Warning: antiretroviral treatment interruption could lead to an increased risk of HIV transmission.

Abstract

Considering the load of antiretroviral therapy in the life of HIV-infected patients in terms of toxicity and the necessity for adherence, in addition to data concerning the interests of interruption in different clinical situations, treatment discontinuation is increasingly being discussed by patients and physicians. We would like to stress one aspect related to the rebound of HIV-1 replication after antiretroviral treatment interruption. Patient A is a 27-year-old man. He was referred to our clinic with symptoms of acute primary HIV infection, characterized by fever, myalgia, erythematous rash and oral candidiasis. The patient knew that he was at risk of HIV because he had had regular unprotected sexual intercourse with a friend followed in our institution for HIV-1 chronic infection (patient B). The diagnosis was confirmed by a positive p24 antigenemia. The results of HIV serologies associated a positivity for the tests Axsym gO (Abbott Diagnostics, Rungis, France) and Genscreen Plus (Biorad Laboratories, Marnes La Coquette, France) and an undetermined Western blot profile with low reactivity for p25 and p34 and negativity for glycoproteins. The plasma HIV-1 viral load using the Monitor 1.5 test (Roche Diagnostics, Meylan, France) was higher than 5.8 log (> 750 000) copies/ml and the CD4 cell count was 516 cells/mm3. Five days later, treatment combining indinavir, ritonavir, didanosine, lamivudine and hydroxyurea was initiated. One month later the Western blot was positive with antibodies against gp160, gp120, p55, p25 and p18, confirming HIV-1 seroconversion, and the viral load had decreased to 3.67 log (4730) copies/ml. Patient B is a 41-year-old man, known to be infected by HIV-1 since 1988 and treated since January 1999. At that time his HIV viral load was 212 000 log (5.32) copies/ml and his CD4 cell count was 288 cells/mm3. He started treatment with lamivudine, stavudine and efavirenz. After 4 weeks, his viral load decreased to below the cut-off value of 200 copies/ml and remained unchanged for 20 months, with an increase in his CD4 cell count to 710 cells/mm3. As he was complaining of fatigue and depression, the decision was taken to discontinue the antiretroviral treatment for one month during his vacation, in order to assess the dependence of the symptoms on his therapy and the evolution of the immunological and virological parameters. In fact, he returned for evaluation after 8 weeks of interruption when his friend was seen for primary HIV infection. His viral load had increased to 242 000 copies/ml and his CD4 cell count had decreased to 405 cells/mm3 without any clinical symptoms. The two patients started their relationship approximately 12 months before patient A seroconverted, and patient A had not had sexual intercourse with any other partner than patient B during this one year period. The nucleotidic analysis of the genes encoding the reverse transcriptase (nucleotides 1–730) and the protease (nucleotides 1–300) showed a homology of 100% between the HIV-1 strains infecting patients A and B, without mutation conferring a resistance. This report leads to some conclusions. It has been reported that the HIV viral load in the plasma is a marker for transmission [1,2], as assessed by comparing levels of HIV in the semen and plasma [3], and subsequently that the risk of sexual transmission was higher during primary HIV infection [4]. The rebound of HIV viral load after antiretroviral treatment discontinuation could be compared with that observed during acute infection in some cases [5,6]. It is very likely that HIV transmission occurred during the viral rebound after patient B's treatment discontinuation. The HIV viral load of patient B after 2 months of discontinuation was similar to the values observed before treatment. This rebound was predictable, as described by many authors [7–10]. It is thus very important to inform patients that ‘undetectable’ viral load is an inappropriate term and does not mean the absence of transmission risk, and to insist on the message that this risk will considerably increase during any antiretroviral treatment discontinuation. We are now facing an increasing number of HIV primary infections related to an increase in unprotected sexual behaviour. The practice of ‘drug holiday’ may dramatically worsen this situation. Roland Tubianaa Jade Ghosna Marcio De-Saa Marc Wirdenb Agnès Gautheret-Dejeanb François Bricairea Christine Katlamaa