No Difference in Cognitive Decline After CABG vs PCI

Stephanie Edwards

June 03, 2021

A study using data from the Health and Retirement Study (HRS) shows no significant difference between coronary artery bypass grafting (CABG) vs percutaneous coronary intervention (PCI) and subsequent memory decline over 5 years among older patients undergoing coronary revascularization.

It has long been thought that PCI should yield improved results in preventing cognitive decline after coronary revascularization because of its less invasive approach, compared with CABG, whether performed on-pump or off-pump, the researchers write.

Conversely, "we found no difference in memory decline between older adults who had CABG and those who had PCI, either before or after surgery or when measured as a pre-to-post-procedure change, or probability of dementia after the procedure, which — given the context for our study — was a big surprise," said lead author Elizabeth L. Whitlock, MD, Department of Anesthesia & Perioperative Care, University of California, San Francisco.

However, what they did find was off-pump CABG, "developed, in part, to avoid the hypothesized cognitive 'hit' from the bypass pump," actually was associated with an increased rate of memory decline post-procedure compared to PCI, while traditional CABG was not, the authors report.

The article was published in the May 18 issue of JAMA.

Comparing Treatment Types

Postoperative neurocognitive disorder (NCD), formerly called postoperative cognitive decline, became a concern for older adults who undergo CABG, after observational data suggested that 42% of patients had demonstrable cognitive decline, considerably higher than a cohort of Medicare beneficiaries. It was hypothesized that exposure to cardiopulmonary bypass pump might be the cause of these neurologic concerns, the authors write.

"Cardiac surgeries, particularly those that require a heart-lung bypass pump, were thought to be one of the most problematic, to the extent that off-pump coronary artery bypass grafting surgeries were developed to avoid using the pump," Whitlock said.

However, it was unknown "whether CABG in older adults causes a meaningful deterioration from patients' presurgical cognitive trajectory and whether off-pump CABG avoids such adverse cognitive effects," the authors write.

PCI, in contrast, treats coronary artery disease but eliminates many of the exposures that were thought to contribute to cognitive decline associated with CABG, they note.

Comparing CABG and PCI, "therefore, offered us the chance to look more directly at the potential contribution of surgery and anesthesia to cognitive decline," said Whitlock.

For this retrospective study, the authors examined a cohort who were 65 years or older when enrolled in the HRS from 1992 through 2010 and underwent CABG or PCI between 1998 and 2015.

A total of 1680 patients were included; 497 had on-pump CABG, 168 had off-pump CABG, and 1015 had PCI, ascertained from Medicare fee-for-service billing records.

The primary outcome was a summary measure of cognitive tests scores and proxy cognition reports performed every 2 years in the HRS, called the memory score, which compared HRS cognitive questionnaire items against Aging, Demographics, and Memory Study samples of HRS participants, the authors write.

The memory score was normalized as a z score (mean of 0, SD of 1 in participants 72 years or older in the 1995 wave). The minimum clinically important difference in memory score was defined as a change of 1 SD of the population-level rate of memory decline, or 0.048 memory units per year.

The primary comparison was the difference-in-differences estimate of the rate of memory score decline from the preprocedure period (that is, 3 years prior to the procedure until the time of the procedure) vs the postprocedure period (defined as 6 months to 5 years after the procedure), comparing CABG vs PCI, the researchers write.

The secondary outcome was the probability of dementia based on the memory score, Telephone Interview for Cognitive Status, and serial-7 subtractions.

In the PCI group, the within-group change was 0.004 memory units per year (95% CI, –0.010 to 0.018) vs –0.011 memory units per year (95% CI, –0.029 to 0.008) in the CABG group. Neither of these differences was statistically significant, and the confidence intervals excluded the potential for a clinically significant difference as well, Whitlock added.

The between group difference-in-differences estimate for memory decline for PCI vs CABG was 0.015 memory units per year (95% CI, –0.008 to 0.038; = .21), slightly favoring PCI, but again eliminating the possibility of a clinically significant difference.

After revascularization, the rate of memory score decline in the CABG group vs PCI group were nearly identical (difference-in-differences, CABG-PCI, 0.0002 memory units per year; 95% CI, –0.018 to 0.01; P = .98).

Table 1. Mean Adjusted Rate of Memory Decline
Time Point PCI Group, Memory Units/Year, (95% CI) CABG Group, Memory Units/Year (95% CI)
Before procedure 0.064 (0.052–0.078) 0.049 (0.033–0.065)
After procedure 0.060 (0.048–0.071) 0.059 (0.047–0.072)

However, they found a statistically significant increase in the rate of memory decline after off-pump CABG vs after PCI (difference-in-differences: mean increase in the rate of decline of 0.046 memory units per year after off-pump CABG, 95% CI, 0.008 - 0.084), but not after on-pump CABG vs PCI (difference-in-differences: mean slowing of decline of 0.003 memory units/year, 95% CI, −0.024 to 0.031, after on-pump CABG).

For off-pump CABG recipients, "this rate increase approached, and the confidence intervals include, 1 SD of the population rate of memory decline which was 0.048 memory units/y," the authors write, suggesting it is likely to represent a clinically significant difference in memory change over time.

Table 2. Mean Rate of Covariate-adjusted Memory Decline: On-pump vs Off-pump CABG
Time Point On-pump CABG, Memory Units/Year (95% CI) Off-pump CABG, Memory Units/Year (95% CI)
Before procedure 0.055 (0.036–0.074) 0.032 (0.001–0.063)
After procedure 0.054 (0.038–0.071) 0.074 (0.052–0.096)

Secondary analysis showed long-term dementia probability was not significantly different between the CABG and the PCI groups. Probability of dementia 5 years after revascularization was 10.5% (95% CI, 8.4% - 13.0%) in the CABG group and 9.6% (95% CI, –2.0% to 3.8%) with PCI; the authors did not analyze dementia probability separately for the on-pump versus off-pump CABG groups.

The authors note that these results show that, overall, "the modality of revascularization is not a strong determinant of subsequent cognitive aging," Whitlock said, adding that "based on these findings, I think it is important for older adults and clinicians caring for them to know that there is no reason to avoid CABG on the basis of concern about cognitive outcomes. They should pursue the procedure which is recommended to them based on the type of heart disease they have and their other medical conditions."

"It's also important for primary care clinicians to know that some older adults will experience greater-than-expected cognitive decline after a procedure, whether it is a minimally invasive one or a major surgery, and to be attentive to concerns about cognitive performance particularly around the time of a major health event," she concluded. An exploratory analysis in the authors' study population identified a small proportion of older adults, regardless of whether they underwent CABG and PCI, who had more cognitive decline after the procedure than was expected, aligning with earlier work showing some patients do experience cognitive decline and/or NCD after a procedure.

The authors note a variety of limitations to their study. First, these models capture population means, and trajectories include some patients who decline and some who improve. Dementia probability and adapted NCD were based on brief neuropsychologic assessments incorporated in the HRS, not a clinical diagnosis. Only patients receiving fee-for-service care were modeled; Medicare Advantage recipients who tend to be younger, part of a minority group, and less likely to receive cardiac procedures were not included. Surgical and interventional techniques have evolved over time, potentially changing this relationship, and the memory score used was weighted toward memory function and may not be sensitive to changes in other aspects of cognition such as executive function, they note.

Patient Risk Factors More Important

In an accompanying editorial, Rebecca F. Gottesman, MD, PhD, Stroke Branch Chief and Senior Investigator, National Institute of Neurological Disorders and Stroke (NINDS) Intramural Program, National Institutes of Health (NIH), Bethesda, Maryland, and Michelle C. Johansen, MD, PhD, Department of Neurology, Johns Hopkins University School of Medicine, Baltimore, note that cardiac risk factors, such as hypertension, diabetes, hyperlipidemia, smoking, and obesity, contribute to the development of coronary artery disease, but they are also known to increase the risk for cognitive decline and even dementia. Because of this, they say, it can be difficult to determine the long-term cognitive effects related to coronary revascularization procedures.

Gottesman and Johansen address the limitations outlined by the authors, saying, "this is an observational study, and although there are rigorous methods used to reduce biases in how the two procedure groups are different, they are not the same as having patients randomized to two different procedures."

Gottesman and Johansen write that there have been several studies on revascularization procedures and cognitive decline that reflect varying results. For example, the SWEDEHEART registry compared patients undergoing revascularization procedures against a control group, and showed comparable results in the development of dementia in older adults. However, some other studies, mainly observational, have found a connection between CABG and cognitive decline, but it is noted that these studies were vulnerable to residual confounding.

"Because the same risk factors lead to cognitive decline, and also to the need for cardiac revascularization, if we can focus on primary prevention, this would reduce both of these outcomes," Gottesman told | Medscape Cardiology.

The editorialists say the data suggest a patient's risk for postprocedural cognitive decline should not influence the optimal procedure choice, although preprocedure cognitive status may influence that decision.

Gottesman and Johansen suggest that "it is likely that the best way to prevent cognitive decline, whether post- or pre-revascularization, is ultimately to prevent, identify, treat and control these risk factors."

They suggest that improving public health messaging about the importance of controlling cardiovascular risk factors is critical for both cardiac and brain health.

Funding for the study was provided by the National Institute on Aging and the National Center for Advancing Translational Sciences, both part of the National Institutes of Health (NIH), and the Foundation for Anesthesia. Whitlock reported receiving grants from the NIH (National Institute on Aging). Disclosures for other authors appear in the paper. Gottesman reported receiving grants from NIH and previously serving as an associate editor for Neurology. Johansen reported receiving grants from the National Institute of Neurological Disorders and Stroke.

JAMA. 2021;325:1941-1942, 1955-1964. Editorial, Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.