- Email: [email protected]
Urinary Lipid Bodies in polycystic Kidney Disease
Urinary Lipid Bodies in Polycystic Kidney Disease Kirk A. Duncan, MD, Francis E. Cuppage, MD, and Jared J. Grantham, MD • Urinary doubly refractile lipid bodies (oval fat bodies) are observed most frequently in patients with heavy proteinuria resulting from glomerular disease. We observed doubly refractile lipid bodies (DRLB) in the urine sediment of 60% of 35 patients with autosomal dominant polycystic kidney disease (ADPKD). All patients had clinical courses typical of ADPKD, and none exhibited the features of a second, unrelated renal disease. DRLB in the urine were correlated with a urine dipstick protein reading exceeding trace. Age, sex, BP, and serum creatinine concentration were not associated with the presence of DRLB in the urine. Examination of cyst fluid obtained from kidneys of six ADPKD patients revealed DRLB in 80% of cyst fluid samples that contained degraded blood (socalled chocolate cysts). The DRLB in cyst fluid were morphologically indistinguishable from those observed in urine, and DRLB from both sources were stained with oil red O. We conclude that urinary DRLB are a clinical feature of ADPKD. INDEX WORDS: Polycystic kidney; cysts; lipiduria; proteinuria.
A. UTOSOMAL dominant f t disease (ADPKD) is a
polycystic kidney relatively common disorder, occurring in approximately 1 in every 500 persons, 1.2 and is the third or fourth most frequent cause of end-stage renal disease. 34 Transmitted in an autosomal dominant fashion, ADPKD is characterized by cysts diffusely scattered throughout the parenchyma of both kidneys. These cysts, some of which may be present at birth, progressively enlarge, ultimately acquiring a size of less than a millimeter to several centimeters in diameter. Patients often remain asymptomatic until the fourth or fifth decades of life, when they may present clinically with massively enlarged kidneys, flank pain, hematuria, hypertension, or azotemia.1.5-7 Most patients develop progressive renal failure and require dialysis in the fifth or sixth decades of life. 1.5-7 We noted the presence of doubly refractile lipid bodies (DRLB, oval fat bodies) in the urine of several of our patients with ADPKD, a finding that had not previously been reported. To investigate further the incidence of DRLB, we did a combined prospective and retrospective study of urine sediments obtained from patients with ADPKD. We have also examined cyst fluid and histopathologic specimens in selected cases. We found DRLB in the urine and in the cyst fluids in a majority of the patients examined. MATERIALS AND METHODS Thirty-five patients attending the Polycystic Kidney Clinics at the University of Kansas Medical Center and the Kansas City Veterans Administration Medical Center were surveyed. All patients had previously had either excretory urography, ultrasonography, or computed tomography examinations, the results
of which were considered diagnostic of bilateral ADPKD.8 Diagnosis was corroborated further by positive family history for ADPKD, associated liver cysts, or arterial aneurysms. Patients were classified as hypertensive if they were receiving antihypertensive agents or if their BP regularly exceeded 140 mm Hg systolic and 90 mm Hg diastolic. Freshly voided urine specimens were obtained. The urine was tested by the dipstick method for semiquantitative determinations of urine protein and urine hemoglobin using dyeimpregnated paper strips (Labstix, Ames Division, Miles Lboratories, Elkhart, Ind). Fifteen milliliters of urine was sedimented in a standard clinical table-top centrifuge for five minutes at ambient temperature, after which the supernatant was decanted. The sediment was resuspended in one or two drops of urine. One drop of sediment was placed on a glass microscope slide with a square coverslip and examined under both low-power (x 100) and high-power (x 450) magnification using a polychromatic light source. The sediment was then examined with polarized light. DRLB were defined as large, dark yellow to dark brown, round to oval bodies with a granular appearance, which varied in size but often had diameters from three to four times greater than an erythrocyte, and which were observed to be doubly refractile in polarized light. The urine sediment exam was considered positive if two or more DRLB were seen under the coverslip. Doubly refractile lipid droplets (so-called free fat droplets), which were smaller than a single erythrocyte, perfectly round, had a clear transparent appearance, and were doubly refractile in polarized light were not equated with DRLB in this study_ In two patients urine specimens were examined after having been frozen for 9 years. Cellular elements were discernible and these patients are included in this study. Chart records from
From the Division of Nephrology, the Department of Medicine, and the Department of Pathology, Kidney and Urology Research Center, University of Kansas School of Medicine, Kansas City. Address reprint requests to Jared J. Grantham, MD, Director, Nephrology Division, University of Kansas School of Medicine, 39th and Rainbow Blvd, Kansas City, KS 66103. © 1985 by The National Kidney Foundation, Inc. 0272-6386/85/010049-05$3.00/0
American Journal of Kidney Diseases, Vol V, No 1, January 1985
49
50
DUNCAN, CUPPAGE, AND GRANTHAM
five patients who had careful urine sediment examinations were evaluated retrospectively. Cyst fluid from polycystic kidneys of six patients obtained during renal surgical procedures or at autopsy were evaluated by a method identical to that for urine examination. Cyst fluid specimens from four patients (19 cysts total) were examined after having been frozen for 6 to 9 years. In each case cellular detail was adequate to permit differentiation of DRLB from other formed elements. Selected urine sediments and cyst fluids were stained with the lipid-soluble stain oil red O. Fifteen-milliliter portions of urine or cyst fluid were centrifuged for five minutes and the supernatant was completely removed. The sediment was resuspended in a 3-to-5-fold dilution of oil red 0 stain (saturated solution in 60% isopropanol). The sediment was stained at room temperature for 20 minutes, centrifuged, washed in water, centrifuged again, resuspended in two drops of glycerol, and examined by standard transmitted light microscopy as before. Lipid-containing structures stained bright red using this technique. The creatinine level in the serum of the patients was measured using a colorimetric assay with alkaline picrate in a standard automated analytical technique. The two-tailed Student's t test for independent means was used to determine level of significance between groups. A P value of less than 0.05 was chosen as significant. Stepwise discriminant analysis was performed to select variables that would differentiate between the two groups of patients (positive versus negative lipiduria). These calculations were performed by computer under the direction of the Department of Biometry at the University of Kansas Medical Center.
RESULTS
The data from 35 patients included in this study are summarized in Table 1. Sixty percent of the patients had DRLB in their urine sediment (positive group), whereas the remaining 40% did not (negative group) (Fig 1). Table 1.
Clinical Features of ADPKD in Relation to Doubly Refractile Lipid Bodies
SP (percent elevated)
Positive DRLS*
Negative DRLS*
Patients (percent of total) 60.0 (21) 40.0 Males (percent of group) 61.9(13) 50.0 Age (yr) 48.1 ± 3.0 (21) 40.8 ± 3.5 BP (percent elevated) 61.9 (21) 64.2 Serum creatinine (mg/dL) 6.6 ± 1.1 (20) 3.6 ± 1.1 Creatinine clearance (mLlmin) 35 ± 10 (14) 57 ± 12 Urine hemoglobin (0-4 +) 0.36 ± 0.11 (14) 0.07 ± 0.07 Urine protein (0-4 +) 1.32 ± 0.20t (19) 0.39 ± 0.15
(14) (7) (14) (14) (13) (9) (7) (14)
'Numbers in parentheses represent the number of patients. tp < 0.001.
Fig 1. Examples of DRLB in urine (A) and in cyst fluid (B and C), with (A and B) and without (C) polarization (x 2,500).
LlPIDURIA IN PKD
The urine sediment of three patients who had DRLB in their urine was stained with oil red O. Examination of all three stained sediments revealed many bight red intracellular lipid droplets filling the doubly refractile bodies, thus confirming the lipid content of these DRLB. The presence or absence of DRLB in the urine was relatively consistent. Ten patients had sequential urine sediment exams over a period spanning several years. In four patients, DRLB were seen on several separate examinations, and in two patients urinary sediment was persistently negative for DRLB. After intervals of 4 and 8 years, two patients who initially had negative exams subsequently showed DRLB in their urine sediment. In two patients who initially had DRLB in their sediment, the elements were not detected on repeat examination 1 to 2 years later. In two families we examined one parent and two children all of whom had ADPKD. In each family, the parent and one child had DRLB in the urine while the remaining child did not. ' The positive and negative groups were not statistically different with respect to sex, age, hypertension, serum creatinine level, creatinine clearance, or hematuria. Although dipstick methods yield only semiquantitative estimates of urine protein concentration, 9 we include the average values and SEM obtained in this study. Urine protein concentration was higher in the positive group (1.32 ± 0.20, scale 0 to 4 + ), than in the negative group (0.39 ± 0.15, scale 0 to 4 + ). The data summarized in Table 1 were also subjected to stepwise discriminant analysis using age, sex, BP, serum creatinine level, and dipstick urine protein as variables for 18/21 positive patients and 13/14 negative patients. Age, sex, BP, and serum creatinine did not discriminate between patients who had DRLB in their urine (positive) and those who did not (negative). The urine dipstick protein determination was the most significant discriminator. Using a semiquantitative cut-off for urine protein between trace and 1 +, patients were accurately predicted to have DRLB in their urine 78 % of the time (a 22 % false-negative rate), and patients having less protein in their urine were accurately predicted to not have DRLB in their urine 85% of the time (a 15% false-positive rate). The mean prediction accuracy of urine protein concentration was 81 %. To search for the origin of urinary lipid bodies we examined sediments aspirated from the renal
51
Table 2.
Lipid Bodies in Cyst Fluid and Urine
Patient
No. of Cysts Examined
1
15
2 3 4
6 6 2
5 6
5
5 2 5 5
39
31
Total
5
No. of Cysts With DRLB
9 5
Urinary DRLB
Absent Absent Present Present Present Present Present 4/6
cysts of six patients with AD PKD. The appearance of fluid obtained from cysts in polycystic kidney disease may vary from clear and amber to a thick viscous, dark-brown fluid resembling melted cho~ colate (so-called chocolate cysts). We examined cyst fluid from four patients with DRLB in their urine and from two patients without DRLB in their urine (Fig 1 and Table 2). Clear cyst fluid had little or no sediment and microscopically showed few erythrocytes and small amounts of amorphous material. Sediment from brown or black cyst fluid contained numerous RBCs (usually crenated), rare leukocytes, and numerous large dark yellow to dark brown, round to oval bodies that were doubly refractile in polarized light. When cyst fluid was stained with oil red 0, these bodies stained red often demonstrating multiple round, red intracel~ lular droplets, confirming their lipid content. The DRLB in these cyst fluid samples were morphologically indistinguishable from the DRLB in the urine sediment of the patients in this study (the positive group). In the six patients, 80% of the cysts contained DRLB (31139 cysts) (Table 2). There was not a direct relation between the presence of DRLB in the cyst fluids and in the urine sediment. DRLB in cyst fluid appeared to increase in number in association with increased brown color and turbidity of the fluid, but we were unable to quantify this relationship. Cyst wall epithelium was stained for lipid content with oil red 0 and examined microscopically in two patients. In both cases, some cyst epithelial cells contained intracellular oil red O-positive lipid droplets. The lipid droplets and enlarged lipidladen cells observed resembled the DRLB found in the cyst fluid. We did not detect oil red O-positive cells in noncystic renal tubules within the same kidneys. DISCUSSION
From among numerous clinical studies of ADPKD and of lipiduria we were unable to find
52
any rigorous studies of DRLB in the urine of these patients.I.5-7.IO-19 DRLB have been reported in ADPKD anecdotally on two occasions. The more recent paper20 appears to refer to an earlier article,21 for which the original reference makes no mention of ADPKD in relation to urinary DRLB. Quinn and Zimmerman found cells in the urine sediment in one patient with polycystic disease that contained a few lipid droplets, but no "true" or classic DRLB were observed. 16 Thus, the current study appears to be the first comprehensive demonstration of DRLB in the urine and cyst fluid of patients with ADPKD. Sixty percent of the patients with ADPKD in the current study had DRLB in the urine. Lipiduria was not related to age, gender, hypertension, or serum creatinine level. However, a urine protein determination exceeding trace appeared to identify patients with DRLB in their urine with about 78 % accuracy. A possible connection between mild proteinuria and lipiduria was not revealed in this study. The DRLB observed in these patients may have originated from noncystic renal tubular epithelium, the epithelium lining the cysts, or other nonrenal sources. DRLB have been observed in the urine of patients with prostatitis, lipoid carcinoma of the prostate, and hypernephroma. 13-19.22 However, we found no evidence that any patient had hypernephroma, and the presence of DRLB in the female patients eliminates the prostate as a sole source of DRLB. DRLB in the urine occur most frequently in systemic disorders affecting the kidneys and in primary renal disorders. There is increased glomerular permeability, proteinuria, tubular epithelial lipid droplet accumulation with cell enlargement, and subsequent shedding of these tubular cells into the urine as DRLB.2326 Our patients had no clinical, physical, or chemical evidence of systemic or primary renal disease other than ADPKD. Hypertension, an acknowledged associate of proteinuria, was present in 63 % of the patients but was equally distributed between both positive and negative patient groups. A primary glomerular lesion has not been reported in ADPKD.1.57.10 Although the exact etiology of ADPKD has not been determined, it is clear that the cysts derive from renal tubules that eventually enlarge to a massive extent. 27 The cysts contain high concentrations of serum proteins. 28.29
DUNCAN, CUPPAGE, AND GRANTHAM
Conceivably, the glomeruli attached to the cystic tubules may leak serum proteins into the cysts, where they are converted into DRLB by the epithelial cells. A similar mechanism has been postulated in the nephrotic syndrome.24.25.30 In future studies the ultrastructural morphology of glomeruli in ADPKD should be critically evaluated for lesions known to be associated with proteinuria and hematuria. Blood may enter the cyst cavity directly due to internal rupture of a fragile vessel stretched across the cyst wall. This happens frequently in patients with ADPKD. Blood is a potential source of lipids and proteins that the cyst epithelium may then reabsorb and convert into lipid bodies. We found lipid bodies in the cells lining cyst cavities, suggesting that the lipid-laden cells aspirated in the cyst fluid may have derived from tubular epithelium. The DRLB gathered in the cysts may enter the urine in two ways. The cyst wall could rupture in response to trauma, spilling the cyst contents directly into the renal pelvis; alternatively, cyst fluid could pass directly into the renal pelvis through an efferent tubule segment. 31.32 In either case the lipid-laden epithelial cells could ultimately be seen as DRLB in the urinary sediment. Patients with ADPKD can develop concurrent glomerulonephritis that cuases hematuria and nephrotic syndrome,12.33 although the simultaneous occurrence of two renal diseases must be quite rare. As shown in the current report, DRLB in the urine of a patient with ADPKD does not necessarily indicate a second coexistent renal disorder. ACKNOWLEDGMENT We would like to express our gratitude to Professor Ruth Hassanein of the Department of Biometry at the University of Kansas Medical Center for assistance with the statistical analysis. Secretarial assistance was kindly provided by Janet Hartley.
REFERENCES I. Higgins CC: Bilateral polycystic kidney disease: Report of ninety-four cases. Arch Surg 65:318-329, 1952 2. Danovitch G: Clinical features and pathophysiology of polycystic kidney disease in man, in Gardner K Jr (ed): Cystic Diseases of the Kdney. New York, Wiley & Sons, 1976, pp 125-150 3. Barnes BA, Bergan 11, Braun WE, et al: The twelfth report of the Human Renal Transplant Registry. JAMA 233:787796, 1975 4. Lowrie EG, Hampers CL: The success of Medicare's
LlPIDURIA IN PKD end-stage renal disease program: The case for profits and the private marketplace. N Engl 1 Med 305:434-438, 1981 5. Braasch WF, Schacht FW: Pathological and clinical data concerning polycystic kidney. Surg Gynecol Obstet 57:467475, 1933 6. Rail IE, Odel HM: Congenital polycystic disease of the kidney: Review of the literature, and data on 207 cases. Am 1 Med Sci 218:399-407, 1949 7. Simon HB, Thompson Gl: Congenital renal polycystic disease. A clinical and therapeutic study of 366 cases. lAMA 159:657-662, 1955 8. Levine E, Grantham 11: The role of computed tomography in the evaluation of adult p~lycystic kidney disease. Am 1 Kidney Dis 1:99-105, 1981 9. Schrier R: Renal and Electrolyte Disorders (ed 2). Boston, Little, Brown, 1980, p 507 10. Braasch WF: Clinical data of polycystic kidney. Surg Oynecol Obstet 23:697-702, 1916 11. Mitcheson HD, Williams G, Castro IE: Clinical aspects of polycystic disease of the kidneys. Br Med 1 1:1196-1199, 1977 12. Ackerman GL: Nephrotic syndrome in polycystic renal disease. 1 Urol 105:7-9, 1971 13. Miloslavich EL: Occurrence of Iipo ids in urine and their diagnostic importance. 1 Lab Clin Med 13:542-546, 1928 14. Brice AT: The incidence of lipoids in urine: Being a report of the micropolariscopic examination of 1470 specimens. 1 Lab Clin Med 15:953-960, 1930 15. Rifkin H, Parker IG, Polin EB, et al: Diabetic glomerulosclerosis: Clinical and pathologic observations with special reference to doubly refractile fatty cells and casts in the urine. Medicine 27:429-457, 1948 16. Quinn lR, Zimmerman HG: Significance of oval fat bodies in urinary sediment. Am 1 Clin Pathol 24:787-795, 1954 17. Walz DV, lames DC: Significance of doubly refractile fat bodies in urinary sediment. Am J Clin Pathol 25:598-600, 1955 18. Neuman M , West M: The relationship between proteinu-
53 ria and fatty elements in the urine sediment. Am 1 Med Sci 241:617-624,1961 19. Comings DE: Anisotropic lipids and urinary cholesterol excretion. lAMA 183:126-131,1963 20. Schumann GB: Urine Sediment Examination. Baltimore , Williams & Wilkins, 1980, p 132. 21. Kurtzman NA, Rogers PW: A Handbook of Urinalysis and Urinary Sediment. Charles C. Thomas, Springfield, Ill, 1974, P 56 22. Meares E lr: Urinary tract infections in men, in Harrison IH, Gittes RF, Perlmutter AD, et al (eds): Campbell's Urology (ed 4), vol 1. Philadelphia, WB Saunders, 1978, pp 509-537 23. Leiter L: Nephrosis. Medicine 10: 135-242, 1931 24. Dunn IS: Nephrosis or nephritis? 1 Pathol Bacteriol 39:1-25, 1934 25. Allen AC: The clinicopathologic meaning of the nephrotic syndrome. Am 1 Med 18:277-314, 1955 26. Zimmer IG, Dewey R, Waterhouse C, et al: The origin and nature of anisotropic urinary lipids in the nephrotic syndrome. Ann Intern Med 54:205-214, 1961 27. Grantham 11: Polycystic kidney disease: A predominance of giant nephrons. Am 1 Physiol 244:F3-FIO, 1983 28. Huseman R, Grady A, Welling D, et al: Macropuncture study of polycystic disease in adult human kidneys. Kidney Int 18:375-385, 1980 29. Gardner KD: Composition of fluid in twelve cysts of a polycystic kidney. N Engl 1 Med 281:985-989, 1969 30. Smetana H, 10hnson FR: The origin of colloid and lipoid droplets in the epithelial cells of the renal tubules. Am 1 Pathol 18: 1029-1049, 1942 31. Lambert PP: Polycystic disease of the kidney: A review. Arch Pathol 44:34-58, 1947 32. Baert L: Hereditary polycystic kidney disease (adult form): A microdissection study of two cases at an early stage of the disease. Kidney Int 13:519-525, 1978 33. Licina MG, Adler S, Bruns Fl: Acute renal failure in a patient with polycystic kidney disease. lAMA 245: 1664-1665, 1981
A. UTOSOMAL dominant f t disease (ADPKD) is a
polycystic kidney relatively common disorder, occurring in approximately 1 in every 500 persons, 1.2 and is the third or fourth most frequent cause of end-stage renal disease. 34 Transmitted in an autosomal dominant fashion, ADPKD is characterized by cysts diffusely scattered throughout the parenchyma of both kidneys. These cysts, some of which may be present at birth, progressively enlarge, ultimately acquiring a size of less than a millimeter to several centimeters in diameter. Patients often remain asymptomatic until the fourth or fifth decades of life, when they may present clinically with massively enlarged kidneys, flank pain, hematuria, hypertension, or azotemia.1.5-7 Most patients develop progressive renal failure and require dialysis in the fifth or sixth decades of life. 1.5-7 We noted the presence of doubly refractile lipid bodies (DRLB, oval fat bodies) in the urine of several of our patients with ADPKD, a finding that had not previously been reported. To investigate further the incidence of DRLB, we did a combined prospective and retrospective study of urine sediments obtained from patients with ADPKD. We have also examined cyst fluid and histopathologic specimens in selected cases. We found DRLB in the urine and in the cyst fluids in a majority of the patients examined. MATERIALS AND METHODS Thirty-five patients attending the Polycystic Kidney Clinics at the University of Kansas Medical Center and the Kansas City Veterans Administration Medical Center were surveyed. All patients had previously had either excretory urography, ultrasonography, or computed tomography examinations, the results
of which were considered diagnostic of bilateral ADPKD.8 Diagnosis was corroborated further by positive family history for ADPKD, associated liver cysts, or arterial aneurysms. Patients were classified as hypertensive if they were receiving antihypertensive agents or if their BP regularly exceeded 140 mm Hg systolic and 90 mm Hg diastolic. Freshly voided urine specimens were obtained. The urine was tested by the dipstick method for semiquantitative determinations of urine protein and urine hemoglobin using dyeimpregnated paper strips (Labstix, Ames Division, Miles Lboratories, Elkhart, Ind). Fifteen milliliters of urine was sedimented in a standard clinical table-top centrifuge for five minutes at ambient temperature, after which the supernatant was decanted. The sediment was resuspended in one or two drops of urine. One drop of sediment was placed on a glass microscope slide with a square coverslip and examined under both low-power (x 100) and high-power (x 450) magnification using a polychromatic light source. The sediment was then examined with polarized light. DRLB were defined as large, dark yellow to dark brown, round to oval bodies with a granular appearance, which varied in size but often had diameters from three to four times greater than an erythrocyte, and which were observed to be doubly refractile in polarized light. The urine sediment exam was considered positive if two or more DRLB were seen under the coverslip. Doubly refractile lipid droplets (so-called free fat droplets), which were smaller than a single erythrocyte, perfectly round, had a clear transparent appearance, and were doubly refractile in polarized light were not equated with DRLB in this study_ In two patients urine specimens were examined after having been frozen for 9 years. Cellular elements were discernible and these patients are included in this study. Chart records from
From the Division of Nephrology, the Department of Medicine, and the Department of Pathology, Kidney and Urology Research Center, University of Kansas School of Medicine, Kansas City. Address reprint requests to Jared J. Grantham, MD, Director, Nephrology Division, University of Kansas School of Medicine, 39th and Rainbow Blvd, Kansas City, KS 66103. © 1985 by The National Kidney Foundation, Inc. 0272-6386/85/010049-05$3.00/0
American Journal of Kidney Diseases, Vol V, No 1, January 1985
49
50
DUNCAN, CUPPAGE, AND GRANTHAM
five patients who had careful urine sediment examinations were evaluated retrospectively. Cyst fluid from polycystic kidneys of six patients obtained during renal surgical procedures or at autopsy were evaluated by a method identical to that for urine examination. Cyst fluid specimens from four patients (19 cysts total) were examined after having been frozen for 6 to 9 years. In each case cellular detail was adequate to permit differentiation of DRLB from other formed elements. Selected urine sediments and cyst fluids were stained with the lipid-soluble stain oil red O. Fifteen-milliliter portions of urine or cyst fluid were centrifuged for five minutes and the supernatant was completely removed. The sediment was resuspended in a 3-to-5-fold dilution of oil red 0 stain (saturated solution in 60% isopropanol). The sediment was stained at room temperature for 20 minutes, centrifuged, washed in water, centrifuged again, resuspended in two drops of glycerol, and examined by standard transmitted light microscopy as before. Lipid-containing structures stained bright red using this technique. The creatinine level in the serum of the patients was measured using a colorimetric assay with alkaline picrate in a standard automated analytical technique. The two-tailed Student's t test for independent means was used to determine level of significance between groups. A P value of less than 0.05 was chosen as significant. Stepwise discriminant analysis was performed to select variables that would differentiate between the two groups of patients (positive versus negative lipiduria). These calculations were performed by computer under the direction of the Department of Biometry at the University of Kansas Medical Center.
RESULTS
The data from 35 patients included in this study are summarized in Table 1. Sixty percent of the patients had DRLB in their urine sediment (positive group), whereas the remaining 40% did not (negative group) (Fig 1). Table 1.
Clinical Features of ADPKD in Relation to Doubly Refractile Lipid Bodies
SP (percent elevated)
Positive DRLS*
Negative DRLS*
Patients (percent of total) 60.0 (21) 40.0 Males (percent of group) 61.9(13) 50.0 Age (yr) 48.1 ± 3.0 (21) 40.8 ± 3.5 BP (percent elevated) 61.9 (21) 64.2 Serum creatinine (mg/dL) 6.6 ± 1.1 (20) 3.6 ± 1.1 Creatinine clearance (mLlmin) 35 ± 10 (14) 57 ± 12 Urine hemoglobin (0-4 +) 0.36 ± 0.11 (14) 0.07 ± 0.07 Urine protein (0-4 +) 1.32 ± 0.20t (19) 0.39 ± 0.15
(14) (7) (14) (14) (13) (9) (7) (14)
'Numbers in parentheses represent the number of patients. tp < 0.001.
Fig 1. Examples of DRLB in urine (A) and in cyst fluid (B and C), with (A and B) and without (C) polarization (x 2,500).
LlPIDURIA IN PKD
The urine sediment of three patients who had DRLB in their urine was stained with oil red O. Examination of all three stained sediments revealed many bight red intracellular lipid droplets filling the doubly refractile bodies, thus confirming the lipid content of these DRLB. The presence or absence of DRLB in the urine was relatively consistent. Ten patients had sequential urine sediment exams over a period spanning several years. In four patients, DRLB were seen on several separate examinations, and in two patients urinary sediment was persistently negative for DRLB. After intervals of 4 and 8 years, two patients who initially had negative exams subsequently showed DRLB in their urine sediment. In two patients who initially had DRLB in their sediment, the elements were not detected on repeat examination 1 to 2 years later. In two families we examined one parent and two children all of whom had ADPKD. In each family, the parent and one child had DRLB in the urine while the remaining child did not. ' The positive and negative groups were not statistically different with respect to sex, age, hypertension, serum creatinine level, creatinine clearance, or hematuria. Although dipstick methods yield only semiquantitative estimates of urine protein concentration, 9 we include the average values and SEM obtained in this study. Urine protein concentration was higher in the positive group (1.32 ± 0.20, scale 0 to 4 + ), than in the negative group (0.39 ± 0.15, scale 0 to 4 + ). The data summarized in Table 1 were also subjected to stepwise discriminant analysis using age, sex, BP, serum creatinine level, and dipstick urine protein as variables for 18/21 positive patients and 13/14 negative patients. Age, sex, BP, and serum creatinine did not discriminate between patients who had DRLB in their urine (positive) and those who did not (negative). The urine dipstick protein determination was the most significant discriminator. Using a semiquantitative cut-off for urine protein between trace and 1 +, patients were accurately predicted to have DRLB in their urine 78 % of the time (a 22 % false-negative rate), and patients having less protein in their urine were accurately predicted to not have DRLB in their urine 85% of the time (a 15% false-positive rate). The mean prediction accuracy of urine protein concentration was 81 %. To search for the origin of urinary lipid bodies we examined sediments aspirated from the renal
51
Table 2.
Lipid Bodies in Cyst Fluid and Urine
Patient
No. of Cysts Examined
1
15
2 3 4
6 6 2
5 6
5
5 2 5 5
39
31
Total
5
No. of Cysts With DRLB
9 5
Urinary DRLB
Absent Absent Present Present Present Present Present 4/6
cysts of six patients with AD PKD. The appearance of fluid obtained from cysts in polycystic kidney disease may vary from clear and amber to a thick viscous, dark-brown fluid resembling melted cho~ colate (so-called chocolate cysts). We examined cyst fluid from four patients with DRLB in their urine and from two patients without DRLB in their urine (Fig 1 and Table 2). Clear cyst fluid had little or no sediment and microscopically showed few erythrocytes and small amounts of amorphous material. Sediment from brown or black cyst fluid contained numerous RBCs (usually crenated), rare leukocytes, and numerous large dark yellow to dark brown, round to oval bodies that were doubly refractile in polarized light. When cyst fluid was stained with oil red 0, these bodies stained red often demonstrating multiple round, red intracel~ lular droplets, confirming their lipid content. The DRLB in these cyst fluid samples were morphologically indistinguishable from the DRLB in the urine sediment of the patients in this study (the positive group). In the six patients, 80% of the cysts contained DRLB (31139 cysts) (Table 2). There was not a direct relation between the presence of DRLB in the cyst fluids and in the urine sediment. DRLB in cyst fluid appeared to increase in number in association with increased brown color and turbidity of the fluid, but we were unable to quantify this relationship. Cyst wall epithelium was stained for lipid content with oil red 0 and examined microscopically in two patients. In both cases, some cyst epithelial cells contained intracellular oil red O-positive lipid droplets. The lipid droplets and enlarged lipidladen cells observed resembled the DRLB found in the cyst fluid. We did not detect oil red O-positive cells in noncystic renal tubules within the same kidneys. DISCUSSION
From among numerous clinical studies of ADPKD and of lipiduria we were unable to find
52
any rigorous studies of DRLB in the urine of these patients.I.5-7.IO-19 DRLB have been reported in ADPKD anecdotally on two occasions. The more recent paper20 appears to refer to an earlier article,21 for which the original reference makes no mention of ADPKD in relation to urinary DRLB. Quinn and Zimmerman found cells in the urine sediment in one patient with polycystic disease that contained a few lipid droplets, but no "true" or classic DRLB were observed. 16 Thus, the current study appears to be the first comprehensive demonstration of DRLB in the urine and cyst fluid of patients with ADPKD. Sixty percent of the patients with ADPKD in the current study had DRLB in the urine. Lipiduria was not related to age, gender, hypertension, or serum creatinine level. However, a urine protein determination exceeding trace appeared to identify patients with DRLB in their urine with about 78 % accuracy. A possible connection between mild proteinuria and lipiduria was not revealed in this study. The DRLB observed in these patients may have originated from noncystic renal tubular epithelium, the epithelium lining the cysts, or other nonrenal sources. DRLB have been observed in the urine of patients with prostatitis, lipoid carcinoma of the prostate, and hypernephroma. 13-19.22 However, we found no evidence that any patient had hypernephroma, and the presence of DRLB in the female patients eliminates the prostate as a sole source of DRLB. DRLB in the urine occur most frequently in systemic disorders affecting the kidneys and in primary renal disorders. There is increased glomerular permeability, proteinuria, tubular epithelial lipid droplet accumulation with cell enlargement, and subsequent shedding of these tubular cells into the urine as DRLB.2326 Our patients had no clinical, physical, or chemical evidence of systemic or primary renal disease other than ADPKD. Hypertension, an acknowledged associate of proteinuria, was present in 63 % of the patients but was equally distributed between both positive and negative patient groups. A primary glomerular lesion has not been reported in ADPKD.1.57.10 Although the exact etiology of ADPKD has not been determined, it is clear that the cysts derive from renal tubules that eventually enlarge to a massive extent. 27 The cysts contain high concentrations of serum proteins. 28.29
DUNCAN, CUPPAGE, AND GRANTHAM
Conceivably, the glomeruli attached to the cystic tubules may leak serum proteins into the cysts, where they are converted into DRLB by the epithelial cells. A similar mechanism has been postulated in the nephrotic syndrome.24.25.30 In future studies the ultrastructural morphology of glomeruli in ADPKD should be critically evaluated for lesions known to be associated with proteinuria and hematuria. Blood may enter the cyst cavity directly due to internal rupture of a fragile vessel stretched across the cyst wall. This happens frequently in patients with ADPKD. Blood is a potential source of lipids and proteins that the cyst epithelium may then reabsorb and convert into lipid bodies. We found lipid bodies in the cells lining cyst cavities, suggesting that the lipid-laden cells aspirated in the cyst fluid may have derived from tubular epithelium. The DRLB gathered in the cysts may enter the urine in two ways. The cyst wall could rupture in response to trauma, spilling the cyst contents directly into the renal pelvis; alternatively, cyst fluid could pass directly into the renal pelvis through an efferent tubule segment. 31.32 In either case the lipid-laden epithelial cells could ultimately be seen as DRLB in the urinary sediment. Patients with ADPKD can develop concurrent glomerulonephritis that cuases hematuria and nephrotic syndrome,12.33 although the simultaneous occurrence of two renal diseases must be quite rare. As shown in the current report, DRLB in the urine of a patient with ADPKD does not necessarily indicate a second coexistent renal disorder. ACKNOWLEDGMENT We would like to express our gratitude to Professor Ruth Hassanein of the Department of Biometry at the University of Kansas Medical Center for assistance with the statistical analysis. Secretarial assistance was kindly provided by Janet Hartley.
REFERENCES I. Higgins CC: Bilateral polycystic kidney disease: Report of ninety-four cases. Arch Surg 65:318-329, 1952 2. Danovitch G: Clinical features and pathophysiology of polycystic kidney disease in man, in Gardner K Jr (ed): Cystic Diseases of the Kdney. New York, Wiley & Sons, 1976, pp 125-150 3. Barnes BA, Bergan 11, Braun WE, et al: The twelfth report of the Human Renal Transplant Registry. JAMA 233:787796, 1975 4. Lowrie EG, Hampers CL: The success of Medicare's
LlPIDURIA IN PKD end-stage renal disease program: The case for profits and the private marketplace. N Engl 1 Med 305:434-438, 1981 5. Braasch WF, Schacht FW: Pathological and clinical data concerning polycystic kidney. Surg Gynecol Obstet 57:467475, 1933 6. Rail IE, Odel HM: Congenital polycystic disease of the kidney: Review of the literature, and data on 207 cases. Am 1 Med Sci 218:399-407, 1949 7. Simon HB, Thompson Gl: Congenital renal polycystic disease. A clinical and therapeutic study of 366 cases. lAMA 159:657-662, 1955 8. Levine E, Grantham 11: The role of computed tomography in the evaluation of adult p~lycystic kidney disease. Am 1 Kidney Dis 1:99-105, 1981 9. Schrier R: Renal and Electrolyte Disorders (ed 2). Boston, Little, Brown, 1980, p 507 10. Braasch WF: Clinical data of polycystic kidney. Surg Oynecol Obstet 23:697-702, 1916 11. Mitcheson HD, Williams G, Castro IE: Clinical aspects of polycystic disease of the kidneys. Br Med 1 1:1196-1199, 1977 12. Ackerman GL: Nephrotic syndrome in polycystic renal disease. 1 Urol 105:7-9, 1971 13. Miloslavich EL: Occurrence of Iipo ids in urine and their diagnostic importance. 1 Lab Clin Med 13:542-546, 1928 14. Brice AT: The incidence of lipoids in urine: Being a report of the micropolariscopic examination of 1470 specimens. 1 Lab Clin Med 15:953-960, 1930 15. Rifkin H, Parker IG, Polin EB, et al: Diabetic glomerulosclerosis: Clinical and pathologic observations with special reference to doubly refractile fatty cells and casts in the urine. Medicine 27:429-457, 1948 16. Quinn lR, Zimmerman HG: Significance of oval fat bodies in urinary sediment. Am 1 Clin Pathol 24:787-795, 1954 17. Walz DV, lames DC: Significance of doubly refractile fat bodies in urinary sediment. Am J Clin Pathol 25:598-600, 1955 18. Neuman M , West M: The relationship between proteinu-
53 ria and fatty elements in the urine sediment. Am 1 Med Sci 241:617-624,1961 19. Comings DE: Anisotropic lipids and urinary cholesterol excretion. lAMA 183:126-131,1963 20. Schumann GB: Urine Sediment Examination. Baltimore , Williams & Wilkins, 1980, p 132. 21. Kurtzman NA, Rogers PW: A Handbook of Urinalysis and Urinary Sediment. Charles C. Thomas, Springfield, Ill, 1974, P 56 22. Meares E lr: Urinary tract infections in men, in Harrison IH, Gittes RF, Perlmutter AD, et al (eds): Campbell's Urology (ed 4), vol 1. Philadelphia, WB Saunders, 1978, pp 509-537 23. Leiter L: Nephrosis. Medicine 10: 135-242, 1931 24. Dunn IS: Nephrosis or nephritis? 1 Pathol Bacteriol 39:1-25, 1934 25. Allen AC: The clinicopathologic meaning of the nephrotic syndrome. Am 1 Med 18:277-314, 1955 26. Zimmer IG, Dewey R, Waterhouse C, et al: The origin and nature of anisotropic urinary lipids in the nephrotic syndrome. Ann Intern Med 54:205-214, 1961 27. Grantham 11: Polycystic kidney disease: A predominance of giant nephrons. Am 1 Physiol 244:F3-FIO, 1983 28. Huseman R, Grady A, Welling D, et al: Macropuncture study of polycystic disease in adult human kidneys. Kidney Int 18:375-385, 1980 29. Gardner KD: Composition of fluid in twelve cysts of a polycystic kidney. N Engl 1 Med 281:985-989, 1969 30. Smetana H, 10hnson FR: The origin of colloid and lipoid droplets in the epithelial cells of the renal tubules. Am 1 Pathol 18: 1029-1049, 1942 31. Lambert PP: Polycystic disease of the kidney: A review. Arch Pathol 44:34-58, 1947 32. Baert L: Hereditary polycystic kidney disease (adult form): A microdissection study of two cases at an early stage of the disease. Kidney Int 13:519-525, 1978 33. Licina MG, Adler S, Bruns Fl: Acute renal failure in a patient with polycystic kidney disease. lAMA 245: 1664-1665, 1981