Published in Legal

FDA scrutiny of unapproved CXL trials reveals critical lapses in patient safety and regulatory adherence

This is editorially independent content
21 min read

Over the last 4 years, the FDA has spotlighted troubling practices in ophthalmology by issuing warning letters to at least four ophthalmologists (MDs) and their clinical practices.

The core issue: Conducting clinical trials using unapproved cross-linking (CXL) procedures without adhering to mandated federal regulations—particularly those regarding investigational new drug (IND) applications.

A major cause for concern: The use of CXL methods other than Glaukos Corporation's iLink system—currently the sole FDA-approved corneal CXL procedure in the United States—on both adult and pediatric patients, often without the necessary IND in effect.

Lots to unpack here. First, a refresh on warning letters.

A warning letter is not a first-step action the FDA takes after conducting an internal inspection of a facility—it’s actually the second.

Let us explain: Once the agency finishes its inspection, if it uncovers significant violations of federal regulations (such as the Food, Drug, and Cosmetic [FD&C] Act), they will send a Form 483 to the prospective investigator or company listing each observation of violations and give them an opportunity to respond with next steps for remediation.

  • If, upon receipt of a response, the FDA deems that the investigator or company’s reasoning fails to convincingly address the agency’s concerns and propose a clear plan for preventing future violations, then a warning letter is issued.

Now explain the contents of a warning letter.

In general, a warning letter details significant violations of federal regulations, such as:

  • Poor manufacturing practices
  • Inaccuracies with claims of a product’s indication
  • Incorrect direction for use

The letter will also mandate that the recipient address the issue(s) with an agency-set timeframe and directions provided for them to notify the FDA of their plans moving forward.

  • To note: This response must include plans for correction, which the agency would then check to ensure such changes are adequate (similar to the response required in Form 483).

See here for details on the different types of warning letters.

Gotcha. And what’s the reason for those four letters you mentioned earlier?

The crux of this legal drama: The letters were sent to the aforementioned MDs due to each of them performing a CXL procedure—not the Glaukos iLink procedure, keep in mind—on a select number of patients under the guise of an apparent clinical study.

The issue with this: None of the physicians submitted an IND application before initiating these studies.

  • Therefore: These physicians were in direct violation of FDA regulations (under the FD&C Act, to be specific, but we’ll get to that later).

Interesting … back up a moment and talk about the controversy with CXL.

In the United States, Glauko’s iLink is currently the only FDA-approved CXL procedure for use on patients with progressive keratoconus.

Quietly simply, as the company states on its website, “If it’s not iLink, it’s not FDA approved.”

  • And if any other drug and device combination is used for CXL other than iLink—“it’s not just off-label, it’s unapproved.”

Give me an example of an unapproved CXL procedure.

Consider: Peschke GmbH, a Swiss company specializing in CXL technology and equipment.

Its Peschke CXL system uses ultraviolet (UV) light and specialized riboflavin eye drops that penetrate the cornea without removing the corneal epithelium.

  • However: In 2022, the FDA issued import alerts (see here and here) for Peschke’s CXL devices and riboflavin solution.

Refresh on import alerts: These are issued under the FD&C Act and used to protect consumers against products with a history of known violations—and can also detain shipments of a product without the agency needing to test or physically examine it.

Go on …

To clarify: While Peschke’s CXL devices are not FDA-approved—meaning, they do not meet regulatory standards—for either import and sale in the United States, they are allowed in one specific scenario (and this is key):

For use in clinical investigations that are part of an FDA-approved investigational new drug (IND) application.

Wait—but isn’t this procedure approved elsewhere?

Yes, it is. However, the key stipulation here is FDA-approved”—as in the United States.

As we all know: Although a drug, device, or procedure may be approved in another country, that doesn't mean it can be utilized in the United States without regulation.

So is there really a difference between approved and unapproved CXL?

Yes, there really is … in fact, Glaukos has honed in on this by providing a complete breakdown of how unapproved CXL treatments stack up against iLink.

Among its points are a few warnings associated with the use of unapproved CXL, including:

  • No FDA-reviewed safety and efficacy data
  • No FDA-verification of safety, effectiveness, or quality of unapproved compounded drugs
  • Compounded drugs are not subject to Current Good Manufacturing Practice (cGMP) requirements
  • No riboflavin solutions are currently FDA-approved for epithelium-on procedures
  • Unapproved devices may be modified or illegally imported with no postmarket surveillance
  • Liability insurance carriers may not cover potential risks associated with commercial use

What else?

A key difference separating these approved and unapproved procedures is that the safety and efficacy of the iLink are supported by pivotal trial data (available via ClinicalTrials.gov)

In comparison: Unapproved CXL procedures should only be conducted “with American IRB (Institutional Review Board) oversight and an FDA-approved IND application.”

  • Take note: An IRB is responsible for reviewing and monitoring human subject-involved biomedical research and has the authority to approve, request modifications for, or reject research.

And in the case of the iLink?

As stated on the Glaukos wesbite: “Physician-sponsored studies (with compounded riboflavin) being listed on ClinicalTrials.gov does not ensure that the study is part of an FDA-approved IND.”

Plus, another major component to this: Informed consent by patients.

Specifically: Glaukos noted that, if using an alternative, unapproved option like riboflavin, then “patients should be notified of the unapproved status of the procedure and [be made aware] that there is an FDA-approved option available.”

Makes sense.

Right? Unfortunately, the issue is that those four doctors we mentioned earlier:

  1. Used an unapproved CXL procedure on patients as part of a clinical investigation
  2. Submitted an IND application for the aforementioned CXL procedure but failed to wait for approval before conducting the study
  3. Conducted their clinical investigation under the guise that a clinical study approved by the IRB translated to the approval of their IND (in case you were wondering: it does not).

Yikes. Give me the rundown on these situations.

The four letters listed below were issued to physicians after the FDA conducted an investigation as part of its Bioresearch Monitoring Program (BIMO; read up on that) and sent the physicians a Form FDA 483 (which indicated violations were found as a result of the investigation).

After receiving the letter, the doctors were given 15 days to reply with specific and detailed actions they would take to prevent similar violations in the future.

A crucial note the agency stressed to the physicians: An IND does not go into effect upon FDA’s receipt of the IND submission—rather, it generally goes into effect 30 days after the FDA receives the application (or potentially earlier, if the FDA gives notification).

FDA Inspection DateDoctor/PracticeIND-Related InfactionsAdditional Infractions
Nov. - Dec. 2021Physician A Location: San Francisco, CaliforniaPhysician A initiated enrollment and treatment of four patients with an investigational new drug product before its IND was submitted to the agency and in effectFailure to obtain IRB-approved “informed consent and assent” of six patients (aged 11-17) involved in his research
Apr. 12 - 20, 2022Physician B Location: Orland Park, IllinoisPhysician B conducted two clinical trials involving a combination product consisting of investigational drugs prior to IND being submitted and in effectPatients were enrolled after IRD approval and IND approval, but before the IND went into effect
Sept. 12 - 19, 2022Physician C Location: Dallas, TexasPhysician C’s practice (the sponsor) conducted two clinical investigations prior to IND submission and end of 30-day waiting periodN/A
Sept. 19 - Oct. 26, 2022Physician D Location: Highland Park, IllinoisPhysician D’s practice (the sponsor) conducted two clinical investigations involving a combination product of investigational drugs before submitting an IND application and allowing it to take effectFailure to maintain adequate records showing the receipt, shipment, or other disposition of the investigational drug

Let’s talk about the importance of an IDE.

An IDE (which all physicians should have had in place prior to conducting trials) essentially allows an investigational device to be used in a clinical study so safety and effectiveness data can be collected.

  • Why this is important: The FDA has noted that “all clinical evaluations of investigational devices, unless exempt, must have an approved IDE before the study is initiated.”

For clinical evaluations of devices that have not yet been cleared for market use: These require a few key criteria to be met—including an investigational plan approved by an IRB and (if the study involves a “significant risk device”) the FDA.

Got it. And this investigational product applied to that stipulation?

Indeed (according to the FDA). The agency stated the product used in the clinical study “comprises both drug and device components and is therefore a combination product.”

As such: An IND was required for any clinical investigation of this combination product.

  • And just to recap: An IND was not submitted.

Talk about using combination products with investigational drugs in trials.

The FDA noted that a “combination product consisting of investigational drugs [used during the clinical investigations] is not a lawfully marketed product in the United States.”

  • Keep in mind: The only way for this specific drug to be IND-exempt would be if it were already commercially available.

Therefore: The agency stated that such a combination drug would “not qualify for an IND exemption,” and the practice in question is actually required to submit (and have in effect) an IND before initiating and conducting their investigations.

  • Let’s also keep in mind: As the clinical investigation’s sponsor, a practice (not the IRB) would be responsible for complying with—and also, really, just being aware of—federal IND regulations.

Got it. Now to some specifics of these trials conducted prior to IND acceptance.

In the case of Physician D: The FDA noted during its inspection that federal investigators reviewed the practice’s conduct of two clinical investigations:

  • Protocol PXL-330: Safety and Effectiveness of the PXL-Platinum 330 System for Corneal Cross-Linking in Eyes with Corneal Ectasia
    • What this involved: The PXL Platinum 330 system, an investigational combination product comprised of riboflavin 0.25% transepithelial solution and riboflavin 0.1% hypotonic solution (Peschke TE), in combination with ultraviolet-A (UVA) light
  • An investigational combination product consisting of an IND

Quick note: Remember, no riboflavin solutions are currently FDA approved (so any use of them must be done under investigational conditions that have received regulatory clearance [an IND]).

And what were the corresponding violations?

The FDA noted that the practice failed to adhere to such “applicable statutory regulations” in regards to not just the conduct of the clinical investigations but also the “protection of [32 participants].”

To be more precise, the agency stated: “You failed to submit an IND for the conduct of a clinical investigation with [INDs] that is subject to [code of federal regulations].”

This lack of follow-through included Physician D’s practice:

  • Initiating and conducting a clinical investigation on the combination product (riboflavin 0.25% and 0.1% in combo with UVA light) IND without submitting and having an IND in effect.

“No FDA records indicate that you submitted an IND before initiating and conducting this clinical investigation,” according to the FDA.

Isn’t record-keeping critical for this type of trial?

Yes, it is. Because this specific clinical trial was conducted under Protocol PXL-330 (an unapproved device), the investigators were required to maintain records consisting of a few key criteria.

  • However, the FDA specified: Investigators failed to maintain adequate records (see here for specifics) of the riboflavin solutions—both also investigational drugs.

Got it. Now, let’s get into the legality of these cases.

First up: The legal ramifications of using an investigational drug without an IND.

Broadly speaking, potential consequences vary for a clinician and an institution.

For a clinician:

  • Suspension or termination from treating patients in any and all FDA-regulated studies

For an institution (clinical practice), if it fails to provide clinicians with guidance and establish clear procedures:

  • Investigators (or a company / institution) may be disbarred from performing future research
  • Civil money penalties may be imposed
  • The FDA may seize products or issue injunctions
  • In extreme cases: criminal prosecution may be possible, including jail time or other fines

Take note: While there are special circumstances in which using an unapproved drug is not illegal—including in the cases of “compassionate use,” “emergency use,” or “expanded access” for seriously ill patients—there are still federal guidelines and protocols to follow.

  • If a clinician fails to do this: From a legal standpoint, it could be considered practicing medicine without a license or violating patient safety regulations.

What do the experts have to say about this?

We spoke with Josh Hanen, JD, OD, FAAO, in practice at River Lake Clinic in Minneapolis, Minnesota, regarding the legal and ethical concerns surrounding the use of unapproved drugs without proper study protocols in place.

He referenced five potential concerns that may have lasting impacts on patients and clinicians.

  • Financial motivation
    • Clinicians might use compounded drugs (an unapproved medication customized by a pharmacist for a specific patient) to reduce costs and potentially compromise patient care for financial gain
  • Lack of informed consent
    • Patients are not informed that a drug is not FDA-approved or part of a properly documented and protocol-regulated study
  • Regulatory noncompliance
    • Clinicians might (knowingly or unknowingly) bypass required regulatory steps (like obtaining an IND application) intended to protect patient safety
  • Potential patient risks
    • With no proper regulatory study protocols in place, potential side effects or long-term treatment impacts on patients cannot be systematically tracked
  • Ethical breach
    • The use of unapproved drugs under the guise of a study with no proper documentation is considered “fundamentally unethical”

Zero in on that regulatory noncompliance part.

A key aspect to this is the concept of follow-through and enforcement—on the side of both a clinical practice and a federal agency.

“An agency may have a rule in place that says clinicians have to follow it one way, but whether the agency actually enforces that rule fully is another story,” Dr. Hanan said.

“But that doesn’t mean you should just go ahead and ignore the regulations,” he added. “Usually, when it becomes a big enough problem [such as the aforementioned CXL violations], you don’t want to be on the wrong side of that.”

  • Plus, he noted: “It just makes sense to focus on helping your patients in ways that are legal and proper and follow the rules, and not trying to bend them to benefit in other ways.”

And in the case of clinicians who claim ignorance of such protocols to follow?

Dr. Hanen emphasized that while clinicians may claim good intentions going into such situations as a clinical study—without following through in securing an IND application—simply believing they are doing the right thing doesn’t absolve them of their responsibility to follow established regulatory guidelines.

As such, he advised that clinicians should always “think prospectively.”

“Let’s say a treatment doesn’t agree with my patient; What should I have in place to make sure that I did all I could to advise and warn the patient? That would mean following all the rules,” Dr. Hanen said.

Based on these letters, it seems these physicians were unclear on proper protocols.

Indeed it does—and they wouldn’t be the first clinicians to be confused, according to Philip J. Desjardins, an FDA regulatory lawyer and partner at Arnold & Porter in Washington, DC.

  • Some background: Desjardins’ expertise lies within the area of medical devices after spending nearly a decade at Johnson & Johnson in its MedTech business, as well as at the FDA in its Center for Devices and Radiological Health—focusing on pre-market and post-market issues.

However, in speaking on the “confusing and challenging” situation involving the use of an already-FDA-approved drug for a new, unapproved indication—such as if the iLink therapy were to be used for an indication other than progressive keratoconus and corneal ectasia—Desjardins noted that physicians’ “naivete in that area is not generally a good defense.”

“The FDA would say that one of your responsibilities as an investigator is to fully understand the regulatory requirements around these products and apply the regulations appropriately,” he said.

Speaking broadly, what should clinicians do if violations are uncovered?

In that case, it’s advisable for clinicians to respond both promptly and comprehensively to any FDA correspondence that’s issued by:

  • Taking Form 483s and warning letters seriously
  • Providing detailed corrective action plans
  • Demonstrating an understanding of potential violations

And as Desjardins previously emphasized, it’s crucial to understand that an apparent “lack of knowledge of regulations” is not a strong legal defense.

  • “Oftentimes, respondents to FDA letters or 483s will start with ‘What's the minimum I can do to have this go away?’ But there's really no guarantee that the FDA is going to accept it,” he said. “So I generally start with what the FDA would like to see in this situation and what we think would be reasonable to go back to them with.”

What to keep in mind: Thinking through not just what the next step is, but “what the conclusory step the FDA could take, and how to avoid that as best as possible.”

And the big question: What can clinicians do to avoid a potential situation like this?

Desjardin noted a few actions clinicians can put into practice when working on clinical trials involving INDs—with the most basic being: Understanding FDA regulatory requirements.

Other measures include:

  • Taking proactive steps prior to beginning research:
    • File proper investigational applications (IND or IDE)
    • Confirm regulatory compliance for specific studies
    • Consult with regulatory experts if unclear on federal guidelines
  • Carefully documenting and tracking all research activities:
    • Maintain comprehensive records
    • Ensure complete transparency with federal regulatory entities
    • Implement strong internal compliance mechanisms
  • Implementing organization-wide compliance strategies
    • Create systemic approaches for regulatory adherence
    • Trains staff on regulatory requirements
    • Develop preventative mechanisms (especially to avoid any repeat violations)

The ultimate goal: To prioritize patient safety and maintain the high, rigorous research standards that align with FDA guidelines.  

*Disclaimer: The information provided in this article does not and is not intended to constitute legal advice; instead, all information, content, materials available herein are for general information purposes only.

This article was updated on May 29, 2025.

How would you rate the quality of this content?