CNS Involvement in Herpes Zoster

abstract & commentary

Synopsis: CNS abnormalities are commonly present in patients with herpes zoster.

Source: Haanpaa M, et al. CSF and MRI findings in patients with acute herpes zoster. Neurology 1998;51: 1405-1411.

Fifty immunocompetent patients, 22 men and 28 women, with acute herpes zoster but without clinically evident CNS complications including signs of meningeal irritation, encephalitis, or myelitis underwent CSF study (n = 46), MRI (n = 16), or both (n = 16) to correlate laboratory abnormalities with clinical findings and outcomes. CSF analysis included measurement of cell count, total protein, IgG index, oligoclonal banding (OCB), IgG and IgM anti-varicella zoster virus (VZV) antibodies, and PCR assay for VZV DNA. MR imaging comprised cranial (n = 12) or cervical studies (n = 4), for cranial or cervical VZV eruptions, respectively, but due to flow artifacts, thoracic and lumbar cord MRI scans were not performed. Statistical analysis included Fisher’s exact test, log-linear modeling, and logistic regression.

CSF pleocytosis was present in 46% (21/46) overall, and was seen in the first sample in 17 of the 21 patients, was present in the first sample in seven of seven patients with HZ-related MR abnormalities but in only one of four patients with a normal MRI (P = 0.01). OCB was present in 9% (4/43), two of which had HZ-related MR findings. Approximately 25% each had elevated CSF protein, anti-VZV IgG antibodies, or VZV DNA, but this did not correlate with clinical outcome or MR abnormalities. IgG index was abnormal in only one of 42 patients. Overall, CSF analysis demonstrated at least one abnormality in 61% (28/46), and MRI findings attributable to HZV were seen in 56% (9/16). None of the MRI abnormalities were enhanced with gadolinium, implying only mild inflammation or necrosis unlikely to lead to permanent changes. Interestingly, five of nine patients with abnormal MR scans developed post-herpetic neuralgia (PHN) within three months. PHN affected none of seven with normal MR imaging. The numbers are too small to be predictive but if extended to larger studies, may influence the initiation of treatment for HZ and prevention of PHN.

Comment by michael rubin, MD

This study confirms the high frequency of CNS involvement in patients with herpes Zoster. Of interest is the apparent association of MRI abnormalities with the development of the most frequent disabling complication of this infection, PHN. Pain control in PHN remains challenging. Standard therapy includes amitriptyline (AT) but lack of efficacy (up to 50%) and unpleasant side effects limit its usefulness (Watson CPN. Neurology 1995;45(suppl 8):58-60). In a double-blind, randomized cross-over trial of 31 patients suffering moderately severe PHN for at least half-a-day for three or more months, amitriptyline was compared to nortriptyline (NT), its major noradrenergic metabolite, for efficacy and side effect profile. Patients with cardiac disease, seizures, brain damage, alcoholism, and severe depression were excluded. Treatment for five weeks, with a two week wash-out period before cross-over to the other agent, was initiated at 10-20 mg and increased by 10 mg increments every 3-5 days. Primary outcome measures, taken weekly, included pain evaluation, pain relief, and sleep (all measured using a visual analogue score), depression (using the Beck Depression Inventory), side effects, disability, and overall satisfaction.

AT and NT equally controlled pain in about 50% showing no significant difference with respect to mood, disability, satisfaction, or use of concomitant analgesic medication. Mean dose was 58 mg and 75 mg for AT and NT, respectively, among responders, and 68 mg and 97 mg among non-responders. Side effects were more common in the AT group, including dry mouth, constipation, and drowsiness, but tolerability was comparable in both groups.

In a similarly designed double-blind crossover study of 38 PHN patients (including 16 from the AT trial!), controlled-release oxycodone, a semi-synthetic opioid analgesic, 10 mg (with titration up to a possible 30 mg) twice-a-day was found to be significantly more effective than placebo for relief of steady pain, paroxysmal brief pain, allodynia, disability, global effectiveness, and patient preference. Constipation, nausea, and sedation were more common with oxycodone but did not overly contribute to drug discontinuation. Oxycodone and tricyclic antidepresssants appear comparable in their analgesic efficacy for PHN. (Dr. Rubin is Associate Professor of Clinical Neurology, New York Presbyterian Hospital-Cornell Campus, New York, NY.)