The phase III win in March testing oral CFTR corrector tezacaftor and approved CFTR potentiator Kalydeco (ivacaftor) in cystic fibrosis (CF) patients left investors and analysts predicting success in early stage testing for triple-combo regimens featuring ivacaftor, tezacaftor and one of four possible next-generation CF correctors. But the phase I and phase II data unveiled late Tuesday by Vertex Pharmaceuticals Inc. simply blew away expectations. (See BioWorld Today, March 30, 2017.)
One aim of the triple-combo approach is to expand treatment to a wider swath of CF patients, particularly those with one F508del mutation and one minimal function mutation (F508del/Min), who have difficult-to-treat CF and currently no approved therapies that tackle the underlying cause of the disease. Judging by the phase II data, Vertex has a good chance of changing that. The F508del/Min patients in Vertex's phase II studies combining one of two next-generation correctors – VX-152 or VX-440 – with tezacaftor/ivacaftor (tez/iva) showed mean absolute improvements in percent predicted forced expiratory volume in one second (ppFEV1) of 9.7 and 12 percentage points from baseline for the triple-combination regimens with VX-152 (200 mg q12h) or VX-440 (600 mg q12h), respectively. Initial data from a phase I study showed a mean absolute improvement in ppFEV1 of 9.6 percentage points from baseline for the regimen of another next-gen corrector, VX-659, in combination with tez/iva in F508del/Min CF patients.
But that wasn't all. Early data also showed improvements in mean absolute ppFEV1 of 7.3 and 9.5 percentage points, respectively, when VX-152 or VX-440 was added in people with two copies of the F508del mutation, who were already receiving tezacaftor and ivacaftor.
"Wow. Just wow," wrote J.P. Morgan analyst Cory Kasimov in a research note, adding that the "mean absolute improvements in ppFEV1 were Kalydeco-like or better, which represents the best-case scenario."
Shares of Vertex (NASDAQ:VRTX), which traded at nearly 10 times the normal volume Wednesday, ended the day at $159.69, up $27.53, or 20.8 percent.
"We now assume a 100 percent probability of success for the triple combo in both populations [up from 80 percent previously]," wrote Leerink analyst Geoffrey Porges, "which we believe is supported by the strong results, emerging safety and tolerability, and the previously documented robust correlation of phase II to phase III data for Kalydeco, Orkambi [ivacaftor/lumacaftor] and tezacaftor/ivacaftor."
Early but promising
The 47-patient, randomized, double-blind phase II study testing VX-440 at 200 mg and 600 mg every 12 hours in combination with tez/iva in CF patients, ages 18 and older, showed mean absolute within-group change in ppFEV1 at day 29 in the F508del/Min patients of 10 and 12 percentage points (200 mg and 600 mg, respectively) vs. 1.4 for the tez/iva plus placebo group (p < 0.0001 for both.) Mean absolute within-group change in sweat chloride was -27 mmol/L and -33.1 mmol/L (for 200 mg and 600 mg, respectively) vs. +1.6 mmol/L for the placebo group (p < 0.0001 for both.) Sweat chloride is a biomarker of CFTR activity.
A secondary endpoint measuring the mean absolute change in the respiratory domain of CFQ-R, a patient-reported outcome tool, at day 29 showed mean absolute improvements of 18.3 points for the 200-mg group, 20.7 points for the 600-mg group and 2.2 points for the placebo group.
The triple combo was generally well-tolerated in F508del/Min patients. There was one discontinuation in the VX-440 600-mg group due to elevated liver enzymes and one additional patient had elevated liver enzymes seen at the final day of dosing. In both cases, the enzyme levels returned to normal after discontinuation or treatment completion.
Data from the second part of the VX-440 study, which added the next-gen corrector to tez/iva to patients with two copies of the F508del mutation following a four-week tez/iva lead in period, showed mean absolute within-group change in ppFEV through day 29 of 9.5 for the triple-combo (p < 0.0001) vs. -2.5 for the placebo plus tez/iva group (p = 0.2755). Mean absolute within-group change in sweat chloride was -31.1 (p < 0.0001) for the VX-400 regimen vs. +2.1 (p = 0.7385) for the placebo regimen. Safety follow-up in those patients is ongoing.
Initial two-week data from the ongoing phase II study testing VX-152 plus tez/iva in F508del/Min CF patients, 18 and older, showed mean absolute within-group change in ppFEV1 of 5.6 percentage points (p < 0.0135) for the 100-mg VX-152 group and 9.7 percentage points (p < 0.0017) for the 200-mg VX-152 group vs. -0.9 for the placebo group (p = 0.6245). Sweat chloride decreases were -19.6 for the 100-mg group (p = 0.0004) and -14.1 for the 200-mg group (p = 0.0219) vs. +1 for the placebo group (p = 0.5659).
The VX-152 triple regimen was generally well-tolerated, with one discontinuation due to pneumonia. The study is ongoing and will test a higher 300-mg dose of VX-152 in combination with tez/iva.
In patients with two copies of the F508del mutation, the VX-152 200-mg triple-combo regimen showed mean absolute within-group change in ppFEV1 at day 15 of 7.3 percentage points (p = 0.0254) vs. -1.4 percentage points for the placebo group (p = 0.2773). On the sweat chloride endpoint, the VX-152 group showed a change of -20.9 mmol/L (p = 0.0010) vs. +3.4 mmol/L for the placebo group (p = 0.1212).
A phase I study testing next-gen CF corrector VX-659 plus tez/iva for two weeks in F508del/Min CF patients showed a mean absolute improvement in ppFEV1 of 9.6 percentage points from baseline in those receiving the triple combination regimen and a mean decrease in sweat chloride of -41.6 mmol/L. For those receiving placebo, there was a mean absolute decrease in ppFEV1 of -0.4 percentage points and a mean decrease in sweat chloride of -11 mmol/L.
The VX-569 regimen was generally well-tolerated.
"These results are extraordinary," said Jeff Chodakewitz, executive vice president and chief medical officer at Vertex. "We've proven what we hoped and believed would be true," that adding a next-gen corrector to tez/iva would bolster treatment effect and offer a potential therapy to F508del/Min patients.
But the results are early. "We have much more to learn" about the VX-440, VX-152 and VX-659, he said.
There's also a fourth possibility, VX-445, in the mix as well.
'A good problem to have'
Since gaining approval of Kalydeco in 2012 – initially only for patients with the G551D mutation of the CFTR gene, or about 4 percent of the CF patient population – Vertex has worked to expand the patient base. Kalydeco's label now includes a host of CF gene mutations, including G1244E, G1349D, G178R, G551S, S1255P, S549N, S549R and R117H, and is used in patients 2 and older. Then came Orkambi, approved for patients 12 and older with homozygous F508del mutations, the largest CF subset. (See BioWorld Today, Feb. 1, 2012, and July 6, 2015.)
Currently, Vertex estimates that roughly 30,000 patients are eligible for its CF treatments. Future label expansions – Orkambi for patients under 12 and Kalydeco for patients with residual function mutations – could raise the number as high as 44,000. But a triple-combination therapy could push the firm's CF treatments to 90 percent coverage.
With that in mind, the Boston-based biopharma has advanced four next-gen correctors in parallel, with the goal of advancing the most promising candidate – possibly two candidates – into pivotal triple-combo testing in the first half of 2018. The fact that so far all have yielded impressive data means Vertex has more work to do.
"We feel a great sense of urgency to bring a triple-combo regimen" to market, Chodakewitz told investors and analysts on a late Tuesday conference call. But there's also a sense of responsibility to choose the best option. "We want to have as much data as possible."
Vertex is speeding up work, with VX-445 in phase II testing, and VX-659 set to start phase II testing early next month. VX-445 and VX-659 will be evaluated with tez/iva in people with one F508del/Min mutations and in people with two copies of the F508del mutation who are already receiving tezacaftor and ivacaftor. Data from both of those trials are expected in early 2018.
Once further details emerge, Vertex will be looking for the best benefit-risk profile. For now, though, it's too early to tell.
"Based on the number of patients that were involved [in the trials just reported], what's impressive is that they all produced double-digit efficacy and they all produced sweat chloride changes of 30 to 40 millimolars," said CEO Jeff Leiden. "But our ability to distinguish between them at this point is really not there."
While analysts tried to tease out which compounds might be early favorites – some noted preclinical toxicology studies with VX-440 that could affect women of childbearing potential – Leiden remained steadfast. "All four molecules have profiles that we could easily move into phase III," he said, at least according to the data available. "It's a good problem to have," he added.
In the meantime, the firm will work with regulators to discuss trial design, including size and duration. It also has established a steering committee comprising CF experts and clinical trial investigators ahead of the pivotal study program.
"Remember, these are medicines patients are going to take for their entire lives, so we feel a tremendous responsibility not only to move quickly but to make sure we take the best regimens to move quickly on," Leiden said.
A fast and furious route to approval might not be the best option, however, suggested H.C. Wainwright analyst Andrew Fein. He noted in a research report that the consistency of the triple-combo data with Vertex's next-gen correctors suggest that the efficacy benefit is a "class phenomenon, and we should expect similar results from any competitor add-on agent out there."
Potential competitors include Galapagos NV, which is working with Abbvie Inc. in an alliance to develop potentiators and correctors targeting the CF gene, also with the aim of creating a triple-combo therapy. There's also Proteostasis Inc., which is advancing a triple-combination of three CFTR modulators, PTI-428, PTI-801 and PTI-808.
One way Vertex could differentiate its triple-combo program would be by "flexing its muscle and leveraging its economies of scale when it comes to pivotal programs," Fein wrote. Rather than accelerating the pivotal program, which would open "a quick and dirty" approval pathway to all, Vertex could pursue "a more traditional and rigorous phase III program similar to [Orkambi trials] TRAFFIC and TRANSPORT (because it can)," a move that "might just put more distance between its triple and the competition, experimentally and commercially."
Vertex, which expects to release its second-quarter earnings next week, has about $1.4 billion on its balance sheet as of March 31.