A prospective study of the predictive power of spiral computed tomographic angiography for defining renal vascular anatomy before live-donor nephrectomy.

OBJECTIVE

To determine the accuracy of spiral computed tomography (CT) imaging of donor venous anatomy by comparing CT angiography (CTA) and operative findings, for both laparoscopic (LDN) and open donor nephrectomy.

PATIENTS AND METHODS

LDN presents unique surgical challenges, particularly with complex venous or arterial anatomy. The limitations of surgical access, poor visibility of the superior and posterior borders of the renal vein during LDN, and the variability of venous anatomy in this region, contribute to the difficulty of LDN, underlining the importance of imaging beforehand. Forty live donors (mean age 46 years, sd 11; 65% female) were assessed by CTA before donation. Scans were reported by the same radiologist. The number and diameter of 'predicted' renal arteries, veins and renal vein tributaries were documented. The donor kidney was removed by two consultant surgeons, and after back-table perfusion the same details were recorded and taken as the 'reference' findings. Tributaries of <1 mm diameter were not recorded. The right kidney was retrieved in seven patients; 25 of the 40 kidneys were retrieved by LDN and the other 15 by open surgery.

RESULTS

In all, 48 actual renal arteries were identified at nephrectomy; of these, 47 were predicted by CTA. Likewise, 41 actual renal veins were found at nephrectomy, 40 of which were predicted. The overall accuracy of spiral CTA in predicting the presence or absence of renal vein tributaries was 83% for gonadal and adrenal veins, and 75% for lumbar veins. There were seven false-negative lumbar veins found at nephrectomy; in these cases the CTA films were retrospectively examined, and five of these seven veins were identified. The predicted renal vein tributary diameter correlated poorly with the measured diameter at nephrectomy.

CONCLUSIONS

Assessing potential renal donors before surgery with spiral CTA provides an accurate prediction of the presence or absence of the gonadal and adrenal vein, but is less accurate for predicting lumbar veins. This is especially pertinent as the posterior lumbar tributaries have the most intra-individual variation, and are the most difficult to display and control at LDN. This highlights the need for meticulous dissection of the renal vein, particularly along its posterior wall.