Tipranavir falls between two stools in naïve trial

An analysis of a trial comparing the protease inhibitor tipranavir (Aptivus) with lopinavir/ritonavir (Kaletra) as a first-line therapy was stopped after 60 weeks of what was intended to be a three-year trial after one dosage of tipranavir was found to be inferior to lopinavir at reducing HIV viral load while the other dosage was found to be more toxic.

The BI 1182.33 trial was an open-label comparison of lopinavir in its familiar ritonavir-boosted co-formulation with 500mg of tipranavir boosted with two different doses of ritonavir – 100mg and 200mg, or one or two capsules a day.

Tipranavir is licensed to be given with 200mg of ritonavir for treatment-experienced patients but manufacturers Boehringer Ingelheim hoped that boosting with half this dose would prove adequate for wild-type virus and reduce toxicity.

A total of 558 patients joined the trial. Three-quarters of them were male, and they had fairly advanced HIV disease, with a median HIV viral load of just over 100,000 and a median CD4 count of 209. Eleven per cent had a CD4 count under 50.

At 48 weeks, the date which was pre-defined in the protocol when the ‘primary endpoint’ or key virological results were to be determined, the drugs appeared to be performing fairly equally. At this point the proportion of patients with a viral load under 50 on an intent-to-treat analysis (that is, with ‘drop-outs’ indicating failure), was 69.2% for lopinavir, 66.7% for tipranavir/ritonavir 200mg and 65.8% for tipranavir/ritonavir 100mg. These results were statistically equivalent.

However the definition of virological success was that patients had to have two consecutive viral loads under 50 and by week 48 15 patients in all three treatment arms had only had one measurement under 50. So results were re-analysed at 60 weeks.

All three drugs showed a slightly better performance once these ‘late responders’ were included. The proportion with viral loads under 50 was now 72.4% for lopinavir, 69.9% for tipranavir/r200 and 67.9% for tipranavir/r100. However this made the crucial difference: by now the difference in performance between lopinavir and tipranavir/r100 was statistically significant, so this dosage of the drug was now officially less potent or, as presenter David Cooper put it, “the requirement to achieve non-inferiority was not met.”

In terms of its potency the tipranavir/r200 dosage was still statistically equivalent to lopinavir, but here the degree of toxicity was significantly worse.

Discontinuations from the trial ran at 17.4% in the tipranavir/r100 arm, of which 7% were due to adverse events; 14.8% in the tipranavir/r200, of which 10% were due to adverse events, and 11.4% in the lopinavir arm, of which only 2.7% were due to adverse events.

Adverse events defined as ‘serious’ by the investigators were twice as common on tipranavir as they were on lopinavir, at 13%, 12% and 6% for tipranavir/r100, tipranavir/r200 and lopinavir respectively.

However what really proved the death knell for the tipranavir/r200 dose was liver toxicity. By week 48 the proportion of patients with serious elevations in their liver enzymes (grade 3 and 4) was 17.7% in patients taking this dose. This was three times the rate on tipranavir/r100 and over five times the rate on lopinavir, which were 5.9% and 3.4% respectively. This was highly statistically significant (p=0.001).

So tipranavir has fallen between two stools and is now certain not to be licensed as a drug for naïve patients.

“Tipranavir/r100 did not satisfy the non-inferiority requirements, and tipranavir/r200 is not suitable for naïve patients with wild-type virus, due to raised liver enzyme levels,” was how David Cooper put it.