Radiographic Validation of the Inferior Mesenteric Artery Tie Level in Rectal Cancer Surgery

Overview

The inferior mesenteric artery is the feeding vessel for tumours in the rectum. When performing surgery for these tumours, the surgeon can cut the vessel close to the aorta or after the vessel bifurcates to the superior rectal artery and the left colic artery. A close division is termed a high tie (and the other, a low tie) and might entail a better lymph node extraction, possibly removing metastasis, but can also lead to nerve damage and e.g. bowel dysfunction. There is no clear evidence favouring either tie level, and large amounts of data are needed to establish superiority as any effects is likely to be small. One such method is to use national registries with prospectively collected data on e.g. level of tie and cancer relapse. However, it is not always easy to determine the level of tie while in the operating room and registries might also contain erroneous data. In order to determine the validity of such data, comparisons to objective measures are needed. This study is an attempt to correlate radiographic imaging to the suggested tie level, as indicated by the surgeon in the operative report and by the nationwide Swedish Colorectal Cancer Registry. If the registry variable tie level has a high correlation with imagining, researchers can more reliably use the registry to establish the benefits and drawbacks with high tie in rectal cancer surgery.

The level of tie of the inferior mesenteric artery in rectal cancer surgery has been a controversy for decades. Proponents argue that lymph node harvest and bowel length can be improved, while detractors state that nerve damage is likely and that blood perfusion to an anastomosis is compromised. Few randomized clinical trials are available and these are underpowered for long-term outcomes, mainly stating no difference. Observational studies usually indicate no differences either, while lymph node harvest might be increased (though not concerning node metastasis). Long-term outcomes such as survival and recurrence have not been investigated with proper sample sizes to detect small effects. In effect, large scale trials are unlikely to be conducted, and only population-based studies are likely to accumulate adequate statistical power. One way would be to use nationwide registry data, such as the Swedish Colorectal Cancer Registry, in which tie level as well as recurrence and survival data are recorded. However, it is not always easy to determine the level of tie while in the operating room and registries might also contain erroneous data. In order to determine the validity of such data, comparisons to objective measures are needed. This study is an attempt to correlate radiographic imaging to the suggested tie level, as indicated by the surgeon in the operative report and by the nationwide Swedish Colorectal Cancer Registry. If the registry variable tie level has a high correlation with imagining, researchers can more reliably use the registry to establish the benefits and drawbacks with high tie in rectal cancer surgery. The RAVAL study plans to include 100 patients from the University Hospitals in Umeå and Örebro, Sweden. Patients have been operated for rectal cancer and are contacted by telephone or at the postoperative visit, before their planned 1-year radiological follow-up. All patients sign informed consent for additional imaging with arterial phase contrast-enhanced computerized tomography (CT), performed at the same time as their planned follow-up (for detection of metastasis). A dedicated study radiologist interprets the images, in conjunction with the routine preoperative imaging, and determines the tie level that was performed at the rectal cancer operation. Additional parameters such as vessel anatomy and the length of the artery stump are recorded. Clinical data from operative reports and histopathology will also be noted. When the study inclusion is completed, radiological determination of tie level will be compared with the surgeon's impression of tie level as well as the tie level in the Swedish Colorectal Cancer Registry. The primary objective is to validate the level of tie in the registry, using sensitivity, specificity and Cohen's kappa. The secondary objectives include validation of tie level in comparison to operative reports, as well as artery stump length in relation to lymph node harvest, use of laparoscopy, and cancer recurrence (using appropriate techniques such as linear and Cox regression). The sample size is calculated per the primary objective, and assumes 90% sensitivity and specificity, which requires 86 and 58 patients, respectively, with a 10% margin of error. Attrition is expected, why 100 patients is determined to be adequate.