A 19% overall response rate for Celldex Therapeutics' (Nasdaq: CLDX) CDX-011 versus a 14% for chemotherapy in breast cancer patients that have failed other therapies is nothing to get too excited about. But that's for all patients enrolled in the trial and the biotech was clearly trying to use its phase 2b trial to test how far it could push the drug.

The biotech figured it out.

CLDX-011 is a targeted therapy that uses an antibody that attacks glycoprotein NMB, or GPNMB, attached to a cell killing chemotherapy. Celldex licensed the antibody drug conjugate technology from Seattle Genetics (Nasdaq: SGEN). GPNMB promotes the movement and invasion of tumor cells. The protein is expressed in breast cancer and other tumor types, but to varying degrees.

In tumors where more than 25% of tumor cells expressed GPNMB -- as opposed to greater than 5% in the full data set -- CDX-011 produced an overall response rate of 32% versus 13% for patients who got chemotherapy. That's exactly what you'd expect to see, targeted therapies like Seattle Genetics' Adcetris and even antibody drugs without toxic payloads like Bristol-Myers Squibb (NYSE: BMY) and Eli Lilly's (NYSE: LLY) Erbitux or GlaxoSmithKline's (NYSE: GSK) Tykerb should only work if the protein they target is there and functioning.

In patients who have very aggressive tumors, called triple-negative because they don't require estrogen, progesterone, or a protein called Her2 to grow, the data looked even better. These patients don't typically respond to later rounds of chemotherapy, as evidenced by the 0% overall response rate in the chemotherapy arm. Triple-negative patients taking CDX-011, on the other hand, produced a 21% overall response rate, which jumped to 36% for triple-negative patients who also expressed a high level of GPNMB.

The only downside to this data is that when you pull out patients, you get fairly small numbers. Celldex enrolled 122 patients and 81 got CDX-011. Of those, only 25 were high GPNMB expressers. There were 24 triple-negative patients who got the drug. And only 11 patients who got CDX-011 were both high expressers and triple-negative.

Obviously that's not enough to get the drug approved -- that wasn't the point of the trial -- but the data give Celldex a testable hypothesis and should give investors a little more confidence that the next trial will be positive.

While you're waiting for more data on CDX-011, take a look at what Fool analysts believe is the next rule-breaking multibagger. Get the free report by clicking here.