Kidney International, Vol. 38 (1990), pp. 529—543

NEPHROLOGY FORUM

Renal biopsy Principal discussant: MICHAEL P. MADAIO New England Medical Center and Tufts University School of Medicine, Boston, Massachusetts

After extensive discussion with the patient, a renal biopsy was performed to determine whether a renal source of the hematuria could be identified. The biopsy demonstrated mild mesangial expansion on light microscopy; immunofluorescence microscopy showed IgA depos-

Editors JORDAN J. COHEN JOHN T. HARRINGTON JEROME P. KASSIRER NICOLAOS E. MADIAS

its in a focal distribution in all glomeruli. Because of this information, he

did not donate a kidney to his sister. Patient 2. A 47-year-old woman with systemic lupus erythematosus

was admitted to the NEMC for renal biopsy. The initial clinical manifestations of lupus occurred 30 years previously in the form of thrombocytopenic purpura and proteinuria. Thrombocytopenia im-

Managing Editor

proved with high-dose prednisone therapy, but proteinuria (200—600 mg/24 hr) persisted. The serum creatinine was less than 1.0 mg/dl on numerous occasions. Fifteen years ago, she developed Sjogren's syn-

CHERYL J. ZUSMAN

drome but was otherwise asymptomatic. At that time, the 24-hour urinary protein excretion was 6.2 g, the BUN was 9 mg/dl, and the

State University of New York at Stony Brook

serum creatinine was 0.6 mg/dl. Urinalysis disclosed: specific gravity, 1.014; 3+ heme; 4+ protein; and many red blood cells with occasional granular casts. An LE prep was positive. The Sjogren's syndrome was treated symptomatically. Eight years prior to admission, she developed cutaneous vasculitis of

and Tufts University School of Medicine

the arms and legs, bilateral pulmonary infiltrates, and anemia. The

serum creatinine was 1.0 mg/dl, the BUN was 15 mg/dl, and an Case presentation Patient 1. A 44-year-old man was evaluated as a potential renal transplant donor for his sister, He had been in good health except for upper gastrointestinal bleeding 4 years previously. At that time, endoscopy demonstrated gastritis and a duodenal ulcer; he was treated with ranitidine for 3 months, and the symptoms did not recur. He was an alcoholic, but he had not used alcohol in more than 5 years; he attended Alcoholic Anonymous meetings regularly and he was employed fulltime as a truck driver. He had no known allergies.

anti-nuclear antibody titer was 1:65,000. Urinary protein excretion was 700 mg/24 hours. An open lung biopsy revealed pulmonary hemosiderosis; renal biopsy showed membranous nephropathy with mild mesangial proliferation. Prednisone was administered (60 mg/day). Hypertension

required therapy with cionidine and hydrochlorothiazide. The prednisone was slowly tapered over the next 2 years, she became normotensive, and all of the extrarenal manifestations of lupus resolved. Four years prior to admission, the patient had another flare of lupus erythematosus characterized by lymphadenopathy, splenomegaly, pul-

monary infiltrates, and anemia. At that time, her blood pressure was

Physical examination in the Nephrology Clinic at New England normal, serum creatinine was 0.8 mg/dl, and 24-hour urinary protein Medical Center (NEMC) disclosed: weight, 100.6 kg; blood pressure, 110/80mm Hg; pulse, 68 beats/mm. Laboratory examination was within

normal limits: hemoglobin, 14.0 mg/dl; creatinine, 0.8 mg/dl; and normal liver enzymes. Urinalysis revealed a specific gravity of 1.030;

excretion was 1.4 g. Renewed therapy with 60 mg of prednisone daily resolved the extrarenal disease activity. The prednisone was slowly tapered over the next year; extrarenal manifestations did not recur. Over the past 18 months, the patient has had pedal edema for the first

pH, 6.0; trace protein; and negative heme. The urine sediment con- time, and she has gained 12 pounds. Her blood pressure became tained 0—5 epithelial cells, no red blood cells, and an occasional white blood cell/high-power field. He was eager to become a transplant donor,

elevated, and furosemide was administered. She was referred to NEMC

and he was referred to a urologist and a social worker for further

complaints. Physical examination revealed: blood pressure, 120/80 mm Hg; pulse, 88 beats/mm; and 2+ pedal edema. The rest of the examination was normal. Urinalysis disclosed a specific gravity of 1.015; 4+ protein; and pH, 5.5. Microscopic examination of the urine revealed 0—2 white blood cells, 0—3 red blood cells, occasional granular casts, and many oval fat bodies; urinary protein excretion was 5.3 g/24 hr. The serum creatinine was 1.2 mg/dl; serum albumin, 2.2 g/dl; anti-dsDNA antibody level, 3.8 mg/dl; CH50, 168 U (normal, 150—200 U); C3, 0.85 mg/mi (normal, 0.87—2.20 mg/mI); and C4, 0.17 mg/mI (normal, 0.15— 0.54mg/mi). Renal biopsy demonstrated membranous glomerulonephritis with segmental glomerulosclerosis and interstitial fibrosis, consistent with class-V lupus nephritis. Patient 3. A 23-year-old man who previously had been in excellent health presented to the emergency ward of a local hospital with a 2-day history of nausea, vomiting, chills, and myalgias, and a one-day history of bilateral flank aching and dark urine. He had no history of hematuria, renal disease, hypertension, diabetes, or substance abuse, nor had he had recent symptoms of an upper respiratory or gastrointestinal illness.

evaluation. Repeat urinalysis by a urologist revealed a specific gravity of 1.014; 2+ heme; and trace protein. Microscopic examination showed 5—10 red blood cells and 0—1 white blood cells/high-power field and no casts.

Intravenous pyelogram and cystoscopy were normal, and neither indicated the source of the hematuria. The patient returned to the Nephrology Clinic, and repeat urinalysis on two separate occasions demonstrated no hematuria. Review of his record from another hospital revealed that he had had trace to 1 + hematuria on dipstick examination when hospitalized for peptic ulcer disease.

Presentation of the Forum is made possible by grants from Pfizer, Incorporated; Merck Sharp & Dohme International; and Sandoz, Incorporated.

© 1990 by the International Society of Nephrology

for further evaluation. Other than the edema, she had no acute

529

530

Nephrologv Forum: Renal biopsy

At the onset of symptoms, he had been attending an engineering scopic appearance of the hypertensive kidney from autopsy conference on Cape Cod. material with the usual postmortem changes, that at the beginOn examination his blood pressure was 130/70 mm Hg and the pulse ning of this study some of the findings due to the early fixation was 76 beats/mm. He had mild flank tenderness and no edema. The examination was otherwise nOrmal. Urinalysis disclosed: specific grav- (a matter of seconds) seemed so conspicious that they were at ity, 1.012; pH, 5; 3+ protein; 4+ heme; and no glucose. The urine first believed to be pathologic" [8]. sediment contained 20—30 red blood cells. 4—5 renal tubular epithelial The percutaneous approach became routine after 1950 with cells, and no casts or crystals. The hematocrit was 44%, white blood the reports of Perez-Ara [9], Iverson and Brun [10], and Alwall cell count, 18,000/mm3 with a normal differential; platelet count,

[11]. In 1954 Kark and Muehrcke modified the procedure by positioning the patient in the prone rather than in the upright electrolytes were: sodium, 139 mEq/liter; potassium, 4.6 mEq/liter; chloride, 110 mEq/liter; bicarbonate, 24 mM/liter; albumin, 3.8 gldl; position, using an exploring needle to localize the kidney prior calcium, 5.5 mg/dl; phosphorus, 6 mg/dl; and creatine phosphokinase, to the insertion of a biopsy needle, and utilizing the Franklin1141 U/liter. A renal ultrasound examination demonstrated no hydro- modified Vim-Silverman needle [12]. These modifications renephrosis; the left kidney was 12.3 cm and the right kidney 11.9 cm in length. A renal scan showed blood flow bilaterally but diminished sulted in a greater likelihood of procurement of tissue and a larger core of tissue for microscopic evaluation. Since these excretion. On further questioning the patient reported that 2 days previously, initial reports, technical innovations principally have focused while walking along the beach, he became thirsty and drank from a around more accurate localization of the kidney prior to biopsy rinsed jug that previously had contained antifreeze. A serum osmolality was measured at 299 mOsm/liter. The serum also was evaluated for and on immediate microscopic inspection of the tissue obethanol, methanol, and ethylene glycol, but they were not found. He tained. I will discuss these aspects in a moment, but before proceeding I would like to consider the circumstances in which underwent emergency hemodialysis for possible ethylene glycol poisoning, and he received 1 g of methyiprednisolone on 3 successive days as percutaneous renal biopsies generally have been considered therapy for possible rapidly progressive glomerulonephritis. 'contraindicated." Five days after the onset of symptoms, he was transferred to NEMC. 277,000/mm3; BUN, 31 mg/dl; and serum creatinine, 6.7 mgldl. Serum

On admission his physical examination was unremarkable; blood pressure was 120/80 mm Hg. Laboratory studies revealed: BUN, 71 mg/dl; serum creatinine, 12.2 mg/dl; hematocrit, 38%; and creatine phosphokinase, 866 U/liter. Urinalysis disclosed: specific gravity, 1.016; pH, 5; 3+ protein, 3+ heme; 10—20 white blood cells, 100 red blood cells, and 0—3 granular casts/high-power field. Renal biopsy demonstrated acute tubular necrosis with many calcium oxalate crystals in the interstitium. The patient underwent three additional hemodialysis treatments, and his renal function gradually improved. Three months later the patient's serum creatinine was 1.0 mgldl, and his urinalysis and sediment were normal.

Contra indications A major contraindication to performing percutaneous renal

biopsy is a bleeding diathesis. In most circumstances, the

bleeding disorder can be corrected prior to and for a few days after the biopsy. If the disorder cannot be corrected and the biopsy is deemed indispensable, an open surgical biopsy should be considered because the site can be visualized directly and bleeding can be controlled locally [13, 14]. Open biopsy can be performed with the patient under general or local anesthesia. Chodak and colleagues utilized mild sedation, local anesthesia, Discussion and direct visualization in 123 high-risk patients [14], including DR. MICHAEL P. MADAIO (Division of Nephrology, New patients with bleeding abnormalities (37), uncontrolled hyperEngland Medical Center, and Associate Professor of Medicine, tension (26), anatomic abnormalities (19), previous technical Tufts University School of Medicine, Boston, Massachusetts): failure (19), uncooperative behavior (13), pregnancy (5), and The use and interpretation of the renal biopsy evoke strong unspecified conditions (21). Adequate tissue was obtained in all opinions and confrontations among even the most docile ne- patients (mean number of glomeruli = 54). Complications were phrologists [11. The decision to biopsy the kidney raises many observed in 6 patients: 3 had wound hematomas, 2 had wound issues, including indications and contraindications, optimal infections, and one required surgery for repair of a colocutanetiming, choice of technique, complications, interpretation of the ous fistula; no deaths were reported. Transvenous renal biopsy may be another alternative in material, and utilization of information. Several textbooks are devoted to various aspects of these issues [2—5]. In this Forum, high-risk patients. In a preliminary report, Mal and coworkers I will confine my remarks to problems that the nephrologist performed renal biopsies through the internal jugular vein using commonly encounters in routine practice. I will discuss se- a 9 F catheter with a 66 cm needle under vacuum aspiration in lected technical and clinical aspects of the renal biopsy proce- 36 patients [15]. Of the 36, 12 had a severe clotting disorder, 4 dure and the informational content of the biopsy. I also will had chronic respiratory failure, 2 had uncontrolled hypertenconsider indications for renal biopsy in patients with several sion, 12 needed a liver and kidney biopsy, and conventional major renal syndromes, including isolated hematuria, mild to biopsy failed in 6. Adequate tissue (mean number of glomeruli SD = 9.5 5) was obtained in 25 of the 36, and the moderate proteinuria, nephrotic syndrome, glomerulonephritis, acute renal failure, and chronic renal failure. I will not discuss complications were minor (macroscopic hematuria in 3 pathe use of renal biopsy in renal transplantation, in pregnant tients, hematoma in 2). Until more experience accumulates, and women, or in children. In addition, I will not consider the a better yield of tissue is obtained, however, the modified open histologic interpretation of biopsy specimens in various disease procedure described here seems more prudent in high-risk states. patients. A second major contraindication to percutaneous renal biThe first renal biopsies, reported in 1923, were open surgical procedures [6]. Ball described the first percutaneous needle opsy is the inability of a patient to comply with instructions. biopsy in 1934 [7]. In 1943 Castleman and Smithwick system- Mild sedation often improves compliance in children [16, 17]. atically evaluated 100 surgical biopsies of hypertensive patients On rare occasions, general anesthesia is necessary to obtain and commented, "Pathologists are so accustomed to the micro- renal tissue in the patient who cannot cooperate. When per-

Nephrology Forum: Renal biopsy

forming an open biopsy under general anesthesia, a core or cup of tissue should be obtained with a needle or forceps. Wedge resection should be avoided, because it can yield inappropriate material for histologic examination [14]. For example, if an area of fibrosis is biopsied, it might not provide diagnostic information.

Severe hypertension or hypotension, renal abscesses, pyelonephritis, hydronephrosis, severe anemia, uremia, large renal tumors, and cysts are relative contraindications to renal biopsy [18, 19]. However, if blood pressure is restored to normal, if

531

reduced in size [20—22]. Most biopsy series of patients with chronic renal failure have had relatively small numbers of

patients, however, and these results must be viewed with caution. Furthermore, although renal biopsy establishes a diagnosis in the majority of such patients, the likelihood of obtaining information that will affect future therapy is minimal. Therefore a renal biopsy in these patients is warranted only infrequently.

Procurement of tissue In their initial report in 1951, Iverson and Brun described

urinary infection or obstruction is eradicated prior to the localization of the lower pole of the right kidney using intraveprocedure, and if uremia or severe anemia is corrected, biopsy can be carried out safely [19—22]. Some nephrologists have argued that patients with only a single functioning kidney should not be subjected to percutaneous biopsy, and that if the biopsy must be done, it should be an open biopsy under direct visualization [18, 19, 23]. I think these beliefs require reevaluation. Let us begin by examining the risk of nephrectomy after renal biopsy. Using data from a questionnaire, Welt performed a respective analysis and reported at the first meeting of the American Society of Nephrology that only 5 of 8081 patients who were biopsied (1/1616) required a nephrectomy [24]. Other investigators have estimated the incidence to be 1/2000 to 1/5000 [18, 19, 21, 23]. The risk of death from general anesthesia, which may be required for an open biopsy,

is available for comparison and varies with the American Society of Anesthesia (ASA) physical status class. The ASA physical status categories range from class 1 (no physiologic or pyschologic disturbance) to class 5 (the moribund patient who has minimal chance for survival). For class-2 patients (mild to moderate systemic disturbances, that is, essential hyperten-

nous pyelography and a lead mark on the skin [10]. Nevertheless, many large series of renal biopsies report a high degree of success using the "blind" technique based on anatomic land-

marks as described by Kark and Muehrcke in 1954 [12]. According to Welt's questionnaire, 6393 renal biopsies had been performed by the "blind" technique, compared with 1398 done under guidance of image-intensification fluoroscopy [24]. Over 90% of procedures performed by either method yielded

renal tissue. A 95% success rate (5 or more glomeruli) was reported by Diaz-Buxo and Donadio in 1000 patients; 176 procedures were performed under fluoroscopic guidance, and the rest were performed by the "blind" technique [21]. Most biopsies today are performed using one of a variety of techniques for localization of the kidney, including image amplification fluoroscopy [32, 33], radionuclide scan [34, 35], retrograde pyelography [36], ultrasound [37], and computerized tomography [38]. Despite the large variety of imaging procedures for visualizing the kidney during biopsy, only a few reports have compared the adequacy of samples using these procedures. Most of these studies provide limited information

because of the small number of patients included. In a study of 48 patients, Bolton and coworkers compared radionuclide scan, specifically attributed to general anesthesia is higher; it is 1.8% intravenous pyelography, and ultrasonography to image-amplifor class-3 patients and 7.8% for class-4 patients [25—28]. In fication fluoroscopy, regarded by these authors as the standard most situations, therefore, the risk of death from general [37]. Radionuclide scan correlated best with fluoroscopically anesthesia is comparable to the risk of nephrectomy following determined kidney length (9.9 0.2 cm versus 9.6 0.2 cm, P closed biopsy. All patients, including those with two kidneys, = NS) and width (5.7 0.2 versus 5.8 0.2, P = NS), whereas are exposed to a similar risk from anesthesia. Furthermore, the kidney appeared significantly larger by ultrasound (length, 0.3, P < 0.5; width, 6.2 0.2, P = NS) and by biopsies of transplanted kidneys are performed routinely with 10.8 excellent results and without important complications [29]. I do pyelogram (length, 13.8 0.3, P < 0.001; width, 7.2 0.2, P < not believe, therefore, that a solitary kidney is an absolute 0.001) (as compared with fluoroscopy). Ultrasound localized contraindication to renal biopsy. In patients at higher risk for the lower pole most accurately, and it correlated best with the bleeding complications (those with small kidneys, uremia, fluoroscopically determined depth of the kidney. The kidney bleeding diathesis), an open procedure under local anesthesia could be localized in all patients by ultrasound, but renal sometimes is more prudent, however. dysfunction limited localization in 10 patients by scan, 5 paIn my opinion, special caution should be used in the biopsy of tients by intravenous pyelogram, and 4 patients by fluoroscopy. patients with chronic renal disease and small kidneys (end-stage The complication rate among the different methods has not renal disease). The risk of serious complications is substantially been evaluated systematically. More recently Birnholz et al increased in this group [13, 18, 19, 30, 31]. Kropp et al reported described an ultrasound guidance technique using a probe that 5 of 29 patients (17%) with end-stage renal failure who positioned along the mid-axillary line [39]. As the biopsy needle underwent percutaneous biopsy experienced significant compli- is advanced, its route is monitored by continuous ultrasound cations [30]. Similar observations have been reported by others guidance. Using this technique, they obtained renal tissue in 29 [31], and most authors report an increased incidence of bleeding of 31 patients; after some practice, they could perform the [19, 21, 23]. Of course, chronic renal insufficiency is not entire procedure within 20 minutes. Similar results have been equivalent to end-stage renal failure. Renal biopsy appears to be reported by others [40]. However, the procedure has not been relatively safe (that is, it produces no apparent increase in major systematically compared with the other techniques I described. A further modification of the ultrasound guidance technique complications) in patients with a gradual, progressive decline in GFR (as opposed to acute renal failure) and a serum creatinine is the utilization of an automated biopsy instrument [41]. This greater than 2.0 mg/dl, provided that the kidneys are not spring-loaded device, containing an 18 gauge biopsy needle (17 sion, mild diabetes, or anemia), the death rate is 3/1000 [25—28].

For patients with more severe disorders, the mortality rate

532

Nephrology Forum: Renal biopsy

.60

Probability that all glomeruli in biopsy specimen are normal

Involvement in kidney 10%

35%

I

I

.40

50%

I

.0 '5 .0

0

°- .20

0

1

2

3

4

5

6

78

9 10

Number of abnormal glomeruli in biopsy of 10 glomeruli A

Involvement in kidney .60

Fig. I. Probability of finding only normal glomeruli in a biopsy speci-

.02

Diabetes Mellitus, Munksgaard, Copenhagen, 1965).

mm specimen notch), is passed through a guide on a transducer

35%

50%

.40

men of n glomeruli, when the true relative frequency of normal glomeruli in the kidney is known, Graphic presentation for varying values of n. (With permission, from: A. C. Thomsen: The Kidney in

10%

B

.20

051015

20

Number of abnormal glome ruli in biopsy of 20 glomeruli

to the capsule. The authors state: "Activation causes the Fig. 2. Binomial distributions of the number of abnormal glomeruli obdurator to advance 23 mm into the kidney followed immedi-

which would be found in biopsy samples obtained from kidneys with an

ately by the cannula" [41]. The actual time the biopsy needle is actual percentage of abnormal glomeruli of 10, 35, and 50%. Biopsies within the kidney is reduced. Furthermore, the procurement of containing 10 giomeruli (A) and 20 glomeruli (B) are shown. The tissue requires less dexterity compared with use of the Tru Cut horizontal axis is the number of abnormal glomeruli seen in the biopsy,

or Franklin modified Vim-Silverman needle. Whether or not these modifications will result in a reduction in the rate of serious complications, however, has not been evaluated. My preference is to use ultrasound to localize the kidney, because it does not require radiocontrast material, and the kidney usually can be localized even if the patient has renal insufficiency. From the limited data available, the yield of renal tissue obtained using ultrasound is at least comparable to that obtained with other methods. An experienced ultrasound technician is an essential member of the biopsy team. I find it useful for the technician to outline the lower pole of the kidney and the precise point of entry on the surface of the skin and to provide advice regarding the orientation of the needle. The procedure described by Kark and Muehrcke in 1954 for procurement of tissue is remarkably similar to that used today,

and the vertical axis is the probability of having exactly that number of abnormal glomeruli. (From Ref. 44.)

vital signs change before then. Gross,hematuria usually clears promptly; if so, the patient can be discharged from the hospital the day after the biopsy. The patient should be instructed to avoid heavy physical exertion for the next few weeks.

Characteristics of the sample The features of the tissue sample that influence the accuracy of diagnosis and prognosis, and consequently the appropriateness of therapeutic approaches, have been analyzed by Thomsen [43] and by Corwin et al [44]. Let us use the diagnosis of a focal glomerular lesion to illustrate these concepts. First, conand I refer the reader to the original text for this description sider the recognition of focal diseases. For the diagnosis of a [12]. 1 prefer the disposable Tru Cut needle (Travenol Labora- focal disease, such as focal segmental glomerular sclerosis, the tories, Deerfield, IL) to the Vim-Silverman needle described by finding of a single abnormal glomerulus establishes the diagnoKark [42]. Usually it provides adequate cores with minimal sis. However the probability that focal sclerosis is not present distortion of tissue, and the cutting edge does not need to be in a patient (that is, someone with idiopathic nephrotic syndrome) depends both on the actual fraction of abnormal gbsharpened. Following the procedure, the patient should assume the meruli and on the number of glomeruli obtained in the biopsy supine position for the next 12 to 24 hours and should undergo specimen. This point is illustrated in Figure 1 [43]. IfS gbomeruli frequent assessment of vital signs. During this interval, fluids are present in the biopsy specimen, and the actual glomerular should be administered liberally (assuming renal function is involvement is 20%, there is a 35% chance that all glomeruli in adequate), and the color of the urine should be monitored. I the biopsy specimen will be normal. By contrast, if the actual usually obtain a hematocnt on the following morning, unless the glomerular involvement is 20% and 10 glomeruli are obtained,

533

Nephrology Forum: Renal biopsy

Table 1. Minimum number of abnormal glomeruli that must be present in a biopsy containing a certain number of glomeruli to infer with 95% confidence that the disease process involves 80%, 50%, or 20% of the kidneya

100 90

Total number of glomeruli in biopsy

80

8

7t C

60 50

40 C

30

80%b

50%b

20%b

8 (100)"

7 (88)

3 (38)

10

10(100)

8(80)

4(40)

12

12 (100)

9 (75)

5 (42)

15

14(93)

11(73)

6(40)

20

19 (95)

14 (70)

7 (35)

25

23 (92)

17 (68)

9 (36)

30

28 (93)

20 (66)

10 (33)

35 32(91) 23(66) 11(31) 40 12 (30) 36 (90) 26 (65) a p = 0.05 (one tailed). (From Ref. 44.) b Percentage of involvement in the kidney sought. Figures in parentheses are percentages of abnormal glomeruli in a biopsy.

20 10

10 20 30 40

Extent of disease involvement in the kidney

50

60

70

% Involvement in renal biopsy

the minimum number of glomeruli necessary to confidently Fig. 3. Binomial distribution of percentage of abnormal glomeruli in biopsy samples from kidneys with mild (10%), moderate (35%), and severe (60%) involvement, (From Ref. 44.)

assume a given level of involvement can be calculated (Table 1).

Sampling issues are especially important in the design of clinical trials. If 50% glomerular involvement is an entry crite-

rion for a study, for example, the glomerular involvement should be based on the sample size to assure that all patients meet this requirement. Meeting the 50% cutoff criterion would require that 7 glomeruli are involved in a specimen containing 8 however, the actual percentage of involvement is only 10% and glomeruli. With larger samples, of course, the required percent10 glomeruli are sampled, the probability of finding all normal age of involvement is less. Thus, to assure a given level of glomeruli is 88%. involvement with confidence (that is, 95%), the requirements Next consider how sampling influences our assessment of the should vary depending on the number of glomeruli in the proportion of glomenilar involvement, again in a patient with a specimen. focal glomerular lesion. Unless all glomeruli within the kidney These concepts have additional strategic implications for are abnormal and affected equally, the number of abnormal study design. When classifying patients into disease categories, glomeruli in a given biopsy specimen will vary. The probability the number of patients with inappropriate assignments should that the observed involvement accurately predicts actual in- be considered. Corwin et al give the example of patients with volvement is related to both the number of glomeruli in the systemic lupus erythematosus (Fig. 4) [44]. Distinctions that specimen and the actual fraction of glomeruli affected [44]. rely on small differences in the percentage of abnormal glomer(This probability is based on the assumption that the disease is uli (that is, mild versus moderate, or moderate versus severe randomly distributed throughout the kidney and the specimen). involvement) have little meaning because there is a large This principle has a number of practical implications (Figs. overlap among patient groups. To assign patients accurately to 2—5). If a biopsy containing 10 glomeruli discloses 3 abnormal either mild or moderate disease categories, more than 100 glomeruli, the probability that 30% of glomeruli are actually glomeruli per specimen are required! Given that the usual abnormal is approximately 25%. There is a 10% chance that the number of glomeruli per specimen usually is substantially less actual involvement is 50%, and a 5% chance that the involve- than this, a distinction based on greater differences in the ment is only 10% (Fig. 2A). If, however, the sample contains 20 groups would be much more meaningful (that is, 50% involvement). of either 10% or 50% involvement is reduced to less than 5%. One should assess whether adequate tissue has been obtained The probability that the glomerular involvement is within a during the procedure itself. A decision whether to obtain more given range (confidence interval) also can be derived from the tissue should be based on an estimate of the number of number of glomeruli in the biopsy and the percentage of glomeruli obtained and on the nature of the disease process(es) involvement (Fig. 3). For example, in a biopsy with 10 glomer- suspected. At least two cores of tissue are obtained for routine uli and 30% involvement, there is a 95% probability that actual pathologic examination. Each specimen should be examined glomerular involvement is between 7% and 65%. With 30 directly by light microscopy with tenfold magnification for the glomeruli and 30% actual involvement, the 95% confidence presence of glomeruli [42], and the number of glomeruli should interval is reduced to 15% to 50%. Based on these principles, be estimated. Using this procedure, one can obtain adequate

the chance of finding all normal glomeruli is only 10%. If,

534

Nephrology Forum: Renal biopsy

of patients, hematuria persists for 2 to 3 weeks [21]. Occasionally, gross hematuria first occurs several days after the biopsy, -C>-

but it too usually resolves within a few days with rest [19]. moderate >20

5) -

Transfusions are necessary in 0.1% to 3.0% of patients [19, 21, 23, 48]. Surgery for persistent or massive bleeding is required in less than 0.2% of patients [19, 21]. Perinephric hematomas are common. If computerized tomog-

5) .0

raphy is carried out one day after the biopsy procedure,

'50

C

severe >50

o o..C

perinephric hematomas are detected in 57% to 85% of the patients [49—51]. Most of these hematomas are clinically silent, although occasionally an otherwise unexplained fall in hemato-

crit is observed [13, 21, 47]. Approximately 1% to 2% of

patients bleed sufficiently to develop hypotension, a fall in the hematocrit, and a flank mass. Rarely, the onset of hemorrhage is delayed for as long as 2 months [19, 52]. Hematomas usually resolve within 3 months; they rarely become secondarily in80 100 fected and rarely require treatment with parenteral antibiotics Actual % abnormal glomeruli in kidney and surgical drainage [53]. Fig. 4. Classification of patients into mild, moderate, or severe disease Arteriovenous fistulas are also common; by arteriography on the basis of biopsy sample (y-axis) as prevalence of abnormal they can be demonstrated in as many as 15% to 18% of patients glomeruli in kidney is varied (x-axis). Misclassification into greater severity is indicated by gray area and lower severity by hatched mark area. (From Ref. 44.)

[54—57]. Such lesions usually are clinically silent, and more than

95% resolve spontaneously within 2 years [54]. In rare instances, surgical correction of the fistula is required because of

severe hypertension, persistent hematuria, congestive heart tissue for pathologic examination in 90% to 95% of patients. Needless to say, the condition of the patient also is a critical determinant: if the patient is unable or unwilling to continue, the procedure must be stopped. Proper interpretation of renal biopsies requires examination by light microscopy (LM), immunofluorescence (IF), and electron microscopy (EM). The tissue must be prepared and fixed in

failure, or hydronephrosis [56—59]. Post-biopsy aneurysms oc-

cur in less than 1% of patients [19, 54, 60, 61]. Infections are unusual except in the presence of active pyelonephritis. Of 65 patients with pyelonephritis prior to renal biopsy, 7 developed

fever after the biopsy, and 4 of these patients developed

bacteremia [62]. There are many less common complications of

renal biopsy, including ileus; lacerations of the liver, spleen, pancreas, intestine, gallbladder, and subcostal or visceral artersimple, and the solutions are inexpensive and stable [45, 46]. If ies; puncture of the renal pelvis leading to urinoma; dislodged complete examination (LM, IF, and EM) cannot be performed renal stones; pancreatitis; pneumothorax; and dispersion of at the hospital where the biopsy is obtained, the tissue should carcinoma [19, 20, 45, 63—681. As I indicated previously, the risk of nephrectomy is between be processed and submitted to a laboratory where the appro1/2000 to 1/5000 following renal biopsy. The risk of surgery for priate evaluations can be performed. When the tissue is to be submitted to another pathologist, the nephrectomy or to repair a laceration due to biopsy is approxsample for electron microscopy should be divided into smaller imately 1/500 to 1/1000 [19, 21, 47, 48, 63]. The risk of death (1 mm) cubes and placed into Trump's fixative [45]; the tissue associated with surgery is related to the general health of the for immunofluorescence should be cut into 2—4 mm cubes and patient. immediately "snap-frozen" in a precooled (—20°C) solution of Indications either isopentane, dry ice and acetone, or freon, and stored at —70°C; if the specimen cannot be snap-frozen, it can be stored Ideally, the information derived from a renal biopsy should in Michael's medium [46] for as long as one month. The identify a specific diagnosis, accurately reflect the extent of remaining tissue for light microscopy can be placed in any of a disease activity, and provide the basis for a decision about number of common fixatives; we currently use Van der Grill's specific therapy that could not otherwise be made. Unfortu[46]. On rare occasions, when the biopsy procedure can no nately, the pathologic findings often are not specific, and a longer be continued, fewer than 5 to 10 glomeruli are obtained. definitive diagnosis cannot always be made. Furthermore, In this situation, the nephrologist must indicate the most likely although the extremes of disease activity can be defined accuclinical diagnosis to the pathologist, so the specimen can be rately from the biopsy, the presence of either severe or minimal disease usually can be predicted from the patient's clinical processed in the most efficient manner. features just as effectively. Between these extremes, structuralComplications functional correlations are poorly defined, and as a conseHematuria is the most common complication of percutaneous quence there often is considerable difference of opinion about biopsy [19]. Microscopic hematuria occurs in virtually all the clinical significance of intermediate levels of pathologic patients, whereas gross hematuria occurs in 5% to 9%. The activity. Thus, the prognosis and/or likelihood of response to presence of uncontrolled hypertension or uremia increases the therapy cannot always be predicted with confidence. These incidence of hematuria twofold [47]. Hematuria usually re- considerations complicate both the use of renal biopsy in solves spontaneously within 2 days, but in approximately 0.5% clinical practice, and its application to clinical research.

the proper solutions; preparation of the tissue is relatively

Nephrology Forum: Renal biopsy

An important consideration when performing a renal biopsy is how the information will be utilized. Does identification of the

535

which it is difficult to determine the type of renal disease from the urine sediment. For example, if the urine sediment examination reveals red blood cells, renal tubular cells, and granular

type of renal disease and estimation of severity influence the diagnosis, prognosis, or therapy? Is tne information derived casts but no red blood cell casts, reasonable possibilities to from renal biopsy useful in patient management? Is the utility of explain the acute renal failure include a proliferative glomeruthe information greater than that which can be derived from lonephritis, an inflammatory tubulointerstitial disease, a disease clinical features alone or from other, less risky tests? Unfortu- of renal arterioles, or a combination of diseases. If the urine nately few studies of renal biopsy adequately address these sediment examination reveals only mild abnormalities, the possibilities would include acute or chronic tubular injury or a questions. Paone and Meyer retrospectively examined 100 consecutive disease of large or medium-sized renal vessels. Ganeval et al renal biopsies to determine whether physicians' judgments reported results of biopsies from 95 patients in whom acute regarding treatment were altered by the biopsy results [69]. renal insufficiency did not appear to be due to volume depletion, Although a definite or probable diagnosis was ascertained by urinary tract obstruction, or acute tubular necrosis [77]. Gbbiopsy in 77% of patients, therapy was altered in only 19%. merulonephritis occurred in 34% of their patients, interstitial None of the patients with acute or chronic renal insufficiency or nephritis in 13%, vascular lesions in 16%, and a variety of with hematuna as the indication for biopsy had an alteration of lesions in the remainder. Beaufils, Richet, and colleagues treatment; changes in therapy were confined to the patients reported on 218 biopsies from hospitalized patients with acute with proteinuria. Similarly, Cohen and colleagues [70] and renal insufficiency [78—80]. The patients did not have clinical Turner and coworkers [71] analyzed the influence of renal features of volume depletion, urinary tract obstruction, or acute biopsy on physicians' judgments regarding diagnosis, progno- tubular necrosis. Glomerulonephritis was present in 24%, acute sis, and treatment in patients with diverse types of renal interstitial nephritis in 14%, vascular lesions in 30%, and acute disease. These two groups reported that changes in judgment tubular necrosis in 32%. A renal biopsy may be required to determine the type and occurred predominantly in patients with heavy proteinuria or systemic disease. Whiting-O'Keefe et al retrospectively ana- extent of severity of disease in patients with acute renal failure, lyzed the case histories of 30 patients who underwent renal and this information dictates therapy. In this regard, the results biopsy for severe lupus nephritis [72, 73]. Knowledge of the of the renal biopsy in Patient 3 were very helpful, because they biopsy result did not improve the accuracy of the physicians' obviated the need for further therapy, and they were predictive prognosis, including prediction of renal insufficiency or renal of a good prognosis. Given these considerations, I recommend failure, worsening proteinuria, or prescription of more aggres- prompt renal biopsy in patients with acute renal failure, in whom the diagnosis cannot be made by routine clinical and sive immunosuppressive therapy. These studies have several shortcomings. Most important is laboratory data. Chronic renal failure. By contrast with acute renal failure, it that they measure only physicians' judgments regarding the type and severity of renal disease, not the outcomes of the is frequently difficult to determine the cause and type of chronic decisions they make regarding treatment. As demonstrated by renal disease on the basis of clinical criteria alone. Kropp et al the results of decision analysis on the necessity of renal biopsy reported the renal pathology on 29 consecutive patients who for idiopathic nephrotic syndrome [74—76], differences in phy- underwent renal biopsy and on 17 consecutive patients subsicians' judgments do not necessarily translate into differences jected to nephrectomy in preparation for renal transplantation in patients' survival or quality of life. Without data on out- [30]. The clinical diagnosis was confirmed in 48%, a different comes, including survival, development of renal failure, and diagnosis was made in 43%, and no definitive diagnosis could be complications from treatment, conclusions based on physi- made in 9%. Similar observations have been made more recians' judgments alone must be considered of limited value. cently by others [20, 22, 69]. Rarely, an unsuspected diagnosis In summary, there are no absolute indications for renal (such as oxalosis) can influence further clinical management biopsy. Nevertheless, virtually all nephrologists agree that the [46]. As I said previously, if both kidneys are small, the risk of information derived from the renal biopsy does have utility. In renal biopsy is increased (see previous discussion). Because it the following discussion I will elaborate on my reasoned opin- is unlikely that assessment of either the type or severity of renal ions about the indications and the utility for renal biopsy in disease will alter prognosis, therapy, or patients' outcomes, I various clinical presentations of renal disease. To do so I will believe that renal biopsy usually is not warranted in this first consider the role of the biopsy in determining the type of circumstance. Asymptomatic hematuria. Renal biopsy is abnormal in aprenal disease in the major renal syndromes and then consider the studies that have attempted to assess the severity of renal proximately 75% of patients with isolated hematuria (that is, patients with no proteinuria, no other abnormalities of the diseases as a guide to prognosis and therapy. Acute renal failure. In most patients with acute renal failure, urinary sediment, normal renal function, and no structural the cause and type of renal disease can be determined without abnormalities of the urinary tract) [81, 82]. In such patients, IgA a renal biopsy; obstruction, reduced renal perfusion, acute nephropathy is the most common lesion, occurring in approxitubular necrosis, and systemic diseases associated with acute mately 50% of patients [81]. Less frequent findings include glomerulonephritis are usually evident from clinical, serologic, hereditary nephritis and other abnormalities of glomerular renal ultrasound, and urine sediment examinations. In a rela- basement membrane, isolated deposits of C3 in vessel walls, tively small number of patients, however, a clear diagnosis and tubulointerstitial diseases [81—86]. Remarkably, 25% of cannot be made from evaluation of clinical findings alone. The patients have normal biopsies [81, 82]. In the absence of proteinuria, decreased renal function, a most difficult diagnostic problems arise in circumstances in

536

Nephrology Forum: Renal biopsy

family history of renal disease, or structural abnormalities of the urinary tract, hematuria is likely due to mild glomerular abnormalities. The likelihood of a renal biopsy elucidating the source of hematuria is high, but the results are not likely to affect therapy. A renal biopsy should be performed only if defining the source of hematuria is necessary to obviate further expensive or hazardous evaluations, to define the therapeutic

biopsy provides information regarding the presence and extent

patient. The first patient presented today exemplifies a circum-

ment membrane are each greater than 90% [99, 100], but the test

of inflammation and immune deposits. This information is especially important for rapid diagnosis and initiation of therapy

in patients with glomerulonephritis and rapidly progressive renal insufficiency. As a possible alternative to renal biopsy, serologic tests have been proposed to distinguish among patients with linear, granular, or no immune deposits. The sensiapproach (including giving no therapy), or to reassure the tivity and specificity of serum antibodies to glomerular base-

stance in which the information was useful in determining is available in only a few specialized laboratories. Anti-neutrowhether the individual was a suitable kidney donor. phil cytoplasmic antibodies (ANCA) are highly specific for Asymptomatic proteinuria. Few studies are available on the patients with active Wegener's granulomatosus (>98% speciusefulness of renal biopsy in patients with mild to moderate ficity by either indirect immunofluorescence or ELISA); how-

proteinuna (0.5 to 2.0 glday). In a series in Singapore of more ever, the sensitivity varies with disease activity [101—1051. With than 68,000 asymptomatic military recruits without a history of active systemic disease, the sensitivity is 93% to 96%; with hypertension or renal disease, less than 1% had persistent active regional disease, the sensitivity is only 60% to 67%, and proteinuria as the only urinary abnormality [871. Of 45 patients with inactive disease it is further reduced to 32% to 40% in whom a renal biopsy was performed, the most common [102—105]. False-positive results can occur in patients with lesion was focal glomerular sclerosis, which occurred in 33 systemic lupus, but these can be distinguished from true posipatients. None of the lesions identified was considered serious tives by indirect immunofluorescence [1031. Less information is enough to require specific therapy, and the clinical significance available regarding ANCA in patients with necrotizing and of these lesions remains uncertain. Although glomerular lesions cresentic glomerulonephritis without systemic manifestations of other than focal sclerosis that typically are associated with vasculitis. In these patients, the presence of ANCA has varied nephrotic syndrome were not seen in this study, other studies have documented that, with the exception of minimal-change from 0% to 75% of patients with active disease [102, 103, 1051. In my opinion, the initial approach to the diagnosis of acute disease, all glomerular diseases can cause mild to moderate glomerulonephritis is based on the patient's clinical and seroproteinuria as their only manifestation [88, 89]. Tubulointerstilogic findings. I reserve renal biopsy for patients in whom these tial diseases also can cause mild to moderate proteinuria, but findings do not identify a specific systemic disease. Because renal biopsy generally is not useful in identifying a specific accurate and prompt diagnosis is essential in selecting the type cause. Most often the biopsy reveals nonspecific findings such as inflammation and fibrosis. Occasionally, infiltration of the of treatment in patients with the nephritic syndrome and rapidly interstitium with eosinophils, deposition of immune complexes, progressive renal insufficiency, I regard renal biopsy as the or immunoglobulins in tubules suggests a specific diagnosis definitive diagnostic test in this setting. In most patients with systemic lupus erythematosus, I think [90—961. initial decisions regarding immunosuppressive therapy can and In summary, renal biopsy in patients with asymptomatic should be based on an assessment of the severity of the disease, proteinuria usually identifies a specific morphologic type of glomerular disease. Treatment for idiopathic glomerular dis- as reflected in the urinalysis, an estimation of GFR, and eases with non-nephrotic-range proteinuria is not well estab- measurements of urinary protein excretion. In patients in whom lished, however; hence the information is unlikely to affect the the severity of glomerular involvement, and hence the likelitherapeutic approach. Furthermore, if proteinuria is tubular in hood of a response to treatment, cannot be ascertained from origin or represents overproduction proteinuria, renal biopsy these findings alone, renal biopsy can be helpful in determining generally is not helpful in making a specific diagnosis. The basis the need for initial immunosuppressive therapy. Nephrotic syndrome. Routine clinical and serologic examinafor deciding whether to perform a biopsy in these patients therefore is similar to that in patients with asymptomatic tion of patients with nephrotic syndrome usually allows the hematuria. physician to determine whether a systemic illness is present Nephritic syndrome. Few studies have been published re- [106]. In patients with primary (idiopathic) nephrotic syndrome, garding the diagnostic utility of renal biopsy in patients with the however, the clinical features overlap to a great extent. Alnephritic syndrome (as defined by the presence of hematuria, though the use of multiple laboratory tests enhances the probproteinuria, and red blood cell casts). Because of substantial ability of distinguishing inflammatory from noninflammatory overlap in pathologic features among various glomerulonephrit- lesions [107], it is not possible to predict the glomerular ides secondary to a systemic disease, the diagnosis of a specific pathology with confidence. Thus renal biopsy is the only disease cannot always be determined by renal biopsy alone [97]. method for determining the type of renal disease in patients The cause of secondary glomerular diseases frequently can be whose clinical features do not clearly reveal a cause. Most identified from the patient's clinical attributes without a renal physicians assume that if an adequate specimen is obtained, the biopsy [98]. Nevertheless, when the cause of glomerular dis- pathologic features will either be pathognomonic for a systemic ease is not known, the information provided by renal biopsy can disease or will be typical of one of four well-described primary be essential for making the diagnosis. If the cause is known, the renal diseases: membranous nephropathy, minimal-change disrenal biopsy still can be useful in estimating the severity of the ease, focal and segmental glomerular sclerosis, or membranodisease. proliferative glomerulonephritis (MPGN). In two series of In patients with primary glomerular disorders, the renal adults with nephrotic syndrome, however, as many as 20% of

Nephrology Forum: Renal biopsy

537

patients had proliferative glomerulonephritis or other morpho- ers observed that glomerular cellularity correlated with inulin clearance [112]. logic changes of uncertain classification [69, 107]. The relationship between renal structure and function has The decision of whether to use biopsy-guided therapy or empiric treatment with high-dose, alternate-day steroids in all been studied more recently in patients with diabetes mellitus, patients with idiopathic nephrotic syndrome has been previ- membranous nephropathy, hypertension, and systemic lupus ously addressed in this Forum and elsewhere [74—76]. These erythematosus. In diabetes, Mauer, Steffes, and coworkers analyses were based on the assumption that high-dose, alter- measured mesangial volume and capillary surface area within nate-day steroid therapy is the optimal treatment for idiopathic glomeruli and total glomerular volume by tracing glomeruli onto membranous nephropathy and minimal-change disease, and the digitizing tablets [113—116] and used a point grid system to model encompassed only patients with serum creatinine con- determine total and fractional volumes. They found that both centrations less than 1.5 mg/dl. These studies concluded that increased fractional mesangial volume per glomerulus and both empiric therapy and biopsy-guided therapy are equally diminished capillary filtration surface per glomerulus were effective and appropriate strategies. Thus, in my opinion, the strongly associated with proteinuria, decreased GFR, and hyapproach selected should be based on individual patient con- pertension [117]. Notably, microalbuminuria did not correlate siderations. To the extent that the physician believes that with these glomerular structural measurements, and glomerular Ponticelli's combination therapy approach of steroids and chlo- lesions were present in patients with normal urinary albumin rambL'cil [108] is more appropriate for patients with idiopathic excretion rates. More recently, they used these morphologic membranous nephropathy, then biopsy-guided therapy might assessments to compare patients who received pancreatic prove more appropriate. In clinical practice, I discuss these transplants following successful renal transplantation with diaissues with the patient, and we jointly decide the diagnostic and betic patients who only received renal transplants [118]. Altherapeutic course. In most of my patients, the approach though the creatinine clearances of both groups were similar, mesangial volume and glomerular volume increased signifiutilized is biopsy-guided therapy. Severity of disease. Accurate assessment of the severity of cantly in the diabetics who had received only renal transplants. disease is essential for determining both the extent of disease In membranous nephropathy, proteinuria can persist despite activity and the probability that the disease will respond to steroid and/or immunosuppressive therapy [119—123]. Repeated specific therapy. These considerations are especially important morphologic analyses of biopsy specimens indicate that in when recommending potentially toxic treatments and for strat- many patients, persistence of proteinuria is unassociated with ification of patients in clinical trials. Accurate histologic evalu- active immune deposit formation [124, 1251. To the extent that ation of renal biopsy specimens should identify those patients effective treatment is developed to prevent immune deposit with a poor prognosis. Ideally, the results of such histologic formation, the biopsy could be helpful in identifying such evaluation also should serve as end points in clinical trials, so occurrences, and potentially toxic therapy could be withheld the effect of therapy can be assessed prior to the development during periods when immune deposits are not forming. In a of end-stage renal failure. Many indices have been derived, but separate study of patients with membranous nephropathy, considerable disagreement exists regarding their value. In re- Shemesh et al observed that both decreased circumferential cent studies, attempts have been made to develop reproducible length of the glomerular capillary loops and reduction in the quantitative measures of disease activity and reversibility. Such number of filtration slits per unit of capillary loop length correlated with the extent of proteinuria [126]. Similar morphoindices are intended to enhance clinical decision-making. For such an index to have clinical utility, the severity of a metric studies in patients with hypertension demonstrated that histologic lesion would have to correlate both with the severity intimal thickening and arteriolar hyaline changes in small vesof functional alterations and with the patient's prognosis. Sur- sels correlated with the level and duration of hypertension prisingly, even in patients with chronic glomerular disease, the [127]. Richardson and coworkers recently examined the reladegree of renal functional impairment correlates better with the tionship between glomerular structure and function by comparextent of tubulointerstitial and vascular damage than with the ing conventional histopathology (light and electron microscopy) extent of glomerular damage. Redson, Sloper, and de Wardener with hemodynamic measurements (including calculation of KF) found a significant correlation between tubular atrophy and in glomeruli isolated from the biopsies of patients with renal creatinine clearance in patients with idiopathic and chronic diseases of diverse cause [128]. The KF correlated directly with glomerular disease (nephritis or nephritic syndrome) [109]. increasing glomerular size, glomerular hypercellularity, glomerStriker and colleagues evaluated structural-functional correla- ular capillary damage (scored semi-quantitatively), and urinary tions in patients with diverse renal diseases, including acute and protein excretion, and correlated inversely with percentage of chronic glomerulonephritis, interstitial nephritis, and nephro- senescent glomeruli and epithelial foot process widening. They sclerosis [110, 111]. They found a significant correlation be- found no correlation, however, between KF and serum creatitween total interstitial and tubular damage and either renal nine.

Whether the information derived from renal biopsy adds importantly to the assessment of the severity of renal disease tion correlated with GFR, but interstitial edema did not. The has been studied extensively in systemic lupus erythematosus extent of vascular disease correlated reasonably well with renal [73, 74, 129—141]. The extent of cellular proliferation and plasma flow. By contrast, a statistically significant relationship capillary wall thickening of the glomerulus is the basis of the between either renal plasma flow or GFR and the extent of World Health Organization (WHO) histologic classification glomerular disease was not found. In patients with acute system. The WHO classification correlates histologic findings proliferative glomerulonephritis, however, Parrish and cowork- with prognosis [89], but this classification is no more predictive

plasma flow or GFR, regardless of the underlying disease

process. Both interstitial fibrosis and inflammatory cell infiltra-

538

Nephrology Forum: Renal biopsy

sis. Patients with an activity index of 12 had a significantly worse prognosis than did patients with an activity index of less than 12, but the chronicity index was a better predictor of outcome than was the activity index. Similar findings have been

reported in children with severe lupus nephritis [144]. The results of a recent multicenter treatment trial of patients with severe lupus nephritis raise questions about these conclusions,

however [145]. In this latter study, patients were retrospectively divided into groups after completion of the study according to the presence or absence of adverse outcome (death, renal failure, 40% increase in serum creatinine). Neither the activity

index nor the chronicity index at study entry distinguished patients who ultimately had a poor outcome. These differences highlight some of the major difficulties underlying the use of renal biopsy in clinical trials. Although agreement is universal that knowing the extent of histologic activity is essential to randomizing patients to treatment groups in clinical studies of lupus nephritis, the histologic parameters that determine outcome are not fully recognized. As I indicated previously, distinctions that rely on small differences in a given parameter (such as a single cutoff value for the chronicity index) may have little meaning because a large overlap exists between patient groups. Furthermore, conclusions often are based on retrospective analysis alone. To validate the initial conclusions, these markers should influence allocation of patients into treat-

ment groups in future studies. Other factors such as age, duration of disease before treatment allocation, clinical activity at study entry, and previous therapy also should be considered concomitantly.

More recent studies of patients with lupus nephritis have

used more precise quantitative assessment of glomerular structural abnormalities for assessment of disease activity. Chagnac et a! performed morphometric studies of glomerular capillary area on serial biopsy specimens from patients with systemic lupus erythematosus [146] (Fig. 5). They reported evidence of progressive loss of glomerular capillary surface area with minFig. 5. Representative digitized tracings of the outline of 2 glomerular cross-sections. The upper tracing (A) is from a healthy control, the imal or no change in proteinuria, GFR, or serum creatinine. lower (B) from a patient with active diffuse proliferative lupus nephritis. Whether quantitative assessment of severity of renal disease in Shaded areas depict patent capillary lumina, and dots the nuclei of SLE from renal biopsy findings is more sensitive than assessendocapillary cells. (From Ref. 146.) ment from either clinical features alone or current histologic evaluation, however, needs further investigation. of prognosis than are clinical features alone. In a retrospective Questions and answers analysis of biopsy specimens from patients with acute lupus nephritis, investigators at the U.S. National Institutes of Health DR. NJCOLAOS E. MADIAS (Chief, Division of Nephrology, and the Walter Reed Army Medical Center examined the level New England Medical Center, Boston, Massachusetts): What of histologic activity and chronicity in renal biopsies at study does adequacy of tissue mean for diagnosis of vascular disease entry, and then they correlated the histologic data with progres- or interstitial disease? DR. MADAI0: As with glomerular disease, "adequacy of sion to end-stage renal failure [129, 140—143]. Semiquantitative scores were assigned on the basis of glomerular and tubular tissue" is highly dependent on both the extent and distribution abnormalities. An activity index (a scale of 0 to 24) was based of the disease. With vasculitis the size and the freqency of on attributes of glomerular structure (including cellular prolif- affected vessels are important factors. For patients with vascueration, fibrinoid necrosis, cellular crescents, hyaline thrombi, litis involving primarily large and medium-size vessels, the and leukocyte infiltration) and tubulointerstial abnormalities probability of making a definitive pathologic diagnosis from the (including mononuclear cell infiltration). A chronicity index (a renal biopsy specimen is remote, because the biopsy needle is scale of 0 to 12) also was determined from glomerular and intentionally directed away from these vessels. The probability interstitial parameters (based especially on the extent of fibrosis is higher for patients with vasculitis involving smaller vessels. and tubular atrophy). Patients with a chronicity index of 1 In practice, therefore, arteriography is more often the preferred had a significantly greater likelihood of developing renal failure diagnostic test for patients suspected of having large or medithan did those with a chronicity index of less than 1 [140—143], um-size vessel involvement. The diagnosis of Wegener's granand patients with a chronicity index 4 had the worst progno- ulomatosis is usually made after histologic examination of

Nephrology Forum: Renal biopsy

539

tissue from the respiratory tract. Vasculitis involving small major criteria that are relevant to the patient would be the effect vessels often can be diagnosed by renal biopsy; however, in of biopsy on selection of therapy and the effect of therapy on many cases, examination of more accessible tissue (such as survival or quality of life. It also would be important to study what value the patient places on information about prognosis skin) usually provides the diagnosis. For patients suspected of having vasculitis, I initially obtain 2 that does not directly affect therapy. to 3 cores of tissue. If. after histologic examination of the tissue,

DR. MARY H. FOSTER (Division of Nephrology, New England

the specimen is inadequate, the patient can be biopsied again. Medical Center): Could you comment on the variability in Although interstitial nephritis is almost always a focal process, interpretation of biopsy specimens by different pathologists? in my experience, 2 to 3 cores of tissue usually provide an DR. MADIA0: This is an important issue. Regarding the adequate quantity of interstitium to detect cellular infiltration. pathologic diagnosis, I don't know of a study that addresses the In rare cases a second biopsy is necessary. issue directly. In my opinion, it is crucial that the specimen be DR. MADIAS: How long would you wait in patients with acute processed and evaluated by an experienced renal pathologist, renal failure before you proceed to renal biopsy? and I have indicated in my discussion how this can be accomDR. MADAJO: At initial presentation, if prerenal azotemia and plished. Given this consideration, there is usually agreement on urinary tract obstruction are excluded and the clinical factors the diagnosis. However, when a disagreement arises regarding usually associated with acute tubular necrosis are absent, I the severity of disease, the dispute can have important progbiopsy the patient immediately to determine whether a disease nostic and therapeutic implications. The pathologist therefore amenable to therapy is present. The third case is a good must be as quantitative as possible in the description of the example of this situation. The biopsy also can provide useful histologic material. This should include reference to adequacy information in patients with prolonged renal failure presumed to of sample size and condition of the specimen, as well as the be secondary to acute tubular necrosis. Prognosis may be extent of involvement. Ideally, in clinical practice, the nephrolascertained (that is, the presence of cortical necrosis), and ogist should review the slides and photographs with the patholoccasionally an unsuspected disease is discovered. ogist. For clinical trials, it is important that each specimen be DR. JOHN T. HARRINGTON (Chief of Medicine, NewtonWellesley Hospital, Newton, Massachusetts): What else should analyzed in a similar manner, specific pathologic parameters be we be doing with the tissue obtained during the biopsy? Are defined both for entry criteria and study end points, and the specimens be reexamined by a committee of pathologists. I there new histologic techniques that we should be utilizing? refer you to the report of the Lupus Nephritis Collaborative DR. ANGELO Ucci (Department of Pathology, New England Medical Center): I think that many new techniques are still just Study Group [147] for an outstanding example of this practice. DR. JEROME P. KASSIRER (Associate Physician-in-Chief, research tools; their clinical utility is doubtful currently. In an New England Medical Center): Is the risk of biopsy increased academic center, I believe that immediate clinical benefit is not for patients with comorbid conditions? our only consideration, however. Some specialized questions DR. MADIAO: The probability of bleeding is not increased; can be answered using lymphocyte surface marker studies on tissue from renal biopsy. We have had several patients with the risks associated with therapy for severe bleeding are higher, acute renal failure whose renal biopsies revealed lymphomatous however. This is because of the greater incidence of therapeutic involvement. We were able to determine the lymphocyte cell intervention and the increased risk associated with the intertype of these lymphomas with tissue procedures that are vention. For example, patients with preexisting anemia and available now. Quantitative measurements of fibril diameters congestive heart failure are more likely to receive blood transcan be useful in distinguishing among various glomerular fibril- fusions than are patients without these conditions. Another loses, and some morphometric techniques might be applicable example is the higher surgical mortality rate in ASA class-4 to estimating the extent of disease. In-situ hybridization tech- patients compared with class-2 patients. If the risk of surgical niques currently are proving useful in evaluating some tissue intervention after a renal biopsy is the same for both groups biopsies for viral infections and other gene products. As such (less than 0.2%), the mortality risk increases from 6/1,000,000 techniques reveal information of diagnostic or therapeutic util- for class-2 patients to 1.6/10,000 for class-4 patients. These ity, they can be rapidly used to more fully evaluate renal considerations should be discussed with the patient prior to the biopsies. biopsy. DR. ANDREW S. LEVEY (Division of Nephrology, New EnDR. MADIAS: Are the risks of renal biopsy increased in gland Medical Center): I would like to add two comments. patients with amyloidosis? First, although renal biopsy is an exquisite tool for defining the DR. MADIAO: The risk of bleeding is not increased in these type and severity of renal disease, it lacks utility in many patients, providing they have normal coagulation and bleeding clinical presentations because of the lack of therapies. If parameters prior to the procedure. However, biopsy of more adequate therapies were available for each type of disease that accessible tissue, such as skin or subcutaneous fat, may provide can be defined by renal biopsy, then of course the biopsy would the diagnosis and obviate the need for renal biopsy [148—151]. have much greater utility. Clearly, as new therapies are discovDR. DEMETRIOS V. VLAHAKOS (Nephrology Fellow, New ered, the utility of renal biopsy in different clinical presentations England Medical Center): Dr. Kassirer, in a state-of-the-art must be re-analyzed. Second, the informational content of renal lecture at the 1989 meeting of the American Society of Nephrol-

biopsies is rarely judged by criteria that are relevant to the ogy, a distinguished nephrologist showed the title of your patient. In most studies, the criteria by which the informational Nephrology Forum, "Is renal biopsy necessary for optimal content is judged are the physician's opinions regarding the management of the idiopathic nephrotic syndrome?" and anpatient's diagnosis, treatment, and prognosis. However, the swered this rhetorical question "yes." His conclusion seems to

540

Nephrology Forum: Renal biopsy

be at variance with the studies that you, Dr. Levey, and Dr. Pauker have published [74, 75]. Would you comment?

12. KARK RM, MUEHRCKE RC: Biopsy of kidney in prone position. Lancet 1:1047—1049, 1954

DR. KASSIRER: Unfortunately, the title was used out of context. The talk was centered around research in renal disease, not around clinical practice. We have never strayed from

13. BOLTON WK, VAUGHAN ED JR: A cooperative study of open

the notion that biopsies are warranted—indeed, mandatory—as

Kidney mt 24:804—806, 1983 15. MAL F, MEGRIER A, CALLARD P. ALTMAN JJ, KLEINKNECHT D, BEAUGRAND M, FERRIER JP: Transvenous renal biopsy: a prelim-

part of properly designed and executed clinical investigative efforts. We have only argued that biopsies are not needed in the

routine care of patients outside such studies. Indeed, I made this distinction quite forcefully in the discussion of the very paper whose title was shown at the meeting. I wrote, "What are the implications for the research efforts currently underway to

determine the optimal approach to INS? Even if a generic analysis shows that the choice between biopsy and 'blind' treatment is a toss-up for all patients with INS {Note: it did [75]}, the conclusion that the biopsy-directed approach offers no

benefit over 'blind' treatment would be applicable only to patients being treated by physicians not engaged in this research. Research efforts to identify optimal therapeutic approaches must continue because more effective treatment regimens will be devised through such efforts [74]." Is renal biopsy necessary for optimal management of the idiopathic nephrotic syndrome (my original title) ? No. Is renal biopsy necessary for optimal study of the idiopathic nephrotic sydrome? Yes.

Iconoclastic views are not always popular, but both the structure of our analysis and the data we employed have withstood the test of time. When solid data on the efficacy of new therapeutic approaches for the various histologic forms of

INS become available, our decision-analysis model could readily be used again to assess whether renal biopsy offers any

surgical and percutaneous renal biopsies, J Urol 117:696—698, 1977 14. CHODAK GW, GILL WB, WALD V, SPARGO B: Diagnosis of renal

parenchymal disease by a modified open kidney biopsy technique.

inary report in 36 patients (abstract). Kidney mt 37:279, 1990

16. WHITE RHR: Observations on percutaneous renal biopsy in children. Arch Dis Child 38:260—266, 1963 17. CARVAJAL HF, TRAVIS LB, SRIVASTAVA RN, DEBEUKELAER

MM, DODGE WF, DUPREE E: Percutaneous renal biopsy in children—An analysis of complications in 890 consecutive biopsies. Tex Rep Biol Med 29:253—264, 1971 18. HEALTH AND PUBLIC POLICY COMMITTEE, AMERICAN COLLEGE

OF PHYSICIANS: Clinical competence in percutaneous renal biopsy. Ann Intern Med 108:301—303, 1988 19. WICKRE CG, GOLPER TA: Complications of percutaneous needle biopsy of the kidney. Am J Nephrol 2:173—178, 1982 20. BACH D, HANRATHS M, MAAR K, GRABENSEE B: Ultrasound guided percutaneous biopsy or operative biopsy in patients with renal impairment. mt Urol Nephrol 20:519—523, 1988 21. DIAZ-BUXO JA, DONADIO JV JR: Complications of percutaneous

renal biopsy: An analysis of 1,000 consecutive biopsies. Clin Nephrol 4:223—227, 1975 22. BACH D, HANRATHS M, GRABENSEE B: Ultrasonic-guided percu-

taneous renal biopsy in patients with renal insufficiency. Ultrason Med 8:149—151, 1958

23. GAULT MH, MUEHRCKE RC: Renal biopsy: Current views and controversies. Nephron 34:1—34, 1983 24. WELT L (cited in KARK RM): Renal biopsy. JAMA 205:220—226, 1968 25. TINKER JH, ROBERTS SL: Anesthesia risk, in Anesthesia (2nd ed), edited by MILLER RD. New York, Churchill Livingstone, 1986, pp 359—380

projects.

26. GOLDMAN L, CALDERA DL, NUSSBAUM SR. SOUTHWICK PS, KROGSTAD D, MURRAY B, BURKE DS, O'MALLEY TA, GOROLL AH, CAPLAN CH, NELAN J, CARABELLO B, SLATER EE: Multi-

Reprint requests to Dr. M. Madaio, Renal Electrolyte Section, Hospital of the University of Pennsylvania, 700 Clinical Research

27. MARX GF, MATEO CV, ORKIN LR: Computer analysis of postanesthetic deaths. Anesthesiology 39:54—58, 1973

special benefit for patients not involved in clinical research

factorial index of cardiac risk in noncardiac surgical procedures. N

Building, 422 Curie Boulevard, Philadelphia, Pennsylvania 19104, USA

Engl J Med 297:845—850, 1977 28. TIRET L, DE5MONT5 JM, HATITON F, VOURC'H G: Complications

associated with anaesthesia: A prospective survey in France. Can

References 1. KASSIRER JP, POLLACK V: Open discussion: Kidney biopsy is indicated in the management of lupus nephritis. Am Soc Nephrol, 1983, p 84 2. BRUN C, OLSEN S: Indications and technique, in Atlas of Renal Biopsy. Coopenhagen, Munksgaard, 1981 3. STRIKER GE, QUADRUCCI Li, CUTLER RE: Use and Interpretation of Renal Biopsy. Philadelphia, Saunders, 1978

4. HEPTINSTALL RH: Pathology of the Kidney. Boston, Little, Brown, 1983 5. WOLSTENI-IOLME GEW, CAMERON MP: Renal Biopsy. Clinical

and Pathological Sign(ficance. Boston, Little, Brown, 1961

6. GWYN NB: Biopsies and the completion of certain surgical procedures. Can Med Assoc J 13:820—823, 1923 7. BALL RP: Needle (aspiration) biopsy. J Tenn Med Assoc 27:203— 206, 1934 8. CA5TLEMAN B, SMITH WICK RH: The relation of vascular disease to the hypertensive state. JAMA 121:1256—1261, 1943 9. PEREZ-ARA A: La biopsia-punctural del rinon no megalico-consid-

erationes generales y aportacion de un nuevo metodo. Bol Liga Contra Cancer 25:121, 1950

10. Ivisor' P, BRUN C: Aspiration biopsy of kidney. Am J Med 11:324—330, 1951

11. ALWALL N: Aspiration biopsy of the kidney. Acta Med Scand 143:430—435, 1952

Anaesth Soc J 33:336—344, 1986 29. CROKER BP, SALOMON DR: Pathology of the renal allograft, in Renal Pathology with Clinical and Functional Correlations, edited by TISCHER CC, BRENNER BM, Philadelphia, Lippincott, 1988, pp 1518—1554

30. KROPP KA, SHAPIRO RS, JHUNJHUNWALA iS: Role of renal biopsy in end-stage renal failure. Urology 12:631—634, 1978 31. FAJERS CM, HOLM J, LINDQVIST B: Percutaneous renal biopsy in

the diagnosis of renal disease in uraemia. Scand J Urol Nephrol 4:153—154, 1970

32. LUSTED LB. MORTIMER GE, HOPPER ii: Needle renal biopsy under image amplifier control. Am J Roentgenol 75:953—955, 1956 33. HADDAD JK, MANI RL: Percutaneous renal biopsy. An improved method using television monitoring and high-dose infusion pyelography. Arch Intern Med 19:157—160, 1967 34. TELFER N, ACKROY DAE, STOCK SL: Radio-isotopic localisation for renal biopsy. Lancet 1:132—133, 1964 35. KELLER HI, MALLOY JP, SAUER GF: The renal scan. An aid to renal biopsy. JAMA 188:1085—1086, 1964 36. COLGAN JR, BISCHEL M, MORROWN JW: Retrograde catheter localization for percutaneous renal biopsy. JAMA 217:824, 1971 37. BOLTON WK, TULLY RI, LEWIS EJ, RANNIGER K: Localization of the kidney for percutaneous biopsy. Ann Intern Med 8 1:159— 164, 1974

38. RAO K: Percutaneous renal transplant biopsy under CAT scanner guidance. Nephron 38:273—275, 1984 39. BIRNHOLZ JC, KASINATH BS, CORWIN HL: An improved tech-

541

Nephrology Forum: Renal biopsy

following aspiration of renal cell carcinoma. J Urol 118:865-867,

nique for ultrasound guided percutaneous renal biopsy. Kidney tnt 40.

27:80—82, 1985 RAPACCINI GL, POMPILI M, CATURELLI E, AMADEI E, ALIOTTA

A, GRATrAGLIANO A, SAvINI E, ANTI M: Real-time ultrasound guided renal biopsy in diffuse renal disease: 114 consecutive cases. Surg Endosc 3:42—45, 1989

1977

69. PAONE DB, MEYER L: The effect of biopsy on therapy in renal

disease. Arch Intern Med 141:1039—1041, 1981

41. MORRISON RT, BARR JG, WILKOWSKI Mi, CAMPBELL JA, NOLAN

70. COHEN AH, NAST CC, ADLER SG, KOPPLE JD: The clinical usefulness of kidney biopsies in the diagnosis and management of renal disease (abstract). Kidney In! 27:135, 1985

CR: A new technique for percutaneous renal biopsy (abstract).

71. TURNER MW, HUTCHINSON TA, BARRE PE, PRICHARD S, JOTHY

Kidney mt 37:281, 1990 42. KARK RM: Renal biopsy. JAMA 205:220—226, 1968 43. THOMSEN A CHR: Representativity of the renal biopsy (chap 4), in The Kidney in Diabetes Me/titus, Copenhagen, Munksgaard, 1965

44. Cotwn'i HL, SCHWARTZ MM, LEWIS EJ: The importance of sample size in the interpretation of renal biopsy. Am J Nephro/ 8:85—89, 1988

45. SCHEEHAN DC, HRAPCHAK BB: Theory and Practice of Histotechnology (2nd ed), St. Louis, Mosby, 1980 46. TISHER CC, CROKER BP JR: Indications for and interpretation of renal biopsy: evaluation by light, electron, and immunofluorescence microscopy, in Diseases of the Kidney, edited by SCHRIER RW, GOTTSCHALK CW, Boston, Little, Brown, 1988, pp 527—556

5, PARFREY PS, SEELEY iF: A prospective study on the impact of

renal biopsy in clinical management: C/in Nephrol 26:217—221, 1986 72. WHITING-O'KEEFE Q, HENKE JE, SHEARN MA, HOPPER J JR,

BIAVA CG, EPSTEIN WV: The information content from renal biopsy in systemic lupus erythematosus. Ann Intern Med 96:718— 723, 1982 73. WHITING-O'KEEFE Q, RICCARDI PJ, HENKE JE, SHEARN MA, HOPPER J JR. EPSTEIN WV: Recognition of information in renal biopsies of patients with lupus nephritis. Ann Intern Med 96:723— 727, 1982

74. KssiREi JP: Nephrology Forum: Is renal biopsy necessary for optimal management of the idiopathic nephrotic syndrome? Kid-

47. BRUN C, RAASCHOU F: The results of 500 percutaneous renal

ney tnt

24:561—575, 1983 J, PAUKER SG, KASSIRER JP: Idiopathic ne-

biopsies. Arch Intern Med 102:716—723, 1958 48. KARK RM, MEUHREKE RC, POLLAK yE, PIItNI CL, KIEFER JH:

75. LEVEY AS, LAU

An analysis of five hundred percutaneous renal biopsies. Arch

phrotic syndrome. Puncturing the biopsy myth. Ann Intern Med

Intern Med 101:439—451, 1958 49. GINSBURG JC, FRANSMAN SL, SINGER MA, COHANIM M, M0R-

76. HLATKY MA: Is renal biopsy necessary in adults with nephrotic

RIN PAF: Use of computerized tomography (CT) to evaluate bleeding after renal biopsy. Nephron 26:240—243, 1980

107:697—713, 1987

syndrome. Lance! 2:1264—1268, 1982 77. GANEVAL D, DANIEL F, LHOSTE BOUCHARD B: Problems diag-

50. ROSENBAUM R, HOFFSTEN PE, STANLEY RJ, KLAHR 5: Use ofnostiques et therapeutiques au course l'insuffisance renal aigue.

computerized tomography to diagnose complications of percutaneous renal biopsy. Kidney tnt 14:87—92, 1978 51. ALTER AJ, ZIMMERMAN 5, KIRACI-IAIWANICH C: Computerized

tomographic assessment of retroperitoneal hemorrhage after percutaneous renal biopsy. Arch Intern Med 140:1323—1326, 1980 52. HAMPERS CL, PRAGER D: Massive bleeding ten days after renal biopsy. Arch Intern Med 114:782—783, 1964 53. SAMELLAS W: Death due to septicemia following percutaneous needle biopsy of the kidney. J Uro/ 91:317—319, 1964 54. BENNETT AR, WIENER SN: Intrarenal arteriovenous fistula and

55.

Actualités Nephro!ogiques de /'hôpital Necker 304—327, 1976 78. BEAUFILS M: Nephrology Forum: Glomerular disease complicating sepsis. Kidney tnt 19:609—618, 1981 79. KOURILSKY 0, SRAER JD, KANFER A, MIGNON F, WHITWORTH J, MOREL-MAROGER L, RICHET G: Intéret de Ia ponction-biopsie rénale dans les insuffisances rénales aigues (abstract), in VIIth mt Congr Nephro/, Montreal, 1978, pp 1—8 80. RICHET G, SRAER JD, KOURILSKY 0, MOREL-MAROGER L: La

ponction-biopsie rénale au cours des insuffisances renales aigUes. Rev Prat 28:3769—3774, 1978

aneurysm: a complication of percutaneous renal biopsy. Am J

81. COPLEY JB: Idiopathic hematuria: a prospective evaluation. Arch

Roentgenol 95:372—382, 1965. EKELUND L, LINDHOLM T: Arteriovenous fistulae following percutaneous renal biopsy. Acta Radio! 11:38—48, 1971

tuna: role of renal biopsy and investigative radiology. Br Med J

56. GRAu JH, G0NICK P, WILSON A: Post-biopsy intrarenal arteriovenous fistula. J Urol 122:233—236, 1979 57. BLAKE S, HEFFERNAN5, MCCANN P: Renal arteriovenous fistula after percutaneous renal biopsy. Br Med J 1:1458—1460, 1963 58. O'BRIEN DP III, PARROTT TS, WALTON KN, LEWIS EL: Renal arteriovenous fistulas. Surg Gynecol Obstet 139:739—743, 1974 59. LEITER E, GRIBETZ D, COHEN S: Arteriovenous fistula after percutaneous needle biopsy—surgical repair with preservation of renal function. N Engl J Med 287:971—972, 1972 60, LOZANO JE, BEACH PD, CARLTON CE: latrogenic renal vascular injury, Urology 12:347—350, 1978 61. IMMERGuT M, PLOTKIN A: Hemorrhage following needle biopsy of the kidney: diagnosis by renal arteriography. JAMA 211:827— 828, 1970 62. JACKSON GG, POIRIER KP, GRIEBLE HG: Concepts of pyelone-

phritis: experience with renal biopsies and long-term clinical observations. Ann Intern Med 47:1165—1183, 1957 63. MyTH RG: The safety of percutaneous renal biopsy: an analysis of

500 consecutive cases. J Uro/ 94:1—3, 1965 64. SLOTKIN EA, MADSEN P0: Complications of renal biopsy; incidence in 5,000 reported cases. J Urol 87: 13—15, 1962 65. GALLOWAY DC, TYTLE T: Laceration of a mesenteric artery: unusual complication of percutaneous renal biopsy. South Med J 73:1413—1414, 1980 66. ALMKUIST RD, BUCKALEW VM: Techniques of renal biopsy. Uro! C/in North Am 6:503—5 17, 1979 67. RIVER GL, DOVENBARGER WV, NIKOLAI TF, MOFFAT NA:

Unusual complications of kidney biopsy. J Urol 103:15—17, 1970 68. GIBBONS RP, BUSH WH, BURNETT LL: Needle tract seeding

Intern Med 147:434—437, 1987 82. MICHAEL J, JoNEs NF, DAvIES DR. TIGHE TR: Recurrent hema1:686—688, 1976

83. SINNIAH R, PWEE HS, LIM CM: Glomerular lesions in asympto-

matic microscopic hematuria discovered on routine medical examination. Clin Nephro! 5:216—228, 1976 FB, Ro 5, NOE N, JERKINS G: Hypercalciuria in

84. STAPLETON

children with hematuria. N

Eng/ J Med 310:1345—1348, 1984

85. COPLEY JB: Isolated hematuria in the adult. Am J Med Sci 29:101—111, 1986 86. TRACHTMAN H, WEISS RA, BENNETT B, GREIFER I: hematuria

Isolated

in children: Indication for a renal biopsy. Kidney Int

25:94—99, 1984 87.

SINNIAI-I R, LAW CH, PWEE HS: Glomerular lesions in patients with asymptomatic persistent and orthostatic proteinuria discovered on routine medical examination. C/in Nephrol 7:1—14, 1977

88. GLA5SOCK RJ, ADLER SG, WARD HJ, COHEN AH: Primary glomerular diseases, in The Kidney (3rd ed), edited by BRENNER BM, RECTOR FC Jr. Philadelphia, Saunders, 1986, pp 929—1013 89. GLASSOCK Ri, COHEN AH, ADLER SO, WARD Hi: Secondary glomerular diseases, in The Kidney (3rd ed), edited by BRENNER BM, RECTOR FC ir, Philadelphia, Saunders, 1986, pp 1014—1084 90. Drriov J, WEIDMANN P, BERNSTEIN M, MASSRY SG: Methicillin nephritis. Medicine 56:483—491, 1977 91. LINTON AL, CLARK WF, BRIEDGERAA, TURNBULL DI, LINDSAY

RM: Acute interstitial nephnitis due to drugs. Ann Intern Med 93:7735—7741, 1980

92. VAN YPERSELE DE STRIHOU C: Nephrology Forum: Acute oligunc interstitial nephritis. Kidney tnt 16:751—765, 1979 93. LYONS H, PINN VW, CORTELL 5, COHEN ii, HARRINGTON iT: Allergic interstitial nephritis causing reversible renal failure in four

542

Nephrology Forum: Renal biopsy

patients with idiopathic nephrotic syndrome. N Eng! J Med 94. BORDER WA, LEHMAN DH, EGAN JD, SASS Hi, GLADE iE, WILSON CB: Anti-tubular basement membrane antibodies in methicillin-associated interstitial nephritis. N Engi J Med 291:381—384, 1974 95. BERGSTEIN J, LITMAN N: Interstitial nephritis with anti-tubular basement membrane antibody. N Engi J Med 292:875—873, 1975

96. MOREL-MAROGER L, BEAUFILS M, RICHET G: The kidney in dysproteinemias, in Nephrology, edited by HAMBURGER i, CROSNIER J, GRUNFELD J-P, New York, Wiley, 1979, p 711 97. FREY BM, FREY FJ, ZIMMERMAN A, ARQUINT N, FLURY B, WEGMULLER E, HODLER J: Prediction of the histological type of glomerulonephritis. Nephron 25:276—279, 1980 98. MADAIO MP, HARRINGTON JT: The diagnosis of acute glomerulonephritis, N Engi J Med 309:1299—1302, 1983 99. Biuccs MA, JOHNSON JP, TIECHMAN S, YEAGLER HC, WILSON CB: Anti-glomerular basement membrane antibody mediated gbmerubonephritis and Goodpasture's syndrome. Medicine 58:348— 360, 1979 100. BOWMAN C, LOCKWOOD CM: Clinical application of a radioimmunoassay for autoantibodies to glomerular basement membrane.

J C/in Lab Immunol 17:197—202,

101.

of mean glomerular volume in patients with insulin-dependent diabetes mellitus. Kidney mt 32:930—932, 1987

288:124—128, 1973

115. BASGEN JM, RICH SS, MAUER SM, STEFFES MW: Measuring the

volume density of the glomerular mesangium. Nephron

50:182—

186, 1988 116. ELLIS EN, MAUER SM, SUTHERLAND DE, STEFFES MW: Gbmerular capillary morphology in normal humans. Lab Invest 60:231—236, 1989

117. MAUER SM, BiLous RW, ELLIS E, HARRIS R, STEFFES MW: Some lessons from the studies of renal biopsies in patients with insulin-dependent diabetes mellitus. J Diabetic Complications 2:197—202, 1988

lbS. BILOUS RW, MAUER SM, SUTHERLAND DER, NAJARIAN iS, GOETZ FC, STEFFES MW: The effects of pancreas transplantation

on the gbomerular structure of renal albografts in patients with insulin-dependent diabetes. N Eng! J Med 321:80—85, 1989 119. FRANKLIN WA, JENNINGS RB, EARLE DP: Membranous glomer-

ulonephritis: Long-term serial observations on clinical course and morphology. Kidney mt 4: 36—56, 1973 120. DAvisoN AM, CAMERON iS, KERR DNS, OGG CS, WILKINSON

RW: The natural history of renal function in untreated idiopathic membranous gbomerulonephritis in adults. Gun Nephrol 22:61—67,

1985 VAN DER WOUDE FJ, RASSMUSSEN N, LOBATTO S. WIIK A, PERMIN H, VAN ES LA, VAN DER GIESSER M, VAN DER HEM

1984 121. NOEL LH, ZANETTI M, DROZ D, BARBANEL C: Long-term prog-

GKTH: Autoantibodies against neutrophils and monocytes: tools for diagnosis and marker of disease activity in Wegener's granulomatosis, Lancet 1:425—429, 1985

66:82—90, 1972 122. GLUCK MC, GALLO G, LOWENSTEIN J, BALDWIN DS: Membra-

102. SAVAGE COS, WINERLS CG, JONES S. MARSHALL PD, LOCKWOOD CM: Prospective study of radioimmunoassay for antibodies

against neutrophil cytoplasm in diagnosis of systemic vasculitis.

Lancet 1:1389—1393, 1987

103. FALK Ri, JENNETTE C: Anti-neutrophil cytoplasmic autoantibodies with specificity for myeloperoxidase in patients with systemic vasculitis and idiopathic necrotizing and crescentic glomerulone-

phritis. NEngIJ Med 318:1651—1656,

1988

104. NOLLE B, SPECKS M, LUDEMANN J, ROHRBACH MS. DEREMEE

RA, GROSS WL: Anti-cytoplasmic autoantibodies: Their immunodiagnostic value in Wegener's granulomatosis. Ann Intern Med 111:28—40, 1989 105. COHEN TERVAERT JW, VAN DER WounE FJ, FAUCI AS, AMBRUS

JL, VELOSA J, KEANE WF, MEIJER S, VAN DER GIESSEN M, HAUW THE T, VAN DER HEM GK, KALLENBERG CGM: Associa-

tion between active Wegener's granulomatosis and anticytoplasmic antibodies. Arch Intern Med 149:2461, 1989 106. HRICIK DE, KA5SIRER JP: The nephrotic syndrome. Disease-aMonth 28:1—56, 1982

107. TOMORA 5, TSUTANT K, SAKUMA A, TAKEUCHI J: Discriminant

analysis in renal histological glomerular diseases. C/in Nephrol 23:55, 1985 108. PONTICELLI C, ZUCCHELLI P. PASSERINI P, CAGNOLI L, CESANA B, Pozzi C, PASQUALI 5, IMBASCIATI E, GRAssI C, REDAELLI B,

SASDELLI M, LOCATELLI F: A randomized trial of methyiprednis-

olone and chlorambucil in idiopathic membranous nephropathy. N Engi J Med 320:8—13, 1989

109. RID50N RA, SLOPER JC, DE WARDENER HE: Relationship be-

tween renal function and histological changes found in renal

nosis of idiopathic membranous glomerubonephritis. Am J Med

nous glomerulonephritis. Evolution of clinical and pathologic features. Ann Intern Med 78:1—12, 1973 123. COLLABORATIVE STUDY OF THE ADULT IDIOPATHIC NEPHROTIC SYNDROME: A controlled study of short-term prednisone treatment in adults with membranous nephropathy. N Eng! J Med 301:1301—1306, 1979 124. TORNROTH T, HONKANEN E, PETTERSSON E: The evolution of

membranous glomerubonephritis reconsidered: new insights from a study on relapsing disease. C/in Nephrol 28:107—117, 1987 125. DONADIO JV JR. TORRES VE, VELOSA JA, WAGONER RD, HOLLEY KE, OKAMURA M, ILSTRUP DM, CHU CP: Idiopathic mem-

branous nephropathy: The natural history of untreated patients. Kidney mt 33:708—715, 1988 126. SHEMESH 0, Ross JC, DEEN WM, GRANT OW, MYERS BD: Nature of the gbomerular capillary injury in human membranous glomerubopathy. J C/in Invest 77:868—877, 1986 127. KATAFUCHI R, TAKEBAYASI-II 5: Morphometrical and functional

correlations in benign nephrosclerosis. Clin Nephrol 28:238—243, 1987 128. RICHARDSON WP, HASSANEIN R, PINNICK RV, SAVIN Vi: Ultrafiltration coefficients of glomeruli from human biopsies. Kidney Int 34:845—852, 1988

129. AUSTIN III HA, MUENZ LR, JOYCE KM, ANTONOVYCH TA, KULLICK ME, KLIPPEL ill, DECKER JL, BALOW iE: Prognostic factors in lupus nephritis contribution of renal histologic data. Am J Med 75:382—391, 1983

130. SCHWARTZ MM, KAWALA KS, CORWIN HL, LEwis EL: The

prognosis of segmental gbomerulosclerosis in systemic lupus erythematosus. Kidney mt 32:274—279, 1987

biopsy specimens from patients with persistent glomerular nephri-

131. HECI-IT B, SIEGEL N, ADLER M, KASI-IGARIAN M, HAYSLETT JP Prognostic indices in lupus nephritis. Medicine 55:163—181, 1976

110. STRIKER GE, SCHAINUCK LI, CUTLER RE, BENDITT EP: Struc-

132. SINNIAH R, FENG PH: Lupus nephritis: correlation between light,

tis. Lancet 2:363, 1968

tural-functional correlations in renal disease. I. A method for assaying and classifying histopathologic changes in renal biopsies. Hum Pathol 1:615—630, 1970 Ill. SCHAINUCK LI, STRIKER GE, CUTLER RE, BENDITT EP: Structural-functional correlations in renal diseases. II. The correlations. Hum Pathol 1:631—642, 1970 112. PARRISH AE, KRAMER NC, HATCH FE, WArr MF, HowE iS:

electron microscopic and immunofluorescent findings and renal function. C/in Nephro! 6:340-351, 1976 133. STAMENKOVIC I, FAVRE H, DONATH A, ASSIMACOPOULOUS A,

DM, GOETZ FC: Structural-functional relationships in diabetic nephropathy. J Gun Invest 74:1143—1155, 1984

CHATELANAT F: Renal biopsy in SLE irrespective of clinical findings: long-term follow-up. C/in Nephrol 26:109—115, 1986 134. LEE HS, MuJAIS SK, KASINATH BS, SPARGO BH, KATZ Al: Course of renal pathology in patients with systemic lupus erythematosus. Am J Med 77:612—620, 1984 135. MAGIL AB, BALLON HS, CHAN \', LIRENMAN DS, RAE A, SUTTON RAL: Diffuse proliferative lupus glomerubonephritis: determination of prognostic significance of clinical, laboratory and pathologic factors. Medicine 63:210—220, 1984

114. BILOUS RE, MAUER SM, BASGEN JM, STEFFE5 MW: Estimation

136. KANT KS, POLLAK yE, WEISS MA, GLUECK HI, MILLER MA,

The relationship between glomerular function and histology in acute glomerulonephritis. J Lab C/in Med 58:197—203, 1961 113. MAUER SM, STEFFES MW, ELLIS EN, SUTHERLAND DER BROWN

Nephrology Forum: Renal biopsy HESS EV: Glomerular thrombosis in systemic lupus erythematosus: prevalence and significance. Medicine 60:71—86, 1981 137. LEE P, URowITz MB, BOOKMAN AAM, KOEHLER BE, SMYTHE HA, GORDON DA, OGRYZLO MA: Systemic lupus erythematosus:

a review of 100 cases with reference to nephritis, the nervous system, infections, aseptic necrosis and prognosis. Q J Med 181:1—32, 1977

138. FRIES JF, WEYL S, HOLMAN HR: Estimating prognosis in systemic lupus erythematosus. Am J Med 57:561—565, 1974 139. BALDWIN DS, GLUCK MC, LOWENSTEIN J, GALLO GR: Lupus nephritis: clinical course as related to morphologic forms and their transitions. Am J Med 62:12—30, 1977 140. CARETTE 5, KLIPPEL JH, DECKER JL, AUSTIN HA, Piorz PH, STEINBERG AD, BALOW JE: Controlled studies of oral immunosuppressive drugs in lupus nephritis. Ann Intern Med 99:1—8, 1983 141. AUSTIN HA III, KLIPPEL JH, BALOW JE, LERICHE NGH, STEINBERG AD, PLOTZ PH, DECKER JL: Therapy of lupus nephritis:

543

JL: Effect of treatment on the evolution of renal abnormalities in lupus nephritis. N EnglJ Med 311:491—495, 1984 144. RUSH PJ, BAUMAL R, SHORE A, BALFE JW, SCHREIBER M: Correlation of renal histology with outcome in children with lupus nephritis. Kidney un 29:1066—1071, 1986 145. SCHWARTZ MM, BERNSTEIN J, HILL GS, HOLLY KS, PHILLIPS

EA, and the Lupus Nephritis Collaborative Study Group: The predictive value of renal pathology in diffuse proliferative lupus

glomerulonephritis. Kidney mt 36:891—896, 1989 146. CHAGNAC A, KLIBERD BA, FARINAS MC, STROBER S, SIBLEY RK, HOPPE R, MYERS BD: Outcome of acute glomerular injury in

proliferative lupus nephritis. J Clin Invest 84:922—930, 1989 147. SCHWARTZ MM, LANS S, BONSIB SM, GEPHARDT GN, SHARMA

HM, Lupus Nephritis Collaborative Study Group: Clinical outcome of three discrete histologic patterns of injury in severe lupus glomerulonephritis. Am J Kidney Dis 13:273—283, 1989 148. METAXAS P: Nephrology Forum: Familial Mediterranean fever

142. AUSTIN HA III, MUENZ LR, JOYCE KM, ANTONOVYCH TT, BALOW JE: Diffuse proliferative lupus nephritis: identification of

and amyloidosis. Kidney ml 20:676—685, 1981 149. GAFNI J, SOHAR E, ZEMER D: Amyloid nephropathy, in Nephrology, edited by HAMBURGER J, CROSNIER J, GRUNFELD J-P, New York, Wiley, 1979, p 689 150. BRANDT K, CATHEART ES, COHEN AS: A clinical analysis of the

specific pathologic features affecting renal outcome. Kidney ml

course and prognosis of 42 patients with amyloidosis. Am J Med

controlled trial of prednisone and cytotoxic drugs. N Engi J Med 314:614—619, 1986

25:689—695, 1984

143. BALOW JE, AUSTIN HA III, MUENZ LR, JOYCE KM. ANTONOVYCH TT, KLIPPEL JH, STEINBERG AD, PLOTZ PH, DECKER

44:955—969, 1968

151. KYLE RA, BAYRD ED: Amyloidosis: Review of 236 cases. Medicine (Baltimore) 54:271, 1975