Volatile Versus Total Intravenous Anesthesia for Cancer Prognosis in Patients Having Digestive Cancer Surgery

A Nationwide Retrospective Cohort Study

Kanako Makito, M.D., M.P.H.; Hiroki Matsui, M.P.H.; Kiyohide Fushimi, M.D., Ph.D.; Hideo Yasunaga, M.D., Ph.D.


Anesthesiology. 2020;133(4):764-773. 

In This Article


This study showed no significant association between total intravenous anesthesia and better overall survival in patients who had elective cancer surgery including esophagectomy, gastrectomy, hepatectomy, cholecystectomy, pancreatectomy, colectomy, and rectal cancer surgery. We also found that total intravenous anesthesia was not significantly associated with better recurrence-free survival by the Cox regression analysis, but that it was significantly associated with better recurrence-free survival by the instrumental variable analysis. This difference in these results may reflect the control for unmeasured confounders by the instrumental variable analysis. However, the adjusted hazard ratio (95% CI) of total intravenous anesthesia for recurrence-free survival was 0.92 (0.87 to 0.98); therefore, the influence of total intravenous anesthesia on reducing cancer recurrence was small, if any.

Many previous experimental studies have revealed a premetastatic effect of volatile anesthesia and a beneficial effect of propofol for various cancer cells. Laboratory studies of prostate or renal cancer cells have shown that isoflurane induces modulation of hypoxia-inducible factor 1-alpha (HIF-1α) or vascular endothelial growth factor, which may affect cancer recurrence after surgery.[1,16] One study of ovarian cancer cells demonstrated increased expression of genes related to metastasis after exposure to sevoflurane, desflurane, and isoflurane, while another study suggested that sevoflurane promoted metastatic potential and chemoresistance in renal carcinoma but not in non–small cell lung cancer.[17,18] In contrast, laboratory studies of propofol have shown antitumor effects in various cancers. A previous study of gastric cancer cells showed that propofol inhibited cell proliferation, invasion, and migration and promoted apoptosis.[19] Another study of non–small cell lung cancer showed that propofol disrupted upregulation of HIF-1α in a dose-dependent manner and therefore reduced the migration and invasion of cancer cells.[20]

Clinical studies have shown conflicting results regarding whether total intravenous anesthesia may have a beneficial effect on cancer prognosis compared with volatile anesthesia. A previous meta-analysis of overall mortality (including three randomized clinical trials and five observational studies) suggested that total intravenous anesthesia might lead to decreased mortality compared with volatile anesthesia.[21] This result may be attributable to one large study that showed higher overall mortality in the volatile anesthesia group than total intravenous anesthesia group. However, this study did not adjust for potential confounders including the cancer stage and preoperative comorbidities.[22] Observational studies have shown inconsistent results. Total intravenous anesthesia was not associated with overall survival or recurrence-free survival in patients having breast cancer surgery.[23] Two other studies showed inconsistent results in terms of overall survival between propofol and sevoflurane in patients having gastric or rectal cancer surgery.[24,25]

Our findings showed little association between the type of anesthesia and cancer prognosis. The advantage of our study is the larger sample size than those in previous studies and the use of instrumental variable analyses to control for unmeasured confounders. Another advantage of the present study was the inclusion of various types of cancer in digestive organs.

In instrumental variable analyses, all individuals in the study population are assumed to be compliers. This is called the "monotonicity assumption."[26,27] In the present study, compliers were those who were likely to receive total intravenous anesthesia in hospitals with a high preference for total intravenous anesthesia, whereas they were unlikely to receive total intravenous anesthesia in hospitals with a low preference for total intravenous anesthesia. Complex decision processes with multiple factors may violate the monotonicity assumption when using the physician's preference as an instrument.[26,27] However, in the present study, the proportion of total intravenous anesthesia use at each hospital as an instrumental variable may not violate the monotonicity assumption because the decision regarding the use of total intravenous anesthesia must be based only on anesthesiologists' preferences, not on patients' willingness; that is, most patients are considered to be compliers.

The current study had several limitations. First, retrospective observational studies are associated with the potential for residual confounding. We therefore performed propensity score–matched analyses, which were designed to balance variables between the two groups and thus reduce the potential measured confounding effect of each variable. In addition, instrumental variable analyses may help to account for unmeasured confounders such as laboratory data and surgical invasiveness. Second, we could identify only patients who died in a hospital in which the patients had cancer surgery; patients who died at home or in another institution could not be followed. Finally, the postoperative follow-up period was short (median of just over 2 yr); a study with a longer follow-up period of more than 5 yr is warranted.

In conclusion, the present study showed no significant difference in overall survival and little difference, if any, in recurrence-free survival between total intravenous anesthesia and volatile anesthesia.