CHICAGO ( TheStreet) -- The efficacy of Synta Pharmaceuticals' ( SNTA) lung cancer drug ganetespib is weakening over time, which should raise even more concerns about the ongoing phase III clinical trial. Snyta shares are down 17% to $6.15 in early Monday trading. The latest update from the phase II "Galaxy-1" study, which compares ganetespib plus docetaxel against docetaxel alone in second-line non-small cell lung cancer, are being presented Monday during the American Society of Clinical Oncology (ASCO) annual meeting: The overall survival hazard ratio for the 252 adenocarcinoma patients is 0.82, which translates into an 18 percent reduction in the risk of death. Median overall survival for ganetespib/docetaxel was 9.8 months compared to 7.4 months for docetaxel. The result was not statistically significant and worse than the previous update of the Galaxy-1 study presented last September when the hazard ratio was 0.69, or a 31 percent reduction in the risk of death. Synta didn't wait for final data from Galaxy-1 before starting to enroll patients in a phase III study of ganetespib in second-line lung cancer. This larger, Galaxy-2 study excludes patients who are deemed to be "rapid progressors" -- defined as patients who responded poorly to first-line therapy. In order to qualify to entry into Galaxy-2, patients have to be stable (no tumor growth) following first-line therapy for at least six months. Call these patients "normal progressors" for lack of a better term. On Monday, Synta released an update on the subset of 176 "normal progressors" from the Galaxy-1 study. Again, the survival trend narrowed. The latest overall survival hazard ratio is 0.61, or a 39 percent reduction in the risk of death. At the median, ganetespib/docetaxale increased overall survival by 4.3 months over docetaxel alone (10.7 months versus 6.4 months.) Last September, the hazard ratio was 0.37, or a 63 percent reduction in the risk of death. Investors have tended to view ganetespib's chances of success with a lot of skepticism because claims the drug would work better in certain targeted subsets of lung cancer patients (elevated LDH or KRAS mutants) have not panned out. The narrowing survival trend from the Galaxy-1 study in data-mined subsets of patients isn't likely to bolster anyone's confidence in Synta's ability to deliver a positive outcome from the pivotal Galaxy-2 study.