Parasitic infections and antiparasite treatment have no effect on HIV levels in HIV-positive adults in Malawi

Parasitic infections particularly helminths were more common in HIV-uninfected than HIV-infected adult outpatients, and neither helminth infections nor anti-parasitic treatment had an impact on HIV viral loads, Malawian and US researchers report in the May 1^st edition of the Journal of Infectious Diseases.

In African countries with an escalating HIV/AIDS problem, co-infections with parasitic helminth and protozoan parasites are common and account for substantial morbidity in people living with AIDS (PLWA). There has been a long-standing concern that helminth co-infections in PLWA might drive the CD4 T-lymphocyte response from a Th1 to Th2 response and cause chronic immune activation. These effects might not only combine to exacerbate the virulence of both infections but might also increase the risk of HIV transmission to sexual partners or unborn babies.

Despite these concerns, the effect of HIV helminth co-infections on the epidemiology of HIV infection remains unclear. Studies of the effect of antihelminthic treatment on HIV viral loads have yielded conflicting results; some demonstrated a reduction while others did not.

It is against this background that a team of Malawian and US researchers investigated the prevalence of enteric and urinary parasitic infections in adult Malawians, estimated the prevalence of specific parasitic infections by HIV infection status, and studied the effect of parasitic treatment on HIV RNA levels among HIV-infected ART-naïve individuals.

The study took place in Lilongwe, Malawi. Individuals visiting HIV counselling and testing services or out-patient clinics at Kamuzu Central Hospital were recruited. At enrolment, patients received HIV counselling and testing including viral load determinations, stool and urine samples were collected for parasitological examinations, and a demographic and clinical questionnaire was completed.

One week after enrolment, all patients returned to receive treatment for the specific identified infections. HIV-infected patients and those with parasitic co-infections returned 4 weeks after treatment. Blood samples were taken again for HIV viral loads, as were stool and urine samples for parasitological examination to ascertain treatment success, and a questionnaire was completed.

HIV diagnosis was based on two rapid HIV tests; a third test (HIV ELISA) was carried out if the two tests gave discordant results. Measurement of HIV RNA levels and CD4 counts and parasitic examination for helminth eggs and other pathogens were carried out using standard procedures. Schistosomiasis was defined as infection with Schistosoma mansoni or Schistosoma hematobium, while the term ‘helminth’ was used to designate both schistosomiasis and other geohelminths. Treatment for specific helminth or protozoan infections was according to national guidelines.

Out of 389 patients, 266 (68 %) were HIV-infected; the PLWA were more likely to be women (76 versus 49 %; p

At baseline, 43 % of all patients (167/389) had evidence of at least 1 parasitic infection; among the parasitized patients, the majority (70 %) had single infections, while double and polyinfections with 3-4 parasites were seen in 25 and 6 %, respectively. The commonest helminth infection in the study population was by hookworms.

Overall, HIV-uninfected patients were significantly more likely to have at least one parasitic infection including a helminth, geohelminth, schistosomiasis, hookworm, or mixed infection by comparison with HIV-infected patients. The prevalence of pathogenic and non-pathogenic protozoan infections were comparable by HIV infection status.

Baseline HIV RNA and CD4 counts were available for 264 patients. Significantly, HIV RNA levels in patients with and without parasitic co-infections were comparable (median 4.89 versus 4.92 log₁₀ copies/ml; p = 0.95). HIV RNA levels were also comparable by type of parasitic infection.

Lower CD4 counts were observed in patients without a helminth or protozoan coinfection. Median CD4 counts in patients with and without helminth infections were 235 (interquartile range 110-392) and 320 (IQR 251-558) cells/mm³, respectively (p 3, respectively (p = 0.002).

Among the 73 HIV-infected patients who had at least one pathogenic parasitic infection, 71 received treatment and 63 returned for follow up evaluation. Baseline and four-week post-treatment HIV RNA levels were available for 57 (80%) patients and were comparable (4.98 versus 4.86 log10 copies/ml; p = 0.89). Even when the analysis was restricted to patients without evidence of parasitic infection during follow up beyond 4 weeks, baseline and post-treatment RNA levels were still comparable. There was no significant change in HIV RNA level by type of parasitic infection, sex, age, or baseline CD4 count.

Hosseinipour et al. point out the major limitations of the study. First, worm burdens were not quantified. Lighter worm burdens might not cause the immune activation or Th1/Th2 shift which is significant enough to impact on HIV replication. Second, other co-endemic infections like malaria and tuberculosis were not investigated and yet they have an effect on HIV RNA levels. Third, the faecal stool smears might have underestimated the prevalence of helminth parasites; a more sensitive antigen detection technique would have been ideal.

Despite these limitations, these findings have important policy implications. It might not be justifiable to de-worm PLWA because this study and others like it do not provide any compelling evidence that antihelmintic treatment will have an impact on HIV transmission or disease progression. But would a policy of withholding antihelmintic treatment from HIV-infected patients be relevant across all African countries?

It might be necessary to carry out a metanalysis of all the studies that have so far examined the effect of helminthic coinfections and antihelmintic treatment on HIV RNA levels. It is possible that the different study sites might have different predominant geohelminth species and worm burdens. It is possible that some geohelminths might exert a more inimical effect on viral loads in PLWA than others. It might also be that the study protocols of such studies in the future will need to be standardized if their results are to inform on antihelmintic treatment policy in PLWA in Africa.