Replies to post #162655 on Biotech Values

[jq1234]

Doesn't LDK's near equal ORR in Criz-resist and naive populations suggest mutations are being missed?

When you evaluate drug candidates, you have to apply same standard to the drug you invested in vs someone else's drug. 113 best ORR in ALK naive was 50% vs 75% in ALK resistant. Does it suggest same thing you just suggested about LDK378?

But of course if AP26113 inhibits most mutants why is ORR at 75% against Chugai's 93%.

First 113 does NOT have 75% ORR, it includes 4 non-confirmed PR, 5 PR waiting to be confirmed out of total 15 PR/CR. You are again comparing not directly comparable numbers. This explains the difference:

Of the 114 patients with ALK-positive NSCLC, 86 (75%) had responses (including unconfirmed responses). Among 79 patients with crizotinib resistance, 62 (78%) responded to LDK378.

When evaluated by RECIST criteria, the confirmed response rate was 58% (66 of 114), 57% in crizotinib-resistant patients (45 of 79), and 60% (21 of 35) of patients with no prior exposure to crizotinib.

http://www.medpagetoday.com/MeetingCoverage/ASCO/39622

Second, Chugai's 802 93% ORR in naive came from Japanese patients, again not directly comparable to 113 or 378 number in ALK naive.

I've heard some predicting '113 ORR will be higher at ESMO.

Not if 113 starts to use confirmed ORR standard. Whoever has that type of expectation will most likely be disappointed. Many people try to make predictions based on data not really comparable, which is why the market doesn't buy that right now. It remains to be seen when data are comparable. Clinical data trump preclinical data, especially in solid tumors.

Regarding Criz I was referring to these studies:
Profile 1005 PH2 (ORR 61%) At least one prior line of chemo, 72% had 2 or more.
Profile 1007 PH3 (ORR 65%) All patients had received one prior chemo.

Just look at the two studies on Xalkori label, vetted by FDA:

http://labeling.pfizer.com/showlabeling.aspx?id=676