Cardiovascular Eligibility Criteria and Adverse Event Reporting in Combined Immune Checkpoint and VEGF Inhibitor Trials

Background Combination therapy with immune checkpoint inhibitors (ICIs) and vascular endothelial growth factor inhibitors (VEGFIs) has improved cancer outcomes and is increasingly used. These drug classes are associated with cardiovascular toxicities when used alone, but heterogeneity in trial design and reporting may limit knowledge of toxicities in patients receiving these in combination. Objectives The aim of this study was to assess consistency and clarity in definitions and reporting of cardiovascular eligibility criteria, baseline characteristics, and cardiovascular adverse events in ICI and VEGFI combination trials. Methods A scoping review was conducted of phase 2 to 4 randomized controlled trials of ICI and VEGFI combination therapy for solid tumors. Trial cardiovascular eligibility criteria and baseline cardiovascular characteristic reporting in trial publications was assessed, and cardiovascular adverse event definitions and reporting criteria were also examined. Results Seventeen trials (N = 10,313; published 2018-2022) were included. There were multiple cardiovascular exclusion criteria in 15 trials. No primary trial publication reported baseline cardiovascular characteristics. Thirteen trials excluded patients with prior heart failure, myocardial infarction, hypertension, or stroke. There was heterogeneity in defining cardiovascular conditions. “Grade 1 to 4” cardiovascular adverse events were reported when incidence was ≥5% to 25% in 15 trials. Incident hypertension was recorded in all trials, but other cardiovascular events were not consistently reported. No trial specifically noted the absence of cardiovascular events. Conclusions In ICI and VEGFI combination trials, there is heterogeneity in cardiovascular exclusion criteria, reporting of baseline characteristics, and reporting of cardiovascular adverse events. This limits an optimal understanding of the incidence and severity of events relating to these combinations. Better standardization of these elements should be pursued. (Exclusions and Representation of Patients With Kidney Disease and Cardiovascular Disease in Drug Trials of the Novel Systemic Anti-Cancer Therapies VEGF-Signalling Pathway Inhibitors Alone or in Combination With Immune Checkpoint Inhibitors; CRD42022337942)

T here is a high prevalence of cardio- vascular disease (CVD) among patients with cancer. 1 The incidence of cardiovascular (CV) events, such as myocardial infarction (MI) and ischemic stroke, is higher in patients with cancer than it is in those without cancer. 2As clinical outcomes for people diagnosed with cancer have improved considerably over the past 2 decades, the competing risks from CV comorbidity and mortality have gained increasing relevance. 3erapies such as immune checkpoint inhibitors (ICIs) and vascular endothelial growth factor inhibitors (VEGFIs) have improved cancer outcomes for patients with a variety of tumor types. 3,4When used alone, ICIs are associated with a range of CV adverse events (CVAEs) including myocarditis, MI, and ischemic stroke. 5,6[9] The use of ICIs and VEGFIs in combination is now a common treatment regimen in various cancer types, including melanoma, renal cancer, cervical cancer, and endometrial cancer. 10This is a consequence of successful trials of combinations of ICIs and VEGFIs conducted over the past 5 years, with more than 90 ongoing clinical trials of ICI and VEGFI combination regimens. 4,11Six combination ICI and VEGFI treatments are currently approved by the U.S.
Food and Drug Administration. 10Given the CVAEs seen with each of these drugs in isolation, understanding the potential for an increased incidence of these effects when the drugs are combined is of major importance.
There is limited understanding of the extent to which pre-existing CVD increases the risk for ICI and VEGFI CV toxicity.To understand these issues, it is imperative to have clarity about the representation of patients with or without pre-existing CVD in trials.
Understanding and limiting heterogeneity among trial populations is required for subsequent robust meta-analysis of CVAEs.Furthermore, consistency and clarity of definitions and trial publication reporting of CVAEs are fundamental to achieving these aims.
We conducted a scoping review of randomized controlled trials of ICI and VEGFI combination therapy in patients with cancer.Our primary interests were trial CV exclusion criteria and the heterogeneity of these across trials.We also examined reporting of baseline CV characteristics and methods by which adverse events (AEs) were defined, adjudicated, and reported in trial results publications.

METHODS
This scoping review protocol was registered with the International Prospective Register of Systematic Reviews (CRD42022337942) and used the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement guidance. 12We used the population, intervention, comparison, and outcome criteria for inclusion (Supplemental Table 1).The registered protocol also included assessments relating to nephrology-related inclusions and trial reporting, and these findings have been published separately. 13As this was a review of publicly available data, no ethics approval was required.OUTCOMES.Key trial characteristics, trial eligibility criteria, and exclusion criteria relating to CVD were extracted.Trial design characteristics relating to the assessment and adjudication of CVAEs and the extent of their reporting within the published paper were recorded.Data were extracted from the original publication, supplemental material, and available protocols from the journal website.Trial registration numbers, identified from the publications, were used to search relevant clinical trial platforms to ensure that all relevant publicly available protocol data were identified if they were not available from the publication.
CVAEs.An AE was defined as a CVAE if it was recorded as a cardiac disorder under the Common Terminology Criteria for Adverse Events (CTCAE) criteria.The CTCAE grade AE severity on a scale of 1 to 5. Grade 1 events are considered "mild" and grade 2 "moderate."Grade 3 events are considered "severe or medically significant but not immediately lifethreatening," while grade 4 events are those with lifethreatening consequences.Death was recorded as grade 5. CVAEs were grouped in similar categories, and of note, MI and "acute coronary syndrome" were reported together under the AE "MI" category.If the AE was not recorded as a cardiac disorder per CTCAE but fulfilled any prespecified trial criteria for CV and stroke endpoints for clinical trials, on the basis of the Food and Drug Administration-endorsed Hicks criteria (such as "sudden death"), it was also classified as a CVAE. 14
CV ELIGIBILITY CRITERIA.Eligibility criteria were available for all 17 trials.CVD trial exclusion criteria were broad, with heterogenous definitions (Figure 2).Of the 4,893 references extracted, 17 studies were included in the final analysis.Full search terms can be found in Supplemental Table 2.There were exceptions to allow the inclusion of participants with prior HF.In 4 trials, participants with HF who did not meet prespecified NYHA exclusion criteria, as well as participants with LVEFs <50%, were eligible to enroll provided they were on a stable regimen that was optimized in the opinion of the physician.
C o r o n a r y a r t e r y d i s e a s e .The 14 trials with LVSD and HF exclusions also excluded patients with histories of recent MI or unstable angina (Table 1).The time frame for exclusion of patients with prior acute coronary syndrome varied from 3 to 12 months prior to screening.In addition to exclusions on the basis of

D u r a t i o n o f A E r e p o r t i n g . Follow-up for CV events
was shorter than the trial duration in all trials (Table 2).
Follow-up for CV events in 5 trials was "the duration of treatment plus 30 days after last dose."In 9 trials follow-up for CVAEs was "duration of treatment plus 30 days or the initiation of new anticancer therapy, whichever came first."Ten trials had extended followup for serious AEs and AEs of special interest (AEOSIs), including CVAEs, ranging from 90 to 120 days.The follow-up period was not specified in 2 trials.

I n c i d e n c e t h r e s h o l d s f o r A E r e p o r t i n g . No phase
3 trial reported all CV events.Fifteen trials reported events when they reached prespecified incidence thresholds (Table 2).The most common threshold reported in the main paper was $10% in 6 trials  4).
CV EVENTS.No trial used the Food and Drug Administration-endorsed, standardized Hicks criteria for reporting of CV events. 14With the exception of hypertension, which was reported in all trials, no trial explicitly stated the absence or occurrence of CVAEs.
In trial papers that did not report CVAEs other than hypertension, it was not clear whether this was because of a true absence of CVAEs or because of their occurrence with an incidence beneath a reporting threshold.
C V d e a t h ."AE deaths of any attribution" were reported in 11 trials (N ¼ 7,203), and 7 of these (N ¼ 4,734) reported the mode of death.In 6 trials (N ¼ 3,110), only deaths that were considered to be treatment related (adjudicated by the investigator) were reported.
Most frequently, CVAEs were described when associated with death.No trial reported the total number of CV deaths.However, 10 trials (N ¼ 7,737) reported AE deaths that would be categorized as CV deaths according to the Hicks criteria (Figure 3).MI.MI was reported in only 4 trials (N ¼ 3,181 [31%]), 2 of which reported only fatal MI (Figure 3  In ICI and VEGFI combination trials, reporting of cardiovascular (CV) adverse events was variable.With the exception of blood pressure, which was reported in all trials, serious adverse events such as myocardial infarction (MI) were reported in only 4 trials and heart failure in 3 trials.
No trial reported the absence of events.
reversible encephalopathy syndrome and another secondary to "uncontrolled hypertension" adjudicated by the investigator.
O t h e r t h r o m b o t i c e v e n t s .Venous thrombotic events were reported in 6 trials (N ¼ 5,309 [52%]), but 4 of these reported only thrombotic events that resulted in death.Four trials (N ¼ 3,599 [35%]) reported arterial thrombotic events, and 3 trials  reported in any primary trial publication.However, a secondary analysis of 1 trial did report the prevalence of baseline CV risk factors. 19In that trial, the baseline prevalence of CV risk factors was low.Only 4% of patients in the ICI and VEGFI arm had dyslipidemia, 9.5% had diabetes, and 3.2% had cerebrovascular disease. 19In addition to potentially stringent trial eligibility criteria, trial recruitment bias toward the inclusion of patients with fewer comorbidities may contribute to a trial population that is not representative of the general population of patients with cancer in whom these drugs may ultimately be used.
Irrespective of these issues of eligibility and potential recruitment bias, the lack of data on baseline CV characteristics means that the baseline CV risk for patients in these trials is unknown.Inclusion of those with comorbidities, when assessed in noncancer trials, only modestly affected the completion of study enrollment, meaning that there could be an increase in the generalizability of trial data with minimal impact on trial completion. 20Without this information, it is impossible to assess the degree to which pre-existing CVD or risk factors may potentiate adverse CV effects of ICI and VEGFI therapy.It also remains possible that an interaction between preexisting CVD and adverse effects of ICI and VEGFI therapy is lower than might otherwise be expected.
These insights are critical for providing patients with CVAE DESCRIPTION AND REPORTING.CVAEs were reported using CTCAE in all trials, and their reporting was based upon incidence thresholds.The threshold that was required to be reached varied from 5% to 25% among trials.Furthermore, only 4 trials used lower reporting incidence thresholds for more severe (CTCAE grade $3).In addition to the standardization of reporting methods, lowering or potentially removing this threshold for reporting in primary trial publications altogether should be considered.
Although the signal-to-noise ratio of grade 1 and 2 events may mean that reporting on the basis of incidence thresholds could be appropriate, we would argue that reporting of all of the more severe AEs may be justified.Irrespective, reporting of events of special interest of any severity should continue for conditions such as myocarditis, for which the most granular information is required to understand whether there may be a potential disconnect between initial CTCAE severity grading and outcomes.
Trial publication reporting of CVAEs, and the clarity of this, was variable.Although many primary trial publications did not report the occurrence of CVAEs, they also did not explicitly state their absence.Reporting of AEs that were specifically SEARCH STRATEGY.The search was conducted in MEDLINE, Embase, and the Cochrane Library on May 20, 2022.All trials published in the public domain until the time of data extraction were eligible for analysis.The search terms are included in Supplemental Table 2. Duplicates were removed.Relevant papers were identified by 2 independent reviewers (B.E.and S.R.).Disagreements were resolved by consensus with a third reviewer (J.S.L.).STUDY ELIGIBILITY CRITERIA.A systematic search of the literature was conducted to identify clinical trials of combination ICI and VEGFI therapy.We included any trial conducted among adult patients with any solid organ cancers who received combination ICI and VEGFI therapy in either the intervention or the control arm.ICIs and VEGFIs that were not approved by the Food and Drug Administration for use as anticancer treatments at the time of data extraction were excluded.Trials using only single dosing or sequential (nonconcurrent) ICI and VEGFI therapy were excluded.INCLUSION AND EXCLUSION CRITERIA.We included all phase 2 to 4 randomized controlled trial with a A B B R E V I A T I O N S A N D A C R O N Y M S AE = adverse event AEOSI = adverse event of special interest disease HF = heart failure ICI = immune checkpoint inhibitor LVEF = left ventricular ejection fraction LVSD = left ventricular systolic dysfunction MI = myocardial infarction VEGFI = vascular endothelial growth factor inhibitor The authors attest they are in compliance with human studies committees and animal welfare regulations of the authors' institutions and Food and Drug Administration guidelines, including patient consent where appropriate.For more information, visit the Author Center.Manuscript received November 3, 2023; accepted December 12, 2023.minimum of 20 participants with available results published at time of extraction.Nonrandomized controlled trials, meta-analyses, review papers, commentaries, subsequent therapy analyses, costeffectiveness analyses, published abstracts, patientreported outcomes, subgroup analyses, and retrospective analyses were excluded.If 2 published papers reported data from the same patient group, such as subgroup analyses and extended follow-up analyses, the original publication was used.

Fifteen
trials (N ¼ 9,389 [91%]) had multiple CV exclusion criteria.Of these, there were specific exclusion criteria for patients with prior HF, MI or unstable angina, hypertension, and stroke in 13 trials (N ¼ 9,283 [90%]).Two of the 15 trials (N ¼ 106 [1%]) had a general exclusion criterion of "clinically significant CVD or impairment."The remaining 2 trials (N ¼ 924 [9%]) did not explicitly exclude patients on the basis of prior CVD but had a general criterion excluding those with "a relevant prior condition that may affect the results of the trial."The interpretation of these more generic criteria was left to the discretion of the investigator.

FIGURE 1
FIGURE 1 Preferred Reporting Items for Systematic Reviews and Meta-Analyses Diagram In the 12 trials reporting eligibility data prior to enrollment, 31% of participants (n ¼ 3,905) were ineligible.Only 1 paper reported reasons for screening failure, and in that publication, 10% of those ineligible were excluded because of CV exclusions (pulmonary embolism or deep vein thrombosis, hypertension, corrected QT interval, and "CV conditions").H F a n d L V S D .Of the 14 trials (N ¼ 9,309 [88%]) with specific exclusions for patients with HF, 7 excluded those with NYHA functional class $II, 6 trials excluded those with NYHA functional class $III, and 1 trial excluded "symptomatic" patients (Table 1).Eight of the trials' HF exclusions specified HF within varying time frames prior to enrollment, ranging from 3 to 12 months prior to screening.Patients with reduced left ventricular ejection fraction (LVEF) were excluded from 8 trials (N ¼ 7,156 [69%]): 5 excluded those with LVEFs <50% (although 3 of these accepted LVEF <50% if the participant was "stable on a medical regimen that was optimized in the opinion of the physician") and 3 excluded patients with LVEFs less than the "lower limit of normal" of the "institutional normal range."Only 4 trials (N ¼ 3,433 [33%]) mandated echocardiography before enrollment for all participants.Three other trials (N ¼ 1,909 [19%]) mandated LVEF assessment prior to enrollment in specific circumstances (for patients with anthracycline exposure in 1 trial and if a patient had "cardiac risk factors or abnormal electrocardiographic findings" in the remaining 2 trials).

FIGURE 2
FIGURE 2 Cardiovascular Exclusion Criteria in Immune Checkpoint Inhibitor and Vascular Endothelial Growth Factor Inhibitor Combination Therapy Trials

FIGURE 3
FIGURE 3 Percentage of Trials Reporting Cardiovascular Adverse Events CV TRIAL ELIGIBILITY CRITERIA HETEROGENEITY.Our review identified that CV exclusion criteria were ubiquitous in these trials.We also identified substantial heterogeneity in the nature of these exclusion criteria and the use of potentially arbitrary CV definitions and exclusion thresholds.It is of note that the Food and Drug Administration recommend the avoidance of "unnecessarily restrictive eligibility criteria" to maximize the generalizability of trial results to the patient population in which the drug may be used in subsequent routine clinical practice.18This recommendation was made particularly to allow trials to inform the net risk/benefit profile.Although we acknowledge that it may be appropriate to include some clinically relevant CV eligibility criteria for trial safety reasons, and although these trials were designed and powered to provide information on cancer treatment effects, potential safety signals may become apparent only when trial populations are combined for meta-analysis.Those insights are currently limited by heterogeneity in eligibility criteria.BASELINE CVD AND CVD RISK FACTORS IN TRIAL PARTICIPANTS.The baseline prevalence of CVD, including CV risk factors or established CVD, was not

J 9 CV
A C C : C A R D I O O N C O L O G Y , V O L .6 , N O . 2 Eligibility and Reporting in ICI/VEGFI Trials the best information relating to potential risks of treatment in the context of pre-existing CVD.

4 CV
considered to have been related to treatment was more frequent than reporting of AEs of any attribution.Reporting of hypertension and, to a lesser extent, myocarditis was common in the context of already well-recognized associations with VEGFIs and ICIs, respectively.However, without consistently robust assessment and reporting of other CVAEs, the ability to discern associations (or the lack thereof) between these drugs and a broader range of potential CVAEs will remain suboptimal.The assessment of CVAE "treatment-relatedness" was by the local investigator, which introduces bias and impedes transparent understanding of AE profiles.One trial included a prespecified subgroup analysis of CV CENTRAL ILLUSTRATION Heterogeneous Cardiovascular Eligibility and Event Reporting in ICI and VEGFI Combination Trials Rankin S, et al.J Am Coll Cardiol CardioOnc.2024;6(2):267-279.In contemporary trials with "state of the art" trial design, such as trials of combined immune checkpoint inhibitor (ICI) and vascular endothelial growth factor inhibitor (VEGFI) therapy, there is marked heterogeneity in definitions of cardiovascular (CV) disease for exclusion criteria and in adverse event reporting.No trial reported CV baseline characteristics or reported the absence of CV events.CV adverse events were reported only when a threshold incidence within the trial population was reached, which is likely to lead to underreporting of CV events.HF ¼ heart failure; HTN ¼ hypertension; MI ¼ myocardial infarction; LVSD ¼ left ventricular systolic dysfunction.Rankin et alJ A C C : C A R D I O O N C O L O G Y , V O L .6 , N O . 2 , 2 0 2 Eligibility and Reporting in ICI/VEGFI Trials

TABLE 1
Randomized Controlled Trials of Immune Checkpoint Inhibitor and Vascular Endothelial Growth Factor Inhibitor Combination Therapy: Exclusion Criteria

TABLE 1
a Within 6 months.b Within 3 months.c Such as aortic aneurysm, dissection or carotid stenosis that requires surgical intervention or stenting, or recent peripheral arterial thrombosis.d Within 12 months.e LVEF <50% acceptable if stabilized on optimal medical therapy in the opinion of the treating physician.CABG ¼ coronary artery bypass graft; CTCAE ¼ Common Terminology Criteria for Adverse Events; DVT ¼ deep vein thrombosis; LLN ¼ lower limit of normal; LVEF ¼ left ventricular ejection fraction; MI ¼ myocardial infarction; PE ¼ pulmonary embolism; QTc ¼ corrected QT; RCT ¼ randomized controlled trial; TIA ¼ transient ischemic attack; VTE ¼ venous thromboembolism.

Table 1
lopathy or crisis.The second trial did not have a specific blood pressure cutoff but excluded participants randomized to receive bevacizumab if they had histories of hypertensive emergency or hypertensive encephalopathy.Any history of hypertensive encephalopathy or crisis was an exclusion criterion in 8 trials (N ¼ 4,463 [43%]).S t r o k e .Previous "cerebrovascular accident" or transient ischemic attack within 3 to 12 months of screening was an exclusion criterion in 14 trials (N ¼ 9,309 [90%]).ceding 6 months" or previous "CTCAE grade 4 venous thromboembolism" in 2 trials.C o r r e c t e d Q T i n t e r v a l a n d a r r h y t h m i a .Patients with arrhythmias were excluded from 10 trials (N ¼ 6,482 [63%])."Unstable" or "hemodynamically significant" arrhythmia was the most common exclusion terminology, but "grade $2," "uncontrolled," and "clinically significant" arrhythmias were used to REPORTING OF AEs.All 17 trials reported AEs using CTCAE definitions and severity grading.CTCAE version 4 was used in 15 trials.AEs were reported by the site investigator with no central or CV specialist event adjudication in 14 trials; this was not specified in the remaining 3 trials.One trial had an independent CV event adjudication committee.AEs were reported either as treatment related or "AEs of any attribution."Treatment-related AEs (adjudicated by the investigator) were reported in all trials.AEs of any attribution were less commonly reported (11 trials, N ¼ 7,458 [72%]).