Returning Incidental Research Findings from FDG PET/CT to Participants: A Survey of Investigators from a Clinical Trial of Rheumatoid Arthritis

Abstract

Objective:

There are limited data on researchers’ attitudes and beliefs on returning and managing incidental research findings from whole body ¹⁸F-fluorodeoxyglucose-positron emission tomography/computed tomography (FDG PET/CT) imaging.

Methods:

Site principal investigators (PIs) who enrolled participants for the Treatments Against RA and Effect on FDG PET/CT (TARGET) trial were surveyed.

Results:

Of the 28 TARGET site PIs eligible for the study, 18 consented to participate (response rate: 64%). Many site PIs returned incidental findings to participants (61%), and the most common finding that was returned was serious (but not life-threatening) and treatable (54.5%). More than half of the investigators believed that adequacy of clinical follow-up (58.8%) and legal liability if incidental findings are not disclosed (55.6%) were extremely important factors in returning incidental research findings from whole body FDG PET/CT. All investigators felt very obligated to return incidental research findings if scans revealed a treatable, high-risk medical condition. Most investigators felt very obligated to disclose incidental findings with important health implications (94.4%), for which proven preventive or therapeutic interventions exist (77.8%), that provide early detection of a health problem (72.2%), if participants ask for their incidental findings (72.2%), and if scans have established validity for a particular medical condition (61.1%).

Conclusion:

Although it is recommended that researchers report and manage incidental research findings, our data show differing views and uncertainties on what and how to return, and the extent of follow up needed to manage, incidental findings from whole body FDG PET/CT; this highlights the need for more specific and standardized guidance.

Introduction

Data regarding the frequency of incidental imaging findings in clinical trials are limited, especially in rheumatology. Potentially serious incidental findings are reported in approximately 3.9% of asymptomatic adults undergoing a magnetic resonance imaging scan of the brain and body.¹ Most of the data on incidental research findings from whole body 18F-fluorodeoxyglucose-positron emission tomography/computed tomography (FDG PET/CT) are in patients with known malignancy^2,3,4,5,6,7 with little corresponding data in patients without known malignancy,⁸ including patients with autoimmune diseases.^9,10

Imaging studies can lead to many incidental findings independent of the type of imaging. In research, a particular imaging method is chosen to best assess a specific research aim. In the process of conducting imaging research, some imaging methods may be more effective in detecting certain types of incidental findings. Additionally, individual diseases can have varying potential manifestations, which may lead to different types of incidental findings even among different imaging modalities. For example, patients with rheumatoid arthritis (RA) can have extraarticular manifestations that, if detected as an incidental finding, can change management (e.g., interstitial lung disease).

With evolving technological advancements in healthcare, we will inevitably discover more incidental findings in research. Investigators often face challenging situations involving incidental research findings: Incidental findings can be medically beneficial (early diagnosis leading to effective treatment) but also potentially detrimental (risks of additional testing, including psychological distress, cost of follow up or unnecessary treatment). The potential need for additional diagnostic workup highlights the researcher’s duty to return incidental findings from imaging trials.¹¹

However, clear guidance on how to explain the risks and benefits of potential incidental findings in the informed consent process and how to organize follow up for these findings is lacking. Management of incidental imaging findings is left up to individual sites and investigators; as a result, there is considerable variability in the management of incidental imaging findings among research sites, and no clear consensus among investigators on how to handle incidental findings in imaging research.¹²

The Treatments Against RA and Effect on FDG PET/CT (TARGET) trial provides an opportunity to gather information on the types of incidental findings found in patients with RA without known malignancy and researchers’ attitudes and beliefs on returning and managing incidental research findings from whole body FDG PET/CT imaging.

Materials and Methods

TARGET Trial

TARGET was a 24-week multicenter randomized controlled trial conducted at 41 centers in the United States between 2015 and 2021, with 28 centers ultimately enrolling participants.¹³ The objective was to determine whether adding a tumor necrosis factor (TNF) inhibitor to low-dose methotrexate resulted in a larger decrease in arterial inflammation compared with adding hydroxychloroquine and sulfasalazine to low-dose methotrexate (triple therapy). Whole body FDG PET/CT was used to assess cardiovascular and joint inflammation. Patients with known malignancies were excluded. After image acquisition, safety reads for incidental findings were read by a local radiologist. Local radiologists assessed for findings that might be of clinical significance that should be shared with the site principal investigator (PI), which included increased focal uptake of tracer that may represent infectious, immune, or malignant processes and any other pattern of uptake that may cause clinical concern. One-hundred thirty-eight participants completed follow up and 115 had paired FDG PET/CT scans that were analyzable for the primary outcome (arterial target-to-background ratio or TBR).

Survey Instrument

A survey was designed to gather information on incidental findings found during the TARGET trial and to assess investigators’ attitudes and beliefs on returning and managing incidental findings from whole body FDG PET/CT. The survey was informed by previously published research^{14,15,16,17,18,19} and had four sections: 1. Management of incidental findings in the TARGET trial, 2. General opinions of investigators on returning incidental findings, 3. Factors that would make investigators more or less likely to disclose incidental findings, and 4. Demographics and work characteristics of the investigators. Race and ethnicity were self-reported from a fixed set of categories with a free-text option. Both quantitative and qualitative data were obtained from this survey using Likert scale, forced choice, rank order, and open-ended questions for free-text responses (see Supplementary Material for the survey instrument). The survey was piloted with three TARGET PIs (JTG, DHS, JMB) and revised with their feedback.

Study Population, Recruitment, and Survey Administration

Twenty-eight TARGET site PIs who enrolled participants were included in the survey sample (PIs who were involved in reviewing the survey and results were not included). In the last quarter of the TARGET trial, all site PIs were emailed the study information and provided a link to access the survey online. If there was no response to an initial email, the participants received a total of two additional email reminders. Consent information for the electronic survey was provided in the survey link, at which point participants could chose to continue with the survey or decline. The online survey was programmed and made available in REDCap (Research Electronic Data Capture). Responses were collected anonymously. Site PIs were compensated $250 for their participation. The study was approved by the Institutional Review Board (IRB) at Columbia University Medical Center.

Statistical Analysis

The final survey data were imported into SPSS (version 25) for statistical analysis. Quantitatively coded response items in the survey (both single choice and “check-all-that-apply” questions) were analyzed by generating frequencies and cross-tabulations. In the event of missing responses to categorical questions, percentages were based on the total number of investigators who provided an answer; no imputation was pursued. The frequencies of responses to questions were evaluated in relation to the characteristics and demographic features of the site PIs (e.g., age, academic rank, and sex) to inform factors that may influence attitudes and beliefs. Answers to rank choice questions were analyzed using weighted averages. Selected free-text responses to open-ended prompts for examples, attitudes, and beliefs were categorized into themes, and representative quotes have been included.

Results

Characteristics of Respondents

Of the 28 TARGET site PIs eligible to participate in the survey, 18 consented to participate (response rate: 64%). Respondents’ characteristics are shown in Table 1; they were mostly female and White and worked at an academic medical center and spent most of or all their time in patient care. Half of the respondents were age ≤50 years.

Table 1.

Baseline Characteristics of Participants

Characteristic	Participants (n = 18) n (%)
Age group
≤ 50 years	9 (50)
> 50 years	9 (50)
Sex
Female	11 (61)
Male	7 (39)
Race
White	16 (89)
Asian	2 (11)
Hispanic/Latino	1 (5)
Study role *
Participant recruitment	17 (94)
Participant consent	13 (72)
Laboratory analysis	4 (22)
Study design	2 (11)
Statistical analysis	0 (0)
Work setting
University/Academic medical center	11 (61)
Private practice	7 (39)
Work location
West	9 (50)
South	5 (28)
Northeast	2 (11)
Midwest	2 (11)
Academic rank
Full Professor	5 (28)
Associate Professor	4 (22)
Assistant Professor	3 (17)
Other†	2 (11)
Not applicable	4 (22)
Time spent on basic science research ‡
Most of the time	1 (6)
Sometimes	1 (6)
Rarely	4 (23)
Never	11 (65)
Time spent on clinical research ‡
All of the time	3 (17.5)
Most of the time	3 (17.5)
Sometimes	9 (53)
Rarely	2 (12)
Time spent on patient care
All of the time	7 (39)
Most of the time	9 (50)
Sometimes	2 (11)
Time spent on Ethical, Legal, and Social Implications (ELSI) research ‡
Sometimes	2 (12)
Rarely	3 (18)
Never	12 (70)
Degrees *
MD	14 (78)
MD, MS	2 (11)
MD, PhD	1 (5.5)
PhD	1 (5.5)
Served on IRB	2 (11)
Prior research participant	8 (44)
Received incidental findings as a research participant	3 (17)

^*Respondents were asked to select all that apply.

^†Clinical Professor (1), Medical Director Clinical Research (1).

^‡One participant did not respond; percentages are based on those who responded.

Return of Incidental Findings from Whole Body FDG PET/CT

Many of the site PIs returned incidental findings to participants (61%), and the most common finding that was returned were those that were serious (but not life-threatening) and treatable (54.5%). Incidental findings were usually returned to participants through the investigator (91%) and in person by the investigator (73%). About half of the site PIs felt that aggregate incidental findings (overall results of the study that relate to study participants as a group, rather than to individual participants) should be returned (56%), with the most favored method of communication being a study newsletter (50%, Table 2). Investigators were given the opportunity to explain why they would not return aggregate incidental findings, which included concerns with participants having difficulty understanding how aggregate findings apply to them or provoking anxiety (Table S1).

Table 2.

Return of Whole Body FDG PET/CT Incidental Findings from the TARGET Trial

	n (%)
Returned any incidental findings to participants (n=18)
Yes	11 (61)
No	5 (28)
Do not know	2 (11)
Type of incidental findings returned to participants* (n=11)
Serious (but not life-threatening) and can be treated	6 (54.5)
Not life-threatening or serious and may be treated	3 (27)
Not life-threatening or serious and may not be treated	3 (27)
Life-threatening and can be treated	1 (9)
Uncertain and cannot be interpreted at the moment	1 (9)
Serious (but not life-threatening) and cannot be treated	0 (0)
Life-threatening and cannot be treated	0 (0)
How incidental findings were returned to participants* (n=11)
Through the investigator	10 (91)
Through the research staff	5 (45)
Through the participant’s physician	2 (18)
How incidental findings were communicated to participants* (n=11)
In person, by the investigator	8 (73)
Phone call	4 (36)
In person, by the research staff	3 (27)
How incidental findings were communicated to participant’s physician* (n=2)
Phone call	1 (50)
Other†	1 (50)
Does the team have plans to return incidental findings to participants (n=7)
Do not know	3 (43)
Yes	2 (28.5)
No	2 (28.5)
How the team plans to return incidental findings to research participants* (n=2)
Through the investigator	1 (50)
Through research staff	1 (50)
How incidental results will be communicated to participants* (n=2)
Phone call	2 (100)
Aggregate incidental findings should be returned to participants (n=18)
Yes	10 (56)
No	4 (22)
Not sure	4 (22)
How would you communicate aggregate incidental findings to participants* (n=10)
Through a study newsletter	5 (50)
Online study updates	4 (40)
Individual letter	4 (40)
Individual email	3 (30)
Phone call	1 (10)
Felt that incidental findings were overall (n=18)
Beneficial	4 (22)
Potentially beneficial	7 (39)
Neither beneficial or detrimental	2 (11)
Potentially detrimental	2 (11)
Detrimental	0 (0)
Not applicable	3 (17)

^*Respondents were asked to select all that apply.

^†The principal investigator was the participant’s rheumatologist.

Site PIs who rarely or sometimes participated in Ethical, Legal, and Social Implications (ELSI) research returned findings that were serious (but not life-threatening) and treatable more than those who never participated in ELSI research (80% vs. 16.7%; p=0.028). Assistant Professors returned incidental results to research participants through the participant’s physician more than Full or Associate Professors (66.7% vs. 0%; p=0.024). Compared with those who did not serve on an IRB, those who served on an IRB returned incidental findings to research participants in person, using study staff (100% vs. 6.7%; p=0.022).

Beneficial vs. Detrimental

Site PIs rated incidental findings from the TARGET trial as beneficial (22%), potentially beneficial (39%), neither beneficial or detrimental (11%), potentially detrimental (11%), or detrimental (0%, Table 2). Although most investigators felt that the incidental findings were beneficial or potentially beneficial, the examples provided to support why findings were potentially detrimental were the most detailed and extensive, emphasizing the additional invasive medical workups that followed or the interruption of RA therapy (Table S2).

Investigators’ Views on Returning Incidental Findings from Whole Body FDG PET/CT

When asked to choose one of three possible scenarios, most investigators favored returning all incidental findings to participants (61%), with the remainder favoring returning some incidental findings to participants (39%, Table S3). General opinions on returning incidental findings from whole body FDG PET/CT were also obtained (Figure 1). All investigators agreed that researchers have an obligation to explain incidental findings to participants. Half of the investigators either agreed or strongly agreed that “identifying incidental findings from whole body FDG PET/CT is burdensome” and “it is the researcher’s responsibility to decide which incidental findings should be reported to a participant.”

Figure 1.

General Opinions Regarding Returning Incidental Findings

Women or those who served on an IRB agreed that “researchers should actively search for incidental findings from whole body FDG PET/CT imaging that are not relevant to the research study” more than men or those who did not serve on an IRB (81.8% vs. 14.3%; p = 0.041 and 100% vs. 46.7%; p=0.036). Additionally, investigators older than age 50 years agreed that “participant preferences should be used to guide which incidental findings from whole body FDG PET/CT imaging should be reported back to them” more than those aged ≤50 years (77.8% vs. 55.5%; p=0.022). Investigators who had been a research participant agreed that “returning incidental findings from whole body FDG PET/CT will increase health care costs” more than those who have not been research participants (87.5% vs. 50%; p=0.029).

Those who were not involved in basic science research or were rarely or sometimes involved in clinical research were more likely to agree that “participants should have an option to choose which incidental findings they would like to receive” compared with those who have some involvement in basic science research or who do clinical research most or all of the time (81.8% vs. 33.3%; p=0.035 and 90.9% vs. 16.7%; p=0.020). Those involved in clinical research all of the time agreed that “identifying incidental findings from whole body FDG PET/CT is burdensome” more than those involved in clinical research sometimes or most of the time (66.7% vs. 50%; p=0.013).

How Incidental Findings Should be Returned and Communicated (Table 3)

Table 3.

How Incidental Findings Should Be Returned and Communicated

	Most to Least Preferred (Weighted Average)
How to Return	Through the investigator (3.0) Through the participant’s physician (1.7) Through the research staff (1.4)
How to Communicate to Participants	In-person (4.6) Phone call (4.0) Report by mail (2.3) Online or web-based tool (2.1) Report by e-mail (1.9)
How to Communicate to Participant’s Physician	Phone call (3.9) Report by e-mail (3.7) Report by mail (3.5) Online or web-based tool (2.1) In-person (1.8)

The most preferred way of returning incidental findings was through the investigator, in-person for participants, and by phone call for the participant’s physician. For communicating incidental findings to participants, investigators aged ≤50 years ranked email higher than those aged >50 years, who were more likely to choose it as their last choice preference (p=0.031). For communicating incidental findings to the participant’s physician, those involved in clinical research some or most of the time or who never served on an IRB ranked using a phone call higher than those involved in clinical research all of the time or who served on an IRB (p=0.008 and 0.001); those who served on an IRB ranked in person higher than those who did not serve on an IRB (p=0.007).

Importance of Various Factors in Returning Incidental Findings (Figure 2)

Figure 2.

Importance of Various Factors in Returning Incidental Findings

More than half of the investigators believed that adequacy of clinical follow-up (58.8%) and legal liability if incidental findings are not disclosed (55.6%) were extremely important factors in returning incidental findings from whole body FDG PET/CT to research participants. Half of the investigators (50%) felt that the financial cost of returning results was not at all important.

More men than women thought “legal liability for adverse outcomes that might occur as a result of returning incidental findings” was a very or extremely important factor (71.4% vs. 54.5%; p=0.011). Compared with those in private practice, more academic investigators felt that the “potential to blur the boundary between research and clinical care” was a very or extremely important factor (36.4% vs. 28.6%; p=0.031). Investigators who were more involved with patient care were more likely to believe that the “clinical utility of the incidental finding” was a very or extremely important factor in returning incidental findings (100% vs. 50%; p=0.043). Those who never served on an IRB felt that “medical costs from follow-up treatments or visits” was very or extremely important more than those who served on an IRB (33.4% vs. 0%; p=0.032).

Obligation to Disclose Incidental Findings (Figure 3)

Figure 3.

Obligation to Disclose Incidental Findings

All investigators felt very obligated to return incidental findings if whole body FDG PET/CT revealed a high-risk medical condition (like a tumor or an aneurysm) that can be treated. Most investigators felt very obligated to disclose incidental findings that have important health implications (94.4%), for which proven preventive or therapeutic interventions exist (77.8%), that provide early detection of a health problem (72.2%), if participants ask for their incidental findings (72.2%), and if the scans have established validity for a particular medical condition (61.1%). Women felt more obligated than men to disclose incidental findings when the “whole body FDG PET/CT has established validity for a particular medical condition” (81.8% vs. 57.2%; p=0.040). If a whole body FDG PET/CT reveals no findings, half of the investigators felt either not obligated at all (38.9%) or slightly obligated to disclose incidental findings (11.1%), with the other half feeling either moderately obligated (5.6%) or very obligated (44.4%).

Most Common Challenges and Strongest Motivations for Returning Incidental Findings

Investigators described in their own words the three most common challenges to returning incidental findings from whole body FDG PET/CT to participants. Five major themes emerged, including: 1. Explaining/Understanding Results, 2. Time/Cost, 3. Follow Up, 4. Emotional Burden, and 5. Data Disclosure (Table S4).

Investigators also described the three strongest motivations for returning incidental findings to research participants. Four major themes emerged, including: 1. Impact on Clinical Care, 2. Right to Data, 3. Prevention, and 4. The Patient Physician Relationship (Table S5).

Discussion

There are a number of publications on participant views on incidental findings in magnetic resonance imaging research^14,20,21 and researcher views on incidental findings in neuroimaging^14,15 and genetic testing.^16,17,19 To our knowledge, this is the first study that examines researchers’ attitudes and beliefs regarding incidental research findings from whole body FDG PET/CT.

The inherent goals in clinical care and research are different: in clinical care, healthcare professionals have a clear duty to disclose clinically relevant findings to patients; in research, the goal is to develop generalizable knowledge about human health and biology, and the duty of the researcher to the research participant is less clear. However, the boundaries of research and clinical care are increasingly blurred.

Most of the investigators returned incidental findings from the TARGET trial and therefore had intimate knowledge of the benefits and challenges of returning those findings. Although 61% felt that the incidental findings found on whole body FDG PET/CT during the TARGET trial were beneficial in some way, there were concerns with returning these incidental findings, including participants undergoing invasive workups that may not lead to clinical benefit or may even lead to clinical or emotional harm. Incidental findings from whole body imaging can lead to invasive procedures, such as biopsies, with most biopsies showing no findings and yielding few malignancy diagnoses, suggesting possible overtesting.²² Overdiagnosis, or the “diagnosis of a medical condition that would never have caused any symptoms or problems,”²³ may also arise; however, it can be challenging to accurately predict the course of a disease, which complicates the idea of overdiagnosis.

Regarding what to report from whole body FDG PET/CT, there was strong agreement in the types of incidental findings that needed to be returned to participants, including those of clinical relevance or that are clinically actionable, including preventive measures, which is consistent with prior recommendations.²⁴ However, for incidental findings that do not provide a clear, clinically actionable benefit or when there are no findings, there was more disagreement among investigators on whether to return these results to participants.

There was also uncertainty around long-term follow up of incidental findings from whole body FDG PET/CT, such as who should ultimately be responsible and to what extent, similar to the views of researchers involved in neuroimaging research.¹⁴ General opinions on returning incidental findings from whole body FDG PET/CT and the importance of various factors in returning those findings also varied.

To balance the potential benefit to the research participant with the burdens on the researcher, it has been recommended that researchers inform participants of incidental findings that have important implications for them, as long as such disclosure does not cause undue burdens to the researcher.²⁵ In their own words, investigators described the most common challenges to returning incidental findings (Explaining/Understanding Results, Time/Cost, Follow Up, Emotional Burden and Data Disclosure), some of which could cause undue burdens on the researcher and be in direct conflict with the need or desire to return incidental findings to participants.

Participant preferences are an increasingly important consideration in returning incidental findings from research. Participants have expressed the desire for a less paternalistic approach, and to be able to make the autonomous choice of receiving their results¹⁴ despite increased stress when being informed of an incidental finding.²⁶ Researchers have also recognized that participants’ expectations for a general checkup should be met, because this can be one of the main motivations for volunteers to participate in a study.¹⁵ In our study, researchers had mixed opinions on the need to actively search for incidental findings from whole body FDG PET/CT, similar to the opinions of researchers involved in neuroimaging trials.¹⁵ One view is that researchers do not have a moral duty to actively look for incidental findings; however, if researchers have a moral obligation to return incidental findings that would benefit participants, one could argue they also have a moral duty to search for these findings.²⁷

Our study has some limitations. Although most of the TARGET investigators did respond to the survey, we did not capture the views and beliefs of the investigators who did not respond; the answers of non-responders may be different than the responders. Our survey used different question types to capture data as fully and accurately as possible, but it still could have missed information that would have been obtained through other qualitative methods such as an interview or focus group. However, we supplemented our survey with free-text portions to obtain qualitative information that may not have been captured from the quantitative survey sections. Although researchers’ views are important, participant views also need to be considered in the reporting and managing of incidental findings from whole body FDG PET/CT and these were not collected in this study.

In terms of regulatory considerations, “the HHS and FDA regulations (45 CFR 46 and 21 CFR 50 and 56, respectively) are silent regarding the return of incidental findings to research subjects, and neither directly require nor disallow this activity.”²⁸ Although it is recommended that researchers report and manage incidental findings from research²⁹ and include the procedures for handling incidental findings in the protocol and consent documents,²⁸ our data show differing views and uncertainties on what and how to return, and the extent of follow up needed to manage, incidental findings from whole body FDG PET/CT. These ambiguities highlight the need for more specific and standardized guidance. One theme that emerged as a strong motivation for returning incidental findings was the effect on the patient-physician relationship. To maintain the public’s trust in the patient-physician relationship and confidence in the research enterprise, investigators should consider participant preferences in receiving their incidental research results, and IRBs may need to provide more oversight and guidance in the return and management of incidental findings from research.

Future Directions

For incidental findings from whole body FDG PET/CT that are not clearly actionable or less straightforward, IRBs may consider requiring a certain level of follow up for different categories or types of incidental findings or require that all incidental findings are reviewed by an independent group that would provide timely recommendations on the most appropriate return and management of those findings. With IRB guidance, very specific and detailed policies and procedures for returning and managing incidental findings should be established for every study, with consistency among the research sites of multicenter trials.

Significance and Innovations.

• To our knowledge, this is the first study that examines researchers’ attitudes and beliefs regarding incidental research findings from whole body 18F-fluorodeoxyglucose-positron emission tomography/computed tomography (FDG PET/CT).

• Although most investigators felt that incidental findings found on whole body FDG PET/CT during the Treatments Against Rheumatoid Arthritis and Effect on FDG PET/CT trial were beneficial in some way, there were concerns with returning these incidental findings, including participants undergoing invasive workups which may not lead to clinical benefit or may even lead to clinical or emotional harm.

• Regarding what to report from whole body FDG PET/CT, there was strong agreement in the types of incidental findings that needed to be returned to participants, including those of clinical relevance or that are clinically actionable, including preventive measures. However, for incidental findings that do not provide a clear, clinically actionable benefit or when there are no findings, there was more disagreement among investigators on whether to return these results to participants.

• Although it is recommended that researchers report and manage incidental research findings, our data show differing views and uncertainties on what and how to return, and the extent of follow up needed to manage, incidental findings from whole body FDG PET/CT; this highlights the need for more specific and standardized guidance.