Before Ziopharm bulls assume success, there are some risks to consider:
Promising phase II data are rarely repeated to the same extent in confirmatory phase III studies. Reproducing the positive outcomes of smaller phase II studies when a drug moves to larger phase III studies is especially difficult in cancer indications.
I can barely come up with a single example where positive phase II data were replicated equally in larger, phase III studies. Doxorubicin and ifosfamide have been the standard of care in STS for the past 20 years for this very reason.
Why would the phase II results of palifosfamide observed in the phase II PICASSO study not replicate in the phase 3 PICASSO-3 study? The answer lies in the different patient populations of the two studies.
PICASSO enrolled a mixture of STS patients. One-third of patients had more advanced STS. These patients were previously treated with chemotherapy and had cancer that was progressive or metastatic. The other two-thirds of patients in the study had less advanced STS and had not been treated with chemotherapy previously. Patients with progressive or metastatic disease following prior treatment generally have a worse prognosis that patients who are newly diagnosed and have not received prior chemotherapy.
PICASSO-3 enrolled only STS patients with metastatic disease who had not received prior chemotherapy. Generally speaking, these patients have less advanced disease, so I'd expect PFS to be greater across both arms of PICASSO-3 compared to what was seen in PICASSO.
Greater PFS across both arms of the PICASSO-3 trial won't necessarily be a concern as long as the difference between the two arms -- the PFS benefit favoring palifosfamide -- remains the same or is larger. But what happens if the control arm in PICASSO-3 performs better than expected?
Looking back at the PICASSO data, the response rate for patients treated with doxorubicin alone was 9%. That's low. According to the medical literature, doxorubicin response rates in STS are usually in the 20-30% range. The response rate for patients treated with palifosfamide-doxorubicin in PICASSO was 23%, consistent with most single-agent doxorubicin studies. So, was the benefit observed in PICASSO due only to the under-performance of patients treated with doxorubicin alone? If that's true, PICASSO-3 might be in jeopardy if patients treated with doxorubicin alone more closely mimic historically higher response rates. [Note: I'm assuming higher tumor response rates correlate favorably with greater progression-free survival rate -- the primary endpoint of PICASSO-3.]