It is perhaps a sign of the unmet need in epidermolysis bullosa that findings in just three patients sent Krystal Biotech’s share price up 41% yesterday. But the small number of recruits was not the only reason to be cautious about the data from the Gem-2 trial of Krystal’s topical gene therapy KB103.
For one, the study had enrolled four subjects, but one dropped out. Krystal did not include this patient in its headline findings; on an intent-to-treat basis the results would have been much less impressive. In addition, KB103 did much better in recurring than in chronic wounds, and the company is still trying to find the best dose for the latter, despite plans to start phase III development by the end of the year.
Krystal’s chief executive, Krish Krishnan, admitted that the company had yet to meet regulators to discuss the pivotal trial design, telling Vantage this was something the group hoped to do in the next couple of months.
Despite this, he was adamant that there would not be a delay to the start of the trial, which he said should include around 15 patients, and would evaluate a mixture of recurring and chronic wounds. Chronic wounds are defined as those remaining open for 12 weeks or more, while recurring wounds can open and close spontaneously.
He also stressed the dropout was due to difficulties that the patient had had travelling to the clinical trial site for follow-up, rather than any safety or efficacy issues with KB103.
Long-distance travel can be hard for epidermolysis bullosa patients, who have extremely fragile skin that is prone to extensive blistering. Mr Krishnan contended that this would not be a problem for KB103 if it reached the market; the gel is designed to be applied by a dermatologist or primary care physician, so the patient could receive treatment close to home.
This is in contrast to autologous gene therapies in development by Abeona and Fibrocell, which involve more complicated procedures that need to be carried out in a hospital (Gene therapies go skin deep to tackle epidermolysis bullosa, March 11, 2019).
Still, even if the dropout was due to logistics, there is no denying that excluding this patient flattered the final analysis.
60% of the time it works every time
In the phase Gem-2 study, patients had three wounds selected, two of which were treated with KB103, and one treated with placebo, so patients effectively acted as their own control. Patients were assessed over three months of on-site visits, followed by a three-month at-home imaging period.
It is the first three months that Krystal reported data on, with four secondary endpoints assessed at this time. The primary endpoint of the trial is wound closure at 24 weeks.
In its press release yesterday Krystal highlighted data at 60 days, when five out of six wounds treated with KB103 showed 100% closure. However, this figure only included the three patients who ultimately made it to the 90-day analysis.
And, according to Krystal's presentation, one chronic wound that was 100% closed at 60 days had changed to 96% closed at day 90.
Shorter-term data were available for the patient who dropped out: adding these, only five of eight KB103-treated wounds, around 63%, had completely closed at 30 days.
|Gem-2 results with Krystal's KB103|
|Number of wounds completely closed|
|Wound type||Day 30||Day 60||Day 90|
|KB103||Recurring||5 out of 5||4 out of 4||4 out of 4|
|Chronic||0 out of 3||1 out of 2||0 out of 2|
|Total||5 out of 8||5 out of 6||4 out of 6|
|Placebo||Recurring||0 out of 4||0 out of 3||0 out of 3|
|Source: Company presentation.|
As recurring wounds are known to heal on their own the real test of KB103’s efficacy in phase III will be its performance in chronic wounds, where, as the table above shows, KB103’s benefit looks questionable.
Perhaps in an acknowledgement of this, Krystal has enrolled two more patients in its phase II trial to test more frequent dosing for chronic wounds.
Mr Krishnan said the group had not decided yet on the dose of KB103 that would be used in phase III, but told Vantage that this could include different dosing schedules for recurring and chronic wounds. He added that around 90% of EB wounds are of the recurring type.
Finding the cash for phase III should not be a problem – Krystal already took advantage of its stock gains to carry out a $100m public offering yesterday. With its finances sorted the next big test for Krystal will be reproducing these results in a larger and longer trial.
This story has been updated with more details on the timepoints at which patients are being assessed.