Cardio-Oncology
Living Guideline

update the stroke-risk framework to CHA₂DS₂-VA and the AF-CARE approach. Anticoagulation is recommended at CHA₂DS₂-VA ≥2 and should be considered at 1, alongside cancer-specific thrombotic risk, bleeding, interactions and patient preferences. This replaces the older sex-specific score wording; it does not make the evidence cancer-specific randomized evidence. Class/level of evidence: retain I C for the cancer-specific recommendation to anticoagulate patients meeting the recommended threshold. Wording update; no cancer-specific evidence-level upgrade.

retain the active-treatment target of <140/90 mmHg. For suitable patients, particularly survivors, incorporate the later general ESC on-treatment systolic target of 120–129 mmHg when tolerated, with individualization for frailty, orthostasis, prognosis and treatment toxicity. EAQ191 adds direct feasibility/safety evidence for more intensive control during VEGFR-TKI treatment: 61 patients were randomized by site to systolic targets <120 or <140 mmHg. It did not establish fewer clinical cardiovascular events. Class/level of evidence: retain the broad active-cancer recommendation at I C. The selected VEGFR-TKI strategy has Level B evidence; proposed IIb B, without a universal <120-mmHg mandate.

replace the reference to the 2021 valve guideline with the 2025 ESC/EACTS valvular heart disease guideline. Continue to incorporate cancer prognosis, comorbidities and patient preferences into intervention decisions. Class/level of evidence: retain I C for cancer-specific management. Reference update; no automatic evidence-level upgrade.

replace the reference to the 2015 pericardial guideline with the 2025 ESC myocarditis and pericarditis guideline. Add the new malignant-effusion evidence described below. Class/level of evidence: retain I C for the overarching cancer-specific recommendation; specific later recommendations must retain their own stated grades.

STOP-CA randomized 300 patients with lymphoma: its LVEF-defined dysfunction endpoint occurred in 9% with atorvastatin versus 22% with placebo. A subsequent meta-analysis of seven RCTs, 887 patients, found less CTRCD, RR 0.46 (95% CI 0.29–0.72). Clinical HF and mortality benefit remain unestablished, and individual trials were not uniformly positive. Evidence-level or recommendation change: proposed IIa B → IIa A, restricted to anthracycline cardioprotection. Do not generalize this upgrade to every cancer therapy or to prevention of clinical HF.

In a prospective, nonrandomized study of 190 patients, six-monthly surveillance was associated with no HF/CV-death events and one asymptomatic CTRCD case at one year. The study excluded prior anthracycline exposure, significant CVD and uncontrolled hypertension. It did not establish randomized noninferiority to three-monthly imaging. Evidence-level or recommendation change: the narrowly defined de-escalation proposal is IIb C → proposed IIa B. Preserve standard three-monthly surveillance outside this selected group. The six-month interval is supported by the study; a universal interval change is not.

HER-SAFE randomized 90 patients. Recurrent moderate/severe CTRCD occurred in 1/45 evaluated after withdrawal versus 0/44 continuing treatment; there were no symptomatic CTRCD events, HF hospitalizations or CV deaths over 12 months. This supports a selected, monitored withdrawal strategy, not long-term safety in all recovered cardiomyopathies. Evidence-level or recommendation change: Figure 37 Class IIa → proposed IIa B. Figure 37 did not print an evidence-level letter, so do not describe this as a literal "C → B" change. Retain the separate I C recommendation for long-term therapy after severe CTRCD; HER-SAFE is insufficient to overturn it.

API-CAT randomized 1,766 patients with active cancer: recurrent VTE was 2.1% versus 2.8%, meeting noninferiority, and clinically relevant bleeding was 12.1% versus 15.6% with reduced versus full-dose apixaban. EVE provides additional randomized evidence, although its bleeding reduction was not significant. Evidence-level or recommendation change: new proposed IIa A option. Individualize for recurrence, bleeding, renal function, interactions and other indications. This does not authorize reduced-dose initial VTE treatment or underdosing for AF.

TREATT added a large post-2022 randomized test: death or WHO grade ≥2 bleeding occurred in 31.7% versus 34.2%, HR 0.92 (95% CI 0.67–1.27). Together with A-TREAT, the evidence does not support routine prophylactic use. Evidence-level or recommendation change: new proposed III A - no benefit. This is an addition, not reversal of an original prohibition. It does not prohibit selected use for active bleeding.

PMAP randomized 50 patients: six-month effusion recurrence was 12% versus 60%, and tamponade 4% versus 40%, with balloon pericardiotomy versus pericardiocentesis/extended drainage. Survival and quality-of-life differences were not significant. The small, under-recruited trial limits generalization. Evidence-level or recommendation change: new Level B evidence. This trial does not upgrade the evidence for a surgical window, which is a different procedure. No new universal first-line mandate is assigned here.

Current prescribing guidance advises prophylactic anticoagulation during the first four months of combination treatment. Incorporate eligibility, contraindications, bleeding and interactions into the plan. Evidence-level or recommendation change: new regimen-specific safety instruction; no ESC class or evidence grade assigned.

the September 2026 FDA label carries a boxed warning for arrhythmia risk when camizestrant is administered with a CDK4/6 inhibitor, particularly ribociclib. Obtain an ECG before treatment, weekly for the first two weeks and periodically thereafter; assess and correct electrolyte abnormalities; avoid additional QT/torsades-risk drugs where possible; and monitor heart rate more frequently during the first 30 days in patients with baseline heart rate <60 bpm or concomitant heart-rate-lowering therapy. In SERENA-6, QTc prolongation occurred in 2.6% and bradycardia in 8%. Follow label-defined treatment-interruption and discontinuation criteria. Evidence-level or recommendation change: regulatory prescribing instruction; no ESC class assigned. The label documents one torsades case with camizestrant–ribociclib and provides the monitoring schedule and QTc thresholds.

serious cardiac dysfunction can occur with belzutifan plus lenvatinib. In LITESPARK-011, grade ≥3 cardiac dysfunction occurred in 6% and an absolute LVEF reduction >20% in 6%. Safety has not been established in patients with baseline LVEF <50%. Consider baseline echocardiography or MUGA imaging, monitor for heart-failure symptoms and signs, reassess LVEF when these develop, and modify or discontinue therapy according to severity. Evidence-level or recommendation change: regimen-specific FDA instruction; no ESC grade assigned.

a formal ESC guideline on cardiac rehabilitation now exists and contains cancer-specific CORE recommendations. Cardio-oncology rehabilitation should be considered in patients with cancer and cancer survivors — particularly those at high CTR-CVT risk — to improve muscle strength, peak VO2 or walking capacity, health-related quality of life and cardiovascular risk profile (IIa, B1). CORE may be considered during cardiotoxic treatment to prevent or mitigate CTRCD, particularly with anthracyclines (IIb, B2). Tele- or hybrid CORE should be considered as an alternative to centre-based rehabilitation during or after cardiotoxic therapy to improve physical functioning and physical activity (IIa, B2). These grades arise from the separate 2026 rehabilitation guideline and its new B1/B2 evidence system, which should not be silently converted to the older A/B/C system; they do not regrade the 2022 cardio-oncology surveillance tables. This replaces the earlier "future panel could consider IIa B" exercise language in Section 8.1.

Cardiac CARE did not show improved LVEF with troponin-triggered candesartan/carvedilol. A newer 100-patient NT-proBNP-guided trial established feasibility/safety, with an exploratory early LVEF signal. Interpret biomarkers with symptoms, imaging and alternative causes; these trials do not establish automatic drug escalation for every isolated elevation. Grading decision: no broad recommendation or monitoring-interval upgrade. A feasibility endpoint does not establish clinical efficacy.

in a contemporary cohort of 677 patients receiving 752 RT courses, eight device malfunctions occurred, all with neutron-producing treatment. This reinforces attention to radiation type/energy and individualized device risk assessment. Grading decision: retain I C for risk assessment and risk-adapted monitoring. Predictors of malfunction do not establish the benefit of a particular interrogation schedule. Appraisal distinction: this Section 15 appraisal retains I C. The earlier Section 9.5 page contains a separate proposed upgrade and has not been rewritten by this appendix update. This Section 9.5 vs Section 15 grading discrepancy is flagged for editorial panel review; no grade has been changed here.