having 33 denervations (eight bilateral and one repeat) with a median follow-up of 39.5 months. Most patients had no identifiable underlying cause although many had initially been diagnosed as having stone disease (42%) or pyelonephritis (25%), but with no corroborative evidence. All patients had been extensively investigated and had normal urine samples and cytology, and no abnormality on intravenous urography, renal tract ultrasonography and isotopic renography. Twenty-four renal denervations (73%) were followed by recurrent ipsilateral pain at a median (range) of 11 (0-120) months after surgery. Nine denervations (25%) in six men and two women were curative (median follow-up 16.5 months). Of those with recurrent pain, nine (38%) proceeded to nephrectomy, of whom three then developed loin pain on the contralateral side and two developed disabling wound pain. The analgesic requirement was less after eight non-curative denervations. There were no significant postoperative complications. CONCLUSIONSRenal denervation has only a 25% success rate for managing pain associated with LPHS and should be used cautiously for this indication. Men had more benefit from the treatment; a third of patients had less requirement for analgesic after non-curative denervation. KEYWORDSrenal denervation, sympathectomy, loin pain haematuria syndrome, chronic pain OBJECTIVETo evaluate the outcome of renal denervation for the treatment of loin pain-haematuria syndrome (LPHS), a rare syndrome of unknown cause associated with debilitating and intractable loin pain. PATIENTS AND METHODSThe case notes of 32 patients having 41 renal denervations were reviewed. Data collected included patient demographics, possible causes, cure or not after renal denervation, time to recurrence of pain after denervation and further operative intervention for managing LPHS. RESULTSFull data were available for 24 patients (13 women; median age 43 years, range

The greatest change in GFR in response to treatment with cyclosporin occurs in the first 3-6 months and the magnitude of the decrement in the first year (or perhaps the first few months) appears to be a vital indicator of future problems. However, the apparent stabilization of renal function, particularly when monitored only by plasma creatinine, can conceal progressive tubulointerstitial injury, and increasing proteinuria is an ominous sign. Although lower doses of cyclosporin and careful monitoring of renal function may be helpful, there is at present no pharmacological intervention to protect or reverse the reduction in GFR that occurs. We believe that the vascular lesion induced by cyclosporin is fundamental, with early and initially reversible cyclosporin-induced vasospasm leading to progressive vascular damage with activation of endothelial cells and increased platelet interactions. Amongst other determinants, the renal response to this vasculopathy will depend on the balance between the presence of vasoactive factors with the vasoconstrictors promoting interstitial fibrosis and the vasodilators inhibiting proliferation. It is likely that the kidneys of heart-transplant recipients are chronically ischaemic and as a consequence their renin-angiotensin systems massively activated, which may further sensitize their kidneys to cyclosporin. Overproduction of angiotensin II, associated with the DD ACE genotype, has already been associated with poor prognosis in diabetic and IgA nephropathy. It is interesting to speculate that this ACE genotype, which is associated with a poor outcome in non-ischaemic heart disease can influence renal sensitivity to cyclosporin and predict the development of morphological injury. Extension of these experimental findings into the clinical arena with a placebo-controlled trial of early introduction of ACE inhibitor therapy in recipients of cardiac transplants would be timely.

We recommend that patients be managed in a regional pain clinic, where a multidisciplinary approach promotes self-management of pain. Patients who were able to accept conservative treatment, and taper or withdraw opiate analgesia, had a better prognosis.

SUMMARY1. Regional heterogeneity of endothelial function exists but its role in the local regulation of vascular tone is uncertain. This heterogeneity may be very important in the control of the glomerular filtration rate (GFR) in which the differential tone in the afferent and efferent arterioles is crucial.2. When an endothelium-independent vasodilator, prostacyclin (PGI2) or nitroprusside, was infused into anaesthetized rabbits there were dose-dependent falls in both mean arterial pressure (MAP) and GFR; PGI2 (04 nmol kg-1 min-') altered MAP and GFR by -18 5 + 3-6 % (mean + S.E.M.) and -37-7 + 13-3 % respectively and nitroprusside (30 nmol kg-' min-') by -29 7+3 1% and -67-0+2-4%. In contrast infusion of an endothelium-dependent vasodilator, acetylcholine (ACh) or substance P, produced dose-dependent decreases in MAP but dose-dependent increases in GFR; ACh (10 nmol kg-1 min-) -15 1 + 2 0 % and + 43 8 + 16-5 % and substance P (30 nmol kg-' min-') -187+ 19% and +45-3+231% respectively. The effects of endothelium-dependent and independent vasodilators on GFR was significantly different (p < 0 005). 3. Simultaneous administration of indomethacin, Methylene Blue or NGmonomethyl-L-arginine (L-NMMA), inhibitors of cyclo-oxygenase and endotheliumderived relaxing factor (EDRF) respectively, attenuated or reversed the effect of ACh (10 nmol kg-' min-') on MAP and GFR.4. These data suggest that endothelium-dependent vasodilatation in the kidney has a heterogeneous effect on the renal microvasculature, exerting a preferential effect on afferent glomerular arterioles and thereby preserving GFR despite the fall in MAP. If correct, this has important implications for the regulation of GFR.

No abstract