Rare complications in cancer surgery

Abstract

Background: Curative cancer surgery aims to reduce mortality and morbidity related symptoms caused by the cancer. The surgery is often done at a time when the patient is presenting lesser or no symptoms and at a time when the cancer is not presenting an immediate threat to the patient’s life. The most illustrative example of this is prostate cancer where in some cases surgery is done to prevent a patient from experiencing morbidity and mortality as much as ten years from the time of surgery. Besides postoperative mortality, there are also complications that could be life threatening and cause morbidity such as venous thromboembolic events and arterial ischemic events. Surgeons must strive not only to cure the patients of their cancer but also to reduce morbidity and mortality caused by interventions with curative intent. This thesis plans to examine the risk of severe postoperative complications, in particular venous thromboembolic events and arterial ischemic events, in cancer surgery. The thesis also aims to examine the postoperative mortality in prostate cancer and its causes.

Material and methods: Data for this thesis derives from multiple Swedish population-based registers, such as the National Patient register, the Cause of Death register, the total Population register, the Cancer registry, the National Quality Registry for Prostate Cancer, and the Prostate Cancer Database Sweden. Patients operated for cancers were extracted from these databases and followed for outcomes such as death, venous thromboembolic events, and arterial ischemic events. Venous thromboembolic events are defined as pulmonary embolism and deep vein thrombosis. Arterial ischemic events were defined as myocardial infarction and ischemic stroke. In one of the studies a medical record review was performed extracting data directly from the medical records of the different regional care providers in Sweden.

Results: In prostate cancer the risk of dying within 90 days of surgery is very low, in fact is it even lower than an age matched background population due to the selection of healthy men for screening, work-up and treatment for prostate cancer. There was no difference in mortality between the different modalities of prostate cancer surgery. The most common causes of death within 90 days of the prostate cancer surgery were myocardial infarction and venous thromboembolic events, this in contrast to the Cause of Death register which has prostate cancer as the most common cause of death. Regarding arterial ischemic events in general, cancer surgery was found to lead to an overall risk increase of death during the first year that was attributable to events occurring during the post-surgery hospitalization period. However, after discharge, the overall risk of myocardial infarction was lower among the cancer surgery patients than among matched comparison participants. Regarding venous thromboembolic events there was substantial heterogeneity between the different cancer surgeries performed both in magnitude and length of lingering risk. Still in most cancer surgeries performed the risk was highest just after the surgery and discharge to fall and reach a plateau at around 90-120 days post-surgery.

Conclusions: Rare occurrences require large population-based datasets such as the ones provided in the Swedish setting to be evaluated properly. Our data suggests that the risk of postoperative prostate cancer mortality is low, and when death occurs within a year after surgery, the likely causes are thromboembolic or cardiac. We highlight the limitations in registry data in specific settings, where the question regarding cause of death within 90 days of radical prostatectomy could not have been described with registry data alone. We also describe the risk of arterial ischemic events and venous thromboembolic events following several different cancer surgeries. We conclude that for arterial ischemic events the surgery is a short-term trigger and for most surgeries the risk falls to that of the background population. For venous thromboembolic events the risk varies substantially between the different cancer surgeries and an elevated level is maintained for most cancers throughout the full postoperative year likely reflecting both the surgery and the elevated risk due to the underlying cancer.