The Optimum Urine Collections for the Detection and Monitoring of Bence Jones Proteinuria

Transcription

1 Vol. 9 No. 5 BRIEF SCIENTIFIC REPORTS 689 time. Comparison of ex vivo measurement and in vitro standardization. Am J Clin Pathol 1985;84: Shapiro GA, Saundra W, Huntzinger MT, Wilson JE. Variation among commercial activated partial thromboplastin time reagents in response to heparin. Am J Clin Pathol 1977;67: Shojania AM, Tetreault J, Turnbull G. The variations between heparin sensitivity of different lots of activated partial thromboplastin time reagent produced by the same manufacturer. Am J Clin Pathol 1988;89: Simon TL, Smith KJ. Heparin assays. In: RW Colman. Disorders of thrombin formation. New York: Churchill Livingstone, 198: Soloway HB, Cornett BM, Grayson JW. Comparison of various activated partial thromboplastin reagents in the laboratory control of heparin therapy. Am J Clin Pathol 197: Twenty patients with malignant disease and Bence Jones (BJ) proteins were studied to determine the optimum urine collections for the detection and monitoring of light chain proteinuria. A 24-hour urine protein collection was followed by individual collections of each sequentially voided specimen over the same time interval. Samples were analyzed quantitatively for protein, and protein electrophoresis was performed on each specimen. Only one patient had BJ protein nondetectable by protein electrophoresis in the early morning specimen. Six patients had one or more random specimens (excluding the early morning specimens) absent for BJ protein on protein electrophoresis. Three patients had nondetectable protein on electrophoresis of the 24-hour specimen despite having some random specimens positive. All random specimens with protein values exceeding.2 g/l had BJ protein visibly detectable on electrophoresis. Thirteen specimens with protein less than.5 g/l still had BJ protein detected by electrophoresis. There was a linear relationship between the early morning protein concentration and the total 24-hour urinary protein production. The authors conclude that early morning specimens or 24-hour urine collections are preferable for the detection and monitoring of light chain proteinuria. These collection methods are not mutually exclusive because there are individual patients who will be negative in one collection but positive in the other. (Key words: Bence Jones protein; Optimum urine collection; Protein electrophoresis) Am J Clin Pathol 199;9: Received May 1, 1989; received revised manuscript and accepted for publication October 19, Address reprint requests to Dr. Brigden: Island Medical Laboratories, Yates Street, Victoria, British Columbia, V8V N1 Canada. 21. Triplett DA. Heparin: Clinical use and laboratory monitoring. In: DA Triplett, ed. Laboratory evaluation of coagulation. Chicago: American Society of Clinical Pathologists Press, 1982: Van den Besselaar AMPH, Meeuwisse-Braun J, Jansen-Griiter R, Bertina RM. Monitoring heparin therapy by the activated partial thromboplastin time. The effect of pre-analytical conditions. Thromb Haemost 1987;57: MM Wintrobe, GR Lee, DR Boggs, TC Bithell, J Foerster, JW Athens, JN Lukens, eds. Wintrobe clinical hematology. 8th ed. Philadelphia: Lee & Febiger, 1981: Zucker S, Cathey MH, Wylie RL. Control of heparin therapy sensitivity of the activated partial thromboplastin time for monitoring the antithrombotic effects of heparin. J Lab Clin Med 1969;7: The Optimum Urine Collections for the Detection and Monitoring of Bence Jones Proteinuria MALCOLM L. BRIGDEN, M.D., EVELYN D. NEAL, B.A., RT, MICHAEL D. D. MCNEELY, M.D., AND GORDON N. HOAG, M.D., PH.D. Island Medical Laboratories, Victoria, British Columbia, Canada DIFFERENCES OF OPINION exist regarding the optimum urine collection for detecting Bence Jones (BJ) proteinuria. Hobbs 6 has implied that random urine collections are satisfactory for paraprotein detection. Alternatively, Kyle 7 is of the opinion that random specimens, however collected, are misleading compared with the precision provided by 24-hour urine collections. This is a matter of clinical importance because detection of BJ proteinuria is used both diagnostically and as an objective index of disease activity. Therapeutic decisions regarding the institution or termination of chemotherapy often are based on the serial quantitation of urinary paraprotein providing renal function is stable. Stable renal function is important because it has been documented that the urinary clearance of light chains is directly proportional to the creatinine clearance. 7 A literature survey revealed that a systematic study comparing the value of random urines versus 24-hour urine collections for the detection and monitoring of light chain excretion had never been performed. Controversy also exists regarding the optimum methodology for assessing the need for further studies in patients with confirmed or suspected plasmaproliferative

2 69 BRIGDEN ET AL. A.J.C.P.-May 199 disorders. Albumin sticks that detect albuminuria will not detect urinary light chain proteins. 26 The heat test described by Putnam has a threshold for detection of 1.45 g/l and misses one-third of cases found by electrophoresis of concentrated urine. 1 Wet chemistry methods are the assays most frequently used to quantitate urine protein but have not been accepted as the best methods for paraprotein detection. Sulfosalicylic acid (SSA) (Exton's reagent) is able to detect urine immunoglobulin light chains' 2 but is relatively insensitive. 6 It also has been found that certain types of BJ protein have delayed reactions to SSA, resulting in false negative test results." The limits for the detection of proteins by the SSA methods are reported to range from.2 to.1 g/l. 15 In this study serial and 24-hour urine collections were compared to determine the optimum urine collection for the detection and monitoring of BJ protein. The relationship of the urinary protein concentration determined by SSA to the presence or absence of urinary light chains on electrophoresis also was investigated. Materials and Methods Patients were preselected based on a known diagnosis of light chain proteinuria associated with a plasmaproliferative or lymphoproliferative disorder as well as the ability to participate in a study requiring multiple collections. Because all patients had been previously diagnosed, the paraprotein and light chain type was not reconfirmed. Most patients had multiple myeloma, Waldenstrom's macroglobulinemia, or another plasmaproliferative disorder. One patient had a poorly differentiated lymphocytic lymphoma. The study included 2 individuals, 6 females, aged 66-9 years, with a mean age of 75 years, and 14 males, aged 49-8 years, with a mean age of 68 years. The patients' ages, sexes, diagnoses, and paraprotein and light chain types are listed in Table 1. Each patient collected a 24-hour urine sample following a standard protocol, and each serial voided urine was collected for a second 24-hour interval in separate 5 ml containers. Each serial collection was performed within two days of the initial 24-hour collection. The serial specimen, provided between 6 and 9 hours, was designated as the early morning specimen. All other serial specimens were classified as random specimens, and these data were analyzed separately from the early morning specimens. Urine specimens were identified by patient name and time of collection. All samples were returned to the laboratory. Total volume, time interval of collection, and, in some cases, specific gravity were recorded. Specimens were either analyzed immediately or kept frozen at -2 C until analysis. Samples were analyzed for urinary protein by the SSA method. 9 Protein quantitation used a blank and a test for each sample. The blank solution was made by diluting one part urine with five parts isotonic sodium chloride (NaCl). The test solution was comprised of one part urine and five parts % (weight/volume) SSA. These remained at room temperature for 1 minutes before being read in an LKB spectrophotometer at 42 nanometers (nm) (LKB, Uppsalla, Sweden). Zero absorbance was set for each specimen using the sample blank. Absorbance of Table 1. Patient Characteristics by Age, Diagnosis, and Type of Paraprotein and Presence in Serum and/or Urine Patient G.G. A.D. WD. M.D. F.T. ED. E.B. M.S. N.T. R.C. L.H. J.R. B.B. J.D. H.S. P.S. W.S. S.A. W.B. J.B. Age (years) Diagnosis Poorly differentiated lymphocytic lymphoma Plasmacytoid lymphoma Waldenstrom's macroglobulinemia Smoldering myeloma Serum Paraprotein Type IgG IgG IgG IgG IgG IgG IgA IgM IgG IgA IgA IgG Urine Paraprotein Type

3 Vol. 9 No. 5 BRIEF SCIENTIFIC REPORTS 691 test against blank was read. The urine protein was calculated by using a calibration curve established with known concentrations of protein (albumin). The sensitivity of our SSA method for protein in urine is.5 g/l. Samples were concentrated using Amicon Minicon-B 15 concentrators (Amicon Canada Ltd., Oakville, Ontario, Canada) before protein electrophoresis. All specimens were concentrated X1 regardless of the total urine protein value. I,i2 Urine protein electrophoresis was performed using the Beckman Paragon electrophoresis system (Beckman Instruments Inc., Fullerton, CA). After electrophoresis of the concentrated urine specimen, the strips were stained with Paragon Blue (Beckman Instruments Inc., Fullerton, CA). The strips were visually examined for the presence of BJ protein by a single observer (EN). Because all patients were known producers of light chains, immunoelectrophoresis was not followed by immunofixation. The sensitivity of this method allows BJ protein to be detected confidently at concentrations of.1 g/l of the original urine. Results The mean number of urine specimens submitted in the serial 24-hour collections was 9, with a range of The 24-hour urine protein values of patients varied from <.5 g/l to 2.1 g/l, with a median value of.4 g/l. The urine protein values for early morning specimens varied from <.5 g/l to 2.4 g/l, with a median of. g/l. The urine protein values for the random specimens, excluding the early morning specimens, ranged from <.5 g/l to.25 g/l, with a median value of.4 g/l. The marked variation in urine protein values noted was largely due to differences in protein excretion between individuals as the protein excretion in samples obtained from the same individual was relatively constant. Overall, the variation in protein concentration of serial samples from individual patients represented less than 15% of the total variation observed between patients. With regard to electrophoresis results, 19 patients had early morning specimens positive for BJ protein, whereas 14 patients had all random specimens positive (excluding early morning specimens) and patients had positive 24-hour specimens. Table 2 documents the various combinations of electrophoresis results. Of the patients whose 24-hour urine was positive on protein electrophoresis for BJ protein, had positive early morning specimens, none had negative early morning specimens, 14 had all random specimens positive, and three had one or more random specimens negative. For the three patients whose 24-hour urine was negative on electrophoresis, two had positive early morning specimens, one had negative early morning specimens, none had all random specimens Table 2. Results of the Electrophoresis of the 24-Hour, Early Morning, and Random Urine Specimens for the Presence or Absence of BJ Protein 24-hour specimens Early Morning Specimens 2 1 Random Specimens All 14 Some positive, and three had one or more random urines negative. Each of the three patients who had one or more random specimens nondetectable on electrophoresis, despite having positive 24-hour collection results, still had a demonstrable band on electrophoresis of the early morning specimen. Figure 1 shows the results of all patients' specimens (random, early morning, and 24-hour), comparing the degree of protein excretion to the presence or absence of BJ protein observed on the electrophoretic strip. Because the sensitivity for protein detection with our SSA method is.5 g/l, cases with nondetectable protein were plotted as protein <.5 g/l. Thirteen specimens had BJ protein present on electrophoresis despite having no protein detected by the SSA method. The boundary between positive and negative specimens on electrophoresis for the majority of the random urine samples was.2 g/l of protein. A substantial variation in the number of specimens makes it difficult to interpret the significance of differences between the other sample types. Figure 2 compares the early morning specimen protein content to the total 24-hour urinary output of protein. A linear relationship was documented for the entire patient population. Discussion In plasmaproliferative disorders the detection of BJ proteinuria is important to establish or confirm the diagnosis. It also is useful for following the response of these patients to a variety of therapeutic interventions.,4,7 Recently it has been appreciated that small amounts of BJ proteinuria (<.2 g/24 hours) are associated with a variety of other B-cell neoplasms such as chronic lymphocytic leukemia and hairy cell leukemia. 14 The ability to accurately detect small amounts of BJ protein may become increasingly important in the monitoring of these conditions. Disagreement exists in the literature as to the optimum method for the detection of urinary light chain proteins in patients.,4,6, Twenty-four hour urine collections are

4 692 BRIGDEN ET AL. A.J.C.P.-May 199 LUES (g/l) OTEIN VA \ S "~ a. >ALICYLIC SULPHOS : i ;.* v... 4aa* 4 4 ^jlimiiisiisb -. Random Specimens Early a.m. Specimens 24 h Specimens PROTEIN ELECTROPHORESIS RESULTS FIG. 1. The urine protein concentration (g/l) for the various specimens (random, early morning, 24-hour) is compared to the presence () or absence (A) of BJ protein on electrophoresis. A solid line indicates the cut-off protein concentration (.2 g/l) for the majority of random specimens positive or negative on electrophoresis. inconvenient for patients, less reliable, and more expensive for the laboratory to supervise. Furthermore, any degradation of proteins due to bacteria, temperature, or time can result in reduced resolution of the protein electrophoretic strip. On the other hand, random urine collections are easy to obtain, allow rapid transfer to the laboratory, and theoretically avoid potential degradation of protein. If it could be shown that the excretion of BJ protein was relatively constant, random collections might be reliably used to estimate tumor burden. However, the synthesis and release of BJ protein may be variable throughout the day. This could cause random urine specimens to be negative at some times and positive at others for the presence of BJ proteinuria. Although the results of this study demonstrated that the protein concentration of samples obtained from the same individual were relatively constant, three patients did have one or more random (excluding early morning) j (A I "S +> o u OI C *C D S ' 1H W 1. < specimens with nondetectable bands on protein electrophoresis despite having positive 24-hour collection results. As all three of these patients had a positive early morning specimen on protein electrophoresis, the data clearly support the early morning specimen as the preferred random sample. To our knowledge the early morning specimen had not been considered as the most appropriate urine collection although it has been used. These data do not support the contention of Hobbs 6 that any random specimen is satisfactory. Analysis of the 24-hour collection results reveals that three patients had nondetectable band on electrophoresis despite having one or more random samples or the early morning specimen positive. Thus, our data do not support the opinion of Kyle 7 and others that 24-hour urines are the optimal preferred specimen. A 24-hour urine collection protocol could miss some patients that will be picked up by a random collection sequence, especially the early morning specimen. Taken overall, the 24-hour, random, and early morning results confirmed the inadequacy of any single urine specimen collection to consistently detect light chains in every instance and probably explain some of the discrepancies reported in the literature. Comparison of thefirstmorning specimen paraprotein content to total 24-hour urinary output of protein showed a linear relationship. Thus, in monitoring patients with known urinary paraprotein excretion, providing potentially confounding factors such as state of hydration and renal function remain stable, either an early morning or 24-hour urine paraprotein collection should provide a re h Urine Protein Values (g/l) FIG. 2. The early morning urine protein concentration (g/l) is compared to the 24-hour urine protein concentration (g/l).

5 Vol. 9 No. 5 BRIEF SCIENTIFIC REPORTS 69 liable assessment. Conceivably, the monitoring of patients who have minimal proteinuria should include both specimens because neither collection method alone assures absolutely reliable detection under all circumstances. Concentration of urine specimens before analysis is routine for the detection of BJ protein. Urine protein electrophoresis, urine immunoelectrophoresis, and immunofixation of light chains all are performed on concentrated specimens. 8 Various authors have advocated a variety of different concentration schedules, and this undoubtedly is a factor in the different sensitivities published for the detection for BJ proteinuria Initially, Perry and Kyle 5 recommended a 1-fold concentration although a more recent recommendation was simply "adequately concentrated." 7 In some laboratories, up to - fold concentration is routinely performed. In this study each specimen was concentrate XI. The sensitivity of this technique in our hands allows BJ protein to be detected at a concentration of.1 g/l of the original urine. It is conceivable that, if urines had been more highly concentrated (X-6), some of the specimens positive for protein but negative on electrophoresis might, in fact, have shown BJ protein. 14 However, the quantitative fluctuations noted in this investigation between the individual serial samples and the 24-hour urine collections still would have been present regardless of the degree of urine concentration. If institutions select a single urine protein concentration as a cutoff value below which they will not proceed to urine concentration and urine electrophoresis, they will miss cases of BJ proteinuria. In this study, patients with random urine specimens and protein concentrations of >.2 g/l invariably had light chains present on electrophoresis. However, 1 specimens had protein present on electrophoresis despite having negative protein detected by the SSA method. While this probably reflects the known sensitivity problem of the SSA method for immunoglobulins, the possibility of delayed SSA reactivity with light chains also exists." Other authors also have documented that urine protein estimation alone does not provide a completely reliable point to ascertain whether a band will be detected. When light chain proteinuria is strongly suspected, the only satisfactory technique is to follow urine concentration by protein electrophoresis and immunofixation. With these techniques minute bands often can be identified even when urinary protein concentration is virtually nondetectable. 7 ' 8112 While immunofixation is more sensitive than electrophoresis, the detection of BJ protein by immunofixation in the absence of visible bands on electrophoresis is of doubtful clinical significance. In summary, this study demonstrates that both a 24- hour urine collection and early morning specimen are preferred for the detection and monitoring of BJ proteinuria. When a urinary paraprotein is suspected, either an early morning specimen or a 24-hour urine specimen may be collected, but both should be performed if either is initially negative. If a total urine protein of.2 g/l determined by the SSA method is taken as a threshold for detecting paraprotein excretion in specimens, paraproteins will be consistently detected. However, specimens with total protein of <.5 g/l may still be associated with paraproteins on urine protein electrophoresis. Acknowledgement. The authors gratefully acknowledge the assistance of the staff and physicians of the Victoria Cancer Clinic, specifically Mary Gaunt, Dianne Ford, and Dr. Ken Wilson. The support of Drs. D. Cuthbert and G. Cantlie also is acknowledged. References 1. Balant LP, Fabre J. Clinical significance of proteinuria. Comprehensive Therapy 1978;4(1): Bowie L, Smith S, Gochman N. Characteristics of binding between reagent strip indicators and urinary proteins. Clin Chem 1977;2: Chronic Leukemia- Task Force, National Cancer Institute. Proposed guidelines for protocol studies II. Plasma cell myeloma. Cancer Chemother Rep 197;4: Gandara DR, MacKenzie MR. Differential diagnosis of monoclonal gammopathy. Med Clin North Am 1988;72: Gyure WL. Comparison of several methods for semiquantitative determination of urinary protein. Clin Chem 1977;2: Hobbs JR. Bence Jones proteins. Essays in Medical Biochemistry 1975;1: Kyle RA. Diagnosis and management of multiple myeloma and related disorders. Progress in Hematology 1986;14: Merlini G, Piro P, Pavesi F, Epis R, Aguzzi F. Detection and identification of monoclonal components: Immunoelectrophoresis on agarose gel and immunofixation on cellulose acetate compared. Clin Chem 1981;27: Meulemans O. Determination of total protein in spinal fluid with sulphosalicylic acid and trichloroacetic acid. Clin Chim Acta 196;5: Nawab RA, Azar HA. The laboratory diagnosis of plasma cell myeloma and related disorders. Orthop Clin North Am 1979; 1: Nahman NS, Murnane MR, Kuehl WD, Hebert LA, Nefl'JC, Bay WH. Light-chain proteinuria: Spurious false-negative reaction to sulfosalicylic acid. Ann Intern Med 1985;12: Oken MM. Multiple myeloma. Med Clin North Amer 1984;68: Perry MC, Kyle RA. The clinical significance of Bence Jones proteinuria. Mayo Clin Proc 1975;5: Pezzoli A, Pascali E. The clinical significance of pure Bence Jones proteinuria at low concentration. Am J Clin Pathol 1989;91: Sun T, Lein YY, Gross S. Clinical application of a high resolution electrophoresis system. Ann Clin Lab Sci 1978;8: Whicher JT, Calvin J, Riches P, Warren C. The laboratory investigation of paraproteinaemia. Ann Clin Biochem 1987;24: Whicher JT. The role of immunoglobulin assays in clinical medicine. Ann Clin Biochem 1984;21: