The suggested modification of our algorithm for managing patients with benign lateralizing haematuria (BLH) to include CDUS before proceeding to helical CT or angiography merits careful consideration. CDUS is safer, less expensive and less time-consuming than helical CT or angiography, especially if it is used at the same time as traditional grey-scale US, and vascular malformations such as small to medium-sized congenital AVMs may be detected. For these reasons, Brown and Matthews proposed its routine use in the investigation of patients with BLH before helical CT or angiography, although they do not specify if it should be before or after endoscopy.
If its use were confined to patients with persistent or recurrent bleeding after endoscopy it may well reduce the requirement for helical CT in those patients with a positive scan. However, these patients are uncommon and we consider that the additional cost of routine CDUS after endoscopy can only be justified if a negative result allows angiography to be safely omitted. However, there is insufficient published evidence to be reassured by a negative Doppler scan. Takebayashi et al. reported the use of CDUS in the diagnosis of congenital renal AVMs in a group of patients presenting with haematuria. We do not consider that this small study alone justifies the omission of angiography if the CDUS is negative. Although the value of CDUS in the diagnosis of iatrogenic AVMs in transplant kidneys is well documented, this experience cannot be transferred to congenital AVMs in native kidneys for two reasons; first because of the different characteristics of these two types of AVM, and second because of the difficulty in clearly insonating the more deeply situated native kidneys, as it is subject to respiratory excursion [1,2].
Routine CDUS and grey-scale US before endoscopy would be cost-competitive only if both endoscopy and helical CT are avoided in patients with a positive scan. However, helical CT and angiography would still be necessary in patients with persistent or recurrent bleeding and a negative CDUS, for the reasons cited above. Furthermore, because most congenital AVMs are located in the lamina propria immediately below the transitional epithelium of the collecting system, it is possible that smaller AVMs are being treated endoscopically . Diagnosis of these lesions by CDUS before endoscopy should therefore lead to a decision between treatment by embolization or whether endoscopic management should be considered. Overall we agree with Brown and Matthews that CDUS warrants a place in the algorithm for the management of BLH and should be used instead of grey-scale US when available. However, we feel that angiography is still indicated in the presence of a negative CDUS and persistent or recurrent bleeding after endoscopy.