Survival after primary breast cancer surgery following propofol or sevoflurane general anesthesia—A retrospective, multicenter, database analysis of 6305 Swedish patients

Abstract

Background: Retrospective studies indicate that the choice of anesthetic can affect long-term cancer survival. Propofol seems to have an advantage over sevoflurane. However, this is questioned for breast cancer. We gathered a large cohort of breast cancer surgery patients from seven Swedish hospitals and hypothesized that general anesthesia with propofol would be superior to sevoflurane anesthesia regarding long-term breast cancer survival. Methods: We identified all patients who were anaesthetized for breast cancer surgery between 2006 and 2012. The patients were matched to the Swedish Breast Cancer Quality Register, to retrieve tumor characteristics, prognostic factors, and adjuvant treatment as well as date of death. Overall survival between patients undergoing sevoflurane and propofol anesthesia was analyzed with different statistical approaches: (a) multiple Cox regression models adjusted for demographic, oncological, and multiple control variables, (b) propensity score matching on the same variables, but also including the participating centers as a cofactor in a separate analysis. Results: The database analysis identified 6305 patients. The 5-year survival rates were 91.0% and 81.8% for the propofol and sevoflurane group, respectively, in the final model (P = .126). Depending on the statistical adjustment method used, different results were obtained, from a non-significant to a "proposed" and even a "determined" difference in survival that favored propofol, with a maximum of 9.2 percentage points higher survival rate at 5 years (hazard ratio 1.46, 95% CI 1.10-1.95). Conclusions: It seems that propofol may have a survival advantage compared with sevoflurane among breast cancer patients, but the inherent weaknesses of retrospective analyses were made apparent.

Is It Definitely Clear That Long-Term Survival after Breast Cancer Surgery Is Not Affected by Anaesthetics?

“…Given the challenge in funding and the hard work during a long time span, we cannot however expect many RCTs. Since retrospective studies have strong limitations [ 61 ], the second best to RCTs would be national quality registers for retrospective analyses of prospectively collected data [ 62 , 63 ]. This author has the result of such a register-based study for breast cancer submitted.…”

“…That said, some of these retrospective studies are small, some extremely so, making the already dubious value of retrospective design even more uncertain. It should also be noted that the largest retrospective study to date ( n = 6305), which examined overall survival from breast cancer, is not included in any of the tables due to its dubious results [ 61 ].…”

“…A fourth study, measuring freedom from recurrence within one year, indicated the same—no difference between the two anaesthetics [ 48 ]. Later, data from seven Swedish hospitals, including 6305 breast cancer patients, were analysed in a new retrospective study [ 61 ]. Two of the participating clinics used propofol, two used sevoflurane and the other three used both drugs alternately in a pseudo-random way.…”

“…Selection bias is an elusive mechanism. Different decisions from anaesthetists, who can evaluate patients differently and thus consider different anaesthetic methods or use different drugs for patients, are often a source of non-adjustable bias, unless a certain hospital has a standard regimen as described for some hospitals in reference [ 61 ].…”

More reviews than RCTs

“…blurred to say the least. We have everything from remarkably large effects of definite clinical significance, 9% better 5-year survival with propofol than with sevoflurane for anesthesia maintenance, hard to believe 18 -this is more than triple the effect on mortality than a new anti-cancer drug-to no effect at all, exemplified by the largest study to date, 196,303 patients. 29 The latter is too new to have been included in any of the three meta-analyses, [14][15][16] which all state a favorable effect for propofol.…”