A Pandemic Template For Swift Approval? Veklury Stability Questions Deferred For Later Study
Executive Summary
The US FDA shaved months from review of Gilead’s Veklury (remdesivir) for COVID-19 with post-approval commitments for multiple stability studies. The Pink Sheet takes a deep dive look at discussions within the agency and with the sponsor.
The way the US Food and Drug Administration approved Gilead Sciences, Inc.’s Veklury (remdesivir) for COVID-19 before the company could generate key drug substance and drug product batch release and stability data may give industry hope that important treatments for other conditions could win approval before all the data are in, even if their approval is not as urgent.
A key challenge the agency overcame to swiftly approve the nucleotide analog RNA polymerase inhibitor for COVID-19 was the need for months of stability testing of batches from each production line of each manufacturing site, required to establish shelf life. The challenge was exacerbated by the rollout of full-scale manufacturing processes in advance of approval to eight facilities, many of which had multiple production lines.
Such rollouts normally get reviewed as a series of technology transfers that occur months and years after the initial approval, but Gilead moved much more quickly in reaction to the pandemic. The company received a US emergency use authorization for the product in May 2020 and full new drug application approval for a narrower population 22 October. (Also see "Gilead’s Veklury Approval Shows US FDA’s Existing Regulatory Tools Are Up To The COVID-19 Challenge" - Pink Sheet, 22 Oct, 2020.)
The FDA in October issued guidance allowing emergency authorization of COVID-19 vaccines based on limited short-term stability data, but that guidance did not apply to approvals. (Also see "Stability Genius: How CMC Section Of COVID-19 Vaccine Guidance Speeds EUAs Safely" - Pink Sheet, 6 Oct, 2020.)
A Critical Juncture
The issue of missing batch release and stability data came to a head on 1 October, FDA documents on its NDA review show, when the FDA’s quality review team agreed that the agency should delay approval until Gilead provides the data.
But the agency never issued the complete response letter the team in its Office of Pharmaceutical Quality recommended that day. Instead, just three weeks later, a second quality assessment conducted in the same review cycle came to a different conclusion: Veklury could be approved if Gilead commits to providing the data later.
It is possible based on Gilead’s experience with Veklury that sponsors may now feel more emboldened to ask the FDA also to approve unrelated but important treatments such as breakthrough therapies before all the stability data are in.
Whether the agency would use the approach it took with the Veklury review in less pressing circumstances remains to be seen. As the Veklury case shows, it can complicate the review process.
A Tight Timeline
Emergency authorization and NDA review documents show how compressed the timeline was for the FDA’s assessment of Veklury.
Gilead had studied Veklury for hepatitis C and respiratory syncytial virus in 2009 and for Ebola, SARS and MERS viruses in 2014, but had never obtained approval of it for any disease, and had only 5,000 doses on hand when the SARS-CoV-2 virus that causes COVID-19 began spreading around the world in January.
By February, the company had resumed production at its La Verne, CA, plant. (Also see "How Gilead Intends To Deliver On Promise Of Remdesivir For COVID-19" - Pink Sheet, 8 May, 2020.)
Just two months later, Gilead was simultaneously asking the FDA to allow emergency and permanent US market access for treatment of the coronavirus.
Gilead submitted its request for emergency use authorization on 16 April. The FDA granted the authorization on 1 May, with revisions on 28 August, 1 October and 22 October to expand the universe of COVID-19 patients who could be treated. (Also see "CGMP Requirements Apply To Remdesivir In COVID-19 Emergency – But Not To Chloroquine Or HCQ" - Pink Sheet, 7 May, 2020.)
Gilead’s rolling submission of its new drug application began 8 April and concluded 7 August. The FDA’s approval came 22 October, well in advance of the 7 April 2021 user fee goal date, but with a slew of post-approval commitments. Among the commitments were 18 stability studies, the list of which took up four pages of the 17-page approval letter.
Two Reviews In One
The reason the FDA’s pharmaceutical quality office gave in quality review documents for its initial complete response recommendation was that Gilead had not provided enough chemistry, manufacturing and controls information during the rolling review. The company had not provided batch data for one of the drug substance manufacturing sites and two of the drug product manufacturing lines. The agency still was waiting for additional stability and leachables data it had requested. In addition, the FDA and Gilead were still discussing proposed shelf life and post-marketing commitments.
With so many elements of the application still outstanding, the integrated quality assessment team recommended writing an addendum to the CMC review, also described as a second CMC review during the same review cycle, or CMC review #2.
A shelf-life analysis by the CMC statistical team in the Office of Biostatistics, which had been advising OPQ on the matter since 1 June, figured prominently in the second review.
The Shelf Life Consult
The purpose of the statistical consult was to evaluate Gilead’s proposed shelf lives for Veklury injection solution and lyophilized dosage forms in the context of the International Council on Harmonization Q1E stability data guideline.
Gilead had proposed one year for the injection solution, refrigerated at 2-8 degrees C, and a redacted period for the lyophilized powder at less than 30 degrees C.
The 14 October statistical review report summarized findings of the biostatistical office’s shelf life consult for each of the eight manufacturing sites. Gilead’s San Dimas, CA, and La Verne, CA, sites manufacture the 5 mg/ml injection solution dosage form. The other six manufacture the 100 mg lyophilized dosage form.
Injection Solution Stability
Stability studies can be completed more quickly if conducted under conditions such as higher temperatures that accelerate the degradation process. However, the ICH Q1E guideline did not allow this for the injection solution made at San Dimas and La Verne, the biostatistics office determined, because a significant change had been observed in the drug product during the first six months of stability testing.
And then there was another problem. The ICH guideline requires at least three batches with 12 months of stability data to support the 12-month shelf life Gilead had requested. However, the San Dimas site had only tested two stability batches for a year. A third batch only had nine months of data.
Similarly, four stability batches at the La Verne site showed significant changes within the first six months, preventing Gilead from relying on the accelerated approach there as well. Needing another three months to complete stability studies at that site, the company agreed to a post-marketing commitment to continue its accelerated studies for another three months and to continue its long-term stability studies for the full 12 months required to support its proposed shelf life.
Lyophilized Product Stability
When the pandemic hit, Gilead had several years of stability data for a 150 mg lyophilized dosage form, but the company had just gotten started with assessing the stability of the 100 mg strength it would produce for COVID-19 treatment.
In October, the agency determined that 18 months of data for three 100 mg batches and four years for one 150 mg batch supported a 30-month shelf life for one of the six lyophilized dosage form sites.
A second site had only 12 months of stability data and it was only for one batch, but the FDA allowed 30 months of shelf life for that site, saying it was “consistent with the analysis” at the other site. The agency declined to analyze data from five other batches there because they had less than a month of long-term stability data.
A third site had five batches of the 100 mg lyophilized dosage form, but only zero to three months of stability data, so Gilead committed to completing the long-term stability studies and conducting six months of accelerated studies post-approval.
A fourth site had three 100 mg batches in their first month of stability testing; the FDA requested – and received – a post-approval commitment to continue the long-term stability studies and conduct six months of accelerated studies.
Three additional lyophilized product manufacturing sites each had a few batches of 100 mg product on long-term stability, which Gilead committed to continuing after approval through the proposed shelf life, even as the agency agreed to six-month accelerated studies as well.
Drug Substance Issues
When the rolling review began on 8 April, there was only one drug substance manufacturing site, Gilead’s plant in Edmonton, Alberta, Canada. The company added two more to the NDA on 12 May.
By 1 October, Gilead had provided data on 15 batches from its Alberta site, but only three batches from one of the new sites and none from the other.
The drug substance review addendum would focus on the registration batch data from the third site, expected on 19 October.
Meanwhile, agency experts determined in the initial review that they would accept Gilead’s post-approval change management protocol, or PACMP, for adding an alternative drug substance manufacturing site 30 days after submitting the supplemental application for the change, known as a CBE-30 supplement, as long as the company met certain stipulations.
Drug Product Manufacturing Issues
CMC Review #2 also covered drug product information that was still missing when the first CMC review was completed, including registration batch release data from various drug product manufacturing sites and lines, as well as updates to stability data and extractables and leachables data.
Gilead removed one drug product manufacturing site from the NDA on 1 October, the day the first CMC review ended, because it was still missing batch release data. Two manufacturing lines at another facility nevertheless remained in the NDA even though they also lacked any batch release data at that point, but the data arrived in time for the FDA to determine in Review #2 that both lines were adequate.
Gilead proposed long-term storage conditions of two to eight degrees Celsius for the Veklury injection solution and below 30 degrees Celsius for the lyophilized powder. After the FDA’s CMC statistical team discussed Gilead’s proposed shelf life, the agency deferred its decision on shelf life to the second review.
The second review also covered a potential post-marketing commitment for analytical method transfer to the Gilead Ireland facility that was abandoned after the facility’s responsibilities were changed.
The agency found Gilead’s drug product specifications acceptable and noted that elemental impurities would not have to be controlled because Gilead’s analysis showed the levels were below the permitted daily exposure.
One specification that attracted the FDA’s attention is pH control, a critical quality attribute for the injectable dosage form that Gilead adjusts with hydrogen chloride or betadex sulfobutyl ether sodium and sodium hydroxide. Gilead tightened the pH range after the FDA submitted information requests for more supporting solubility and stability data.
The company also agreed to post-approval evaluation of remdesivir injection stability at the lower end of the pH range.
Gilead also agreed to provide additional leachables data for both remdesivir formulations after approval, given the multiple stoppers it had proposed and the limited leachables data it had provided for review.
A Heated Temperature Debate
The package insert was found deficient upon initial review due to Gilead’s insistence on storing the lyophilized version at temperatures below 30 degrees C rather than at the USP controlled room temperature conditions of 20-25 degrees C.
Gilead said the 30 degrees figure jibes with its stability data.
The FDA countered that in its efforts to avoid NDA-specific storage requirements, it has only allowed such storage conditions for PEPFAR products.
But when the agency asked Gilead to switch to the USP conditions, the company replied with a list of its commercial products that have the 30 degrees C storage statement, including Sovaldi, Epclusa and Biktarvy.
The agency set the issue aside for further discussion between the company and the agency’s OPQ labeling expert team and resolution in a second labeling review.
Also deferred for the second review were concerns the FDA’s Division of Medication Error Prevention and Analysis raised concerning a discrepancy between the labeling and the prescription insert regarding storage of the lyophilized formulation after it’s reconstituted.
The Post-Approval CMC Commitments Gilead Made
In the end, the company won approval, but the approval letter included extensive CMC post-marketing commitments for:
-
Release data for two additional registration batches of drug substance to be manufactured at two sites.
-
Release data and three months of long-term and accelerated stability data for three drug substance registration batches from another site.
-
Release data for one batch of lyophilized drug product manufactured at one site.
-
Release data for one batch of lyophilized drug product made at one site.
-
Three months of long-term and accelerated stability data for three batches manufactured on one line at a lyophilized drug product site.
-
Three months of long-term and accelerated stability data for two batches manufactured on another line at a lyophilized drug product site.
-
Three months of long-term and accelerated stability data for two batches manufactured on yet another line at a lyophilized drug product site.
-
Six months of long-term and accelerated stability data for three batches on a lyophilized drug product line.
-
Three months of long-term and accelerated stability data for one batch on a lyophilized drug product line.
-
Three months of long-term and accelerated stability data for two batches of lyophilized product manufactured on certain lines.
-
Three months of long-term and accelerated stability data for three batches of lyophilized product manufactured at a certain facility.
-
Three months of long-term and accelerated stability data for three process validation batches of lyophilized product made an undisclosed facility.
-
Three months of long-term and accelerated stability data for one batch of lyophilized product manufactured on a certain line at a certain facility.
Additionally, there were commitments for six months of long-term and three months of accelerated stability data to support a change from the stopper Gilead used for the 150 mg lyophilized product to the stopper it was using for the 100 mg lyophilized drug product for COVID-19 treatment.
Yet another commitment was for six-month time point data to support Veklury’s stability at the lower end of the proposed pH specification range.
There was also a commitment to continue through expiry a leachables study for three lots of each proposed commercial packaging configuration for the injectable and lyophilized dosage forms. The study had only generated one month of data so far.