NSTEMI Patients With CKD See Better Outcomes With Invasive Strategy
The observational study’s results contrast those of ISCHEMIA-CKD, a trial that enrolled patients with stable ischemic heart disease.
NOTE: TCTMD previously retracted this story after learning of concerns about the mortality endpoint used in the analysis. We have now, based on an amended abstract and slides submitted to the TCT abstract committee addressing these concerns, republished the story with clarification on the mortality endpoint and added commentary.
(UPDATED) Compared with medical management, an invasive strategy for treating NSTEMI in the context of advanced chronic kidney disease (CKD) is associated with a lower risk of death on readmission, MI, and major adverse events, according to results of a large analysis.
The observational data, drawn from the Nationwide Readmissions Database (NRD), provide some reassurance, too, when it comes to complications like acute kidney injury (AKI).
“The risk of AKI requiring dialysis and bleeding, as has been shown previously in other studies, was high, but the number-needed-to-harm was also high,” said Ankur Kalra, MD (Cleveland Clinic, OH), who presented the study as a “Key Abstract” online this week as part of a sneak peek at TCT 2021.
In terms of disease severity, the results for mortality and MACE were consistent for CKD stages 3 and 4, as well as end-stage renal disease (ESRD), giving the advantage to invasive management over medical therapy. For patients with CKD stage 5, however, there were no statistically significant differences in the endpoints of in-hospital or 6-month mortality.
Moderator Ziad Ali, MD (St. Francis Hospital & Heart Center, Roslyn, NY), noted that the results are clinically meaningful in light of CKD patients being largely left out of interventional cardiology RCTs. He added that the lack of data on how these patients fare with PCI can result in what he termed “renalism,” in which operators actively avoid treating NSTEMI patients with CKD.
“There are a huge number of patients who need treatment, have risk factors for cardiovascular disease that coincide with risk factors for renal disease, and they're simply left to be medically managed even though you have now clearly shown that there is a benefit for an invasive strategy,” Ali observed.
But Mamas Mamas, BMBCh, DPhil (Keele University/Royal Stoke University Hospital), who raised concerns about the data following its online presentation, pointed out that the NRD cannot be used to ascertain mortality outcomes after hospital discharge because of the specific nature of the data it gathers.
“Whilst the NRD is able to capture annualized readmissions following an index hospitalization, it does not record postdischarge mortality,” he told TCTMD in an email. “The only mortality captured by the NRD is that associated with a hospitalization event,” making it difficult to compare with deaths among medically managed patients. “Comparing readmission mortality rates between two therapeutic strategies . . . would be meaningless, as the causes of readmission would vary by therapeutic strategy and therefore, by virtue of these differences, so would mortality,” he said. “Furthermore, the bigger problem is that a significant proportion of deaths would occur in the community which are not captured by the NRD data set. This would mean that any mortality rates at a given time point reported would fail to capture a significant proportion of deaths, and any comparisons would be difficult to interpret.”
Kalra’s original slide presentation did not specify that the deaths included in the analysis only included readmission mortality; in response to concerns raised by Mamas, Kalra and colleagues submitted an amended abstract and slide set.
"This is a step closer to where we need to be," Ali told TCTMD in an email after reviewing the updated slides. "Obviously, the national readmission database is a highly skewed population, and it is impossible to determine the out-of-hospital mortality and what happens to CKD patients with NSTEMI who are not readmitted. Nonetheless, these patients are definitely undertreated and 'renalism' is real. I have a patient who took two trains and a bus and walked into the ER, because three hospitals wouldn’t treat his chest pain because his creatinine is 6.2 mg/dL."
Consistent Benefit Across CKD Stages
Kalra and colleagues analyzed mortality outcomes both in-hospital (n = 141,052) and among patients readmitted over the next 6 months postdischarge (n = 133,642) among matched cohorts of CKD patients with NSTEMI who either underwent angiography (with or without PCI) or were treated with medical management alone.
For CKD stage 3 patients, in-hospital mortality was 3.8% with invasive management versus 7.7% for medical management. Similarly, in CKD stage 4 patients, the results were 6.3% versus 8.1%, and in ESRD patients they were 6.8% versus 12.5% (P < 0.001 for all comparisons).
Mortality based on hospital readmissions over the next 6 months was 1.9% and 3.1% for the stage 3 group, 2.2% and 3.6% for stage 4, and 3.4% and 4.6% for ESDR with invasive versus medical management (P < 0.001 for all comparisons). Six-month rates of MACE (a combination of mortality, stroke, MI, and readmission for heart failure) also favored the invasive approach in all stages, including CKD 5 (P < 0.001 for all comparisons), which was confirmed in Kaplan-Meier analyses. Lower cumulative incidence of death with invasive management compared with a medical approach also was confirmed in the Kaplan-Meier analyses, this time in all groups except CKD stage 5 (P = 0.058).
I'm bullish on doing complex PCI in this patient population. Jun-Jie Zhang
For the secondary endpoint of in-hospital AKI requiring dialysis, the number-needed-to-harm with invasive versus medical management was 588 for CKD stage 3 patients and 124 for CKD stage 4 patients. For in-hospital major bleeding, another secondary endpoint, the number-needed-to-harm was 333 for CKD stage 3, 91 for stage 4, and 40 for stage 5. At 6 months there was no difference across CKD groups, including ESRD, in the combined safety endpoint of AKI, major bleeding, vascular complications, and stroke.
Kalra and colleagues also looked at in-hospital differences between patients who were revascularized and those who had an angiogram but no revascularization. In both CKD stage 3 and stage 4 patients, there was no difference in in-hospital rates of AKI requiring dialysis regardless of the management strategy used.
Ali noted that the results stand in contrast with those of the randomized ISCHEMIA-CKD trial, which found no improvement in rates of death or MI in patients with CKD and stable CAD who underwent an invasive procedure versus those who received medical therapy alone (adjusted HR 1.01; 95% CI 0.79-1.29).
“The only way to reconcile it in my own mind is that ISCHEMIA-CKD looked at stable ischemic heart disease, and we predominantly were looking at non-ST-segment elevation MI. Even though it may fall under the same rubric, I truly believe it is a different subset of patients that are at a heightened risk for future cardiovascular events,” Kalra responded. Jun-Jie Zhang, MD, PhD (Nanjing Medical University, China), agreed, adding that there may be other reasons as well.
“In a clinical trial, in my opinion, you exclude the highest-risk patients,” he said. “So, when we do these trials, it's very difficult to find the right balance versus this, [which] is a real-world analysis including everybody. The benefits ideally demonstrate it. So, I think I'm bullish on doing complex PCI in this patient population.”
But Mamas urged caution in comparing this analysis, which captures only readmissions data, to a randomized study where patient were followed regardless of whether they later ended up in hospital or not. “Observational research provides important insights into real-world outcomes of patients often not recruited into randomized trials,” he said,“but it is important that the limitations of databases are considered when designing observational studies.”
Kalra A. Invasive versus medical management in patients with chronic kidney disease and non-ST-elevation myocardial infarction. Presented at: TCT 2021. October 20, 2021.
- Kalra reports no relevant conflicts of interest.