Bringing doctors together

or

A free, private network for verified physicians

Leading voices

Dr. William Aird

Classical Hematology · BIDMC

View

Dr. Cynthia Dunbar

Classical Hematology · NIH

View

Dr. Ariela Marshall

Hemostasis & Thrombosis · University of Minnesota

View

Dr. Philip Schauer

Bariatric Surgery · PBRC

View

Dr. Justin Hovey, FAAP, FACP, DABOM

Obesity Medicine · Noland Hospital Dothan II

View

Dr. Kimberly Gudzune

Obesity Medicine · JHU

View

Dr. Richard Dutton

General Anesthesiology · UMSOM

View

Dr. Sarah Ro

Obesity Medicine · UNC Health Care

View

Dr. Rebecca Andrews

Primary Care · UCHC

View

Dr. Jed Wolpaw

General Anesthesiology · Johns Hopkins

View
  • Discover your community

  • Stay current

  • A profile that represents you

  • Find your people

About RoonPrivacyTerms

    Overview

    Endothelium as a Source of Cardiovascular Toxicity From Antitumor Kinase Inhibitors.

    Aug 15, 2024DOI: 10.1161/atvbaha.124.319864Journal ArticleAHA

    Written by: Richard J TraversAlec StepanianIris Z Jaffe

    Find these authors on Roon

    RT
    Dr. Richard Travers
    Classical Hematology · Tufts

    Introduction

    Cancer and cardiovascular disease, the two leading causes of death, have overlapping risk factors and competing pathophysiologic mechanisms.1 There has been rapid development of new anti-cancer therapies that are oral kinase inhibitors (KIs) directly targeting cellular processes necessary for cancer development and progression.2 Although these more specific drugs have fewer side effects than traditional chemotherapy, cardiovascular side effects remain common. Cancer treatments can exacerbate cardiovascular disease by impacting risk factors or by directly damaging the heart and vasculature.3 Endothelial cells (ECs) line the lumen of all blood vessels and as such are exposed to all systemically administered cancer therapies. This review summarizes recent advances in our understanding of the impact of KIs specifically on ECs and how this may contribute to the cardiovascular toxicities of novel cancer treatments. The two classes of KIs with the most evidence for EC dysfunction manifesting as cardiovascular side effects are the vascular endothelial growth factor receptor inhibitors (VEGFRIs), which induce hypertension, and the BCR-Abl KIs, which cause acute arterial thrombosis.4 As clinical experience with other KIs expands, the evidence base tying drug-induced EC impact to cardiovascular risk is also growing. Current understanding of EC toxicity mechanisms is summarized for the; 1) BCR-Abl KIs, 2) VEGFRIs, 3) Bruton’s tyrosine kinase inhibitors (BTKIs) and 4) emerging new anti-cancer KIs. For each drug class, the clinically evident vascular toxicities are first described and the current understanding of the impact of each KI class on EC function is reviewed, with emphasis on studies from the past 5 years since this topic was last reviewed.5–7.The discovery that chronic myeloid leukemia (CML) is caused by the Philadelphia Chromosomal translocation, producing a constitutively active BCR-Abl kinase that drives leukocyte proliferation, lead to the development of the first molecularly targeted cancer therapy, imatinib. The growing class of small molecule BCR-Abl tyrosine KIs has transformed this once lethal leukemia into a chronic disease, as CML 5-year survival improved from less than 50% to over 90%. As imatinib resistance has developed,8 more potent second (dasatinib, nilotinib, bosutinib) and third generation (ponatinib) agents were developed that more effectively induce cancer remission, and hence are now first line therapies (Table 1). Resistance is typically due to Abl kinase mutations. Indeed, ponatinib was the only therapy for patients with the T315I mutation until asciminib, an allosteric inhibitor, was approved in 2021.While imatinib has an excellent cardiovascular safety profile,9,10 the newer generation BCR-Abl inhibitors, nilotinib, dasatinib, and ponatinib, are associated with a 3- to 4-fold increased risk of life-threatening arterial thrombotic events including heart attack, stroke, and acute limb ischemia. In the PACE trial, 31% of participants randomized to ponatinib experienced arterial thrombotic events,11 and more recently, addition of ponatinib to chemotherapy resulted in several deaths from myocardial infarction, necessitating dose reduction.12 Similarly, nilotinib, has been associated with severe peripheral arterial occlusion in up to 17% of patients.13 Together, a meta-analysis revealed that this increase in cardiovascular risk nullifies the survival benefit from the improved cancer efficacy of these newer BCR-Abl TKIs.14In addition to promoting arterial thrombosis, 30% of dasatinib treated patients experienced pleural effusions. Of these patients, 5% also developed pulmonary arterial hypertension during extended follow-up.15 In addition to dasatinib, ponatinib and bosutinib have also been implicated in pulmonary hypertension.16 There is a known association between endothelial cell dysfunction and pulmonary hypertension that likely contributes to this side effect of some BCR-Abl KIs.17 Furthermore, ponatinib and nilotinib are also associated with increased risk of arrhythmia, specifically atrial fibrillation. Increasing evidence suggests that endothelial dysfunction may contribute to atrial fibrillation which could also further contribute to arterial thrombotic events via embolization.18,19In healthy vessels, ECs form an anti-thrombotic, anti-inflammatory, and pro-vasodilatory monolayer between circulating blood and underlying smooth muscle cells. Exposure to risk factors like smoking, diabetes, and hypercholesterolemia, induces endothelial damage. Endothelial dysfunction is the first step in the development of atherosclerosis which may culminate in plaque rupture and arterial thrombosis, the typical cause of heart attack and stroke. Early studies examining the effects of BCR-Abl TKIs on platelet function appeared contradictory (reviewed in 15). One study demonstrated that ponatinib given to mice in vivo reduced bleeding time and elevated platelet glycoprotein-VI activation, supporting increased platelet activation and reactivity.20 However, another study demonstrated that ponatinib-treatment in vitro reduced platelet activation, spreading, granule secretion and aggregation.21 The apparent discrepancy may be clarified by an elegant study by Alhawiti et al. showing that blood from nilotinib-treated mice demonstrated increased platelet adhesion ex vivo, whereas nilotinib had no effect on platelet aggregation when platelets were directly treated in vitro. These data suggest that other supporting cells are needed for BCR-Abl TKIs to induce platelet activation in vivo, potentially implicating the endothelium as one such mediator.The damaged endothelium expresses platelet-activating factors on the cell surface, including P-selectin. In addition, when ECs die or are denuded, the basement membrane, which includes platelet-activating proteins such as von Willebrand factor (VWF), is exposed to blood components.22 A recent study demonstrated that ponatinib and dasatinib, but not imatinib, nilotinib and asciminib, increase EC expression of P-selectin.23 Ponatinib also induced thrombotic microangiopathy and VWF secretion in ApoE knockout mice.24 Nilotinib treatment of coronary artery ECs induced secretion of tissue plasminogen activator, plasminogen activator inhibitor-1, and VWF in a dose dependent manner.25 In most published studies imatinib does not affect EC viability, apoptosis, or wound healing.5,26–29 Only one publication reported that imatinib induced apoptosis of ECs, but that study used the EA.hy926 cell line30, rather than primary human ECs as in other studies. Bosutinib also did not induce EC damage in vitro31 and has minimal cardiovascular side effects in patients.14 In contrast, ponatinib, nilotinib, and dasatinib reduce EC proliferation, impair wound healing, and induce EC death by apoptosis or necrosis.23,25,28,32 Overall, the evidence suggests that the BCR-Abl TKIs associated with thrombosis in humans, damage ECs leading to a prothrombotic EC phenotype.As the endothelium becomes damaged, the protective effects of EC-derived nitric oxide are lost. When human adipose arterioles were incubated ex vivo with nilotinib or imatinib, there was a shift in the mediators of endothelial-dependent vasodilation from vaso-protective nitric oxide to pro-inflammatory hydrogen peroxide,33 similar to what has been demonstrated in coronary arteries of patients with coronary heart disease.34 Thus, these BCR-Able TKIs may contribute to microvascular dysfunction in the heart and other tissues. As ECs injury progresses, cell adhesion molecules (CAMs) expression is induced, leading to leukocyte recruitment to the vessel that can contribute to atherosclerotic plaque inflammation, the cause of plaque rupture and arterial thrombosis. Ponatinib and dasatinib increase EC expression of ICAM-1 and VCAM-1 in vitro23 and dasatinib-treated rats and patients had increased circulating ICAM-1, VCAM-1, and E-selectin.35 The effects of nilotinib are less clear, with studies showing variable effects on EC adhesion molecule expression in vitro but increased thrombosis and atherosclerosis in vivo in mouse models.23,36,37 One explanation for the disparity may be that only nilotinib induces insulin resistance and dyslipidemia, and thus may impact EC function in vivo secondary to alterations in traditional cardiovascular risk factors, which are not modeled by isolated ECs in vitro.38,39Injured ECs alter cell-cell junctions and become permeable, allowing recruited leukocytes to transmigrate into the sub-endothelial space, but also contributing to vessel leakage. Clinically, only dasatinib induces pleural effusions, which could be secondary to excessive endothelial permeability.40,41 A rat model of pleural effusion revealed dasatinib increases pulmonary endothelial permeability in a ROS-dependent manner and induces apoptosis of pulmonary ECs in vitro.35,42 Dasatinib also induced actin disorganization and activation of the RhoA and rho-kinase pathways in HUVECs and human microvascular ECs in vitro, resulting in pulmonary arterial hypertension in vivo.43,44 Rho-kinase inhibitors have been shown to reduce the dasatini- induced increase in mean pulmonary arterial pressure and lung weight in mice.45 A recent study using human engineered microvessels lined with human umbilical vein ECs (HUVECs) demonstrated that only dasatinib (not nilotinib, ponatinib, imatinib or asciminib) induced actin cytoskeletal rearrangement with loss of cell-cell junctions and increased vessel permeability.23 In contrast, imatinib has been shown to attenuate acute lung injury in a mouse model, and may even be protective in the setting of inflammatory vascular leak.46The understanding that tumor-derived VEGF47 promotes tumor angiogenesis and hence cancer growth and metastasis resulted in the development of anti-VEGF cancer therapies. 48 Multiple small molecule inhibitors of the VEGFR kinase domain are approved to treat solid tumors (Table 1).49 VEGFRIs are used alone and in combination with chemotherapy and immunotherapy.50 Since much of the cancer efficacy of VEGFRIs derives from preventing tumor angiogenesis, which is initiated by ECs, this class is unique in that EC toxicity is integral to their effectiveness against cancer. Indeed, preclinical data shows that axitinib, sorafenib, sunitinib and anlotinib, are all more effective at killing ECs than tumor cells.51–53Hypertension is the most common cardiovascular side effect of VEGFRIs.54–56 Almost all patients treated with VEGFRIs will have an increase in blood pressure (BP). The reported incidence of VEGFRI-induced hypertension varies widely due to differences in baseline BP in the population studied and the criteria used to define hypertension. For example, early studies focused on renal cell carcinoma patients, a group enriched for BP dysregulation due to prior nephrectomy, showed incidences of hypertension as high at 66% when treated with sunitinib.56 More recent meta-analyses including diverse VEGFRIs and cancer types revealed an incidence of new onset hypertension of 23–28% and of high grade hypertension of 1.8–4.4%.54,55 Since patients treated with VEGFRIs frequently have pre-existing hypertension, the combined incidence of new and worsening hypertension is up to 80% in some studies.57Cardiac toxicity due to VEGFRI therapy occurs less commonly, with a reported incidence of 3–8%,49,58 although newer studies that routinely screen for cardiac dysfunction without clinical evidence of heart failure suggest a higher prevalence. A small single center prospective trial of patients on sunitinib showed that 8% of patients developed asymptomatic cardiac dysfunction and more than 30% developed changes in global longitudinal strain by echocardiography, an early marker of systolic dysfunction.59As much is known about the role of VEGF signaling in healthy EC function, substantial mechanistic insight into VEGFRI-induced cardiovascular toxicity has been generated and was recently reviewed.60 Thus, the area is briefly reviewed here with a focus on more recent studies.ECs produce paracrine factors that contribute to BP regulation by modulating resistance arteriolar smooth muscle cell vasoconstriction and vasorelaxation. This impacts arteriolar diameter and myogenic tone to contribute to peripheral vascular resistance, an important determinant of systemic blood pressure.61 EC-derived vasodilators include nitric oxide (NO), prostacyclins, and endothelial-derived hyperpolarizing ions. ECs can also produce reactive oxygen species (ROS), including hydrogen peroxide, which under some circumstances acts as an alternative vasodilator, or superoxide, which can prevent vasodilation by inactivating NO.61 Indeed, the interplay between eNOS activity, ROS, and vascular dysfunction in health and disease is complex and multifaceted (reviewed by Gutterman et al. 62).VEGF signaling in ECs induces endothelial nitric oxide synthase (eNOS) expression and activity which promotes vasodilation, EC survival, and angiogenesis.63 Conversely, VEGFRIs decrease eNOS expression and activity, lower NO production, and increase markers of oxidative stress in ECs.64 Whether this mechanism is sufficient to cause VEGFRI-induced hypertension is less clear. Supporting data includes polymorphisms in eNOS that associated with sunitinib-related hypertension65 and a small case series in which NO donors reduced BP in bevacizumab-treated patients.66 However, in a swine model of sunitinib-induced hypertension, sunitinib was able to increase BP, even after infusion of a nitric oxide synthase inhibitor and ROS scavenging antioxidants.67 In a mouse model, sunitinib reduced EC-dependent vasodilation in mesenteric resistance arteries, while in humans with sunitinib-induced hypertension, there was no change in endothelial-dependent vasodilation in the brachial artery, suggesting that least large vessel EC function remained intact, although more detailed investigations of microvascular function in humans on VEGFRi’s are needed.68 Overall, the available literature supports that VEGFRIs induce endothelial dysfunction characterized by reduced NO availability, but whether that alone explains the development of hypertension remains to be clarified in vivo and in cancer patients.ECs also release and respond to vasoconstrictors including endothelin-1 (ET), angiotensin-II and thromboxane.61 Cancer patients that developed hypertension on sunitinib have increased circulating levels of ET.69,70 Similarly, in swine and rat models, ET receptor blockers reversed VEGFRI-mediated hypertension.67,71,72 ECs also release membrane bound microparticles in response to cell damage and during apoptosis. EC microparticles have been implicated not only as biomarkers of EC dysfunction but also as potential pathogenic mediators of hypertension.73 Patients treated with VEGFRIs have elevated levels of EC-microparticles which when used to stimulate human aortic ECs in vitro, induced hallmarks of VEGFRI-induced EC dysfunction, including increased ET, decreased eNOS activity and increased oxidative stress, all of which were mitigated by treatment with ET receptor blockers.74 EC-derived microparticles have also been found to play a role in the beneficial tumor response to VEGFRI therapy.75VEGFRIs also cause damage to glomerular ECs in the kidney, resulting in proteinuria and impaired renal function, which can also contribute to hypertension (reviewed by Van Wynsberghe et al76). Pre-clinical studies in rats and in human cells reveal that VEGFRIs increase renal oxidative stress, decrease NO production and alter podocyte cytoskeleton formation, resulting in impaired glomerular function. Furthermore, some of these effects were mitigated by ET receptor antagonism.71,77–79 Thus, VEGFRi-induced hypertension may be mediated by similar mechanisms via VEGR inhibition in glomerular and renal arteriolar ECs.While many studies focus on direct cardiomyocyte toxicity of kinase inhibitors including VEGFRIs (reviewed in 80), cardiac EC dysfunction may also contribute to cardiotoxicity of VEGFRIs, independent of development of HTN. ECs are the second most abundant cells in the heart, making up approximately 60% of non-cardiomyocyte cells.81 This rich network of capillary ECs provides oxygen and nutrients and removes waste, processes critical for normal function of highly metabolically active cardiomyocytes.82 ECs can further impact cardiomyocyte function via paracrine mediators.82 As such, EC dysfunction has been implicated in chemotherapy-mediated cardiac dysfunction.83,84 In a mouse model of VEGFRI toxicity, both bevacizumab and sunitinib induced left ventricular dysfunction, cardiac biomarker elevation and increased cardiac oxidative stress as early as one day after the onset of VEGFRI-induced hypertension, likely too soon to be secondary to pressure overload.85 Direct treatment of EC with VEGFRI in vitro induced senescence in human umbilical vein ECs86 and cardiac microvascular EC damage.87 VEGFRIs also impaired coronary microvascular pericyte function, mainly through inhibition of platelet-derived growth factor receptor signaling.88 These studies all support the concept that vascular cell toxicity along with secondary impacts of hypertension on the heart, in addition to direct cardiomyocyte toxicity, may contribute to VEGFRI-induced cardiac dysfunction. Additional exploration of the molecular mechanisms behind EC-mediated VEGFRI cardiotoxicity is warranted.BTK is crucial for B-cell survival and function,89 hence, oral small molecule BTKIs, including ibrutinib and newer more potent agents including acalabrutinib, zanubrutinib, and pirtobrutinib, are used to treat B-cell malignancies, lymphomas and Waldenstrom’s macroglobulinemia (Table 1). The first recognized cardiovascular side effects of BTKIs was atrial fibrillation (AF). A meta-analysis of randomized clinical trials revealed a cumulative incidence of AF of 13.8% at 36 months,90 however this may be an underestimate of the true incidence outside of clinical trials. For instance, one study of 53 patients on ibrutinib at a single cardio-oncology clinic demonstrated an atrial fibrillation incidence of 38%.91 New or worsening hypertension is also observed with BTKIs, although with lower frequency than with VEGFRIs. In a retrospective single center study, 36.6% of ibrutinib-treated patients had at least a 10 mmHg increase in BP which was significantly higher than patients randomized to chemoimmunotherapy.92While data directly implicating BTKI-mediated EC dysfunction in the development of AF is not yet available, multiple studies show that non-specific biomarkers of inflammation (C-reactive protein, tumor necrosis factor, monocyte chemoattractant protein-1) or EC dysfunction (interleukin-6, VWF, soluble p-selectin), are elevated in patients who develop AF with associated thromboembolism.93,94 One overarching question in the field of Cardio-Oncology is whether cardiovascular side effects of cancer therapies are linked to their anti-tumor efficacy, or due to off target effects. A recent study suggested that AF is an off target toxicity by demonstrating that mice with total body deletion of BTK remain susceptible develop inducible AF from ibrutinib exposure.95 Whole heart kinome profiling implicated ibrutinib inhibition of c-terminal SRC kinase as a potential mechanism of increased fibrosis and development of inducible AF. Since the phenotype could be re-capitulated by cardiomyocyte specific c-terminal SRC kinase knockout, the cardiomyocyte likely contributes, but future studies may consider the role of EC dysfunction in this model.Few preclinical studies have examined the direct impact of BTKIs on ECs and none in the context of hypertension. One study compared the effects of each of the approved BTKIs on HUVECs, showing differential EC toxicity. Specifically, ibrutinib caused potent cytotoxicity via apoptosis and downregulation of B-cell lymphoma 2A1 expression and inhibited angiogenesis via increased bone morphogenic protein-4 expression, while zanubrutinib caused less severe EC apoptosis, and acalabrutinib had no impact on the measured components of EC dysfunction.96 Interestingly, these findings in ECs correlate with clinical cardiotoxicity, as the more selective and non-covalent BTKIs, zanubrutinib and acalabrutinib, have less cardiovascular toxicity compared to ibrutinib.97 In vivo studies in zebrafish also show that ibrutinib disrupts vascular development and causes apoptosis of vascular ECs.98 There have been case reports of a pre-eclampsia like thrombotic microangiopathy with ibrutinib, and also with VEGFRIs, suggesting the possibility of shared EC toxicity mechanisms between BTKIs and VEGFRIs.99 These preclinical studies suggest that BTKIs are directly toxic to ECs in vitro and, as with other classes of KIs,31,100 the degree of EC toxicity correlates with the relative clinical cardiovascular toxicity. However the mechanistic link between EC toxicity of BTKIs and their cardiovascular side effects remains to be firmly established.In addition to BCR-Abl TKIs, VEGFRIs, and BTKIs, there are many other kinase inhibitors currently in clinical use (Table 1). Cyclin dependent kinase (CDK)-4/6 inhibitors used to treat breast cancer significantly increase the risk of venous thromboembolism, which is reported at 1–5% in clinical trials but up to 10% in some retrospective studies.101 While generally considered safer than traditional chemotherapy, a large retrospective study showed that almost 25% of patients on CDK-4/6 inhibitors developed new heart failure or cardiomyopathy, which was similar to the incidence in patients receiving anthracyclines.102 BRAF and MEK inhibitors used to treat melanoma show a 30% incidence of hypertension at five year follow-up.103 Similarly, selperectinib, a RET inhibitor (with some VEGFRI activity), commonly induces hypertension.104 While there have been rare cases of vascular toxicities or heart failure associated with the anaplastic lymphoma kinase (ALK) inhibitors crizotinib and alectinib, conduction abnormalities including prolonged QT interval are also well documented.105,106 While epidermal growth factor inhibitors (EGFRIs) like erlotinib are associated with rashes mediated by EC dysfunction and barrier breakdown,107 these drugs demonstrate minimal cardiovascular toxicity. This includes lapatinib, and EGFRI which also inhibits HER2, the target of the known cardiotoxic monoclonal antibody trastuzumab.108While generally understudied, emerging evidence reveals that some of these newer KIs demonstrate direct EC toxicity. For example, the CDK-4/6 inhibitor pablociclib induced HUVEC senescence,109 and the ALK inhibitor crizotinib (like the BCR-Abl KI dasatinib) exacerbated pulmonary hypertension in a rat model.110 Substantial future investigation is needed to determine the impact of emerging therapies on EC health.

    Conclusions and Areas of Further Study

    The explosion of novel cancer treatments specifically targeting oncogenic kinases has improved cancer survival while decreasing adverse side effects compared to traditional chemotherapy. However, with better cancer survival comes a lower tolerance for side effects. Cardiovascular toxicity has emerged as a common and morbid outcome that often necessitates dose reduction and limits quality of life and overall survival in cancer patients. As our understanding of the critical role of EC dysfunction in diverse cardiovascular diseases has continued to advance, the potential for EC dysfunction caused by anti-cancer therapies to mediate cardiovascular side effects has come into focus (Figure 1). Since ECs are exposed to high concentrations of KIs in circulation, these cells are impacted by the on-target and off-target effects of all systemically administered drugs. Despite this, the impact of novel agents on EC function is rarely studied during drug development. Most KIs target the highly conserved ATP binding site, hence these drugs are not entirely specific, and off target effects are potential sources of toxicity. Multiple kinase binding screens have been published attempting to determine if side effects can be predicted by inhibition of a culprit kinase.111,112 Most kinome profiling studies have not been performed in endothelial cells, nor have clear culprit kinases been identified that predict toxicity. For example, when the overlap between the kinomes for the BCR-Abl KIs was determined, a specific kinase could not be identified that was inhibited by all the drugs associated with arterial thrombosis and was not also targeted by at least one agent without vascular side effects.113 This is consistent with an emerging model of distinct toxic effects of each drug on EC function that may culminate in similar clinical side effects. More recently, phospho-proteomic data has begun to be considered which can incorporate the impact of regulatory factors and cellular contexts that are not modeled in traditional kinome studies.114 One phospho-proteomic study performed in ECs identified a phospho-proteomic signature that predicts BCR-abl KI toxicity.31 Future studies combining kinome profiling with phospho-proteomics in ECs are needed to determine if this approach may be refined to predict vascular side effects of emerging KIs.As our understanding of the correlation between impacts on ECs in preclinical models and the side effects seen in humans grows, incorporation of this knowledge into drug development and clinical trials has potential to lead to safer therapies. For situations in which the EC toxicity is an off target effect, identifying agents that protect the endothelium to prevent cardiovascular toxicity is a critical next step to improving outcomes in cancer patients. In contrast, for drugs like the VEGFRIs, for which EC toxicity is a primary mechanism for the anti-cancer efficacy, it may not be possible to find EC-protective agents that preserve anti-tumor activity. Clinical trials comparing anti-hypertensive strategies to identify those that mitigate VEGFRI-induced hypertension without limiting cancer efficacy are needed. Our understanding of the molecular mechanisms behind the cardiovascular side effects of the BTKIs is much less developed, highlighting a gap in knowledge. Exploration of the EC impact of newer KIs such as the EGFR inhibitors may help us understand which molecular mechanisms are related to EC dysfunction versus cardiomyocyte toxicity. Similarly, studies examining EC toxicities early in the clinical development of new KIs could predict potential side effects thereby improving lead molecule prioritization. Closer monitoring for EC toxicity in cancer trials could expose fewer cancer patients to adverse cardiovascular side effects. In addition there is growing understanding that environmental stressors as well as racial and ethnic background can affect endothelial cell health which may play a role in the side effect profiles of various anti-tumor KIs that effect the endothelium.115 Further research into the differential impact of KI side effects by race or ethnicity and, if possible, EC models that account for different racial and ethnic backgrounds are needed to advance the field.116
    RT

    Dr. Richard Travers

    Classical Hematology
    Tufts Medical Center
    richardtravers