Pelvi-ureteric junction obstruction by crossing renal vessels: clinical and imaging features

OBJECTIVE To present the characteristic clinical and imaging findings of pelvi-ureteric junction (PUJ) obstruction caused by crossing renal vessels (CRV), as it presents particular features within the spectrum of congenital hydronephrosis. PATIENTS AND METHODS Between April 1982 and December 2000, 384 children underwent surgery for PUJ obstruction. In 71 (18.5%; mean age 8.5 years, range 2 months to 14 years; 49 aged> 5 years), the obstruction was caused by CRV. The data collected from the medical records of these patients were analysed for their clinical presentation and imaging findings, i.e. ultrasonography (US), intravenous urography (IVU) and diuretic renography. RESULTS The main presenting symptom was recurrent renal colic (pain, nausea, vomiting) in 59%, followed by urinary infection (UTI) in 20%, gross haematuria in 11% and an incidental diagnosis in 10%. By contrast, in the 313 children with intrinsic PUJ obstruction, renal colic was present in only 10.5%. Moreover, from 1991 to 2000, when the use of prenatal US became widespread, hydronephrosis was detected prenatally in 42 of 212 children (20%) with intrinsic PUJ obstruction, but in only two of 31 (6%) with obstruction by CRV. However, in 10 children with CRV operated on during this period, prenatal US had shown mild hydronephrosis (< 15 mm), which during the follow-up decreased until the children became symptomatic after 5-9 years (eight renal colic, two UTI). US during acute symptoms showed significant hydronephrosis (> 25 mm), and colour Doppler US of two patients directly showed the CRV. In all 71 children with CRV obstruction diuretic IVU and renography during the acute symptoms had an obstructive pattern, and in 24 renal colic was reproduced during the examination. The differential kidney function was < 40% in 11 children who presented with UTI; two required nephrectomy and in the remaining 69 an Anderson-Hynes pyeloplasty, after which none had an episode of renal colic or UTI during a mean (range) follow-up of 10.2 (2-20) years. CONCLUSIONS PUJ obstruction by CRV should be suspected in older children presenting with recurrent renal colic and hydronephrosis. Good kidney function is expected in most of these children, despite their age, because the vascular obstruction is intermittent. Mild prenatal hydronephrosis that could decrease postnatally does not exclude the possibility of vascular obstruction, which may later become symptomatic. Imaging (US, diuretic IVU and renography) during an episode of pain is essential and colour Doppler US could help to establish the diagnosis in these cases. Knowing that a child has a CRV is important for choosing an open surgical approach rather than endoscopic pyelotomy, to avoid potential complications.