RENAL BIOPSY: CLINICAL INDICATIONS

Serological abnormalities= hypocomplementemia, autoantibodies, circulating immunocomplexes
* the answer rates between different geographical areas are significantly different (P < 0.05)
° the answer rates between the experts group and all respondents are significantly different

• * = the answer rates between different geographical areas are significantly different (P < 0.05)

• * = the answer rates between different geographical areas are significantly different (P < 0.05)
• ° the answer rates between the experts group and all respondents are significantly different

Transplanted kidneys
After surgery. Failure of the graft to function within one week
Rapid functional deterioration after initial good function, before antirejection therapy
Slow progressive functional deterioration
Onset of nephrotic proteinuria

Statistical analysis.

Data are expressed as percentage of positive answers to each question. To determine differences between geographical areas, the repliers were grouped as follows: 1) North Europe (UK, Germany, France, Belgium, Netherlands, Switzerland, Scandinavia); 2) South Europe (Italy, Greece, Spain); 3) North America (USA, Canada); 4) Antipodes (Australia, New Zealand); 5) Others (China, Hong Kong, Brazil, Japan, Taiwan) (Table1). In addition, to compare the opinions of all with those of colleagues known for their specific experience in the field, each author was asked to indicate twenty names of "experts" taken from the list of nephrologists involved in our study. The following obtained the highest number of "nominations" and were therefore designed as "experts" group: Berthoux F, Broyer MJC, Cameron JS, Coggins CH, Cohen A, Davison AM, Glassock RJ, Hostetter TH, Korbet SM, Lai KN, Madaio M, Meyrier A, Ponticelli C, Remuzzi G, Striker G, Striker L, Suki WN, Teitelbaum I, Van Ypersele C, Winearls CG.

The comparisons between the opinions of the "geographical" groups as well as between "experts" group and all repliers were made by Fisher’s exact test.

We received back 166 forms from the following doctors: Cochran M, Dawborn JK, Elder GJ, Fassett R, Greg P, Healy H, Hodson E, Hutchinson BG, Ibels LS, Nanra RS, Nicholls K, Nicmans, O' Donnel D, Perry GJ, Petrie J B, Regby R, Roger S, Stewart JH, Thomas M, Trevillian P (Australia); Bosmans JL., Tielemans C, Christophe JL, Donck JB, Lamerie NH, Van Ypersele C, Vanholder R, (Belgium). Ruiz Centeno J, Suassuna JHR, (Brazil); Blake P, Cattran DC, Oreopulos G (Canada); Tesar V (Czesh), Sobh M (Egypt) Honkanen E, Baumelou A, Berthoux F, Broyer MJC, Druet P, Fillastre JP, Jacobs C, Laville M, Lesavre P, Mery JP, Meyrier A, Sraer JD (France); Edel HH, Jorres A, Kramer HJ, Neumann KH, Neumayer HH, Querfeld U, Schindler R, Schlille P, Schlondorff D O, Weber M (Germany); Siamopoulos KC, Vosnides G, (Greece); Chang MK, Tong MK , Lai KN(Hong Kong); Keogh JAB (Ireland); Bernheim J (Israel); Albertazzi A, Baggio B, Barbiano di Belgioso G, Cagnoli L, Campanacci L, Cianciaruso B, Cinotti G A, Coratelli P, D'Errico M, De Santo N, Di Maggio A, Di Paolo B, Ferrario F, Fusaroli M, Garibotto G, Giannattasio M, Grasso C, La Greca G, Li Vecchi M, Locatelli F, Lusvarghi E, Manganaro M, Messina G, Palla R, Pasquariello A, Piccoli G, Ponticelli C, Ragni R, Remuzzi G,Rotolo U, Sabbatini M, Sessa A, Slaviero G, Sorice P, Vercellone A, Verzetti G, Zoccali C (Italy); Kurokawa K, Tomino Y,(Japan); Herrera-Acosta J (Mexico); Henè RJ, Monnens LAH, Van Es LA (Netherlands), Bailey R, Lynn KL, Reddy J, Voss D (New Zealand); Jorstad S (Norway); Kokot F (Poland); Morgano T, Quessos J (Portugal); Lavilla Royo FJ, Montoliu-Duran J, Reddy J, Revert L, Sanz Guajardo D, Velo M (Spain); Ahlmen J, Backman U, Westberg G, Westman KWA (Sweden), Schmidli M, Wauters JP (Switzerland); Huang CC, Kao MinTsung (Taiwan); Briggs JD, Cameron JS, Davison AM, Kerr D, MacLean AG, Neild G, Swainson CP, Winearls CG, Wing AJ; Burns A (UK); Alpern RJ, Bennett WM, Bergstein J, Coggins CH, Cohen AJ, Dennis VW, Friedman E.A, Garella S, Glassock RJ, Grabes ML, Guasch A, Hamm L, Helderman JH, Hostetter TH, Korbet SM, Lazarus MJ, Madaio M, Ravenscraft M, Rinner S., Roseman M, Schwab SJ, Sedor JR, Stone DK, Striker L, Striker GE, Suki WN, Teitelbaum I, Toto RD (USA).

The answers rates to the questions are reported in the Tables 1-7.

The answers to our questionnaire provide information on the current indication for performing renal biopsy in 166 nephrology units in the world and help in understanding in which cases there is or not consensus for renal biopsy.

Biopsy in normal renal function. Asymptomatic hematuria in the patient with normal GFR is probably the most frequent presenting complaint in which renal biopsy is considered when other diagnostic techniques fail to identify the source of bleeding. However, the answers to our questionnaire indicate that only a minority of nephrologists consider isolated microscopic hematuria as specific indication to biopsy, even when the erythrocytes in urinary sediment appear to be of glomerular origin.

This trend to avoid renal biopsy is consistent with literature and is presumably based on the consideration that 25% of these patients have normal renal biopsy and that there is no effective therapy for the responsible nephropathy (usually IgA nephropathy, followed by hereditary nephritis and thin basement membrane disease) (9,11): in a prospective study of 276 native renal biopsies, biopsy for isolated hematuria laid to therapy change in only one out 36 patients (12). Yamagata et al, in a long-term study on 850 patients with asymptomatic urinary abnormalities, reported that, during the follow up period, while 44.2% of the 432 patients with asymptomatic pure hematuria showed disappearance of hematuria, 10.6% manifested proteinuria, but without renal insufficiency (13). Thus, the abstention from renal biopsy may be considered a legitimate procedure until an effective therapy of IgA nephropathy will become available (14).

The association of microscopic hematuria with episodes of gross hematuria did not substantially modify the low propensity to perform renal biopsy, unless proteinuria (of any grade) is also present: in this case, in fact, more nephrologists (47%) perform biopsy, probably because the association with proteinuria is a bad prognostic factor. In the Yamagata study (13), in the follow up of 134 patients with asymptomatic hematuria and proteinuria, renal insufficiency occurred in 14.9%.

The choice to perform biopsy in case of isolated proteinuria appeared to be mainly influenced by the level of proteinuria. Only a minority of nephrologists (12.7%) would perform biopsy when proteinuria is below 1 g/24 hours; but the percentage increases to 71.7% (range 52-87.5%) when proteinuria is between 1 and 3 g/24 hours and approximates 100% (range 90.3-100%) in case of nephrotic proteinuria. The answer rates from different areas did not differ significantly, while the percentage of the "experts" performing biopsy when daily proteinuria is lower than 1 g was significantly greater (40% vs 12.7%). On this point there is controversy also in the current literature. In fact, while Rose (9) advises not to perform the biopsy in isolated nonnephrotic proteinuria, because the prognosis is often excellent, Tisher (6) recommends it at the time of initial presentation in asymptomatic patients who remain nonnephrotic. However, in the Yagamata study (13), 10.6% of the 151 patients with asymptomatic proteinuria developed renal insufficiency, so that the authors concluded that renal biopsy should be considered at least in those with proteinuria exceeding 1 g/24 hour in order to evaluate the possibility of a treatment (13).

Biopsy in chronic renal insufficiency. In the patient presenting with chronic renal insufficiency can be difficult to determine the cause of renal disease on the basis of clinical criteria alone. According to Kobrin and Madaio a definitive pathologic diagnosis can often be ascertained by kidney biopsy, that confirms the clinical suspicion only in approximately half of the cases (11); but the same authors also point out that the procedure, in course of chronic renal failure (CRF), has a higher risk of complications, and it is unlikely that leads to modification of therapy. Similarly, Rose (9) suggests that patients with slowly progressive chronic renal failure have not to be biopsied because of little probability of finding a treatable disease. On the other hand, both Ponticelli et al. (15) and Glassock and Massry (14) distinguish between the cases with small, contracted kidneys, in which renal biopsy is of no help, and those with normal or mildly-reduced sized kidneys in which biopsy can identify the cause of functional impairment and sometimes suggest effective treatment.

The opportunity of biopsying patients with CRF was tested in our questionnaire under different clinical conditions. Only a minority of nephrologists would perform a biopsy in case of severe renal insufficiency (GFR<30 ml/min), except in patients in which renal sizes are not markedly reduced (Table 3). In this case, in fact, as many as 59.6% of participants would perform a biopsy, thus witnessing increasing awareness that some forms of advanced CRF of unknown origin hide potentially treatable nephropathies.

Also the decision to biopsy patients with mild to moderate chronic renal insufficiency (GFR 80-30 ml/min) and asymptomatic urinary abnormalities (microscopic hematuria, low-grade proteinuria) appeared to be strongly dependent on the kidney size. Interestingly, the coexistence of clinical and/or laboratory findings of diabetes disease did not reduce significantly the percentage of nephrologists favorable to the biopsy. Since no specific treatment is available for any type of diabetic nephropathy, a possible explanation is that many nephrologists believe that finding a non-diabetic nephropathy in a diabetic patient is not uncommon. Indeed, Gambara et al (16) found that 17 out of 52 patients with type II diabetes overt clinical nephropathy had glomerular disease superimposed on diabetic nephropathy. Similarly, Parving et al, in a prospective study of the prevalence and causes of persistent albuminuria greater than 300 mg/24 hr, reported that 8 out of 35 patients undergoing kidney biopsy had a variety of non-diabetic glomerulonephritis (17). Also a serological positivity for cryoglobulinemia or ANCA- or anti-DNA antibodies, although highly diagnostic, did not modify the high rate of agreement to perform the renal biopsy, probably because the assessment of the type and severity of renal lesions is considered essential to optimize the treatment (18, 19).

The association of renal insufficiency and nephrotic proteinuria, in presence of near normal kidney sizes resulted also to be a strong indication for renal biopsy, except than in case of diabetes, probably because the association of renal insufficiency, nephrotic syndrome and diabetes makes the diagnosis of diabetic nephropathy extremely probable (17).

By contrast, only a minority of nephrologists performs the biopsy in patients with renal insufficiency and very low proteinuria.

Biopsy in acute renal insufficiency. The role of renal biopsy in the diagnosis of acute renal failure (ARF) is also an important and controversial matter. The forms of ARF in which a specific treatment is more frequently indicated are glomerulonephritis, vasculitis and acute interstitial nephritis. Actually, glomerulonephritis and vasculitis are not frequent causes of ARF, accounting just for a 10% of the cases in a large series from 1956 to 1988 (20). However, when only patients selected for a renal biopsy are considered, the incidence notably increases. In a survey of the Italian Registry of Renal Biopsy, in 15,461 renal biopsies the incidence of ARF was 9,2%, with necrotizing vasculitis in 20,1% of cases, crescentic glomerulonephritis in 14% and acute interstitial nephritis in 11.3% (21). The importance of renal biopsy in influencing the treatment is also emphasized by the study of Richards et al (8), reporting that renal biopsy altered the management in 22 out of 31 cases of ARF (71%). Thus, it is not surprising that, in case of organic ARF of unknown etiology, most replied (87.1%) to be favorable to perform renal biopsy, even if 40.3% of them would prefer waiting up to 4 weeks of spontaneous non recovery (Table 4). As the "experts group", we disagree to wait so long, because in many circumstances an early treatment can prevent the progression of the renal lesions (22).

The majority of eases of ARF are due to acute tubular necrosis In a prospective study in adult patients admitted to the 13 tertiary-care hospitals of Madrid (population 4.2 million) in a period of nine months, Liano et al (23) reported 748 cases of ARF with an incidence of 209 cases per million population (pmp) early. The incidence of the different forms of ARF were: ATN 88 cases pmp: prerenal ARF 46 cases pmp: acute-onset chronic renal failure (CRF) 29 cases pmp: obstructive ARF 23 cases pmp; acute interstitial nephritis (AIN) 15 cases pmp; vasculitis 3.5 cases pmp; primary glomerulonephritis 2.8 cases pmp. The most frequent causes of ARF were: ATN 45%, prerenal ARF 21%, acute-onset CRF 12,7%, and obstructive ARF 10% (23). The use of renal biopsy in patients with ARF was usually reserved by the: (a) absence of an obvious cause of ARF. (b) extrarenal manifestations suggesting the possibility of a systemic disease. (c) symptoms suggesting a cause different from ATN, and (d) anuria persisting more than three weeks (23).

In the cases in which the etiology is not known, despite careful history, accurate physical examination and availability of laboratory studies, renal biopsy becomes crucial for a correct diagnosis, for assessment of prognosis and even for deciding on the therapeutic procedure (24,25)

The importance of renal biopsy in influencing the treatment and, consequently. the outcome of several forma of ARF has been emphasized by a recent prospective study of 276 native renal biopsies performed in 266 patients in a single center: in 22 out of 31 cases of ARF (71%), renal biopsy altered the management of the patients: thus, the potential benefit overcomes the risk o(the procedure (24). The diagnoses for which specific treatment is more frequently indicated are glomerulonephritis, vasculitis and acute interstitial nephritis

Actually glomerulonephritis and vaseulitis are no frequent causes of ARF, accounting only for about 10% of the cases in a large series from 1956 to 1988 (27). however, when only patients selected (or renal biopsy are considered, the incidence of glomerulonephritis and vaseulitis becomes elevated. In a survey o( the Italian Registry of Renal Biopsy, in 15,461 renal biopsies performed over a period of seven years (1987 to 1993), the incidence of ARF was 9.2%; 20.1% of the ARF cases were secondary to necrotizing vasculitis, 14% to crescentic glomerulonephritis, 11.3% to AIN and 7,9% to ATN (28).

The histological counterpart of most eases of ARE due to glomerulonephritis and vasculitis is crescentic glomerulonephritis, usually occurring on a background of focal necrotizing glomerulonephritis. Crescent formation is a non-specific response to severe injury to the glomerular capillary wall. The lesion in the capillary wall causes outflow of plasma products into Bowman's space with subsequent fibrin formation, influx of macrophages and T cells. and release of proinflammatory cytokines. The stage of active inflammation is often followed by the development of fibrocellular and fibrous crescents due to fibroblast proliferation driven by fibroblast growth factors and transforming growth factor-beta (29).

The histological assessment of the stage of the glomerular lesions is very important clinically because fibrous crescents suggest that the disease is not likely to respond to immunosuppressive therapy. Crescent formation is common in microscopic renal vasculitis, idiopathic rapidly progressive glomerulonephritis antiglomerular basement membrane disease. systemic lupus nephritis, but it can occur with any form of severe glomerular disease. The diagnosis of crescentic glomerulonephritis traditionally lies on the finding of crescents at least two cells thick affecting more than 50% of the glomerulus in more than 50% of glomeruli (30).

Noncrescentic glomerulonephritis may also cause ARF. This is the case of diffuse proliferative glomerulonephritis (such as acute postinfectious glomerulonephritis, lupus nephritis, acute cryoglobulinemia), IgA nephropathy (in presence of macroscopic hematuria) and disorders associated with glomerular thrombosis (lupus nephritis, hemolytic-urcmie syndrome, thrombotie-thrombocytopenie purpura, disseminated intravascular coagulation). Although in some of these disorders the diagnosis can be confirmed serologically, renal biopsy can help to attests the activity and the severity of the disease, thereby greatly influencing the

Drug-induced AIN is a frequent cause of ARF and commonly has systemic manifestations of an allergic process (31).

First described with the use of methicillin, it has been observed following the use of many drugs, such as penicillins, cephalosporins, rifampicin, sulfonamides (including furosemide, thiazides, trimethoprim-sulfamethoxazole), cimetidine, allopurinol, ciprofloxacin and even nonsteroidal anti-inflammatory drugs (NSAID) (31). The clinical diagnosis can be suspected in presence of (a) acute rise in plasma creatinine temporally related to a drug or infection; (b) fever; (e) rash; (d) urine sediment revealing white cells, red cells, and white cells casts; (e) eosinophilia and eosinophiluria; (f) mild proteinuria; (g) signs of tubular disfunction.

However in some cases, such as the form induced by NSAIDs, most of these symptoms are typically absent or different (proteinuria is often nephrotic in NSAIDs-induced AIN) (32). Indications for biopsy include uncertainty of the diagnosis, advanced renal failure, or lack of spontaneous recovery following cessation of drug therapy [13]. The AIN is not dose-dependent and its onset may occur three to five days after the drug therapy, but many times alter weeks or even months (such as in case of NSAIDs). The major histologic picture is interstitial edema and interstitial infiltrates of T lymphocytes and monocytes, sometimes eosinophils, piasmacells and neutrophils, with minimal changes in the glomeruli. When infection is responsible for AIN, neutrophils are present in great number.

In a recent study we have evaluated the morphology of intrarenal vessels, as assessed by light microscopy, in renal biopsies obtained immediately before removal of the kidney for donation. The age-related changes in basal glomerular filtration rate (GFR) and renal plasma flow (RPF), have been extensively investigated (33-39). All these studies document a significant reduction of both GFR and RPF in the healthy advanced age, even though the absolute values result particularly variable.

A relevant question is whether the decrease in renal hemodynamics in physiologic elderly is irreversible, as a result of morphological changes, or rather functional because of an imbalance between vasodilating and vasoconstricting factors, and, consequently, potentially reversible. This difference is critical since it is likely associated with a different renal adaptation to the acute ischemic injury, which is major cause of acute renal failure in elderly subjects (40).

For the evaluation of ageing on renal vascular pathology, baseline kidney biopsies were reviewed from two groups of young (n=15) and elderly (n=11) healthy kidney donors.

Both young and elderly subjects were recruited in the course of the clinical assessment of suitability as potential kidney donor. Extensive diagnostic evaluation was carried out to exclude unsuspected, and potentially relevant, renal and extrarenal diseases. Inclusion criteria were: (a) absence of personal and family history of diabetes mellitus, hypertension, hyperlipidemia, obesity, cardiovascular or renal disease; (b) normal physical examination; (c) absence of fasting glycemic values greater than 120 mg %, absence of arterial blood pressure levels greater than 140/90 on three separate occasions, plasma levels of cholesterol and triglycerides below 200 and 170 mg/dL, respectively, plasma lipoprotein HDL > 40 mg/dL; (d) absence of hypertensive or atherosclerotic retinopathy; (e) no evidence of cardiac insufficiency or hypertrophy or valvular heart or atherosclerotic vascular disease, as assessed by color Doppler echocardiography ; (f) normal urine chemistry and sediment and negative urine culture; (g) serum creatinine £ 1.2 mg/dL; (h) absence, at both ultrasound evaluation and urography, of renal cyst, nephrolithiasis, pyelonephritis, urinary tract ectasis, residual volume of urine in the bladder (> 50 mL), or prostatic hypertrophy; (i) daily proteinuria < 150 mg; (l) absence of any medication, (m) absence of drug or alcohol-dependency; and (n) no smoker habits, as established according to a recent study (44). All participants gave their informed consent. The subjects were divided into healthy young (age >18 and £ 39 years) and elderly (age ³ 65 years old).

All needle biopsies had been performed during the nephrectomy procedure, immediately before removal of the kidney for donation using an automated punch device (BioptyCut, Radioplast, Bromma, Sweden) with a 18-gauge needle. Specimens were processed for light microscopy as previously described (46, 47). Light microscopical changes had been semiquantitatively estimated according to a scoring system partly based on the Banff schema (48, 49) grading (0-3) the following variables: arteriosclerosis (as), arteriolar hyalinosis (ah), glomerulosclerosis (gsc), interstitial mononuclear cell infiltration (mi), and interstitial fibrosis/tubular atrophy (if/ta). In particular, arteriosclerosis was graded on the basis of the severity of the fibrointimal thickening of arteries. Arteriolar hyalinosis was defined as PAS-positive insudation or hyaline thickening of the arterioles and scored on the basis of the percentage of the circumference affected. In order to increase the sensitivity of the method, the score of arteriosclerosis (as) of each artery within each biopsy was recorded and the sum of artery scores added, giving a ratio of as score/number of arteries in each biopsy. An identical approach was made for the scoring of arteriolar hyalinosis. Glomerulosclerosis (gsc) was defined according to the extent of increase in mesangial matrix. The fraction of globally sclerosed glomeruli (gsg) was calculated in the section with the highest number of glomeruli in each biopsy; a combined score of glomerulosclerosis (gsc) was then calculated combining the grades of sclerosis in preserved glomeruli and the fraction of gsg (46, 47). A combined score of if/ta was given in each case. The chronicity index (CI) of each baseline biopsy in the present paper is defined as the sum of the scores of as + gsc + mi + if/ta.

Table 8 depicts the main aging-induced changes of renal morphology obtained in the groups of young and elderly donors shortly before the removal of the kidney for donation. As expected, the morphologic changes were of low grade. Of note, among the different variables scored, only arteriosclerosis of intrarenal vessels (interlobular and arcuate arteries) and interstitial fibrosis/tubular atrophy were significantly greater in elderly. The grading of the other variables considered led in fact to results that were not statistically different in elderly and young kidney donors. Nevertheless, the sum of the scores, that is, the chronicity index, was significantly higher in the elderly group.

The results of the kidney biopsy study support the hypothesis that the age-related impairment of renal hemodynamics may be primarily dependent on a greater degree of arteriosclerosis in interlobular and arcuate arteries and the associated interstitial fibrosis/tubular atrophy. Of note, this abnormality, although of low grade, mainly accounts for the higher chronicity index; in fact, the score of the other variables examined, such as glomerulosclerosis and interstitial mononuclear infiltration, did not significantly differ in young and old donors.

Taken together, these data contribute to elucidate the independent role of age in the pathophysiology of the enhanced risk of acute renal failure (ARF). It is well known that hypovolemia and renal hypoperfusion are the most prominent risk factors for ARF in elderly patients (40). According to the findings of the current study, the susceptibility to hemodynamically-mediated ARF is enhanced because aging per se is coupled with a physiologic decrease of renal perfusion of organic nature. The presence of cardiovascular risk factors, including chronic cigarette smoking (44), may further increase this susceptibility.

Renal allograft biopsy. The usefulness of renal biopsy in transplanted kidney represents a widely accepted exception to contraindication of percutaneous renal biopsy in a single-functioning kidney. The clinical cases in which the renal biopsy is usually considered in transplanted patients are delayed graft function, acute and chronic renal functional deterioration, and onset of nephrotic proteinuria. Delayed graft function after surgery was considered by the majority as a very early indication to renal biopsy (60.9% would perform it within one week of failure of the graft to function) (Table 5). This is not surprising, because, since acute tubular necrosis is the most frequent cause of the failure of the graft to function in the immediate posttransplantation period (30% of cases, range 10-60%), only renal biopsy allows to diagnose acute rejection nonfunctioning renal allograft.

Also in case of rapid functional deterioration, most nephrologists (68.1%) would perform biopsy before starting antirejection therapy, excepting those included in the subgroups South Europe and "Others", in which 50% and 57.1%, respectively, decide for biopsy only after failure to respond antirejection therapy. A possible compromise could be to empirically treat with pulse corticosteroids pending the biopsy (50).

The slow progressive deterioration of allograft function also represents a clear indication to perform the biopsy for 89.2% of nephrologists (Table 5). This high rate is probably due to the high incidence of cases of chronic cyclosporine nephrotoxicity, since in the other forms of late chronic dysfunction there is no effective therapy.

Even in case of occurrence of the nephrotic syndrome or nephrotic-range proteinuria, the majority (90.8%) was favorable to biopsy. despite no specific treatment is effective in transplant nephropathy and in most cases of recurrent glomerulonephritis. Notwithstanding this, agreement for biopsy is reported even in the current literature (8,15).

The answers to the questions regarding the technical aspects of renal biopsy showed that most nephrologists perform the procedure with automatic needles, under guidance of ultrasound and a fine locating needle, and placing a small sandbag under the patient’s abdomen (Table 6). Only a low percentage (43.4%) check with immediate microscopic confirmation (by transillumination) the adequacy of the specimen, despite this simple and inexpensive precaution helps in avoiding to repeat the entire biopsy procedure in case few or no glomeruli are present.

A specimen for electronic microscopy is routinely obtained in 57.1% of centers – up to 91.7% in Australia and New Zealand – and electronic microscopy is routinely performed by 61% of them, independent of the diagnosis obtained by immunofluorescence and light microscopy (Table 7). This information confirms the growing importance of the electronic microscopy in the examination of renal biopsies (5), as emphasized by Haas (51) in a recent study on 233 native renal biopsies. This study, in fact, showed that in 45% of cases electron microscopy was absolutely needed to make the final diagnosis, in 21% providing important confirmatory data and in 3% adding an unrelated diagnosis. Immunofluorescence is widely preferred to immunoperoxidase, probably because the disadvantages of the latter technique over immunofluorescence (increase in preparation time and added expense) are considered greater than the advantages (greater sensitivity and permanency of the staining) (8). Only in few centers (28.9%) the number of consecutive sections routinely cut for light microscopy is less than 10, confirming a diffuse awareness of the risk of missing focal lesions when an insufficient number of sections are examined (52)

Only a small percentage of nephrologists replied to routinely perform ultrasound control of the kidney after the biopsy. In our opinion this is surprising, since clinically silent perirenal hematomas are extremely frequent (up to 85% of cases) and renal biopsy has been recently proposed as an outpatient procedure (53-56). According to the finding that the simple observation of patients for 8 hours or less is not sufficient to avoid missing more than 20% of complications (57), in our unit we perform routinely ultrasound control of the biopsied kidney soon after the procedure, three hours later, and, in case of perirenal hematoma, 12 hours later. This protocol allowed us in at least three occasions to detect and treat early initially asymptomatic bleeding that subsequently required blood transfusion, thereby avoiding severe systemic consequences. Therefore, we would suggest a routine use of post-biopsy renal ultrasound, particularly if the biopsy is performed in outpatients.

Although the answers to the questionnaire emphasize the primary role of the pathologist in the interpretation of the renal biopsy, it is also important to note that the diagnosis is often made in a biopsy meeting. This is in agreement with the observation of D’Agati (58), that a renal biopsy cannot be correctly interpreted without maintaining an open line of communication between nephrologist and nephropathologist and that the nephrologist’s understanding of the pathologic evaluation is always greatly enhanced by viewing the biopsy with the pathologist.

The data we have collected throughout the world provide novel information on the current indication for renal biopsy. We have tried to summarize them in Table 9, in which the indications were defined as strong, controversial and weak on the basis of the answers rate. Thus, although there is no doubt that the performance of a renal biopsy should always be guided by the individual clinical situations and by careful balancing potential risks and benefits (10), we believe that the results of this survey will represent a reliable guideline that will help clinicians to correctly decide when the evaluation of renal histology can contribute to patient management.