To put it mildly, optimizing the dose has never been a major focus of drug development plans in oncology.

In the context of significant unmet medical need and acute disease burdens, conventional chemotherapy drugs have been highly toxic – and the typical approach has been to push the dose as high as a patient can tolerate. But in an era of targeted therapies and therapies that act indirectly by boosting the immune system rather than killing the tumor, those dynamics have changed.

That is why the FDA and the American Association for Cancer Research (AACR) held a two-day workshop in May to discuss the best practices of dose finding and dose selection for small molecule kinase inhibitors in oncology.

The stated goal of the workshop was to promote “a movement away from conventional dose escalation trial design” and move toward innovative designs. “Ideally, this workshop will propel a movement toward integrating dose finding into the entire life cycle of product development as opposed to confining it to the Phase I, first-in-human trial based on short-term safety measures,” FDA said.

The two-day workshop featured broad discussion on statistical and modeling approaches to early-stage dose-finding studies. The participation of Division of Oncology Products I Acting Deputy Director Geoffrey Kim and medical team leader Amy McKee indicates the issue has received high-level attention within the Office of Oncology & Hematology Products.

Like so much else in oncology development, the workshop reflected the impact of the Breakthrough Therapy Designation program. In the context of dose-optimization, the impact appears to be double-edged.

On the one hand, participants agreed, the excitement about advances in targeted cancer therapy in the “Breakthrough” era has made enrollment much easier for many trials – thereby making it easier to explore different dosing regimens early on. On the other hand, the goal of Breakthrough (and of commercial organizations generally) is to shorten the time to development, which often makes finding the optimal dose a secondary consideration.

Breakthrough: “Now In Confirmatory Trials”...

“One of the issues that oncology development has had in the past with these early phase trials [is that] recruitment can be slow,” one industry stakeholder noted. “That can be a limitation when you’re exploring dose response.” However, “if the drug looks very good hopefully you can explore different doses and the recruitment is good.” One thing is certain, he said: if the drug gets a Breakthrough designation, “recruitment won’t be a problem.”

The benefit of high interest in early trials, however, then introduces the commercial challenge of getting a cutting edge therapy approved by regulators as quickly as possible.

“One of the big challenges with Breakthrough Designation from the company’s perspective is that you are now in confirmatory studies,” said Stuart Bailey of Novartis AG. “You are now at a point where you are now full speed trying to get this drug into the market.” Another industry stakeholder added that “as soon as we hear that word, we’re told to go fast.”

Using a larger dose-finding study as one of the two pivotal trials to support approval was one of the ideas put forward by stakeholders at the meeting. However, FDA officials turned the focus back to the benefits of engagement with the agency as key to appropriate dose-finding designs.

…But Designation Does Not Mean Dose Is Right Or Approval Is Certain

When a breakthrough designation is awarded, FDA’s Rajeshwari Sridhara said, “We don’t say this is the exact dose that should be given. What we are saying is that the activity of the drug is worth calling it a breakthrough therapy for now. But then you have to come up with more data. Certainly it does bring up issues of how you are going to conduct future studies. The designation allows you to come and talk to us in advance and multiple times in designing your trials. I think we should look at all the data that is available to see how we can make the design better.”

“Breakthrough means that you have a clinical signal that you may have a very good product. It doesn’t guarantee approval,” FDA’s McKee said. “The rush to get your product on the market may not have always led to the best development program.”

McKee followed up with the idea that FDA is open to approving multiple doses at once assuming there is solid reasoning behind it. “We’re not opposed to multiple doses in the label. We’re open to that if you support it.”

University of Chicago’s Mark Ratain pointed out a different sort of commercial tension, focused not on time-to-market but on “price-per-milligram.” The “whole conundrum around pricing” is “another piece of this. In our current pricing models, prices are set per mg, or per tablet or per capsule, as opposed to per life years saved or whatever.”

FDA’s Kim, however, suggested that argument also cuts both ways. “On the R&D side, you might be more concerned about studying lower doses early so that you’re not finding out that a lower dose is just as good as a high dose” after the drug is approved—and priced. “That’s right,” Ratain responded. If “you raise the dose after approval, that doesn’t have any commercial harm. I think we have to keep all of this in mind,” since “post-market commitments for drugs to study lower doses” are very challenging for sponsors.

The workshop was held in the context of FDA’s stated desire to stop using historical dose escalation methods for new small molecule oncologics. Over the two days, dose-finding cases discussed either through formal presentations or panel Q&A included: crizotinib (Pfizer Inc.’s Xalkori), dasatinib (Bristol-Myers Squibb Co.’s Sprycel), ceritinib (Novartis’ Zykadia), Roche’s alectinib (breakthrough designation), Ariad Pharmaceuticals Inc.’s brigatinib (breakthrough designation), axitinib (Pfizer’s Inlyta), and vandetanib (AstraZeneca PLC’s Caprelsa).

FDA’s Kim: “We Don’t Like The 3+3”

There are two main categories of dose-finding studies: rule-based designs and model-based designs. “The main characteristic of rule-based designs is that they do not stipulate any prior assumption of the dose–toxicity curve. These designs comprise the so-called “up-and-down” designs because they allow dose escalation and de-escalation,” according to the Journal of the National Cancer Institute. The model-based approach uses statistical models that actively seek a dose level that produces a pre-specified probability of dose-limiting toxicity by using toxicity data from all enrolled patients to compute a more precise dose–toxicity curve, the JNCI explains.

Clearly an important reason for the two-day workshop was for FDA to communicate a move away from the 3+3 design—a rule-based approach—in Phase I oncology trials.

The 3+3 design uses cohorts of three patients; the first cohort is treated at a starting dose that is considered to be safe based on extrapolation from animal toxicology data, and the subsequent cohorts are treated at increasing dose levels that have been fixed in advance, JNCI describes. It requires no modeling of the dose–toxicity curve beyond the classical assumption for cytotoxic drugs that toxicity increases with dose.

“The dose increments become smaller as the dose increases. If none of the three patients in a cohort experiences a dose-limiting toxicity, another three patients will be treated at the next higher dose level. However, if one of the first three patients experiences a dose-limiting toxicity, three more patients will be treated at the same dose level. The dose escalation continues until at least two patients among a cohort of three to six patients experience dose-limiting toxicities. The recommended dose for phase II trials is conventionally defined as the dose level just below this toxic dose level.”

FDA’s Kim laid out several goals moving forward: move away from 3+3, dose limiting toxicity definitions beyond the first 28 days of dosing, more doses being studied, evaluating novel approaches to data integration that include modeling and simulation, and improved lifecycle integration of data among others.

In a presentation on innovation in dose-finding designs, FDA’s Laura Fernandes provided a “wish-list for the future” that asked drug sponsors to consider a number of questions in future dose-finding designs: incorporation of data from multiple cycles from the same patient, long-term safety and late onset or delayed toxicities, how to study dose combinations, and how to think beyond 28/21 day cycles—noting that small molecules are dosed continuously “unlike chemotherapy.”

Johnson & Johnson statistician Jose Pinheiro added that an improper understanding of dose response has resulted in “inadequate dose selection for confirmatory programs remains a critical problem in drug development.” He cited three negative outcomes: the need to repeat Phase II studies, delays or outright rejections of approval, and post-approval changes in dose—typically reductions—that result in loss of revenue. In his concluding remarks, he advocated for model-based approaches with wider dose/exposure ranges (>10 fold) and larger number of doses/regimens (>3) in Phase II.

Genentech Inc. statistician Jin Jin gave a presentation on the complexities that arise when you have combination therapies in development. For a new molecular entity being added to standard of care (SOC), the dose and regimen for the SOC is generally known but the dose and regimen for the NME piece still must be optimized. For NME + NME combos, she called it a “new area with accumulating knowledge” where the optimal dose and regimen for both agents are unknown. Additional complexities include poly-combinations (triplet and quad), combinations of multiple molecule types and multiple novel mechanism types.

“Breakthrough” Dose-Finding In Oncology

More from Archive

More from Pink Sheet