WP 206 Natural History of Syphilis in HIV-Positive Patients in the HAART Era

Tuesday, June 10, 2014
International Ballroom
Karen Peterson, MD, Denver Public Health, Denver Health, Denver, CO and Robert Beum, BS, Denver Public Health, Denver Health and Hospital Authority, Denver, CO

Background:  Syphilis is common in HIV-infected patients and early progression to neurosyphilis may occur. Studies suggest CD4 count may affect this, but most were done prior to use of modern HAART therapy for HIV.  We reviewed our experience of syphilis in HIV-positive patients seen in the Denver Health HIV/AIDS clinics from 2002 to the present to characterize the natural history of syphilis in the HAART era. 

Methods:  Syphilis diagnoses in HIV-positive patients from 2002-2013 were identified from an electronic database, with hand chart review done to find syphilis serologies.  When lumbar puncture (LP) was done to rule out neurosyphilis, symptoms, CD4 count, viral load, and HAART use were also extracted.  Statistical significance was analyzed using Fisher’s exact test, 2-tailed.

Results:  187 episodes of syphilis were identified and fully reviewed.  Titers ranged from 1:8 to >= 1:4096, with a median of 1:128.  No prozone phenomenon was seen.  Regardless of titer, 86% had a >= 4-fold drop in titer by 6 months after treatment.  32 cases were identified where LP was done to rule out neurosyphilis; 15 had neurosyphilis while 17 had normal CSF results.  No HAART use, CD4 <350, and detectable viral load were significantly more common in the neurosyphilis cases (p < .05 for all).  The RPR titer was not predictive.  Symptoms in the neurosyphilis cases included fever/rash (1), gait/balance disturbance (2), hearing loss (1), headache (3), and visual changes (7).  Of 9 asymptomatic patients with high RPR titers, only 1 had neurosyphilis.

Conclusions:  Denver Health has seen early progression to neurosyphilis in HIV-positive patients, with CD4 < 350, no HAART use, and detectable viral load all associated with neurosyphilis in patients where LP was done to rule it out.  The RPR reagent in use by our lab did not require serum dilution to detect even extremely high titers.