Now you see it, now you don’t: Peregrine’s bavi might work after all

Give Peregrine Pharmaceuticals full marks for effort. In true biotech fashion the company has refused to throw in the towel, and yesterday said its review of an earlier review of a shambolic phase II study of bavituximab suggested that the molecule might be working after all.

Investors certainly bought into the hype, sending the company’s stock up 80%, but then again the shares are still well short of their high four months ago. What Peregrine really needs, if it wants to give credence to the view that bavi could be another Avastin in the making, is a lucrative big pharma partnering deal. This currently looks as improbable as it did in September.

The likeliest reason is that no one still knows with any degree of certainty whether what Peregrine claims to have shown is real (Peregrine’s bavi data: too good to be true?, September 10, 2012). Considering the class action lawsuits accusing Peregrine of making false statements, and the multiple rounds of data dredging that have yielded the latest result, the bavi asset is probably still too toxic to handle.

3mg good?

The farrago concerns a 121-patient phase II study in second-line treatment of non-small-cell lung cancer, in which two bavi doses – 1mg/kg and 3mg/kg – were tested against placebo.

While the primary endpoint, objective response rate, was missed, what had caused initial excitement was the revelation in September that the important secondary endpoint of median overall survival strongly favoured bavi 1mg/kg (5.5-month increase, p=0.0286), as well as a post hoc pooling of the two doses (6.5 months, p=0.0154), versus placebo.

It was especially remarkable that, in a relatively small study, bavi seemed to blow away a drug as huge as Roche’s blockbuster Avastin, which in first-line use improves median OS by just two months. But the fairy tale unravelled when a review uncovered “major discrepancies”, leading Peregrine to advise that the data could no longer be relied on.

In the latest twist, a review of those discrepancies has suggested that only the 1mg/kg dose was compromised, in that some vials relating to patients in the 1mg/kg arm had been mixed up. There is no evidence of vial coding irregularities with 3mg/kg, Peregrine says.

In September the company said 3mg/kg had shown a strong numerical but statistically non-significant trend towards an OS benefit of 7.5 months over placebo (p=0.0714). It now says it is taking the unusual step of combining the 1mg/kg and placebo groups into a single control arm against which the 3mg/kg result can be compared.

And although logically this new control group must yield a median OS much higher than the original placebo group’s 5.6 months, the 3mg/kg dose still shows “favourable tumour response rates, progression-free survival and OS” when compared against it, Peregrine insists. The company promises further analyses, but says the new findings do support moving bavi into phase III.


If this story were not complicated enough already, investors might be irked by the recent disclosure that senior management awarded themselves share options while the stock price was depressed at the end of last month. Just over a week later the options, which can be exercised from March 27, are well in the money.

Peregrine ended October 2012 with $24.4m in cash, having raised $16.1m by selling equity under an at-market issuance agreement with McNicoll, Lewis & Vlak; this replaced a $15m loan with Silicon Valley Bank that had been called in for immediate repayment after the bavi phase II snafu (Peregrine crashes to earth after loan default, September 27, 2012).

If – and it is a big if – the phase II study can somehow now be salvaged, a robust phase III programme has yet to be designed and carried out, and this can surely only be funded by a licensing partner. And the lack of any visibility on the partnering front speaks louder than any number of apparently positive post hoc analyses.

Expect Peregrine to tap investors for more cash very soon.

Study Trial ID
121 second-line NSCLC patients NCT01138163

To contact the writer of this story email Jacob Plieth in London at [email protected] or follow @JacobEPVantage on Twitter

Share This Article