April 14, 1997

Association of Tuberculosis Risk With the Degree of Tuberculin Reaction in HIV-Infected Patients

Author Affiliations

From the Centro di Riferimento AIDS e Servizio di Epidemiologia delle Malattie Infettive, Ospedale Lazzaro Spallanzani, Rome, Italy (Drs Girardi, Antonucci, and Ippolito); Tuberculosis Program, World Health Organization, Geneva, Switzerland (Dr Raviglione); Osservatorio Epidemiologico Regione Lazio, Rome (Dr Rapiti); Istituto di Immunologia e Malattie Infettive, Università di Verona, Verona, Italy (Dr Di Perri); Istituto di Malattie Infettive, Università di Sassari, Sassari, Italy (Dr Babudieri). The members of the Gruppo Italiano di Studio Tubercolosi e AIDS and participating institutions appear in the acknowledgment section at the end of this article.

Arch Intern Med. 1997;157(7):797-800. doi:10.1001/archinte.1997.00440280129013

Background:  The risk of developing active tuberculosis associated with a different size of induration to purified protein derivative (PPD) has not been prospectively assessed among individuals infected with human immunodeficiency virus (HIV). The quantification of this risk is important to more appropriately identify candidates for preventive therapy for tuberculosis.

Methods:  A prospective, multicenter, cohort study on tuberculosis in HIV-infected patients was conducted in 23 infectious disease units in public hospitals in Italy. Two thousand six hundred ninety-five HIV-infected patients were enrolled in the study. Of these, 1054 patients who were nonanergic at the time of entry were included in the present analysis. The median duration of follow-up was 102 weeks. The main outcome measure was a diagnosis of active tuberculosis confirmed by the isolation of Mycobacterium tuberculosis in culture.

Results:  Among the 252 patients with PPD reactivity, patients with an induration to PPD of 2 to 4 mm had a median CD4+ lymphocyte count of 0.34X109/L (interquartile [IQ] range, 0.14X 109-0.56X 109), those with a response of 5 to 9mm had a median count of 0.38X109/L(IQrange, 0.24X109-0.56X 109), and those with a response of 10 mm or higher had a median count of 0.37X 109/L (IQ range, 0.23X 109-0.52X109) (P=.38). Compared with the 802 nonanergic PPD-negative patients, hazard ratios of tuberculosis were 2.1 (95% confidence interval [CI], 0.2-18.3) among the 55 patients with a response to PPD of 2 to 4 mm, 5.7 (95% CI, 1.6-19.8) among the 128 patients with a response to PPD of 5 to 9 mm, and 23.1 (95% CI, 7.8-68.6) among the 69 patients with a response to PPD of 10 mm or higher.

Conclusions:  Among nonanergic HIV-infected patients, the degree of response to tuberculin does not appear to reflect the degree of immunosuppression and is strongly correlated with the subsequent incidence of tuberculosis. To identify HIV-infected patients who are at an increased risk of tuberculosis and may benefit from preventive therapy, a response to PPD of 5 mm appears to be an appropriate cutoff point.Arch Intern Med. 1997;157:797-800