Relying on the Society of Thoracic Surgeons Predicted Risk of Mortality (STS-PROM) score to determine surgical aortic valve replacement (SAVR) candidacy may deny many patients safe surgery, as the low-risk score (< 4%) overestimates risk. That’s the conclusion of a Cleveland Clinic review of isolated SAVR procedures that found rates of mortality, permanent stroke, extended hospital stays and other adverse measures to be lower than expected from the STS-PROM model.
Cleveland Clinic is a non-profit academic medical center. Advertising on our site helps support our mission. We do not endorse non-Cleveland Clinic products or services Policy
The research was reported as a plenary presentation at the virtual 2020 annual meeting of the American Association for Thoracic Surgery, held May 22-23.
“Our data support early surgery and provide a benchmark for comparing real-world transcatheter outcomes,” says study presenter Douglas Johnston, MD, a Cleveland Clinic cardiothoracic surgeon.
SAVR vs. TAVR: Never an easy decision
Recommendations for managing aortic stenosis continue to evolve. American Heart Association/American College of Cardiology guidelines for managing patients with valvular heart disease suggest that SAVR is a reasonable option in asymptomatic patients with low surgical risk and decreased exercise tolerance or rapid progression.
But two 2019 randomized trials, the Evolut Low-Risk Trial (N Engl J Med. 2019;380:1706-1715) and the PARTNER 3 trial (N Engl J Med. 2019;380:1695-1705), found short-term equivalence between transcatheter AVR (TAVR) and SAVR in patients with an STS-PROM score of less than 4%.
“SAVR may offer advantages that these trials fail to capture, especially for young, otherwise healthy patients who want to enjoy an active lifestyle for decades to come,” says Dr. Johnston. “Basing the decision on the STS-PROM score may not reflect current real-world SAVR risk.”
Study design and findings
This study included 3,493 adults (mean age, 64 ± 13 years) who underwent isolated SAVR at Cleveland Clinic between January 2005 and January 2017. All had an STS-PROM score of less than 4%, with the median score being approximately 1.2% throughout the study period.
Operative approach and choice of prosthesis were according to surgeon discretion. About 40% of patients underwent a minimally invasive incision. Prostheses implanted with a full root technique (e.g., allografts) were not included, as they are excluded from the STS definition.
In-hospital outcomes were as follows, with observed results listed first followed by results as predicted by the STS-PROM model and the P value for the difference:
- Operative mortality, 15 (0.43%) vs. 55 (1.6%), P < 0.0001
- Permanent stroke, 26 (0.74%) vs. 40 (1.2%), P = 0.02
- Renal failure, 52 (1.5%) vs. 97 (2.8%), P < 0.0001
- Prolonged ventilation, 149 (4.3%) vs. 249 (7.1%), P < 0.0001
- Deep sternal wound infection, 6 (0.17%) vs. 8 (0.23%), P = 0.5
- Reoperation, 116 (3.3%) vs. 230 (6.6%), P < 0.0001
- Major morbidity or mortality (composite adverse event), 278 (8.0%) vs. 449 (12.9%), P < 0.0001
- Prolonged length of stay (> 14 days), 129 (3.7%) vs. 165 (4.7%), P = 0.004
“For all outcome measures, our experience was better than expected by STS-PROM,” Dr. Johnston notes. “This calls into question how well this prediction model reflects current realities for low-risk SAVR.”
Detailed analysis revealed that the observed risks of operative mortality and morbidities were less than expected for each decile of the score, with the differences greater at the higher end of the 0%-to-4% risk spectrum.
Multivariate analysis identified the following significant risk factors for mortality or major morbidity: mitral regurgitation, left ventricular septal thickness, higher bilirubin level, lower creatinine clearance and chronic obstructive pulmonary disease.
Earlier date of surgery was also an identified risk factor, reflecting the length of the study period, during which surgical techniques improved. In the first years of the study, expected and observed outcomes were closely correlated; however, observed morbidity and mortality declined dramatically while predicted risk did not (Figure). Observed mortality was zero in the last 4 years of the study period.
Figure. Temporal trend of observed and predicted STS major morbidity or mortality after SAVR. Reprinted from Johnston D, et al., “Redefining ‘Low Risk’: Outcomes of SAVR in Low Risk Patients in the TAVR Era,” plenary presentation at the virtual annual meeting of the American Association for Thoracic Surgery (AATS), May 22-23, 2020, with permission from AATS.
Long-term survival of patients compared favorably to U.S. census-matched controls. Freedom from reoperation was 95% at seven years.
Conclusion: Early SAVR is safe for low-risk patients
Dr. Johnston highlights the following key takeaways from this study:
- STS-PROM score overestimates risk in contemporary practice. For patients with a score less than 4%, SAVR was found to be associated with extremely low risk. Dr. Johnston says the surgical community’s ability to evaluate risk is continually evolving. “We need an agile, real-time quality assessment of SAVR practice that accounts for ongoing improvements in techniques and technology,” he says, noting that differences between observed and predicted mortality and complication rates are increasing with time.
- Data support early surgery for aortic stenosis in patients with an STS-PROM score < 4%. “SAVR can be considered for patients who may not reach classic criteria for intervention but who want to maintain an active lifestyle or have careers that demand it,” he observes. Early SAVR can also help prevent poor outcomes from delayed surgery, especially as left ventricular hypertrophy was identified as a risk factor. “New strain imaging models may one day contribute to better decisions regarding timing of surgery in asymptomatic patients,” Dr. Johnston adds.
- Patient selection and institutional factors matter. Because this study reflects the experience of a single high-volume institution, outcomes may not be widely applicable. Also, the Cleveland Clinic population tended to be younger than participants in the PARTNER 3 trial, and some comorbidities differed that also may have lowered risk.
“Our intention was not to pit SAVR versus TAVR but to recognize that risk is a moving target as technologies advance,” Dr. Johnston observes. “Our study reassured us that for asymptomatic patients with aortic stenosis, SAVR remains a reasonable intervention with an excellent long-term outcome.”
In comparing outcomes between SAVR and TAVR in the trials of low-risk patients, Lars Svensson, MD, PhD, Chair of Cleveland Clinic’s Heart, Vascular & Thoracic Institute, points out that 26% of SAVR patients in the PARTNER 3 trial underwent an additional cardiac operation, mostly coronary artery bypass. “Thus, the PARTNER 3 data comparing TAVR with SAVR should be interpreted with caution,” he says. “Furthermore, more recent data reported from the PARTNER trials shows a concerning increase in strokes and death in the TAVR research arms when compared with SAVR. Hence, we advocate SAVR at Cleveland Clinic based on our results, particularly for patients younger than age 65 and for patients with a bicuspid valve, an aorta larger than 4.5 cm, or concomitant problems such as coronary disease, other diseased valves (such as the mitral valve) or other cardiac problems.”