The double phase III miss for hypoxia-activated prodrug evofosfamide likely will mean the end of Threshold Pharmaceuticals Inc.'s nearly four-year partnership with Merck KGaA, but the South San Francisco-based biotech isn't backing down from the premise that targeting hypoxia in the tumor microenvironment could tackle tumor progression, metastases and cancer treatment resistance.

"We've been working in this space for 10 years – we've been focused in the tumor microenvironment even longer than that," Threshold CEO Barry Selick told investors on an early Monday conference call. In recent years, in particular, "there's been a resounding swell of interest, as measured just by publications and presentations – in the field of tumor hypoxia.

"So it's not just a Threshold hypothesis that hypoxia is real and that it plays a significant role in treatment failure," he added. "I think it is the overwhelming consensus within the medical community that that is the case."

The overwhelming view of Wall Street, however, seemed more skeptical, with shares of Threshold (NASDAQ:THLD) dropping as low as 51 cents Monday morning, the lowest the shares have traded since late 2008. The stock closed at 60 cents, down $2.74, or 82 percent.

Explaining the failure of evofosfamide to statistically significantly improve overall survival in both studies – the MAESTRO study in pancreatic cancer patients or the TH-CR-406/SARC021 study in soft tissue sarcoma patients – likely will require a deeper dive into the data than Threshold execs could provide Monday, though Selick offered a few thoughts.

"Merck got very close" on the MAESTRO study, he noted. The trial, conducted by Merck under an FDA special protocol assessment (SPA), randomized 693 pancreatic cancer patients to receive evofosfamide plus gemcitabine or gemcitabine alone and resulted in a hazard ratio (HR) of 0.84, at a 95 percent confidence interval (CI), with a "p" value of 0.0589.

"All the endpoints trended in the right direction," Selick said, who called the overall results "encouraging," despite the statistical miss.

The soft tissue sarcoma study was a different story, falling well short of expectations. In that case, Selick said, "we were ambushed by an extraordinarily long median overall survival in the control arm, which at 19 months is "unprecedented [and] speaks to the quality and the evolution of care" sarcoma patients currently are receiving.

Also conducted under an SPA, the TH-CR-406/SARC021 study recruited 640 patients with locally advanced unresectable or metastatic soft tissue sarcoma to receive either evofosfamide in combination with doxorubicin or doxorubicin alone. Results failed to show an overall survival improvement (HR: 1.06; 95 percent CI), with the median overall survival in the doxorubicin-only arm actually besting vs. the combo arm, a finding that even surprised analysts.

Piper Jaffray's Charles Duncan said he was "astonished" by the efficacy shown in the control arm. He also noted in a research reported his surprise "by the lack of a trend here, as nitrogen mustards (a similar mechanism to evofosfamide) have shown efficacy in other studies" of soft tissue sarcoma.

Evofosfamide (previously designated TH-302) is designed to work by activating only under hypoxic conditions, a characteristic of the tumor microenvironment. Once activated, it releases bromo isophosphoramide mustard, a DNA alkylating agent aimed at killing tumor cells. The theory has been that, upon activation, the cell-killing agent could spread to surrounding areas of the tumor, but, to date, evidence of that "bystander effect" has been difficult to prove.

In retrospect, Threshold's chief medical officer, Tillman Pearce, attributed that in part to evofosfamide's lack of a biomarker. The drug is preferentially active in hypoxic regions, but unlike mutant EGFR, for example, there's not a clear driver mutation that can be detected in every tumor cell.

"We're now deploying HX4, which is a hypoxia-imaging tracer," Pearce said. Threshold gained rights to [18F]-HX4 [flortanidazole (18F)], an investigational positron emission tomography imaging agent from Siemens Healthcare in 2013. "And we're hoping that might allow us to better select patients."

The test for Threshold will be resolving what appears to Piper Jaffray's Duncan as a "disconnect between mechanism and ability to leverage into therapeutic benefit." With evofosfamide likely out of the picture – a phase II study in non-small-cell lung cancer (NSCLC) is ongoing, though expectations for that study are virtually nil – the company's focus turns to earlier-stage hypoxia-activated prodrug tarloxotinib.

CEO: INTERESTING DATA BY ASCO

Formerly known as TH-4000, tarloxotinib is designed to release an irreversible EGFR tyrosine kinase inhibitor. While several EGFR-targeted therapies are on the market – Tarceva (erlotinib, Roche AG and Astellas Pharma Inc.), Iressa (gefitinib, Astrazeneca plc) and Gilotrif (afatinib, Boehringer Ingelheim GmbH) – many patients develop treatment resistance due to different mechanisms, one of which may be regions of hypoxia in the tumor microenvironment.

Charles Hart, the firm's senior vice president of biology, said "unique pharmacokinetic and pharmacodynamic" data for tarloxotinib already are emerging, including a "dramatic bystander effect," indicating that it can "drug the rest of the tumor as well as the hypoxic compartment." It also appears to be able to hit the entire EGFR subfamily, "which we think is a key attribute."

Tarloxotinib is in phase II testing in NSCLC and squamous cell head and neck cancer. "We hope to have an interesting amount of data to present by ASCO," said CEO Selick, which he called a "conservative-slash-realistic" timeline. Both trials are evaluating response rate, measured by RECIST, as the primary endpoints.

In the meantime, Threshold has some decisions to make, especially if Darmstadt, Germany-based Merck opts to pull out of the collaboration inked in early 2012. "We believe this is likely and expect to hear something definitive in the coming days," Selick said. (See BioWorld Today, Feb. 6, 2012.)

While there may be some other ways forward for evofosfamide, based on investigator-initiated studies testing it in combination with checkpoint inhibitors or with radiation therapy, "I think we have to step back and look at all the indications for which we've generated additional data," he added. "Then we'll make a data-driven decision whether one or more of those are worth continued pursuit."

Threshold expects to end the year with $40 million to $45 million on its balance sheet – taking into account any financial obligations remaining in the Merck deal – and that cash is expected to carry the company through the end of 2016.