The two-month clinical hold that threw a wrench into Cytrx Corp.'s aldoxorubicin (aldox) phase III trial for soft-tissue sarcoma (STS) last July dimmed hopes, but by late November the company had dug deep enough into the results that it could plan for an NDA filing late this year. As the Los Angeles-based firm gears up for a type B pre-NDA meeting in March with U.S. regulators, investors are weighing odds. "I think our chances are reasonably good," the firm's chief medical officer, Daniel Levitt, told BioWorld Insight. "We're certainly going in optimistic," since data suggest that the compound is as safe and effective "or better than any treatment that was being used at the time we did the study."

Aldox combines doxorubicin with a novel single-molecule linker that binds directly and specifically to circulating albumin. Since tumors concentrate albumin, delivery of the linker molecule with the attached doxorubicin is greater to the cancer site and – in the acidic environment of the tumor, but not the neutral environment of healthy tissues – doxorubicin is released at much greater possible doses with less toxicity.

Cytrx also has the candidate "at a slightly lower dose" in a phase IIb study against small-cell lung cancer (SCLC), Levitt said. "We should have data in the first half of this year. It wasn't a large study, roughly 130 patients, but that should be enough to tell us whether we're looking better than topotecan."

Last summer, the experiment comparing aldox to investigator's-choice therapy in patients with relapsed/refractory STS showed a miss on the primary endpoint of progression-free survival (PFS). The clinical hold had meant a small change to the protocol but led to a substantial delay in the getting the study reapproved by ethics committees and institutional review boards, extending the delay past the hold's duration another month to six months. Conducted under a special protocol assessment, the trial's analysis took place after 191 progression events, and the analysis didn't have enough follow-up for almost two-thirds of patients who entered the study after the hold was resolved and enrollment started again. Almost half of all patients were excluded – censored – from the PFS measure. Specifically, the results in hand as of July turned up a median of 4.17 months and 4.04 months, respectively, for the pair (hazard ratio: 0.91). (See BioWorld Today, July 12, 2016.)

tracking overall survival

Brighter aspects of the July findings included objective response rate (ORR) and disease control rate (ORR plus stable disease ≥ 4 months), a near doubling in the aldox arm compared to investigator's choice, including in patients who'd previously been treated with doxorubicin. Disease control rate for aldox was significantly greater than investigator's-choice therapy in the intent-to-treat population (p=0.048) as well as in patients who received prior doxorubicin (p=0.0415). Patients continue to be followed for overall survival, a secondary endpoint of the trial. Treatment-related adverse events for aldox were as expected, and the drug was not associated with clinically significant cardiac, kidney or liver toxicities.

Better news came around Thanksgiving 2016, when analysis found that aldox hit the mark in PFS among the North American segment enrolled: 312 patients out of the total 433 subjects that signed up. Specifically, the phase III effort met its goal in PFS in 246 patients with what Levitt called "the two L sarcomas," leiomyosarcoma and liposarcoma (p=0.007), the most common forms of STS. They accounted for 57 percent of patients in the trial. "These were subsets that were specified as part of our statistical analysis plan," Levitt said, noting that "one of the subsets is exactly the same as the indication for which Yondelis [trabectedin, Pharmamar SA] was approved." The hazard ratio was 0.62 (95 percent CI 0.44-0.88), representing a 38 percent reduction in the risk of tumor progression for patients receiving aldox when compared to the treatment chosen by investigators.

PFS in North American patients was especially satisfying (p=0.028; HR=0.71, 95 percent CI 0.53-0.97), and for the whole study group, aldox performed better than treatments doctors selected, though it only came near and did not reach statistical significance (p=0.12; HR=0.81, 95 percent CI 0.64-1.06). If the company had chosen just one of the comparators, the data suggested, a significant difference might have appeared in the whole population, but the FDA liked the investigator's-choice route, which was a factor in getting the SPA, "not a requirement but strongly advised," Levitt said. As it was, in the entire study group, aldox achieved a statistically significant improvement with regard to disease control rate (DCR): 29.4 percent vs. 20.5 percent for the patients treated with investigator's choice (p=0.030). In North American patients, those treated with aldox showed a DCR of 32.9 percent, compared to 19.2 percent for patients treated with investigator's choice (p=0.007), an overall improvement of 71 percent. ORR in North American patients also favored aldox, 8.7 percent vs. 3.3 percent (p=0.058). No objective responses surfaced in patients treated with Votrient (pazopanib, Novartis AG), approved by the FDA for STS in 2012. Patients continue to be followed for overall survival.

And what of SCLC? The indication gets little attention, at least as compared with non-small-cell lung cancer, the more prevalent type. Last April, SCLC made headlines last April when North Chicago-based Abbvie Inc. took over Stemcentrx Inc. in a deal that could be worth almost $10 billion, bringing aboard lead late-stage asset rovalpituzumab tesirine (Rova-T), an antibody-drug conjugate (ADC). Abbvie agreed to pay Stemcentrx, of South San Francisco, $2 billion in cash, with the remainder in stock. Shareholders in Stemcentrx could collect another $4 billion if milestones are met. At the American Society of Clinical Oncology (ASCO) meeting in June, early findings from a first-in-human trial showed that the ADC brought promising efficacy against recurrent SCLC, halting tumor growth in 89 percent of patients with high levels of DLL3 in the tumor and shrank tumors in 39 percent. Jefferies analyst Jeffrey Holford said in a research report at the time that the ASCO data "further validate" Rova-T. "The Stemcentrx management team also gave a detailed overview of its stem cell-targeting technology platform [at the meeting], which has already delivered five assets into clinical trials across multiple tumor types," he wrote in a research report. "We see great potential for a significant number of novel anticancer agents to be delivered that could radically improve treatment for a wide range of cancers that could be further potentiated through combination with immuno-oncology agents." (See BioWorld Today, April 29, 2016.)

The SCLC landscape grew more complicated in January, when a preclinical study was published in Cancer Cell suggesting that the disease comes in two types, each of which may need to be treated differently. Research came from labs at the Huntsman Cancer Institute at the University of Utah.

Analyst sees 'optionality'

Although NSCLC gets more attention, SCLC is "more deadly," Levitt said. "The survival is substantially shorter, and even though patients respond very well to first-line treatment, the responses are relatively short in duration."

The only approved second-line therapy is topotecan. "It's pretty toxic, the responses are quite low and they're very short," he said, acknowledging that SCLC "can be a graveyard for a drug" as well as patients. "I don't see in SCLC anything I've read that looks very promising," despite early stage outcomes that seem to presage success. "When you expand these cohorts and look at a broader patient population, [survival benefits] tend to disappear. Really, the major progress you can make in SCLC is to get people to stop smoking." Most SCLC patients are smokers, unlike with NSCLC, in which a "very significant percentage have never smoked," he said. In any case, "it made sense with some of the good responses we've seen in a small number of patients in our early phase trials [with aldox] to pursue SCLC" along with STS.

For Cytrx investors, the question is whether deeper-dive discoveries in STS, plus the data to come in SCLC, will be enough to move aldox into approval territory. H.C. Wainwright analyst Andrew Fein said Wall Street should "get clarity on the outcome of the meeting [with the FDA] and future development plans around April/May, after meeting minutes are issued. In our view, the company's best play is to press regulators towards a label in leiomyosarcoma and liposarcoma," since statistical significance was met there, he wrote in a January report. "Overall, we recognize that this will be a challenging case to argue (the company may need to justify to regulators why the other sarcoma types failed, and would not be included in the label)."

But Fein cited several reasons why gatekeepers might be sold. "The PFS analysis of these sarcoma types was pre-specified, and a survival analysis of the study will eventually follow, making an outright regulatory 'no' quite unreasonable," in his view. In 2015, the FDA approved Yondelis for the same subtypes with prior anthracycline treatment, based on response and PFS data, without seeing survival data. Cytrx noted in the November analysis of the aldox phase III results that its drug performed in a manner "comparable" to Yondelis, though the results could not be compared head-to-head because the latter wasn't yet approved.

Levitt said Cytrx may end up doing a confirmatory study in the two L sarcomas after the drug is cleared for marketing. "That would be something we'll discuss with the FDA as maybe a post-approval commitment; it's what other companies do," he said, pointing to Yondelis and to Indianapolis-based Eli Lilly and Co. with Lartruvo (olaratumab), given the STS nod last October for use in combination with doxorubicin.

If all else fails, aldox might go the route of tamibarotene, especially should SCLC data turn out positive. A retinoid alpha receptor agonist (RARα), tamibarotene failed at the phase IIb stage in NSCLC and ended up with Japan's TMRC Co. Ltd., which then licensed it to Cambridge, Mass.-based Syros Pharmaceuticals Inc. Using its gene control discovery and development platform, Syros pinpointed a cancer dependency to RARα and a biomarker to identify patients who may respond. Dubbed SY-1425, the drug is approved in Japan as Amnolake to treat relapsed or refractory acute promyelocytic leukemia, a form of acute myeloid leukemia (AML) that is driven by a fusion of the RARα gene with other genes. In Syros' hands, tamibarotene has reached midstage clinical development in relapsed/refractory AML; newly diagnosed, older-unfit AML; relapsed/refractory myelodysplastic syndromes (MDS); and lower-risk, transfusion-dependent MDS, with preclinical work underway in breast cancer.

Wainwright's Fein wrote that, "while we continue to believe that tamibarotene is a mediocre drug with biological and intellectual property issues, in our view this history lesson from within the Cytrx pipeline goes to show that the same drug in different hands" could do better, which implies "some business development optionality left in aldox." Levitt said, "I don't know, if the FDA is not positive coming out of the meeting, what the plan would be." He stayed upbeat about the possibilities with regulators, given aldox's safety and efficacy. "We've not seen any clinically significant cardiac problems with this drug," he said, though researchers have been able to dose amounts that "far exceed any amount of drug you could administer with doxorubicin. I think it has a lot of good features. It's certainly easier to administer than Yondelis," which requires pre-dexamethasone therapy followed by a 24-hour infusion. Aldox calls for no premedication with aldox and is given in 30 minutes. Yondelis and aldox are given every three weeks.