Mitochondrial imaging detects early cardiac responses to cancer immunotherapy
H Chae, U Shrestha, Y Seo, J LeviAbstract
Background
Cancer therapy-induced cardiotoxicity, particularly that associated with immunotherapy, is hard to diagnose and manage because current methods lack molecular specificity and are unable to detect early cardiac injury. Mitochondrial dysfunction underlies cardiotoxicity from conventional therapy, whereas T-cell infiltration drives immunotherapy-associated myocarditis. [¹⁸F]F-AraG is a PET tracer that allows noninvasive visualization of mitochondrial biogenesis in cardiomyocytes and activated T cells, enabling early detection of cardiotoxicity associated with cancer therapies, including immune checkpoint blockade (Figure 1).
Purpose
To evaluate whether [¹⁸F]F-AraG PET can serve as an early imaging biomarker of treatment-related mitochondrial alterations in the myocardium.
Methods
Doxorubicin-treated mice were used to model chemotherapy-induced cardiac injury. Early myocarditis detection was evaluated using [¹⁸F]F-AraG PET/CT in poly I:C–treated and experimental autoimmune myocarditis (EAM) mouse models. To assess the translational potential of [¹⁸F]F-AraG PET for assessment of off-target cardiac effects, 26 healthy subjects underwent PET imaging to establish physiological myocardial uptake. Seven patients with stage III melanoma and ten with advanced NSCLC were scanned before and after one infusion of immunotherapy. Myocardial uptake (SUVmax, SUVmean, SUVtotal) was quantified in the left (LV) and right (RV) ventricles, with regional LV analysis using a 17-segment model. Myocardial tracer uptake was compared with ECG findings in two patients with head and neck (H&N) cancer who had available ECG data. Associations between tracer uptake, mitochondrial DNA content, and PGC-1α expression were evaluated.
Results
Cardiac [¹⁸F]F-AraG uptake decreased after doxorubicin treatment (6.4 vs 4.5 %ID/cc, p<0.0001) alongside reductions in mitochondrial DNA copy number. In contrast, uptake increased in myocarditis models (poly I:C: 6.4 vs 9.9 %ID/cc, p<0.0001; EAM: 7.2 vs 9.9 %ID/cc, p=0.007) (Figure 1). Healthy human myocardium showed consistent, spatially uniform uptake (SUVmean COV: 11.2% LV, 12.3% RV), with higher LV uptake corresponding to greater mitochondrial content and PGC-1α expression. Previously treated patients with NSCLC demonstrated higher baseline LV uptake than controls (SUVmean 4.27 vs 3.47, p=0.01). Although mean post-immunotherapy changes were modest, >20% ventricular uptake increases occurred in 4/7 melanoma and 4/10 NSCLC patients; focal increases >40% in regional analysis were observed in 3/7 and 3/10 patients, respectively. Altered uptake patterns were associated with ECG abnormalities in patients with H&N.
Conclusion
[¹⁸F]F-AraG PET detects therapy-associated mitochondrial changes in the myocardium following cancer treatment. These findings support its potential utility as a noninvasive imaging approach for early evaluation of therapy-induced cardiac effects and anti-tumor immunity to guide patient care (Figure 2).