Report 6 -  4th International Workshop on HIV Drug Resistance and Treatment Strategies
Written by Jules Levin
Sitges, Spain, June 12-16 2000

Tipranavir: in vitro resistance data on a protease inhibitor for PI resistance

Tipranavir is a non-peptidic dihydropyrone protease inhibitor that has antiviral activity against broadly PI resistant HIV. At last year's Resistance Workshop I San Diego, Brendan Larder reported that 96/107 highly PI cross-resistant clinical isolates were fully sensitive (< 4-fold resistance or a mean of 2-fold resistance) to tipranavir when tested in vitro. Eight of the 107 isolates had intermediate resistance (4-10 fold resistance), and only 3 clinical isolates had resistance (>10-fold). He suggested tipranavir's activity against PI resistant HIV could be due to a flexible mode of binding in the protease active site. Since last Summer, tipranavir development rights were transferred to Boehringer Ingelheim from Pharmacia & Upjohn. Clinical study development has been slow as a result of the transfer. At this year's meeting Sharon Kemp of Virco reported on additional in vitro resistance testing of tipranavir.

Kemp reported that of 85 clinical isolates that had greater than 10-fold resistance to at least 4 of the current protease inhibitors in a recombinant phenotypic assay, 74 (87%) remained completely susceptible to tipranavir. Limited phenotypic resistance to tipranavir has been seen. The purpose of Kemp's study was to define the patterns of PI resistance mutations associated with TPV resistance.

They constructed 11 site directed mutants containing 6-8 known protease muations which were highly cross-resistant to other protease inhibitors, and >4 fold resistant to TPV. The most comon genotypic mutations were at positions 10, 20, 36, 46, 54, 71, 84, together with 82T and 90M. 20 site directed mutants were constructed, and upon phenotypic testing the site directed mutant with only the backbone mutations had a 2 fold tipranavir resistance. When 82T or 90M or both were added to the secondary mutations a mean 2.4 fold tipranavir resistance resulted. They constructed a series of site-directed mutants with various mutations (up to 10 mutations in one mutant) selected from 4 viruses with >10 fold TPV resistance but were unable to create TPV resistance greater than 4 fold. Often TPV had no more than 2 fold increased IC50. One virus they created included 82A, 54V, and 90M with a background of 10/20/36/71/84. Although these viruses were resistant, and in a number of cases highly resistant, to other protease inhibitors, they were sensitive to TPV.

To further test for resistance they passaged a pre-existing PI resistant but tipranavir sensitive clinical isolate in escalating doses of tipranavir (up to 30uM). This isolate had >10 mutations and was >40 fold resistant to IDV, RTV, NFV, and SQV but was sensitive to TPV.  13-33 fold tipranavir resistance developed in culture with the following mutations developed: I47V, V82L, I85V, T91A in a background of 10I, 20I, 46I, 77I, 84V, and 90M. Thereupon, Kemp created  a site-directed mutant with a 47V mutation added to 82L and 54V to the backbone of 10/20/36/71/84. The mean fold increase in IC50 for TPV was about 5-fold.

Kemp concluded that there does not appear to be an obvious combination of mutations associated with tipranavir resistance. Complex combinations of mutations were observed in PI cross-resistant samples with > 4 fold increase in IC50 to TPV (n=11). Unusual mutations at codon 82 (T/L) appear to play a role in TPV resistance. To date, re-constructed virus mutants have failed to show high level TPV resistance, although resistance development is possible by in vitro selection.

The in vitro resistance data so far generated suggests tipranavir will be helpful as a salvage drug for people with PI resistance. But as you know along the way in drug development obstacles can emerge. So enthusiasm should be reserved until safety and antiviral activity can be further tested in humans with PI resistance.