Angiogram strategy (with a stent, balloon or bypass if needed) vs medicines first after a heart attack, for people aged 70 and over

“symptomatic end points were not studied”

The authors used published trial results, not individual patient data, so they could not look closely at who bled or test effects over time. End point definitions and eligibility varied slightly between trials. For deaths and heart attacks, most of the information came from After Eighty and SENIOR-RITA. The funnel plots were uneven: smaller trials showed fewer deaths with the invasive strategy, while larger trials were closer to no difference or the other way. All the trials were open-label. In SENIOR-RITA, fewer than a third had multivessel PCI, so the authors suggest incomplete revascularisation may help explain the lack of a survival benefit.

Reddy et al., 2025, Methods and Discussion (read in full, PDF supplied).

The journal corrected the P value for major bleeding from .05 to .046 in the abstract, Results and Figure 3.

Error in P value, JAMA Intern Med, 2025 (read in full via Scite).

We read the MOSCA-FRAIL extended report. Its table lists readmission episodes, including repeat admissions of the same person, out of 179 and 188 episodes. Reddy's bleeding count of 16 and 5 matches that table, and so do its counts for heart attacks (23 and 22), later procedures (8 and 8), strokes (7 and 6) and heart failure (39 and 30). Reddy divides these by 84 and 83 people, so one person can be counted more than once. This is our inference from the identical counts; the review does not say so. Its death counts (43 and 50) are people. This trial carries 18.2% of the weight for bleeding. In the review's analysis of trials of people aged 75 or over, a sensitivity analysis it did not prespecify and which leaves out MOSCA-FRAIL, more bleeding was no longer statistically significant. Hamed gives 8 and 1 bleeds for this trial at 1 year. Separately, the death-or-heart-attack analysis lists 16 and 5 again, which cannot be right because there were 43 and 50 deaths, and this went into that analysis's printed result. Figure 1 also lists that analysis with 4 trials, while the supplement shows 5. We do not use that analysis in this fact box.

Our check of Reddy et al., 2025, Figures 1 to 3 and eFigures 1 and 8; Sanchis et al., 2024, Table; Hamed et al., 2025, Figure 3.

The later-procedure rows of the absolute-numbers table (eTable 7) are out by a factor of ten, and the numbers needed to treat with them: the forest plot counts give 208 of 1,508, or 14 in 100, not the table's figure. This fact box uses the forest plot counts. The text gives a different combined value for deaths from Figures 1 and 2 and eTable 6; this fact box follows the figures. The trial table gives MOSCA 2016 the size of another trial (313, not 106).

Our check of Reddy et al., 2025, Results, Figures 1 to 4, Table, eFigure 1 and eTables 6 and 7.

“a worse frailty status was the most powerful indicator of a benefit loss”

The authors of the MOSCA-FRAIL trial agree with the review's main message. They add that their trial, the only one limited to frail people, found no benefit from the invasive strategy, no difference in repeat heart attacks in a later analysis, and more bleeding. They argue for frailty assessment to guide the decision.

Sanchis et al., 2026, JAMA Intern Med (letter; read in full, PDF supplied).

“only reached nominal significance for bleeding”

In their reply, the authors say their review reached only nominal significance for bleeding. They report that in frail people in SENIOR-RITA, heart attacks did not differ and bleeding was numerically higher (11.3% vs 6.7%). They conclude that an invasive strategy should not be routine in frail older adults. Our check: the reply and the paper's Key Points give the review's size as 3,009 participants; its Results and figures give 3,099, which is the number that adds up.

Reddy et al., 2026, JAMA Intern Med (reply; read in full, PDF supplied).

In the extended follow-up of MOSCA-FRAIL (median 1,113 days), there was no clear difference in survival time overall: 28 days more with the invasive strategy, could be from 188 days fewer to 230 days more. People in the invasive group tended to die sooner in the first year, and this shifted to a later benefit. There were 5 deaths from bleeding in the invasive group and none in the conservative group. The authors call these findings exploratory, and events in the extended follow-up were not centrally adjudicated.

Sanchis et al., 2024, JAMA Netw Open (read in full via Scite and Europe PMC).

In an exploratory analysis of the 469 frail patients in SENIOR-RITA, the main outcome occurred in 37.7% with the invasive strategy and 29.4% with medicines only, a difference that could be chance. When frailty was treated as a continuous score, the authors saw a possible signal of harm at the highest levels. The review had no individual patient data, so it did not look at frailty itself (our reading).

Rubino et al., 2026, JAMA Netw Open (main text read via Scite).

After Eighty carries 72.5% of the weight for deaths and SENIOR-RITA 18.2%. After Eighty counts so much because almost everyone in it died during long follow-up (205 and 194 of about 229), which makes its result very precise. For heart attacks, the two trials share the weight (40.2% and 40.5%). Hamed's review of the same 7 trials gave SENIOR-RITA 66.3% for deaths because it used earlier, shorter reports of After Eighty and MOSCA-FRAIL.

Our reading of Reddy et al., 2025, Figure 2 and eFigure 1; Hamed et al., 2025, Figure 2.

Patients and doctors knew the treatment, so a later procedure in the medicines group depended partly on clinical judgement. For these 7 trials, the review's table lists crossover from the medicines group at 4.3% in 80+, 8.7% in RINCAL, 10.8% in MOSCA-FRAIL, 20.3% in MOSCA, 24.2% in SENIOR-RITA and 28.9% in Italian Elderly ACS. It lists none for After Eighty, whose report counted later procedures as study outcomes instead. The review found no sign that crossover changed the result for deaths.

Our reading of Reddy et al., 2025, Table and Results; Berg et al., 2023.

The supplement lists BARC type 2 or greater for SENIOR-RITA, BARC type 3b or greater for RINCAL, TIMI major bleeding for After Eighty, and no definition for MOSCA-FRAIL, which Vats describes as bleeding requiring rehospitalization. SENIOR-RITA carries 56.4% of the weight and counts less serious bleeding than some other trials.

Our check of Reddy et al., 2025, eTable 4; Vats et al., 2025, Table S2.

“without an increase in the risk of major bleeding”

Hamed pooled the same 7 trials but used earlier reports of After Eighty and MOSCA-FRAIL, and included MOSCA 2016 in its bleeding analysis, which Reddy leaves out. It found fewer heart attacks and later procedures, no difference in deaths, and no clear difference in major bleeding.

Hamed et al., 2025, J Am Heart Assoc (read in full, PDF supplied).

“Invasive management increased the incidence of severe bleeding”

This review of 8 trials with 3,887 patients aged 65 or over found more severe bleeding. Its authors call this not robust: it was no longer statistically significant once the one trial with patients under 70 (TACTICS-TIMI 18) was removed. It rated its evidence with GRADE as high certainty for heart attacks and moderate for its other outcomes.

Vats et al., 2025, Front Cardiovasc Med (full text and supplement read, PDFs supplied).

“invasive management reduces the risk of MI and revascularization without increasing mortality risk”

Khan (11 trials, 4,114 patients aged 70 or over, searched to December 2024) describes itself as GRADE-assessed; its plain-language summary says bleeding was not higher. Goyal (7 trials, 2,997 patients) also reports no significant difference in bleeding. Taken together, the reviews agree on heart attacks, later procedures and deaths, and disagree on bleeding.

Khan et al., 2026, Future Cardiol, and Goyal et al., 2025, Catheter Cardiovasc Interv (abstracts only).

• The first MOSCA-FRAIL report (Sanchis 2023), which may give bleeding as people rather than episodes (doi 10.1001/jamainternmed.2023.0047).
• Navarro et al., 2026, Am J Cardiol, a citing paper on revascularisation in patients over 75 (doi 10.1016/j.amjcard.2025.11.014).
• Full texts of Goyal 2025 and Khan 2026.
• The individual trial papers other than SENIOR-RITA, the After Eighty long-term report and the MOSCA-FRAIL extended report.