1351868 - Assessment of Myocardial Injury in Patients Treated with Immune Checkpoint Inhibitors: A Prospective Study Using Biochemical Markers and Advanced Cardiac Imaging

Immune checkpoint inhibitors (ICI) are first-line treatment for various malignancies. Case reports describe fulminant myocarditis, and there is uncertainty regarding the prevalence of subclinical myocardial injury (MI) due to ICI. Hence, we sought to assess prevalence and severity of MI when treated with ICI using biochemical markers and advanced cardiac imaging.

Methods:

In this prospective multi-centre study, ICI naive patients with no prior cardiac disease were enrolled. Prior to ICI initiation, baseline assessed cardiac troponin I (cTnI), N-Terminal pro brain natriuretic peptide (NT-proBNP) and 2D-transthoracic echocardiogram (TTE). These parameters were repeated at 6 weeks, adding cardiovascular magnetic resonance (CMR). The primary outcome was presence/severity of cTnI elevation at 6 weeks. The secondary outcomes were ΔNT-ProBNP, ΔLV global longitudinal strain (LV-GLS) and ΔRV free wall strain (RV-FWS) measured by TTE, assessment of MI by CMR and major adverse cardiac events (MACE); cardiovascular mortality, heart failure, arrhythmias, heart block. 

Results:

Of 79 patients screened, 42 were excluded. Thirty-seven were enrolled,1 withdrew. Mean age was 68(61% male). Eleven (31%) had cancer related mortality prior to 6-week follow-up, thus 25 had follow-up. Prior to ICI initiation, mean cTnI measured 8.3 ng/L±17.8(0-54ng/L), NT-ProBNP was 264 ng/L±253.7(0-124ng/L), LVEF was 62%±5.2%, LV-GLS measured -18.1%±2.7% and RVFWS was -22.1%±5.9%. At follow-up, there was no significant difference in cTnI (Δ+16,p=0.763), NT-ProBNP(Δ+149,p=0.172), LVEF(Δ-1.0%,p=0.476) and RVFWS(Δ-2.0%,p=0.06). However, there was significant reduction in LV-GLS (Δ-8.8%,p=0.003). A 10% decline in LV-GLS and RV-FWS was seen in 42% and 49% of patients respectively. 1(3%) MACE and 1(3%) was clinically diagnosed with ICI related pericarditis. 15(60%) underwent CMR. Eight (53%) had CMR evidence of MI; elevated T1 mapping in 7(47%), elevated T2 mapping in 1(7%) and non-ischemic mid-wall LGE in 8(53%). Elevated T1 and LGE were primarily in basal-mid septal segments. Seven out of 8(87%) with CMR evidence of MI had been treated prior with chemotherapy and/or radiotherapy. Segmental abnormalities on CMR and corresponding LV-GLS values were analyzed. In segments with abnormal T1 mapping/LGE on CMR, there was pre-existing abnormalities in corresponding segmental LV-GLS values at baseline (-15%±0.4%) which did not worsen at 6 weeks follow-up (-14±0.5%,p value=0.731). This may suggest pre-existing subclinical myocardial dysfunction prior to ICI initiation due to prior cardiotoxic chemotherapy and/or radiotherapy. Therefore, CMR findings of MI most likely correspond with pre-existing TTE LV-GLS abnormalities, and not ICI cardiotoxicity. Absence of significant cTnI increase supports this assumption.

Conclusion:

In this prospective study, the incidence of new subclinical MI due to ICI were low and most LV abnormalities detected on imaging were pre-existent. Larger studies are warranted to confirm the cardiac safety of ICI.