Purpose

Clinical Trial Agreements (CTAs), also called “Clinical Study Agreements”, are contracts for the performance of clinical research on human subjects. There are multiple types of CTAs, but most are between an industry sponsor (e.g., a pharmaceutical, biotech, medical device company) or its clinical research organization (CRO) and the research site (e.g., a university, medical institution, hospital, etc.) performing the research on the sponsor’s trial product (i.e., drug, biologic, or medical device) according to the research protocol provided by the sponsor.

This type of work is extremely expensive and involves substantial risks to all involved. Thankfully, it is also heavily regulated in order to protect the privacy, safety, and well-being of the subjects undergoing the trial. As a result of the foregoing, however, CTAs are highly complex agreements that require careful negotiation.

This article aims to shed light on some of the areas of concern from the perspective of both a sponsor and a site, but its focus isn’t necessarily on those provisions most commonly negotiated. Contracting professionals that work with this niche type of contract generally negotiate a large enough volume of CTAs to have these issues well in hand. Rather, this article examines nuances in specific provisions or issues that might be overlooked by busy practitioners.

From the perspective of a Sponsor you will want to be thoughtful of the following:

Most sponsors choose sites based, at least in part, on the identity of the principal investigator (PI) working for the site that will be conducting the trial. The length of a PI’s track record and proven expertise in conducting certain types of research trials is a distinguishing factor as is whether a PI has a substantial body of well-respected published work in the PI’s field of research, which will lend credibility to a study’s results. Publication of the results of a study conducted by such a respected PI is also more likely to be read by other practitioners in the field and publicized at research conferences, gaining traction in the community of physicians that specialize in treating the disease or condition that the trial product is being developed to address. Not surprisingly therefore, sponsors are rightly concerned about sites promising that a study will be conducted by one PI only to have the site switch PIs during the study, so often add CTA language attempting to control a site’s ability to do this.

It is important to be thoughtful, however, about the actual language used in the CTA to establish such a safeguard. For example, sponsors will sometimes use phrasing that is unintentionally problematic for the site, such as “Site shall not remove the PI without Sponsor’s prior written consent.” The problem with this language is that it restricts the site’s ability to remove the PI from leading the study, which is discretion the site needs to retain to avoid potential liability as an employer and medical research facility. Imagine that the site becomes aware of previously unknown conflicts of interest that the PI has with respect to the product or sponsor, the PI’s license to practice medicine is suspended, or the PI has been found to have engaged in serious misconduct with respect to site team members. The site would likely need to remove the PI in each of these situations. Also, the PI may need to withdraw from leading the study to take medical or family leave or resign from employment with the site to accept a position elsewhere. In all of these scenarios, the site must be free to act independently as an employer and medical facility without being hamstrung by a requirement to gain consent from the Sponsor. Moreover, the details of many of these types of matters must be kept confidential by the site in order to comply with privacy laws and/or protect the site’s privileged information. Fortunately, the removal of a PI is not really the issue of concern for the Sponsor: What is problematic is the appointment of a replacement PI that is unacceptable to the Sponsor. Better language could look like “Site may remove from the Trial the individual identified herein as the principal investigator, but shall not appoint another individual as principal investigator of the Trial without Sponsor’s prior written consent.”

Savvy to the fact that removal of a PI does create a certain amount of pressure for the Sponsor to accept a lesser qualified replacement PI given the significant costs involved in shutting down the trial and starting it anew elsewhere, some Sponsors will condition PI substitution on the PI becoming “unwilling or unable to continue conducting the Trial.” Notably, a PI may be willing and able to continue with the trial but the site may not be willing for the PI to continue in that role due to the risky behavior of the PI. Whether such language is acceptable to any particular site given this nuance is a matter of discretion, but arguably removal of the PI should also not be limited to the PI’s discretion given that the other party to the CTA is actually with the site, whose interests don’t align with that of the PI in all cases. One way to avoid such a situation that won’t encounter push back during the contract negotiation is to run the trial at a site that has a large number of qualified PIs in the applicable field of research.

There are a number of important concepts to cover from the perspective of a sponsor with respect to subject enrollment. Firstly, subjects should not be screened for enrollment until the study has IRB approval. Also, subjects that do not meet the eligibility criteria for the study should not be enrolled, but the suitability of prospective study participants should at all times be the independent medical judgment of the PI. Along these lines, the subject enrollment goal for the site should be stated, if possible; the site should be obligated to make reasonable efforts to achieve that goal within a reasonable time of the commencement of the study at the site; and the sponsor should have an express right to terminate the trial at the site (and at all sites, if it is a multi-site study) if the enrollment goal is not achieved or it is not achieved by a specific date.

Unlike a non-disclosure agreement, the confidentiality and non-use provisions of a CTA should define the Sponsor’s confidential information to include more than just the non-public information provided to the site by or on behalf of the Sponsor. It should also include, at minimum, any arising inventions assigned to the Sponsor. If these are not kept confidential their value could be destroyed, particularly given that the patent laws in the United States have switched from a “first to invent” to a “first inventor to file” system. If the site grants the sponsor the option to license its rights in any “other inventions,” usually defined as those inventions arising during the performance of the study that are unrelated to the sponsor’s product or confidential information, those should also arguably be subject to the confidentiality provision until such time as the sponsor’s option has expired. Finally, it is recommended that sponsors carefully consider to what extent Trial Data should be included in the definition of confidential information, bearing in mind any language in the CTA licensing the use of Trial Data to the site for specific purposes and how “Trial Data” is itself defined in the CTA.

The primary work product from the performance of a clinical trial is research data, which is typically recorded in a sponsor-provided Case Report Form completed by the site for each subject enrolled and in study reports, which analyze and evaluate the trial data. In the United States trial data is generally owned and controlled by the entity that records and captures the data in a readable format. It is therefore important that ownership of trial data be assigned to the sponsor in the CTA so that it can use such trial data for its own business purposes, such as compiling the data for submission to the Food and Drug Administration, consistent with the study subjects’ wishes as defined in the informed consent and HIPAA authorization forms, which may or may not be incorporated into a single document.

To reach agreement on this issue it is helpful to understand the concerns of the site in properly defining what constitutes “Trial Data”. For example, the definition should explicitly exclude “source records,” which are original documents or data, including subject medical records, laboratory notebooks, data provided or entered by trial subjects (often into electronic systems), as well as healthcare providers’ records from pharmacies, laboratories and other facilities involved in the clinical trial, and data from automated instruments, such as wearables and sensors. Medical providers are legally obligated to maintain these records, so must retain ownership of them.

Additionally, sites generally expect to be granted a license to use the Trial Data for publication purposes and other internal institutional purposes, such as patient care, education, and non-commercial research. This expectation is reasonable as many institutions need such a license in order to maintain their accreditation and tax-exempt status or to comply with the conditions of certain funding vehicles such as municipal bonds issued with the proviso that the funds will be only used for a public purpose. Universities and medical centers in particular, with their defined missions of education, research, and clinical care, are precluded from accepting contractual restrictions that conflict with these missions.

In most industry-sponsored trials the sponsor provides the protocol, the product, its confidential information, and payment for the service. Assignment language is therefore vital to any sponsor’s ability to both maintain its freedom to operate and earn a return on its research development investments. Yet the value of any legally effective assignment language is only as great as the value of the property being assigned, which is why a large part of negotiating these types of provisions focuses on the definition of the intellectual property being assigned. For example, a CTA might presently assign to the sponsor “all inventions, discoveries, developments, improvements, processes, compositions of matter, formulations, or methods of use or delivery related to the Study Drug or Sponsor’s Confidential Information, whether patentable or not, that are invented by Site’s personnel (whether solely or jointly with others) as a part of and during the direct conduct of the Trial.” While seemingly substantial, this language is not ideal for the sponsor and some of the reasons why are explored below.

One thing that drafters sometimes fail to remember, including the authors of the above example definition, is that though the product and the sponsor’s confidential information are only supposed to be used to conduct the trial, the misuse (i.e., for a purpose outside the scope of the protocol) of either can also lead to the generation of intellectual property. And while the inclusion of a contractual prohibition in the CTA against such misuse would provide a remedy for breach of contract, ownership of the intellectual property generated during the prohibited use would still arise in the inventor(s). Sponsors should therefore strongly consider defining the intellectual property to be assigned to include any generated through the use of the sponsor’s product and confidential information, regardless of whether its use is part of “the direct conduct of the trial” to ensure all intellectual property arising from the use of its property is presently assigned via the CTA.

Finally, the sponsor may see language, like that above, where the definition of intellectual property assigned is limited to that arising “during the conduct of the trial”. However, it is after study completion that investigators and/or sponsors clean the data, derive additional variables, adjudicate endpoints, lock the data set to create the full analyzable data set, carry out pre-specified and additional analyses using software programs, and prepare manuscripts for publication. Investigators may also use parts of the analyzable data set to prepare analyses for presentations, data exploration, and hypothesis generation. Therefore, it is more encompassing for the Sponsor to be assigned any invention that “arises out of” the use of the study product, the sponsor’s confidential information, or the performance of the study.

From the perspective of a Site you will want to be thoughtful of the following:

Sponsors have a legitimate concern about sites using subcontractors in the performance of a study, but it is important that sites maintain some flexibility in order to ensure that they can negotiate acceptable language with subcontractors. As such, language in the prime CTA stating that any subcontractors must execute subcontracts “containing the same terms and conditions as those in the CTA” should generally be eschewed. Even if the site believes the terms and conditions in the CTA are reasonable, it may be seeking to work with subcontractor sites that are more constrained in what they can agree to. This is particularly the case when the potential subcontractors include public institutions that are subject to statutory limitations regarding matters such as liability, indemnification, and subject injury. Many public institutions also require certain language regarding intellectual property rights and data licensing in order to protect their accreditation and non-profit status. Language that is more flexible and often still acceptable to sponsors would be something along the lines of “substantially similar terms and conditions as those in the CTA” or “terms consistent with this CTA.”

Protocol deviations happen. Sometimes this is due to a failure on the part of the study team or sponsor, and sometimes it is due to unforeseen adverse reactions requiring the PI to deviate from the protocol for the safety and well-being of one or more study subjects. Similarly, there may come a point in the conduct of a trial when it must be halted or terminated because it is unsafe for the subjects to proceed. While protocol deviations and trial suspension and/or termination are generally dealt with in separate sections of a CTA, the same issue permeates both— independent medical judgment. It is imperative a PI retain the sole discretion to make decisions based on the PI’s independent medical judgment to deviate from the protocol or stop the trial to protect the trial subjects.

Notably, some sponsors are known to utilize a variety of provisions to attempt to constrain this discretion and/or influence the PI’s decisions as much as possible, even though they have an inherent conflict of interest. Examples of such provisions include requiring the PI to seek consent to deviate from the protocol even where such deviation is for emergency care or urgently necessary to protect subject safety, qualifying the PI’s right to exercise medical judgment such that it must be “reasonable” and “in line with standard medical practice” (even though this is an experimental product), splitting hairs as to the circumstances under which the Sponsor will pay for site services related to a protocol deviation, and requiring a PI to first reach a consensus with the Sponsor as to whether a trial should be suspended and/or terminated for safety concerns before being able to act accordingly.

And while some of these measures may seem innocuous, they may intimidate PIs into, at the very least, second guessing their own medical judgment as to what their patients need to remain safe. Finally, it is worth pointing out that while Sponsors generally have physicians on staff that may monitor and/or liaise with the PI regarding the trial, they are not on staff with the site, may not be licensed in the jurisdiction in which the site is located, and do not have a physician-patient relationship with the subjects. As such, strong language in these provisions that ensures these medical decisions are made solely based on the unconditioned independent medical judgment of the PI is paramount to preserve the rights of the subjects and to avoid future disputes about liability and/or patient safety matters.

On occasion sites may see language stating that the sponsor will not pay for services resulting in invalid data. For example, one publicly available CTA between UCLA and the CRO giant IQVIA states that IQVIA will not pay UCLA for “any Study Subject whose enrollment in the Study materially deviates from the Protocol’s eligibility criteria or . . . from whom Study Data cannot be analyzed because of . . . material Protocol deviations . . . .” While the first exception to payment is reasonable, the latter could very well result in unfairness to the site.

Protocol deviations (material and otherwise) may occur for a number of reasons that are either beyond a PI’s control or are mandated by law. Severe weather, holidays, and other force majeure events may prevent the site’s study team from performing the protocol as written. Also, subjects may deviate from the protocol through no fault of the site by failing to or incorrectly self-administering a test agent, being unable to attend a scheduled study visit, forgetting to take medication on time, etc. Finally, PIs have an ethical, legal, and professional obligation to meet the standard of care for medical practice with respect to their study subjects. This means deviating from the IRB-approved protocol when, and to the extent necessary, to eliminate apparent immediate hazards to study subjects. Unlike a mere failure to adhere to the protocol out of neglect, this a legitimate reason for protocol deviations recognized by the FDA, DHHS, and ICH-GCP. And because this is experimental science, conducted for the very purpose of studying the effects of the product on humans because they aren’t adequately known, unanticipated problems including serious adverse events requiring an immediate deviation do occur.

The risk of these types of deviations should not be shifted onto the site and can often be lessened by a sponsor more thoughtfully designing the protocol. Consider a scenario where a sponsor’s study on a product to treat Alzheimer’s Disease involves lengthy directions for subjects to follow at home, yet the sponsor’s draft of the CTA excludes both payment for services and reimbursement for subject injury on the basis of a patient’s failure to follow instructions. Would it be surprising if the site found those proposed terms to be unacceptable? No, of course not. A fairer way to draft such language is to simply state that the sponsor will pay for all services performed in the conduct of the study, except to the extent such services result in unusable data due to a breach of the CTA.

Sponsors often want to limit what site personnel communicate with regulatory authorities both pursuant to requests for information and during audits. Examples of such language include “Site shall not communicate with any regulatory authority about the trial without the prior consent of the Sponsor”. Obviously, this is highly problematic language as the site is an independent contractor, not an agent of the Sponsor, and must be able to communicate freely with its own regulators. Whether a site will agree to provide the sponsor with advance copies of proposed written communications to the FDA in response to its inquiries or any right of comment is at the discretion of the site, but many of the higher regarded academic centers only provide a copy of what they have already sent and received from the FDA pertaining to the study. These institutions choose to prioritize their academic integrity and avoid any conditions that could be construed as allowing sponsors to improperly influence them or their research.

Overreaching sponsors may also want to be alerted to and attend any audit of the site that “may” have an impact on the study, which is vague, speculative, and problematically overbroad. A site that is a large medical research center may not be able to anticipate all of the audits that could impact the study, nor is it necessarily able to permit a sponsor to attend all of them. For example, a PI could be leading multiple trials for a variety of sponsors within the same disease group and any audit of a PI’s trial might impact all of the PI’s studies if it reveals that the facilities in which the PI conducts them or the manner in which they are being conducted violate laws or regulations. However, the confidentiality obligations imposed in any well-written CTA would obviously prohibit the sponsor of one study from attending the audit of another sponsor’s study.

Many sites are accredited by the Association for Accreditation of Human Research Participant Protection Programs (AAHRPP), which requires those institutions to include specific language in the research agreements they execute, including among other things, the right to independently publish their research results. Furthermore, tax-exempt sites must ensure that they maintain their status as mission-based entities, where academic and research institutions have as part of that mission the public dissemination of scientific research. As such, failure to properly address the right of publication in the CTA may jeopardize a site’s accreditation and tax-exempt status. While it is reasonable to condition the site’s freedom to publish on the implementation of reasonable procedural protections to protect the sponsor’s confidential information and any intellectual property for which it wishes to pursue legal protection, the site should not be required to discuss and negotiate changes to the substantive results or study conclusions of the publication. Typically, these procedural protections include 30-60 days for the sponsor to review the proposed manuscript for confidential information prior to its publication, which if found must be removed by the site, and another 30-60 days of delay if the sponsor identifies patentable intellectual property in the manuscript so it can file for legal protection.

Furthermore, this section of the CTA should also address multi-center publication rights if the study being conducted at the site is just one part of a larger trial being conducted at multiple sites. Specifically, it should be made clear that the first publication of the Study results will be the joint multicenter publication with all sites contributing data, study analyses, and comments. In order to ensure that sponsors cannot prevent the publication of unfavorable results through an indefinite delay of the joint publication or the sponsor’s decision to not publish, a deadline should be set for such joint publication after which each site is free to publish on its own. A reasonable deadline from the point of view of most academic research institutions is the earlier of (a) notice by the sponsor that it has decided not to publish the joint publication, or (b) twelve months after the conclusion, abandonment, or termination of the study at all sites. In order to effectuate this right, language specifying that the site, upon request, will be provided access to the data for its own publication should also be included in the CTA. Finally, authorship of multi-center publications is also generally addressed with reference to the authorship rules of the International Committee of Medical Journal Editors (ICMJE).