is infiltration by HIV-infected mononuclear cells that secrete neurotoxic factors, including cytokines, possibly in conjunction with neurotoxic astrocyte factors. A significant amount of scientific support exists for this paradigm. Transgenic mice that express HIV gene products in oligodendrocytes develop clinical and histologic features that resemble the human disease.
Although direct HIV infection of astrocytes and neurons is reported in the brain and dorsal root ganglia, it is not a major feature in vacuolar myelopathy.
The impaired ability to utilize vitamin B-12 as a source of methionine in transmethylation metabolism for myelin maintenance in the spinal cord may be a contributing factor.
Epidemiology and Prognosis
Before the introduction of highly active antiretroviral therapy (HAART), vacuolar myelopathy was seen in 5-20% of adult HIV patients in clinical studies and in 25-55% of adult HIV patients in histologic studies. Since the introduction of HAART, it is estimated that fewer than 10% of AIDS patients develop HIV myelopathy.
The prognosis in HIV patients with vacuolar myelopathy is poor. Most patients die within 6 months of developing symptoms.
Vacuolar myelopathy typically presents as a slow progression of painless leg weakness, stiffness, sensory loss, imbalance, and sphincter dysfunction. Relapsing-remitting courses have also been described. 
Back pain is not a prominent feature. Arm function is usually normal except for advanced vacuolar myelopathy.
Vacuolar myelopathy is often associated with AIDS dementia complex and peripheral neuropathy. In such cases, patients have cognitive decline and distal limb pain and numbness.
The following may be noted on physical examination:
Slowly progressive spastic paraparesis
Hyperreflexia and extensor plantar responses
Rarely, asymmetric features and involvement of upper extremities
A discrete sensory level is usually absent; if present, this strongly suggests other causes of myelopathy.
The differential diagnosis includes neurosyphilis. Other problems to be considered in the differential diagnosis of HIV-related vacuolar myelopathy include the following:
Herpes simplex virus infection
Cervical disk syndromes
HIV-1 associated conditions (eg, Kaposi sarcoma, lymphoma, toxoplasmosis) in the spinal canal
Mycobacterium tuberculosis infection
Compressive myelopathy (eg, tumor, abscess, herniated disc, arteriovenous malformation)
Human T-cell lymphotropic virus type 1 (HTLV-1)–associated myelopathy/tropical spastic paraparesis (can co-exist with HIV infection)
Spinal epidural abscess in HIV-infected parenteral drug users
Electrophysiologic tests can confirm a clinical diagnosis of vacuolar myelopathy. Somatosensory evoked potential (SSEP) may be a valuable tool in the diagnosis of AIDS-associated myelopathy, particularly when myelopathy and peripheral neuropathy coexist. 
Other studies are used to identify or rule out other potential causes of myelopathy. CSF analysis can exclude infection with cytomegalovirus, varicella-zoster virus, herpes simplex virus, HTLV-1, and HTLV-2. CSF results are usually normal in HIV-1–associated vacuolar myelopathy.
Serum studies can determine vitamin B-12 and folic acid levels. In patients with borderline low B-12 levels, elevated homocysteine and methylmalonic acid levels are better indicators of a deficiency. B-12 levels are usually normal in vacuolar myelopathy.
A Schilling test, hematologic studies, and CD4+ lymphocyte counts may be indicated.
CT scan results are often noncontributory but may reveal unsuspected coexisting conditions such as extramedullary or intramedullary infections, neoplasms, degenerative disk disease, or degenerative joint disease of the spine.
Magnetic Resonance Imaging
MRI scans are often noncontributory but may reveal unsuspected coexisting conditions such as extramedullary or intramedullary infections, neoplasms, degenerative disk disease, or degenerative joint disease of the spine.
Spinal cord atrophy is the most common abnormal finding involving the thoracic cord with or without cervical cord involvement. T2-weighted MRI often shows symmetric nonenhancing high-signal areas, which are present on multiple contiguous slices and are usually symmetrical; these may result from extensive vacuolation. [3, 4, 5] Lesions may be confined to the posterior columns, especially the gracile tracts, or may be diffuse. (See the images below.)
Histologic findings may indicate multifocal, occasionally asymmetric vacuolation and myelin pallor involving the dorsal and lateral tracts more than the anterior and anterolateral tracts and involving the cervical and thoracic segments more than the lumbar segments or the brainstem, accompanied by astrogliosis. (See the images below.)
HIV-infected, activated, lipid-laden macrophages and microglia expressing interleukin-1 and/or tumor necrosis factor-alpha may be seen. [6, 7] On electron microscopy, intramyelinic or periaxonal vacuoles and, rarely, disrupted axons may be seen.
The histology resembles subacute combined degeneration from vitamin B-12 deficiency. 
Treatment & Management
Care for patients with HIV-associated vacuolar myelopathy is primarily supportive. Although no specific treatment is currently approved for this syndrome, viral control tailored to the individual patient's medical and viral history is important, as case reports showing clinical and radiologic improvement with highly active retroviral therapy (HAART) have been described. [9, 10, 11, 2]
A pilot study showed improvement in patients treated with L-methionine.  However, a randomized, double-blind, placebo-controlled study of 56 patients found that L-methionine was of no benefit. 
Refer the patient to a physical medicine specialist for spinal cord treatment and follow-up care.