Skip to main content
July 27, 2021

Update: Use of Real-World Evidence in FDA Approvals and Product Promotion


Real-world evidence (RWE)—clinical evidence on the use and benefits or risks of a medical product derived from real-world data (RWD)—has become an important focus of industry and regulatory authorities alike, with the hope that the generation and use of such data can reduce the cost and speed of drug development and provide better insights into actual patient experiences with biomedical products. Recent FDA actions in this area suggest a growing FDA reliance on RWE in product approvals, as well as increasing scrutiny of the use of such data in promotional communications. As detailed in this advisory, such developments present important opportunities—and risks—for biopharma and medical device companies.

Historically, the Food and Drug Administration’s (FDA) review divisions showed limited interest in industry RWE generation activities outside of restrictive policy statements cautioning companies against using RWE to attempt to substantiate marketing claims. This was in part because of doubts about whether such studies could meet the “adequate and well controlled studies” standard required to support prescription drug or biologic product safety or efficacy claims.1, 2The 21st Century Cures Act of 2016 3 offered much needed clarity to the disused Food and Drug Administration Modernization Act (FDAMA) sections governing pharmacoeconomic information relying on RWE, and also led FDA to issue guidance to industry regarding how RWE and real-world data (RWD) more broadly could be used for new or supplemental drug approvals.4 Moreover, use of RWE is now increasingly a focus in accelerating confirmation of the effectiveness of products approved under the accelerated approval framework for drugs and biologics.

First Amendment developments have also required FDA to rethink the boundaries of what manufacturers can address in promotion. This includes guidance conceding that promotional communications that are “consistent with” FDA-required labeling and “scientifically appropriate and statistically sound to support the representations or suggestions made”—including potentially RWE-based communications—are lawful if appropriately framed.5

FDA’s review of the use of RWE in various settings is increasingly being revealed in product-specific enforcement and decisions. In July, FDA took modest steps to apply its developing approach to the use of RWE, giving some insight to industry on the situations in which FDA will view RWE as sufficient to support product safety and efficacy claims. First, the FDA Office of Prescription Drug Promotion (OPDP) issued an Untitled Letter to Amgen regarding a professional banner advertisement6 for Neulasta® (pegfilgrastim) injection—which has four approved biosimilars—for allegedly making false and misleading claims comparing the product to biosimilars utilizing data from an RWE study.7 Second, FDA approved a new indication for Prograf® (tacrolimus) based on a non-interventional (observational) study providing RWE of effectiveness.8 FDA has made other recent approvals of new indications based on real-world data.9

Untitled Letter: RWE-Based Claims

Earlier this month, OPDP cited Amgen for disseminating a professional animated banner advertisement for Neulasta® (pegfilgrastim) injection. According to the Amgen Untitled Letter, Amgen had submitted the banner advertisement under Form FDA 2253. The Amgen Untitled Letter notes that FDA “also received complaints” through FDA’s Bad Ad Program regarding other “promotional communications with similar claims and presentations” as the cited banner advertisement. Neulasta, which can be delivered to patients via an on-body injector (OBI) known as Onpro®, is the reference product for all FDA-licensed biosimilar pegfilgrastim products, which are only available as a prefilled syringe (PFS).

OPDP alleges that Amgen made “false or misleading representations about the benefit of Neulasta” when administered through OBI compared to a PFS. Amgen cited “data on file” from an RWE study to support these claims.10 Specifically, the banner advertisement presented claims that:

  • “In a Real-World Study with nearly 11,000 patients Pegfilgrastim PFS resulted in a significantly higher risk of FN [febrile neutropenia] vs. Onpro®
  • Across all cycles of chemotherapy, the incidence of FN associated with prefilled syringe (PFS) was 1.7% (n = 455) vs 1.3% (n = 126) for Neulasta® Onpro®” (frame 1)
  • A large presentation of an upward arrow containing the claim, “31%* *p = 0.01” (frames 1 and 2)
  • “With PFS, FN incidence increased by 31% vs Onpro®” (frame 2)11

According to OPDP, these “claims and presentations create a misleading impression regarding the benefit of the product by stating that there is a statistically significant higher risk of febrile neutropenia (FN) when pegfilgrastim is administered via the prefilled syringe (PFS) compared to the” OBI. The claims were misleading, according to OPDP, because the “multiple limitations of the cited study[] preclude the drawing of such [comparative] conclusions … [based] on delivery method” of Neulasta.12 Specifically, OPDP took issue with the data on file and RWE study because:

  • “the study was not designed to ensure that patients with FN were appropriately identified for inclusion in the analysis” because the “study use[d] an unvalidated algorithm[] to identify participants.”
  • “[T]he study was not designed to ensure that the PFS and OBI patient populations were adequately balanced or controlled for potential [selection] bias” because the study “did not control for factors other than the delivery device that may influence the incidence of FN in the compared groups.”13 Accordingly, OPDP stated its view that “reporting a p-value for the claims and presentations … is misleading” since “[i]t cannot be ruled out that selection bias is entirely responsible for the observed risk difference.”14

OPDP stated that Amgen’s inclusion of “two limitations to the study,”15 which OPDP stated were the “two major deficiencies of the study design,” did “not mitigate the misleading claims and presentations in the banner” advertisement.16 In the end, FDA appears to have focused on the “multiple limitations of design and analytic strategy” to reach the conclusion that use of the RWE to support the noted comparative claims in the banner advertisement layout was not appropriate.

The Untitled Letter also states that the “use of the proper name (i.e., nonproprietary name) of Amgen’s PFS product, on the one hand, and the proprietary name of its OBI product, on the other, could result in healthcare providers failing to understand that Amgen’s Neulasta was used in both arms of the [RWE] study.”17 OPDP explained that this could cause healthcare providers to “conclude that a biosimilar for pegfilgrastim product delivered via PFS is not as effective as Amgen’s OBI product.” In an FDA press release announcing the Amgen Untitled Letter, FDA stated that Amgen’s “false or misleading statements [] could undermine confidence in FDA-licensed biosimilar products [and] potentially slow the progress and uptake of these important therapies” “including their potential to create misperceptions about the safety and effectiveness of FDA-licensed biosimilars.”18 The statements about naming echo concerns FDA raised about biosimilar and innovator naming practices when the agency issued its final guidance on Nonproprietary Naming of Biological Products.19

FDA Approval of New Indication Based on RWE Study

While the Untitled Letters reflect FDA’s reservations about use of RWE to support promotional claims in certain contexts, FDA’s July 2021 approval of Prograf® (tacrolimus) based on a non-interventional (observational) RWE study provides some insight on when RWE can be used to support a clinical claim. The Prograf supplemental New Drug Application (sNDA) approval marked “the first approval of an immunosuppressant drug to prevent rejection in adults and pediatric patients who receive lung transplants.”20 FDA first approved Prograf in 1995 as an oral and injectable treatment21 to prevent organ rejection in patients receiving liver transplants, which was later approved to prevent organ rejection for kidney and heart transplants as well.

In announcing the approval FDA explained that it considered the “fit-for-purpose real-world data [], when compared with a suitable control, [to be] adequate and well-controlled under FDA [new drug approval] regulations.” One of the reasons FDA found the data to be sufficient was because the study used RWD from the US Scientific Registry of Transplant Recipients (SRTR), supported by the Department of Health and Human Services. The data were collected on all lung transplants in the US and were supplemented by information from the Social Security Administration’s Death Master File as a trusted repository of mortality data.22

Notably, the RWD submitted by the sponsor was apparently not the sole basis for the new indication. In addition to the RWD, FDA also considered; (1) data from “randomized controlled trials of Prograf used in other solid organ transplant settings [as] confirmatory evidence of effectiveness”; and (2) “clinical trial evidence from research publications supports the independent contribution of Prograf as part of a multidrug immunosuppressive regimen.”23 FDA explained that the RWD showed a “dramatic improvement in outcomes [] observed among lung transplant patients receiving Prograf as part of their immunosuppression medications compared to the well-documented natural history of a transplanted drug with no or minimal immunosuppressive therapy.”24

*   *   *

These developments provide some insight into FDA’s efforts to modernize its approach to the use of RWE data to support product development and approvals, as well as in product promotion. Given the rapid expansion of high quality RWE resources and associated digital tools, and the use of artificial intelligence to probe such data, the value and importance of RWE in supporting drug development and approval and promotional communications is growing rapidly. These developments present enormous opportunities, but also significant risks if companies to not ensure the quality of the data and adherence to rigorous methodologies in RWE studies, or fail to ensure appropriate transparency and balance in communications. We anticipate that the development and use of RWE will remain a significant focus in the upcoming reauthorization of FDA user fees, and in agency policymaking and enforcement in the Biden-Harris Administration.

© Arnold & Porter Kaye Scholer LLP 2021 All Rights Reserved. This Advisory is intended to be a general summary of the law and does not constitute legal advice. You should consult with counsel to determine applicable legal requirements in a specific fact situation.

  1. Enactment of the Food and Drug Administration Modernization Act (FDAMA) of 1997, Pub. L. No. 105-115, 111 Stat. 2296, marked a formal recognition that pharmaceutical companies could promote the economic implications of on-label product use and substantiate such pharmacoeconomic claims under a lower evidentiary standard than conventional safety or efficacy claims, notably the competent and reliable scientific evidence (CARSE) standard.

  2. While the Center for Drug Evaluation and Research (CDER) and Center for Biologics Evaluation and Research (CBER) have required NDA/BLA sponsors to establish long-term patient safety data registries to study the real-world safety risks from products of potentially higher risk, in general, FDA’s recognition of RWE as a potential tool to support drug or biologic approval or marketing claims has been limited until recent years. That said, FDA has used RWE extensively to probe prescription drug product safety concerns. In the medical device industry, however, such real-world data has for years been utilized more widely as part of the review of higher risk implantable and life-sustaining devices and an important data point for Medicare coverage decisions. See e.g., FDA, Examples of Real-World Evidence (RWE) Used in Medical Device Regulator Decisions, (Mar. 2021).

  3. Pub. L. No. 114-255, 130 Stat. 1033.

  4. See e.g., FDA Draft Guidance, Submitting Documents Using Real-World Data and Real-World Evidence to FDA for Drugs and Biologics, Docket No. FDA-2019-D-1263, (May 2019); FDA Guidance, Use of Electronic Health Record Data in Clinical Investigations (July 2018); FDA, Framework for FDA’s Real-World Evidence Program (Dec. 2018).

  5. See, e.g., FDA Guidance Medical Product Communications That Are Consistent With the FDA-Required Labeling — Questions and Answers (June 2018) (describing circumstances and various examples for which FDA would consider an advertisement or promotional statement to be consistent with the FDA-approved label, and thus, unlikely to warrant regulatory scrutiny if otherwise consistent with the Guidance).

  6. See Amgen Banner Ad.

  7. FDA, Untitled Letter to Amgen (July 7, 2021) (“Amgen Untitled Letter”).

  8. FDA, Press Release, FDA Approves New Use of Transplant Drug Based on Real-World Evidence (July 16, 2021) (“Prograf Press Release”).

  9. See e.g., Zachary Brennan, Pfizer Uses EHR Data to Support Expanded Indication for Breast Cancer Drug, Regulatory Focus (Apr. 5, 2019).

  10. Amgen Banner Ad. According to the citation in the banner advertisement, the “retrospective study {was} designed to compare the incidence of FN associated with Neulasta® Onpro® vs Neulasta® PFS among patients receiving myelosuppressive chemotherapy.” The source of the data is “MarketScan® Commercial Claims and Encounters/Medicare Supplemental and Coordination of Benefits Databases.” Id. at p. 3, frames 6-7.

  11. Amgen Untitled Letter at 2 (emphasis in original).

  12. OPDP notes that the “data on file w{ere} provided by Amgen in response to OPDP’s request for the information.” Amgen Untitled Letter at 2, n.2.

  13. Id. at 3. OPDP expressed concern that “{e}ligible patients for the study had highly diverse clinical characteristics, and the study report did not include any information on the baseline comorbidity or risk factors for FN of the two exposure groups or on design or analytic strategies to minimize the risk of selection bias. Selection bias{} is a key concern for this study because even slight differences in populations (e.g., risk factors for FN, age, chemotherapy regimen and dosing, etc.) could substantially impact the incidence of FN and the conclusions stemming from the analysis.”

  14. Id.

  15. Amgen’s cited limitations for the study were “Retrospective analysis that did not control for additional variables that may influence the incidence of FN” and “Database was not sufficient to understand root causes for observed lower rate of FN for patients receiving Onpro®.” Amgen Banner Ad at frame 7.

  16. Amgen Untitled Letter at 3.

  17. Id. at 4 (e.g., the claim “Pegfilgrastim PFS … vs Onpro”). OPDP notes that “frame six of the banner states, less prominently and in smaller font than the claims and presentations set forth in frames one and two, that the retrospective study evaluates Neulasta Onpro vs. Neulasta PFS, but this statement is not sufficient to mitigate the more prominent presentation of Pegfilgrastim PFS vs. Neulasta Onpro and Onpro.” Id.

  18. FDA Press Release, FDA notifies Amgen of misbranding of its biological product, Neulasta, due to false or misleading promotional communications about the product’s benefit (July 14, 2021).

  19. FDA Draft Guidance, Nonproprietary Naming of Biological Products (Jan. 2017). Specifically, former FDA Commissioner Scott Gottlieb highlighted that separate naming or a distinguishing suffix for a biosimilar “should {not} be used to advance the{} goals” of “promot{ing} competition, lower prices and { } greater access.” FDA, Press Release, Statement from FDA Commissioner Scott Gottlieb, M.D., on FDA’s steps on naming of biological medicines to balance competition and safety for patients receiving these products (Mar. 7, 2019).

  20. Prograf Press Release.

  21. FDA approved an oral suspension version of Program in 2018.

  22. Prograf Press Release.

  23. Id.

  24. Id.