June's meeting in Sitges had a wealth of posters addressing the spread of drug-resistant HIV-1 in different patient populations. For the first time, evidence was presented that the rates of transmission of drug resistant virus are becoming heterogeneous, with higher rates of drug resistance seen in some cities like New York City, Seattle and Montreal, and declining rates in Switzerland and San Francisco. The reasons for these differences have yet to be determined.

Data on 74 seroconverters from the Netherlands showed that 8/74 (11%) had AZT resistance mutations [1]. Four of the patients had sequential samples obtained where T215Y was overgrown by T215D/S/T. These variants rapidly reverted to T215Y in vitro when AZT or d4T was added to the culture media. These data suggest that T215Y/F/D/S/T should all be considered evidence of the transmission of AZT resistance in newly infected patients. A similar analysis of 110 patients from the US and UK showed that isolates from 11/110 primary infections during 1997-8 had T215Y/F/D/S/N mutations, suggesting a transmission rate of drug resistant virus of 10% [2]. The CDC presented data from 72 newly infected patients studied from 10/97 to 9/99 [3]. All of the drug-resistant isolates in this cohort came from 53 patients in Seattle. For this group 4/53 (8%) had AZT resistance mutations and 2/53 (4%) had evidence of NNRTI resistance. They suggested that there may be different prevalence rates of drug resistant HIV-1 in naive patients from different locations. Overall, these data suggest that AZT-resistant virus transmission rates remain stable at 8% to 10% of newly HIV-infected persons in the United States and Europe.

Studies in two cities with well-defined HIV seroconverter cohorts suggest that drug resistance transmission rates are going down, with decreased prevalence of nucleoside resistance but increases in NNRTI and PI resistance [4,5]. The San Francisco cohort looked at 118 subjects seen at SFGH between 6/96 and 12/99 [4]. During that period nucleoside resistance in newly infected persons decreased from 15.4% to 6.3% (P=0.08), while NNRTI resistance increased from 0 to 6.3% (P=NS) and PI resistance mutations also increased from 0 to 6.3% (P=NS). Similar data were presented on 198 newly infected persons from Switzerland followed from 1996 to 2000 [5]. The overall rate of drug resistance transmission was 16/198 (8%). Of note, the rates of drug resistance to NRTI and PI both peaked in 1997 at 10% and 8.6%, respectively. Rates of NRTI- and PI-resistant virus have subsequently declined to 3.4% and 2%, respectively, in 1999. It is suggested that potent combinations of antiretroviral drugs combined with increased clinical monitoring may be decreasing the rates of drug-resistant virus transmission.

Similar decreases in drug-resistant transmissions do not appear to be occurring in New York City [6] and Montreal (AIDS. 2000;14:F17-F23). The New York cohort looked at 41 newly infected persons seen between 4/99 and 3/2000. Three of 31 viruses assessed phenotypically had drug-resistant virus and 2 viruses were multidrug resistant. Genotypic testing revealed that 10/36 viruses (27%) had RT resistance mutations while 5% had the PI mutation L90M. The reason for apparently higher levels of drug resistant virus transmission in Seattle, New York and Montreal, but declining rates in Switzerland and San Francisco, are uncertain and need to be evaluated closely.

Data were presented for two maternal-infant pairs from Spain where drug resistant virus transmission was documented [7]. In both instances the mothers had high viral loads at term: 150,000 copies/mL and 55,000 copies/mL, respectively. The authors suggested that resistance testing should be performed on both maternal and infant blood samples since one infant only received some of the mutations that were present in his mother's virus.

The CDC looked at 61 needle stick source patients and genotyped 48 samples where there was a detectable viral load [8]. Forty percent of source patient samples had detectable resistance mutations with rates of 67% in patients on treatment and 12% in patients not on therapy. Five patients had mutations not associated with their current drug regimen. No transmission events occurred in this small cohort. The authors recommended that the source patient's antiretroviral treatment history be obtained prior to initiating post-exposure prophylaxis.

Emerging public health data suggest that a very heterogeneous pattern of drug-resistant HIV-1 transmission is developing in North America and Europe. Data is too limited to determine whether this is due to different populations of at-risk patients, different practice styles or simply random chance due to small numbers of patients who are evaluated at each site. There is clearly a need to develop surveillance systems for HIV-1 similar to current surveillance of drug-resistant pneumococcus and tuberculosis to allow local practitioners to determine the need for resistance testing of newly infected persons based on local prevalence rates. The CDC has recently proposed a new initiative to develop a wider surveillance network for primary HIV infections in the US.