Author + information
- Published online December 18, 2017.
- aDivision of Cardiology, Duke University Medical Center, Durham, North Carolina
- bDuke Clinical Research Institute, Duke University Medical Center, Durham, North Carolina
- ↵∗Address for correspondence:
Dr. Daniel B. Mark, Division of Cardiology, Duke Clinical Research Institute, PO Box 17969, Durham, North Carolina 27715.
On the 40th anniversary of Andreas Gruentzig performing the first human percutaneous transluminal coronary angioplasty, it is worth reminding ourselves that, in 1977, chronic angina due to obstructive coronary artery disease (CAD) was the cause of an enormous amount of suffering and a number of premature deaths (1). Coronary artery bypass graft (CABG) surgery was used in those days primarily to treat refractory disabling angina, although understanding of its additional potential for prognostic benefits in 3-vessel and left main disease was slowly emerging. The vision Gruentzig articulated for percutaneous transluminal coronary angioplasty was to provide a “comparable functional outcome” to CABG, but with “considerably less suffering” and earlier return to work/normal functioning. After substantial evolution in the technology of percutaneous coronary intervention (PCI), Gruentzig’s dream of outcomes comparable to CABG appears to have arrived in a most unexpected way: in the treatment of left main disease.
Obstructive atherosclerotic disease of the left main coronary artery has long been the 800-pound gorilla of CAD, the form of the disease that no clinician wanted to misdiagnose and that, once diagnosed, caused more clinician anxiety than almost anything else in cardiovascular medicine. Recently, meta-analyses of clinical trial data involving about 4,500 patients have concluded that major clinical outcomes with PCI and CABG are prognostically equivalent (aside from repeat revascularization) and the choice for patients with low to intermediate complexity of associated CAD is essentially a toss-up (2,3). Having achieved this milestone, some commentators presume, as Gruentzig did, that patients would uniformly prefer and their physicians would routinely recommend the path of “considerably less suffering” and “faster return to usual living” (1,4). However, others are more cautious, suggesting some, particularly younger, lower-risk patients, might prefer the proven long-term durability of CABG. In the modern vision of patient-centric medicine, individual patient preferences are expected to provide the deciding element. However, for patients to develop a personal preference for complex mixtures of unfamiliar medical outcomes requires both an understanding of their likelihood of such outcomes as well as the likely effects on symptoms, functioning, and subjective well-being associated with each (5).
This is a tall order, and it typically starts with an examination of the evidence around 2 types of treatment outcomes: survival and major morbid events, such as stroke, on the one hand, and quality of life (QOL) on the other. In this issue of the Journal, Baron et al. (6) provide new data on the patient’s perspective regarding revascularization with PCI using everolimus-eluting stents versus CABG in left main CAD from the EXCEL (Evaluation of XIENCE versus Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization) trial. EXCEL enrolled 1,905 left main patients with low to intermediate associated coronary disease severity (SYNTAX [Synergy Between Percutaneous Coronary Intervention with Taxus and Cardiac Surgery] scores ≤32) (7). The primary composite outcome (death, stroke, myocardial infarction [MI]) at 3 years occurred in 15.4% of the PCI group and 14.7% of the CABG group (hazard ratio: 1.0), and the individual components did not differ significantly by treatment group.
Given the interpretation of the clinical EXCEL outcomes as a “toss-up,” Baron et al. (6) examined whether QOL differences, as measured with the Seattle Angina Questionnaire might aid physicians and patients in understanding the trade-offs involved and in reaching a preference. This important work fills an evidence void on the patient-reported outcomes for patients with left main disease facing a revascularization choice. The adjusted mean differences by intention-to-treat in the Seattle Angina Questionnaire scales at 12 and 36 months for angina frequency, physical limitation, and disease perception QOL were all very similar. As has been reported in several earlier trials, at 1 month, CABG patients had substantially worse results on the physical limitation scale, reflecting the slower recovery pattern characteristic of a major cardiac surgical procedure. Although no QOL measurements were made in EXCEL between 1 month and 12 months, a similar pattern was seen in the SYNTAX trial, in which CABG and PCI physical function scores reached parity at 6 months (8). In EXCEL, improvement in dyspnea symptoms was slower with CABG than with PCI, but at 1 year, the proportion free of dyspnea was equivalent (46% to 49%). At baseline, 21% of patients had depressive symptoms significant enough to qualify as clinical depression. CABG patients showed significantly less improvement than PCI patients did at 1 month, but by 1 year, the treatment gap in depressive symptoms had almost completely closed.
Primary clinical composite endpoints typically give equal weight to death and nonfatal adverse events. However, MI and strokes can be very heterogeneous in terms of their effects on patient functioning and other QOL domains. The effect of nonfatal MI on QOL is often modest, but as such patients also tend to have lower baseline QOL scores, the incremental effect of the MI is challenging to isolate (9). Because strokes and MI occurred so infrequently in EXCEL (about 3% and 8%, respectively, at 3 years) they likely contributed little information to the average QOL scores for each treatment group.
In the early development of PCI and CABG, disabling angina was very prevalent and a primary indication for intervention. In EXCEL, only 10% of the study cohort reported daily angina at randomization, about 30% had weekly angina, and the remaining 60% of subjects had monthly angina or none (6). When viewed at the overall trial level, the EXCEL data do appear to demonstrate parity in angina relief and QOL between PCI and CABG out to 36 months. However, this may simply be an artifact of averaging heterogeneously responding patient subgroups. Over a decade ago, Spertus et al. (10) reported that the primary determinant of improved QOL after PCI was pre-procedure angina frequency. If the patient did not have angina before the PCI, revascularization was quite unlikely to improve QOL. The largest magnitude improvement was seen in patients with daily angina. In addition, the COURAGE (Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation) trial showed that 80% or more of the improvement in Seattle Angina Questionnaire scores seen with the PCI strategy was also seen in the control/medical therapy arm (11). Thus, due to changes in the populations being treated and the increased effectiveness of medical therapies, the opportunity space for differences in revascularization effectiveness to demonstrate incremental QOL benefits has shrunk (Figure 1). Based on the findings from earlier studies, if an angina-relief difference—and associated QOL benefits—from revascularization choice were to be seen in EXCEL, one would expect the clearest signal in the subjects with the highest angina burden. This pattern of QOL treatment response was seen in the SYNTAX trial (8). With the daily and weekly angina subjects pooled (due to small numbers in the former group), a trend in favor of greater angina relief with CABG does suggest itself in EXCEL, although the precision of the subgroup estimates is insufficient (due to sample size) to have strong confidence in this finding.
With our portfolio of outcome evidence regarding the choice of revascularization for left main disease enhanced with this new data on patient-reported outcomes, how do we help the patient develop and articulate preferences so as to reach a patient-centric treatment choice? The published data on shared decision making are a bit vague on this step (12). What the patient wishes to know is the overall treatment experiences and responses “for people like me.” Unfortunately, our trial and meta-analysis data instead provide complex, disaggregated information on different facets of the expected treatment outcomes for a hypothetical average patient (representing the admixture of all the subjects enrolled in the trial).
Compounding the treatment choice problem, patients often arrive at the therapeutic encounter with pre-existing preferences based on unrealistic beliefs about the benefits revascularization will offer, beliefs that are not easily corrected, even with structured educational efforts to do so (13,14). Patients are not the only source of unfounded beliefs about efficacy. Clinicians often use physical metaphors to simplify the problem (e.g., “fixing the plumbing,” “opening the blockage”), with little explicit discussion of incremental risks and benefits (15). The “oculostenotic reflex” may be a function of cardiologists taking their own metaphors too literally. What patients want, however, is not cosmetically attractive coronary arteries but restoration of their health so they can be “free of symptoms” and “live a normal life” to the extent possible.
At present, to understand what therapy would work best for them, patients must immerse themselves in the technical jungle of evidence-based medicine. No wonder that many give up the struggle and choose whatever the doctor recommends. However, if we are really serious about moving to a patient-centric model of medicine, we need to meet patients where they are in understanding and answer the existential questions that matter most to them. Although much work on this has already been done (16), EXCELlence is still a long way off.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. Mark has received research grants to Duke University from Eli Lilly and Company, Gilead, AstraZeneca, Bristol-Myers Squibb, Merck, and Oxygen Therapeutics. Dr. Patel has received research grants to Duke University from Bayer, Janssen, and AstraZeneca; and has served as an advisory board member for Bayer, Janssen, and AstraZeneca.
- 2017 American College of Cardiology Foundation
- Nerlekar N.,
- Ha F.J.,
- Verma K.P.,
- et al.
- Giacoppo D.,
- Colleran R.,
- Cassese S.,
- et al.
- Ruel M.,
- Verma S.,
- Bhatt D.L.
- Mark D.B.
- Baron S.J.,
- Chinnakondepalli K.,
- Magnuson E.A.,
- et al.,
- on behalf of the EXCEL Investigators
- Stone G.W.,
- Sabik J.F.,
- Serruys P.W.,
- et al.,
- for the EXCEL Trial Investigators
- Kaul P.,
- Ohman E.M.,
- Knight J.D.,
- et al.,
- for the TRILOGY ACS Investigators
- Spertus J.A.,
- Salisbury A.C.,
- Jones P.G.,
- Conaway D.G.,
- Thompson R.C.
- Ting H.H.,
- Brito J.P.,
- Montori V.M.
- Goff S.L.,
- Mazor K.M.,
- Ting H.H.,
- Kleppel R.,
- Rothberg M.B.