International Study of Comparative Health Effectiveness With Medical and Invasive Approaches: Primary Report of Clinical Outcomes

Presented during the AHA Scientific Sessions 2019 by Judith S Hochman (NYU Sch of Med, New York, NY; ISCHEMIA Research Group) .

Quality of Life data were presented by John A. Spertus (Mid America Heart Inst, Kansas City, MO; ISCHEMIA Research Group)

Introduction and methods

The ISCHEMIA trial aimed to answer the question whether in stable patients with at least moderate ischemia on a stress test, there is a benefit of adding cardiac catheterization and, if feasible, revascularization to optimal medical therapy.

Stable patients with moderate or severe ischemia either underwent a blinded CCTA or no CCTA when this was not required, for instance when eGFR was 30 to <60 or when coronary anatomy was previously defined, and were then randomized to an invasive strategy (INV) or a conservative strategy (CON). INV consisted of optimal medical therapy (OMT) plus catheterization plus optimal revascularization and CON of OMT only, with catheterization reserved for OMT failure.

The primary endpoint was time to CV death, myocardial infarction (MI), hospitalization for unstable angina, heart failure (HF) or resuscitated cardiac arrest (RCA). Various methods were used to assure complete ascertainment and reporting of events. 5179 Patients were randomized. Median follow-up for survivors was 3.3 years in both the INV and CON groups. Proportion of follow-up completed was 99.4% and 99.75 for the respective groups. Baseline characteristics were well balanced among groups. On the qualifying stress test (core lab interpretation), baseline inducible ischemia was severe in 53%, moderate in 34% and mild/none in 12% of those in the INV group and 55%, 32% and 12%, respectively in the CON group.

Quality of life was assessed at randomization, and throughout the follow-up period of up to 36 months. SAQ Angina Frequency Scale scores were used. A higher score reflects less angina.

Main results

Clinical outcomes

• An adjusted HR of 0.93 (95%CI: 0.80-1.08, P=0.34) was seen for the primary outcome in INV vs. CON. The events curves crossed at around year 2 or follow-up, with INV first showing the highest incidence and later the lowest.
• Because of the crossing curves, the absolute difference in event rates of INV vs CON was calculated at two time points. At 6 months the difference was 1.9% (95%CI: 0.8-3.0%) and at 4 years it was -2.2% (95%CI: -4.4 to 0.0%).
• A similar pattern of curves was seen for the major secondary outcome of CV death or MI, with an overall adjusted HF of 0.90 (95%CI: 0.77-1.06, P=0.21). Absolute risk difference at 6 months was 1.9% (95%CI: 0.9-3.0%) and at 4 years -2.2% (95%CI: -4.4 to -0.1%).
• The net clinical benefit endpoint considered CV death, MI, unstable angina, HF, RCA and stroke and showed an HR of 0.95 (95%CI: 0.82-1.10, P=0.50).
• The risks of CV death, all-cause death and MI were not affected by INV vs CONV.
• A pre-specified Bayesian analysis suggested that the probability of at least a 10% relative risk reduction of INV on all-cause death was <10%.
• The risk of procedural MI was significantly higher in INV vs CON (HRadj: 2.98, 95%CI: 1.87-4.74, P<0.01), whereas risk of spontaneous MI was lower (Hradj: 0.67, 95%CI: 0.53-0.83, P<0.01). Hospitalization for UA was lower in INV (HRadj: 0.50, 95%CI: 0.27-0.91, P=0.02) and hospitalization for HF was higher (HRadj: 2.23, 95%CI: 1.38-3.61, P<0.01). RCA and Stroke were not significantly different between groups.
• Subgroup analyses did not reveal heterogeneity of the treatment effect.

Quality of Life (QoL) outcomes

• At baseline, in both groups, about a third of patients had no angina, about 44% had it several times per month and about 20% had daily/weekly angina.
• During follow-up, the INV group reported higher SAQ summary scores, as well as higher SAQ angina frequency score and SAQ QoL score.
• At any level of baseline angina frequency, those in the INV group showed a higher probability of not having angina than those in the CON group, with a factor 3 difference (45% vs. 15% probability of no angina) at an angina frequency score of 50. The difference was lower in less symptomatic patients.

Conclusion

These data show that overall, in initial INV strategy as compared with an initial CON strategy, did not show a reduced risk over a median 3.3 years for the primary composite endpoint and the major secondary endpoint of CV death or MI. The probability of at least a 10% benefit of INV on all-cause mortality was less than 10%.

Based on the QoL-data, it was concluded that patients with stable CAD and moderate to severe ischemia had significant, durable improvements in angina control and QoL with an invasive strategy if they had angina (daily/weekly or monthly). In patients without angina, an invasive strategy led to minimal symptom or quality of life benefits, as compared with a conservative strategy.

When interpreting these data, it should be noted that it was an unblinded trial, without a sham procedure. Also, due to the exclusion criteria, these data do not apply to patients with acute coronary syndromes within 2 months, highly symptomatic patients, patients with left main stenosis and those with LVEF <35%. Also, as centres were specifically trained for low complication rates, the findings may not be generalizable to centres with higher procedural complications rates. Completeness of revascularization has not yet been assessed. It should also be noted that women were enrolled in the trial, but they were more often excluded from randomization than men, due to less ischemia and more non-obstructive CAD.

Regarding the QoL-data, it is important to note that enrolment was skewed towards less symptomatic patients.

- Our reporting is based on the information provided during AHA Scientific Sessions 2019 -

Read our summary of the results of the ISCHEMIA-CKD trial Watch a video by Donald Lloyd-Jones about the ISCHEMIA trial