Vol. 114, September Printed in U.S.A.

THE JOURNAL OF UROLOGY

Copyright© 1975 by The Williams & Wilkins Co.

UROLOGIC MANIFESTATIONS OF PAROXYSMAL NOCTURNAL HEMOGLOBINURIA VICTOR BRAREN, * STEPHEN A. BUTLER, ROBERT C. HARTMANN

AND

DAVID E. JENKINS, JR.

From the.Department of Surgery, Division of Urology, Department of Medicine, Division of Hematology and the Elliot V. Newman Clinical Research Center, Vanderbilt University Medical Center, Nashville, Tennessee

ABSTRACT

A long-term study of 17 patients with paroxysmal nocturnal hemoglobinuria revealed an unexpectedly high incidence of functional and anatomic renal abnormalities. All patients demonstrated varying degrees of hematuria and proteinuria distinct from hemoglobinuria. All patients also had granular casts in multiple urinalyses. Evaluation of renal function revealed hyposthenuria, abnormal tubular function and declining creatinine clearance. Radiologically, one or more of these demonstrated enlarged kidneys, renal cortical infarcts and thinning, papillary necrosis, acute renal atrophy, retroperitoneal hematoma and ureteral infarction, which were confirmed by autopsy studies. Hypertension developed in 7 patients. Urinary tract infection was uncommon and no patient had a clinical history compatible with chronic or acute pyelonephritis. Contrary to usual opinion our studies clearly showed evidence of frequent and widespread renal pathology in paroxysmal nocturnal hemoglobinuria most likely due to repeated microvascular thromboses similar to the venous thrombosis involving other organs in this disorder. Since most of these patients present initially to urologists knowledge of this entity is mandatory. Paroxysmal nocturnal hemoglobinuria is a rare, acquired hemolytic disorder of unknown etiology characterized by 2 major clinical features: chronic intravascular hemolysis with episodes of gross hemoglobinuria and diffuse venous thrombosis. The clinical course of these patients cannot be predicted and may vary from severe hemolysis and thrombosis to years of relative quiescence. 1- 3 Factors precipitating hemolysis include infections, reactions to drugs and immunizations, and exercise but frequently the cause remains unknown. Renal failure has been reported in paroxysmal nocturnal hemoglobinuria but usually has been attributed to concomitant renal disease or transfusion reaction. In an extensive review of the subject Crosby stated that "when renal failure has occurred in paroxysmal nocturnal hemoglobinuria it was not because of siderosis but a consequence of repeated episodes of pyelonephritis". 1 He also stated that " ... in paroxysmal nocturnal hemoglobinuria there is no scarring of the kidney or destruction of nephrons unless infection supervenes". Marked hemosiderin deposits in the proximal renal tubule are common in all autopsy and biopsy Accepted for publication January 17, 1975. Read at annual meeting of American Urological Association, St. Louis, Missouri, May 19-23, 1974. Supported in part by Grants HL 03509, FR 05424 and RR-95 from the United States Public Health Service. * Requests for reprints: Division of Urology, Vanderbilt University, Nashville, Tennessee 37232. 430

reports dealing with paroxysmal nocturnal hemoglobinuria. 2 It has been repeatedly stated that these deposits have no effect on renal function. Dacie stated, "It is a remarkable fact that kidney function is seldom severely impaired in paroxysmal nocturnal hemoglobinuria despite continuing intravascular hemolysis and the heavy siderosis of the kidney" .2 There are reports of paroxysmal nocturnal hemoglobinuria patients suffering acute renal failure after transfusion•-s but little documentation of chronic renal failure.2· 4 There are no reports of long-term followup of renal function in patients with paroxysmal nocturnal hemoglobinuria, although Dies and associates reported a defect in urine concentrating ability in a patient known to have renal hemosiderosis and considered this a sign of early renal failure. 7 Our report concerns a review of the renal status of 17 patients followed from 1957 until the present. The case numbers are the same as in previous publications. 3 • 8 • 9 Contrary to usual opinion our studies clearly show evidence of widespread nephrologic and urologic abnormalities in paroxysmal nocturnal hemoglobinuria in which the major factor is probably repeated thromboses. MATERIALS AND METHODS

The diagnosis of paroxysmal nocturnal hemoglobinuria was confirmed using the acid hemolysis test 2 and the sucrose hemolysis test. 10 Patient age

PAROXYSMAL NOCTURNAL HEMOGLOBINURIA

431

prolonged venous phases, suggesting microvascular obstruction in the renal cortices (fig. 3). One patient (case 18) had papillary necrosis and gross hematuria and proceeded to infarct the left ma! ureter (fig. 4). All 5 patients who did not have IVPs (cases 2, 3, 9, 11 and 15) had abnormal renal function tests. Renal function studies. Urinalysis: During the many years of observation all patients except case 12 had episodes of mild to severe gross hemoglobinuria. When diligently sought and carefully delineated from hemoglobinuria, all patients had microscopic hematuria on most occasions and 3 initially presented with gross hematuria. Microscopic hematuria was seen remote from episodes of gross hemoglobinuria. Intermittent and granular casts were seen at some time in most patients. The proteinuria (for example 3.4 gm. per day) was frequently more than could be accounted RESULTS for by the presence of (occult) hemoglobin in the Radiologic. Of 12 patients in whom IVP studies urine. 9 Bacteriology: Urinary tract infections were not were done 7 had l or both kidneys measuring more than 14 cm. in length. Six had radiologic evidence prominent feature in our patients. Four patients of cortical infarction (fig. 1) and 1 of cortical thin- had positive urine cultures at some time. One ning. Five patients had normal IVPs but the IVP patient (case 4) had an Escherichia coli in 1 of these (case 8) was at the time of the original tract infection associated with benign prostatic diagnostic studies and was not repeated during the hypertrophy. One patient (case 8) gave a history of subsequent 9 years prior to death in 1970. One Trichomonas prostatitis and 2 (cases 7 and 15) had patient (case 4) subacutely infarcted the right E. coli urinary tract infections. However, no pakidney. An IVP showed a kidney that had de- tient had a clinical history compatible with chronic creased in size by 4 cm. as compared to a year or acute pyelonephritis. Concentration tests: Fishberg concentration previously. An arteriogram showed a contracted .right kidney and failure to visualize the cortical tests were done repeatedly in 9 patients and all arteriolar structures (fig. 2). Three other patients demonstrated inability to concentrate the urine to had arteriograms and all showed strikingly a specific gravity greater than 1.016. Normally 1 urine specimen should have reached a gravity of 1.025 (800 mOsm. per I.). One (case 15) had a urine osmolality of only 480 mOsm. per 1. after 24 hours of dehydration and after :JO hours of dehydration had a specific gravity of 1.019. Most of these patients had abnormal Fishberg tests within 5 years of the onset of paroxysmal nocturnal hemoglobinuria. The higher numbered patients did not have concentration tests because of our increasing concern regarding the association of thrombosis and dehydration. Phenolsulfonphthalein excretion: The most sensitive indication of abnormal tubular excretion is the first 15 minutes of this test. Normally an average of 35 per cent of the injected is excreted within this period with a minimum excretion of 28 per cent. Of 9 patients studied 7 had defective tubular function as measured by a low excretion of phenolsulfonphthalein (less than 25 per cent) within the first 15 minutes. Again, most of these patients had abnormal phenolsulfonphthalein excretion rates within 5 years after the onset of paroxysmal nocturnal hemoglobinuria. Glomerular filtration rate: Of 16 patients tested 13 had reduced creatinine clearance rates. Eleven had creatinine clearances of less than 70 ml. per Frc. 1. Case 18. Segmental cortical infarction of right minute and 8 less than 60. The lowest creatinine kidney.

at the time of diagnosis ranged from 14 to 67 years. Time of diagnosis was arbitrarily set at the onset of gross hemoglobinuria. One patient (case 12) never had gross hemoglobinuria and onset was timed from the discovery of positive serologic tests for paroxysmal nocturnal hemoglobinuria. There were 10 female (4 dead) and 7 male patients (5 dead). Of the patients who died 7 were autopsied. Renal studies were carried out in the ch:ronic, steady state, that is remote from episodes of gross hemoglobinuria. Renal evaluation consisted of 1) Tadiologic-excretory urogram (IVP) and renal arteriogram; 2) renal function studies-urinalysis, bacteriology, concentration test (Fishberg) and urine osmolality, phenolsulfonphthalein excretion, glomerular filtration rate by creatinine clearance, and effective renal plasma flow by para-aminohippurate clearance and 3) autopsy findings.

432

BRAREN AND ASSOCIATES

FIG. 2. Case 4. Arteriogram demonstrates renal infarction of lower half of right kidney following episode of gross hematuria and flank pain.

clearance was 17. Two patients (cases 3 and 16) had reduced creatinine clearance rates after episodes of acute tubular necrosis. Effective renal plasma flow: Para-aminohippurate clearances were used to measure the effective renal plasma flow in 4 patients. Only 1 patient (case 12) had a normal effective renal plasma flow. This patient had relatively mild paroxysmal nocturnal hemoglobinuria with no episodes of gross hemoglobinuria or major thromboses. Renograms: Renograms were done on 5 patients (cases 5, 12, 14, 17 and 18). Three patients (cases 5, 12 and 17) demonstrated delayed time to peak. One patient (case 18) demonstrated decreased parenchymal phase and an obstructive pattern of the remaining ureter (after left nephrectomy). Autopsy findings. Eight of our patients have died and autopsies were done in 7 (cases 3, 4, 6, 7, 8, 11 and 13). One patient (case 18) had a left nephrectomy after renal hemorrhage and infarction of the proximal third of the ureter. General: Hepatic venous thrombosis was the cause of death in 4 patients (cases 4, 6, 11 and 13) and an associated finding in 2 others (cases 7 and 8). Two patients (cases 11 and 13) also had thrombosis and obstruction of the inferior vena cava below the entrance of the hepatic vein. Pulmonary thrombi or emboli were documented in 5 cases (cases 4, 6, 8, 11 and 13). Diffuse submucosal hemorrhage of the gastrointestinal tract, ureter and bladder was noted in 5 patients (cases 4, 6, 8, 11 and 13). Renal: All patients had moderate to heavy hemosiderin deposits in the proximal renal tubules. Kidney weights ranged from 150 to 380 gm. per kidney with an average weight of 258 gm. (compared to a normal average of 150 gm.). Four patients had renal infarcts (cases 4, 5, 11 and 18), 3 had cortical thinning (cases 7, 8 and 18) and 3

FIG. 3. Case 17. Arteriogram demonstrates lack of visualization of cortical vessels.

had papillary necrosis (cases 3, 11 and 18). Additionally, 4 patients had marked hyalinization of the interstitium with scarring (cases 3, 4, 7 and 11). One patient (case 11) had uric acid uropathy manifested by uric acid casts in the collecting tubules with generalized intrarenal hydronephros1s. Adrenal: Five patients (cases 3, 4, 7, 11 and 13) had atrophy of the adrenal cortex and 1 (case 8) had lipid depletion. One patient (case 4) also had a blunted response to adrenocorticotropic hor-

433

PAROXYSMAL NOCTURNAL HEMOGLOBINURIA

FIG. 4. Case 18. Retrograde pyelograms demonstrate A, papillary necrosis (September 12, 1973) and B, ureteral infarction with stricture and hydronephrosis (October 26, 1973).

mone stimulation. Three patients (cases 7, 8 and had never been on adrenocorticosteroids. Whether the prolonged androgen (mainly fluoxymesterone) therapy administered to some of these patients bore any relationship to adrenal atrophy remains unknown. DISCUSSION

In a previous publication we have presented an in depth description of the nephrologic and hematologic abnormalities of paroxysmal nocturnal hemoglobinuria. 11 A brief urologic review is presented herein. We stress the etiologic role of venous thrombosis resulting in diffuse microinfarctions. We agree with Stri.ibing that this is probably caused by sludging of erythrocyte ghosts in the venous side of the capillary system. 12 We also have stressed the role of hemoglobinuria and hemosiderosis in the production of the pathology while stressing that the major insult is brought about by microinfarction. We have focused much of our attention on renal failure in paroxysmal nocturnal hemoglobinuria. We believe that acute renal failure is brought about hemoglobinuria, whereas chronic renal failure is the result of microinfarction. A review of our group of patients clearly demonstrates the progressive nature of chronic renal failure. Hypertension was demonstrated in 7 of 17 patients, although renal vein renins were not obtained in any patient. We are currently studying the patients in our series for this parameter. It should be stressed that all patients had abnormal urinalyses. The finding of enlarged kidneys with IVP evidence of gross renal infarction coupled with the abnormalities of urinalysis will hopefully bring to mind the possibility of diagno-

sis of paroxysmal nocturnal hemoglobinuria. The addition of a renal biopsy may assist in the diagnosis. Under such conditions we should that definitive studies with acid hemolysis tests and/or sucrose hemolysis tests would be undertaken to screen patients presenting in this manner. It is interesting to note the rare phenomenon of ureteral infarction seen in 1 of our patients 18). We believe that here also venous thrombosis and microinfarction of the ureteral vessels were responsible. REFERENCES

1. Crosby,

2.

3.

4.

5.

6. 7.

8.

W. H.: Paroxysmal nocturnal hemoglobinuria; relation of clinical manifestations to underlying pathogenic mechanisms. Blood, 8: 769, 1953. Dacie, J. V.: Haemolytic Anaemias-Congenital and Acquired, 2nd ed. New York: Grune & Stratton, Inc., part IV, p. 1128, 1967. Hartmann, R. C. and Kolhouse, J. F.: Viewpoints on the management of paroxysmal nocturnal hemoglobinuria. Ser. Haematol., 5: 42, 1972. Heitzman, E. J., Campbell, J. S. and Stefanini, M. Paroxysmal nocturnal hemoglobinuria with hemosiderin nephrosis. Amer. J. Clin. Path., 23: 975, 1953. Heffernan, C. K. and Jaswon, N.: A case of paroxysmal nocturnal haemoglobinuria associated with secondary haemochromatosis, lower nephron nephrosis, and megaloblastic anaemia. J. Clin. Path., 8: 211, 1955. Rubin, H.: Paroxysmal nocturnal hemoglobinuria with renal failure. J.A.M.A., 215: 433, 1971. Dies, F., Lisker, R. and Rivera, A.: Nocturnal paroxysmal hemoglobinuria. II. Change of renal function in a clinical case. Rev. Invest. Clin., 13:. 219, 1961. Peytremann, R., Rhodes, R. S. and Hartmann, R. C.: Thrombosis m paroxysmal nocturnal hemo-

434

BRAREN AND ASSOCIATES

globinuria (PNH) with particular reference to progressive, diffuse hepatic venous thrombosis. Ser. Haematol., 5: 115, 1972. 9. Hartmann, R. C., Jenkins, D. E., Jr., McKee, L. C. and Heyssel, R. M.: Paroxysmal nocturnal hemoglobinuria: clinical and laboratory studies relating to iron metabolism and therapy with androgen and iron. Medicine, 45: 331, 1966. 10. Hartmann, R. C., Jenkins, D. E., Jr. and Arnold, A.

B.: Diagnostic specificity of sucrose hemolysis test for paroxysmal nocturnal hemoglobinuria. Blood, 35: 462, 1970. 11. Butler, S. A., Braren, V., Hartmann, R. C., Jenkins, D. E., Jr., Hutchison, J. K., Rhamy, R. K. and Arnold, A. B.: The kidneys in paroxysmal nocturnal hemoglobinuria. Blood, in press. 12. Striibing, P.: Paroxysmale haemoglobinurie. Deutsche Med. Wchnschr., 8: 1, 1882.

Tm;

Vol. 114, September

JouRNAL oF UROLOGY

Copyright© 1975 by The Williams & Wilkins Co.

Printed in US.A.

EDGE ENHANCEMENT OF RENAL X-RAY STUDIES ARTHUR N. TESSLER

AND

JEFFREY KOSSOW

From the Department of Urology, New York University Medical Center, New York, New York

Edge enhancement is a process in which margins in a radiogram or photograph are emphasized and backgrounds are eliminated for easier recognition of boundaries, lines and fine structures. It has been used in industry for the detection of cracks and and the measurements of holes, mechanical clearances and density of materials. In urology, it can be an adjunct in the interpretation of excretory urograms (IVPs), ne"phrotomograms and angiograms.

enhanced and an irregularly shaped soft tissue density is again noted. The margins of the lesion are well delineated and the difference in tissue density between the lesion and the rest collecting system is better appreciate,-,. In A caliceal diverticula are demonstrated In the enhanced mode, figure 5, B, the the diverticula is better appreciated.

1

DISCUSSION

The purpose of edge enhancement is to augment pyelography and angiography in our armamentarium in order to better define various renal lesions. The main advantage of the is that various densities of the kidney can recorded in greater detail accentuating the margins of renal lesions. The also offers increased resolution of the finer structures within the kidney and permits magnification of the nal x-ray for more detailed inspection. It represents

METHODS AND MATERIALS

To produce and enhance the magnified picture the radiogram must first be illuminated by a light box and scanned by a high performance television ,camera. High speed analogue computer circuits then calculate the density gradient. The results of · this computation, the enhanced picture, are then displayed on a television screen and a photograph is taken. The equipment used for this study was an edge enhancer model 401. * EXAMPLES

In figure 1, A the nephrographic phase of a renal arteriogram demonstrates a radiolucent cyst in the upper pole of a kidney. The cyst is enhanced in figure 1, B. The margins of the cyst wall are sharply defined and the differences in density between the cyst and parenchyma are better appreciated. In figure 2, A a large mass in the lower pole of the left kidney is demonstrated with renal arteriography. In figure 2, B the left kidney is delineated in the enlarged enhanced mode. In this picture the margins and extent of the lesion can be seen much more readily than on the original radiogram. The enhancement process accentuates the outline of the lesion and the abnormal tumor vessels, which become more obvious with this technique. In figure 3, A an irregularly shaped filling defect in the renal i.s noted on an IVP. In the enhanced mode, figure 3, B, the extent of the lesion is better delineated and enlarged making the diagnosis of transitional cell carcinoma more apparent. Another example of transitional cell carcinoma of the renal pelvis is seen in figure 4. In part A a large irregularly shaped filling defect of the renal pelvis is demonstrated. In part B the lesion is Accepted for publication January 10, 1975. Read at annual meeting of New York Section, American Urological Association, The Royal College of Surgeons, London, England, September 19-27, 1974. * Spatial Data Systems, Goleta, California.

FIG. 1. A, radiolucent cyst with splaying of calices seen on IVP. B, edge enhancement of radiolucent cyst.

435