598 
Possible Mitochondrial Dysfunction in a Large U.S Cohort of HIV-infected Children
Marilyn Crain*1, M Chernoff2, J Oleske3, S Brogly2, K Malee4, P Borum5, W Mitchell6, H Fort-Chatterton7, R Van Dyke8, and G Seage, III2
1Univ of Alabama at Birmingham, US; 2Ctr for Biostatistics in AIDS, Harvard Sch of Publ Hlth, Boston, MA, US; 3Univ of Med and Dentistry of New Jersey, New Jersey Med Sch, Newark, US; 4Children`s Mem Hosp, Chicago, IL, US; 5Univ of Florida, Gainesville, US; 6Children`s Hosp Los Angeles and Keck Sch of Med, Univ of Southern California, US; 7Frontier Sci and Tech Res Fndn, Amherst, NY, US; and 8Tulane Univ, New Orleans, LA, US
Background: Perinatal HIV infection has been
reported to be associated with mitochondrial dysfunction, both in association
with disease progression and in response to ART. Partly due to the absence of a
specific diagnostic algorithm, mitochondrial dysfunction has not been widely
studied in children with HIV infection. Studies conducted to date have included
a small number of HIV-infected children.
Methods: Participants were enrolled in the Pediatric
AIDS Clinical Trials Group (PACTG) protocols 219 and 219C between 1993 and
2004. We used 2 clinical case definitions of mitochondrial dysfunction: the
Enquête Perinatale Francaise (EPF) criteria developed for HIV-uninfected
children perinatally exposed to both HIV infection and ART, and the
Mitochondrial Disease Classification (MDC), developed to identify children in
the general population with respiratory chain disorders. Applicable MedDRA
terms were assigned for each clinical definition. Persistence of conditions for
6 months was required, and laboratory abnormalities were graded using the
National Institutes of Health Division of AIDS 2004 toxicity tables for adverse
events. Associations between recent single and dual NRTI use and possible mitochondrial
dysfunction were estimated using multiple logistic regression.
Results: Of 2931 perinatally infected
children, 982 (34.4%) met the case definition of mitochondrial dysfunction by
one or both classifications, and 135 of the cases (13.7%) died. Of the MDC cases,
80% were incident, compared with 64% of the EPF cases. Overall concordance for mitochondrial
dysfunction case identification between the EPF and MDC was 86.3%, of which 10.8%
met both criteria, and 75.5% met neither. The mortality rate varied between mitochondrial
dysfunction case definition subgroups and was highest among incident cases meeting
both criteria (22.1%) vs 14.9% of incident mitochondrial dysfunction cases
overall. Exposure to stavudine (d4T) was most strongly associated with mitochondrial
dysfunction in the adjusted analyses using both case definitions (EPF OR 2.74;
95%CI 1.97 to 3.82; MDC OR 2.78; 95%CI 2.07 to 3.73). Of NRTI combinations, lamivudine
(3TC)/d4T, d4T/didanosine (ddI), and zidovudine (ZDV)/3TC had adjusted odds
ratios >1.5.
Conclusions: These results are consistent with previous
studies associating NRTI with presumptive mitochondrial dysfunction. Additional
studies are needed to determine potential mechanisms for these findings and to
confirm that these clinical signs are indeed related to mitochondrial
abnormalities.
|