Urinary Excretion of Interferon, Albumin, and /?2-Microglobulin During Interferon Treatment1

[CANCER RESEARCH 44, 3599-3603, August 1984)

Urinary Excretion of Interferon, Albumin, and /?2-Microglobulin during Interferon Treatment1

Bauer E. Sumpio,2 Marc S. Ernstoff, and John M. Kirkwood

Departments of Surgery [B. E. S.J and Medicine [M. S. E., J. M. K.Â¡,Yale University School of Medicine, New Haven, Connecticut 06510

ABSTRACT The aim of this study was to attempt to characterize the Serum and urinary levels of albumin, ÃŸ2-microglobulin, and proteinuria which occurs during IFN treatment and to attempt to correlate these findings with the serum and urine levels of IFN in Interferon were determined in ten patients undergoing Interferon the patient group. therapy. The pharmacokinetics during a phase I trial of interferon administration intramuscularly is presented. Only trace amounts of interferon activity are found in the urine, even during peak MATERIALS AND METHODS serum interferon activity. Serum ft>-microglobulin levels in creased after interferon treatment, especially at the higher dosing Patient Selection. Ten patients with histologically verified metastatic levels. Urinary excretion of 02-microglobulin increased due to the carcinoma (6 with melanoma, 3 with colorectal adenocarcinoma, and 1 with osteogenic sarcoma) were enrolled in this study (Table 1). All relatively low affinity of the transport system. Saturation, com petition, or inhibition of the absorption process for 02-microglob- patients had measurable or Ã©valuablemetastatic cancer and a perform ance status <2 on the Eastern Cooperative Oncology Group scale, had ulin was not attained. Measurement of the urinary albumin/ been off of all therapy for 4 weeks or more than 6 weeks for nitrosourea urinary 02-microglobulin ratio reveals no glomerular or tubular and mitomycin C treatment prior to this study with no residual toxicity lesion, and we conclude that interferon therapy does not result from previous therapy, and had no serious secondary effects of their in a clinically significant nephrotoxicity. cancer (e.g., paraneoplastic, hormonal, or metabolic syndromes). Exclu sion criteria included pregnancy, central nervous system metastasis, previous Â¡nterferon exposure, and abnormal laboratory studies (WBC INTRODUCTION <4000/ml; granulocytes <1500/ml; platelets <100,uOO/ml; hemoglobin The IFNs3 are a group of glycoproteins first discovered be <10 g/dl; serum calcium >11 mg/dl; serum creatinine >2 mg/dl; and aspartate aminotransferase, alanine aminotransferase, or alkaline phos- cause of their potent antiviral activity (11). Subsequent obser phatase greater than twice the upper limits of normal). vations have suggested that IFN may, in addition, have significant IFN Treatment Protocol. Recombinant IFN a-2, prepared, purified, immunological, antiproliferative, and antitumor activity against and assayed as reported by Goeddel et al. (B), was supplied by Schering- virally and chemically induced tumors, as well as spontaneously Plough Corporation (Bloomfield, NJ). In brief, nucleic acids obtained from occurring tumors (2, 12). A range of toxicities was encountered virally challenged human lymphocytes were spliced into Escherichia coli during the initial clinical studies of IFN in humans, but the majority DNA by means of plasmid vectors. The product was assayed for IFN have been tolerable and virtually all reversible on withholding of activity and found to contain 1.0 to 2.4 x 108 units of antiviral activity the drug (14). Dose-limiting toxicity included leukopenia, eleva per mg of protein. Limulus assay testing for endotoxin (5) revealed less tion in transaminases, changes in central nervous system func than 5 ng of lipopolysaccharide/1.0 x 106 units. Cultures for bacteria tion, and fatigue. Recently, the appearance of protein in the urine and Mycoplasma were negative. of patients treated with increasing doses of recombinant IFN a- IFN Â«-2was administered i.m. at doses of 3, 30, 50, and 100 x 106 units per day for 28 days or to tolerance (Table 1). Patients were seen 2 has been found during a phase I trial (6). In addition, the for administration of each dose, and each patient had serum and timed administration of high doses of exogenous and endogenous IFN urine collections on the day prior to initiation of IFN therapy and on the to newborn mice has been demonstrated to cause a progres day after the end of their treatment protocol. Urine was checked for pH, sively fatal glomerulonephritis which can be ameliorated by anti- and an aliquot was stored at -70Â° with a simultaneous serum specimen. IFN globulin treatment (9, 10). These observations suggested Analysis. Serum and urine creatinine and albumin levels were mea that IFN administration, in relatively high doses over long periods sured in duplicate by the clinical laboratory, the latter by a nephalometric of time, might lead to a clinically important nephrotoxicity. technique. The serum samples were assayed for IFN by the Department i82-Microglobulin is a small protein with a molecular weight of of Microbiology, Schering Corporation. A radioimmunoassay procedure 11,700 and is normally found in low concentrations in serum and was used.The analysts did not have access to the random treatment in all biological fluids (7). In the kidneys, albumin and /32-micro- code until all assays were completed. The radioimmunoassay for the globulin are filtered at the glomerulus and subsequently reab determination of IFN levels in serum at 0 hr and following administration of IFN Â«-2was a modification of the Celltech Interferon Immunoradio- sorbed in the proximal tubules (19, 20). An increased urinary metric Assay. Following treatment of the serum to remove interfering excretion of /32-microglobulin associated with unchanged or only substances, appropriate dilutions of the sample are mixed with an 125I- slightly increased albumin excretion has been demonstrated to anti-a IFN monoclonal antibody and a polystyrene bead coated with a occur primarily in tubular lesions; on the other hand, increased sheep polyclonal anti-a IFN antibody. This mixture is then incubated at albumin excretion with unchanged 02-microglobulin excretion ambient temperature for 3 to 4 hr, followed by overnight storage at 4- suggests a glomerular lesion (16). 6Â°.During this period, IFN in the serum sample reacts with both antibod ies, thereby linking the labeled monoclonal antibody to the bead. The 1Supported in part by NIH Grants CA 09200 and CA 08341. 2To whom requests for reprints should be addressed, at Department of Surgery, amount of radioactivity bound to the bead is then determined. The level Yale University School of Medicine, 333 Cedar St., New Haven, CT 06510. of IFN is determined by comparing this amount of radioactivity to that 3The abbreviations used are: IFN, interferon; GFR, glomerular filtration rate. on a standard curve. This standard curve is determined using serum Received January 30,1984; accepted April 30,1984. samples spiked with known amounts of IFN a-2 treated and assayed at

AUGUST 1984 3599

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1984 American Association for Cancer Research. S. E. Sumpio et al. the same time as the study samples under the same conditions. Controls Serum fo-microglobulin levels rose after IFN treatment in all of of IFN tt-2 at low concentrations are treated in the same manner as the the patients, but only in 6 of these were the values above the samples and must have a standard deviation equal to or less than 10% reported upper limits of the normal population. for the assay to be considered valid. The level of IFN is expressed as ID/ Although there was no detectable rise in serum albumin after ml. The above procedure can detect IFN at a level as low as 5 ID/ml with IFN treatment, 3 of the 10 patients excreted significant amounts a precision of >90%. ,c?2-microglobulin was quantitated in duplicate by of albumin in the urine. One of the patients, G. T., however, radioimmunoassay using a commercial kit (Phadebas Pharmacia, Upps ala, Sweden). already had an increased urine albumin level prior to initiation of therapy. Serum creatinine levels remained unaffected by the IFN treatment. The urinary albumin/fe-microglobulin ratio is given in RESULTS Table 2 and is plotted in Chart 1. The levels of /tf2-microglobulin, albumin, and creatinine mea The clearance values for the proteins, calculated from the sured from the serum and urine of the patients in the study are product of the urine/plasma ratio of the protein and the urine presented in Table 2. The "pre"-treatment values reflect serum flow rate, is given in Table 3. GFR is determined from the and urine levels of the proteins obtained just prior to initiation of creatinine clearance. The fractional clearance of the proteins, IFN therapy. The "posf'-treatment values were from samples clearance/GFR, is also given in Table 3 and is a more accurate taken close to or at the end of the treatment protocol. Pretreat assessment of the renal handling of the proteins independent of ment serum and urine levels of 02-microglobulin, albumin, and the GFR. As seen in Table 3, the fractional clearance of 02- creatinine were within normal limits for all patients except in 3 microglobulin and albumin in these patients remained at very low instances. Patient J. M. had an elevated serum j32-microglobulin levels, despite IFN therapy. level, R. P. had a slightly increased urinary excretion of ÃŸ2- The relationship between the filtered load of ft>-microglobulin microglobulin, and G. T. had a slightly elevated albumin excretion and its urinary excretion rate and tubular absorption rate in these in the urine. IFN treatment led to an increased urinary excretion patients was assessed. Filtered load is a product of the glomer- of 02-microglobulin in 9 of the 10 patients. In 8 of these patients, ular filtration rate, plasma protein concentration, and glomerular the values were above the reported range of normal values. sieving coefficient of the particular protein. The glomerular siev ing coefficient is an index of the relative permeability of the Table 1 substance with respect to a freely permeable glomerular marker. Patient demographics and IFNdosage schedule The sieving coefficient for 02-microglobulin is 0.94 as measured 10*)DiagnosisColorectal (units x in the rat (18,20) and is consistent with the known permselective PatientR. properties of the glomerulus (3). Urinary excretion rate is deter B.J. adenocarcinomaMelanomaMelanomaMelanomaColorectalmined by the product of the urine concentration of the protein M.P.M.A. and the urine flow rate. The tubular absorption rate is calculated from the difference between the filtered load and the urinary E.E.G.S. adenocarcinomaMelanomaColorectal excretion rate. Chart 2 shows the relationship between filtered B.G.T.R. load and tubular absorption rate and urinary excretion rate for adenocarcinomaMelanomaMelanomaOsteogenic /32-microglobulin in these patients, plotting both "pre"- and "post"- P.G. C.W. IFN treatment values. The straight line is the line of identity. This P.SexMMMFFMMFFMAge(yr)54582351595964652443DosagesarcomaDaily333303030305050100Cumulative8184455104204205104506501100"renal titration" curve for /32-microglobulin obtained in this manner

Table 2 Serum levels and urinary excretion of fa-microglobulin and albumin Normalranges were: serum tfz-microgtobulin,<3.0 >Â¡g/mÃ(7);urineft>-microglobulin,<0.21 mg/24 hr (16);urinealbumin, <240mg/24hr(16). /32-microglobulinR.

creatinine O^g/ml)1.892.944.046.721.791.872.424.411.652.522.315.881.897.772.523.151.946.302.002.30Urine(mg/24hr)0.130.060.081.170.220.520.130.210.081.020.204.730.253.150.639.450.100.220.160.50AlbuminSerum(mg/ml)37.4033.6040.5041.5048.8039.9044.9035.6047.7046.1043.1044.9045.6241.0039.4034.3036.1033.6038.9042.90Urine(mg/24hr)23.0011.045.787.5113.8649.4212.3013.239.2431.8012.15472.54488.70375.3045.00252.0091.4836.1412.5615.48Serum(mg/dl)0.80.81.41.30.80.81.01.00.70.71.21.31.71.70.80.80.70.80.80.8UB'/UBÂ»Â«177184626639595631163161100195511971279151647931 B.J.

M.P.M.A.

E.E.G.S.

B.G.T.R.

P.G.

C.W.

P.PrePostPrePostPrePostPrePostPrePostPrePostPrePostPrePostPrePostPrePostSerum

"UALB,urineconcentration of albumin; UBJ^,urine concentration of frmicroglobulin.

3600 CANCER RESEARCH VOL. 44

"PROTEINURIA" Table 3 2000r Renal clearance of fa-microglobulin and albumin Clearance of the protein was calculated from the product of the urine/plasma ratio of the protein and the urine flow rate. GFR was determined from the clearance of creatinine. Fractional clearance of the protein, C/GFR, is a more accurate assessment of the renal clearance of the protein independent of changes in GFR. When this value equals 1, it implies free filtration through the glomerulus and no tubular absorption. When the fractional clearance approaches 0, it signifies either hindrance of filtration by the glomerular barrier or complete reabsorption of any filtered protein.

1500- x x10-*)3.94.62.01.71.03.33.85.01.74.72.889.8124.8132.04.658.621.616.32.11.9 10-4)4.02.03.015.05.08.07.06.04.024.082.068.016.058.010.0231.05.05.03.011.0CAUJ/GFR(units PatientR. (iJ/min)0.530.270.100.130.200.830.210.280.130.550.207.367.446.360.795.271.760.800.360.25GFR(ml/min)13459487919825155567811772826048171908249171130CB^/GFR(units B.J.

M.P.M.A.

E.E.G.S. 1000-

B.G.T.R.

P.G.C.W.

500 - P.PrePostPrePostPrePostPrePostPrePostPrePostPrePostPrePostPrePostPrePostCBZ-MGJ/min)4713141218919338323528159558932821742083372453149CALB

a GB?Â«,clearance of ft-microglobulin; CALB,clearance of albumin.

PRE POST Chart 1. Effect of IFN therapy on the urinary albumin/urinary ft,-microglobulin (/32M) ratios. Ranges of values of "glomerular," "tubular," or "normal" proteinuria

were obtained from Ref. 16. ABSORPTION RATE

is consistent with that obtained for other low-molecular-weight (/Â¿g/min) proteins (15, 18, 20). Even at the highest filtered loads of the protein, the transport process was not saturated (high capacity). Nevertheless, 02-microglobulin appeared in the urine, even at the lowest filtered loads, albeit in small amounts, because of the relatively low affinity of the tubular absorption process. Chart 3 illustrates the pharmacokinetics of IFN serum levels in 100 - the 4 different dosage groups. Mean serum IFN values are calculated from pharmacokinetic data collected on Days 1 and 22. Accumulation of IFN on Day 22 prior to that day's dose administration was seen, and it was dose-dependent. Peak serum levels occurred at 4 hr consistently in all dose groups, with little detectable IFN activity seen at 24 hr on Day 1. There FILTERED LOAD of B2M (/Â¿g/mm) Chart 2. Tubular absorption rate and urinary excretion rate of frmicroglobulin was no detectable IFN activity in the urine samples of all patients. (/32M)as a function of the filtered load. Filtered load (ordinate) is a product of the serum concentration of /32-microglobulin, the glomerular sieving coefficient (0.93), and GFR. Urinary excretion rate (O) is the product of the urine concentration of /32- DISCUSSION microglobulin and the urine flow rate. Tubular absorption rate (â€¢)was measured from the difference between the filtered load and the urinary excretion rate. The The results of the present study demonstrate that treatment solid line is the line of identity. with IFN leads to a uniformly increased appearance of /32-microglobulin in the urine. Although albumin excretion was increased and in proteinuria secondary to renal disease, Peterson ef al. in some patients, their values were low in comparison to those (16) have noted a ratio of urinary albumin to urinary ^-micro- seen in glomerular diseases and in the nephrotic syndrome (16). globulin of 1,100 to 14,200 in glomerular proteinuria, 33 to 163 In addition, from studies of the urinary excretion pattern of in normal proteinuria, and 1 to 13 in tubular proteinuria (16). albumin and ft-microglobulin, in patients with normal proteinuria From the data in Table 2 and Chart 1, there are 2 extreme values

AUGUST 1984 3601

models and in humans to have a transport maximum (20). The results of the present study do not rule out the possibility, **o- _.:Â».>-.~~-- however, that much higher plasma levels are required to produce 30 million units filtered loads capable of saturating the ft,-microglobulin transport system. The IFN has a molecular weight in the range of 20,000 and isoelectric points between 5.5 and 7.0 (17). It is thus similar in size and net charge to growth hormone and glucagon, which have been clearly shown to be easily filtered at the glomerulus and subsequently reabsorbed by the cells of the proximal tubule (15). Assays of urine of the patients in this study indicate that, even when high plasma concentrations of IFN are attained, very little IFN activity is detected in the urine. This suggests that IFN, like growth hormone or glucagon, is being degraded by the j 0124 8 24 kidney (1). It does not exclude the possibility that in vivo, in the TIME (hr) presence of urine, IFN may be inactivated (4). Chart 3. Pharmacokineticsof i.m. administered IFN a-2. In summary, the present study demonstrates that IFN treat ment may lead to an increased serum level of ft-microglobulin for the urinary ratio of albumin/ft-microglobulin. In the one in in some patients. This high serum level of 02-microglobulin results stance in which the ratio approached 2000 in a pretreatment value and was in the "range" for the glomerular proteinurias, in increased filtered loads of the protein and, because of the low affinity of the protein to its transport mechanism, it leads to Patient G. T. also had a significantly increased pretreatment urine spillage into the urine. No glomerular or tubular lesion needs to albumin value which was not worsened by IFN therapy. Similarly, be implicated to explain the ft>-microglobulinuria. In addition, the Patient J. M. displayed a low urinary albumin to /32-microglobulin absence of IFN activity in the urine, even at high serum IFN ratio, in the "range" of the tubular proteinuria, but he was also levels, confirms that the kidney is a major site for the degradation the only patient to have an abnormally increased pretreatment of the IFN. serum level of /32-microglobulin. If we exclude those 2 patients, estimation of the urinary ratios of albumin to ft-microglobulin in ACKNOWLEDGMENTS this study suggests the pattern of normal proteinuria. Further more, the very low values obtained on calculation of the fractional We would like to thank Carol A. Davis, R.N., M.S.N., for her nursing assistance clearance of /32-microglobulin (clearance of /32-microglobulin/GFR) and Lynn C. Qambardellafor preparingthis manuscript. (Table 3), support the hypothesis that tubular reabsorption of the protein is unimpaired during IFN treatment. This IFN-induced REFERENCES proteinuria is an example of the so-called "normal" proteinuria 1. Bino, T., Madar, Z., Gertler, A., and Rosenberg,H. The kidney is the main site which occurs in normal kidneys in the face of increased plasma of interferon degradation. J. Interferon Res., 2:301-308,1982. concentrations of a protein (15). In this scenario, low-molecular- 2. Borden, E. C. Interferons:rationalefor clinicaltrials in neoplastiadisease.Ann. weight plasma proteins appear in the urine either when the Intern. Med., 91: 472-479, 1979. filtered load exceeds the protein's tubular reabsorptive capacity 3. Brenner, B. M., Hostetter, T. H., and Humes, H. D. Glomerularpermsetectivity: barrier function based on discrimination of molecular size and charge. Am. J. or, even before that level is reached, because of the relatively Physiol.,234: F455-F460,1978. low-affinity property of the protein transport process. As can be 4. Cesano,T., and Tilles,J. G. Inactivationof human interferon by urine.J. Infect. Dis., J27: 311-314, 1973. seen from Chart 2, the latter is probably responsible for the ft>- 5. Coteman,D., Emstoff, M. S., Ryan,J., Eckstein, H., and Kirkwood, J. M. Effect microglobulinuria, since no saturation of tubular absorption of ft>- of intravenous a-2 interferon on monocyte Fc dependent phagocytosis in patients with malignant melanoma.Interferon Res., in press, 1984. microglobulin was attained. 6. Emstoff, M. S., Rudnick, S. A., and Kirkwood, J. M. Toxicity and antitumor The mechanism of the increased /32-microglobulin serum levels effects of daily recombinant human interferon a-2. Proc. Am. Federation of ClinicalResearch(Eastern Section),30. A694, 1982. is not entirely clear, but it seems to be causally related to higher 7. Evrin, P. E., and Wibell, L. The serum levels and urinary excretion of /32- IFN doses or longer intervals of IFN treatment. Recent reports microglobulinin apparently healthy subjects. Scand. J. Clin. Lab. Invest., 29: have suggested that the IFNs may markedly enhance the expres 69-74,1972. sion in vitro of histocompatibility antigens, such as 02-microglob- 8. Goeddel,D. V., Yelverton, E., Ullrich,A., Heyneker,H. L., Miozzari,G., Holmes, W., Seeburg, P. H., Dull, T., May, L., Stebbing, N., Crea, R., Maeda, S., ulin, on human lymphoid cells, as well as in cultured human McLandliss, R., Stoma, A., Tabor, J. M., Gross, M., Familtetti, P. C., and melanoma cells (13, 21). This effect is dose dependent and Pestka, S. Human leukocyte interferon produced by E. coli is biologically active. Nature (Lond.),287.:411-416, 1980. reversible upon removal of the IFN. Increased shedding of cell 9. Gresser, I., Maury, C., Tovey, M. G., Morel-Maroger, L., and Portillon, F. surface antigens such as ft.-microglobulin as a result of IFN Progressiveglomerulonephritisin mice treated with interferon preparations at treatment in vivo would lead to the accumulation in the plasma birth. Nature (Lond.),263. 420-422,1976. 10. Gresser I., Tovey M. G., Maury C., and Choroulinkow, I. Lethality of interferon and may explain the phenomenon outlined here. preparationsfor newborn mice. Nature (Lond.) 258: 76-78, 1975. The results illustrated in Chart 2 suggest that, even at plasma 11. Isaacs, A., and Lindenmann,J. Virus interference. I. The interferon. Pro. R. levels of 02-microglobulin that are twice the upper limits of normal, See. Ser. B., 147: 258-267,1957. 12. Kronenburg, L. H., Rosenblatt, H. M., Bryson, Y., and Merigan, T. C. In: E. R. leading to increased filtered loads of this protein, the tubular Stiehm (moderator), Interferon: immunotnologyand clinical significance. Ann. absorption process for ft-microglobulin is not saturated. This is Intern. Med. 96: 80-93, 1982. 13. Uao, S. K., Kwong, P. Câ€žKhosravi,M., and Dent, P. B. Enhancedexpression in contrast to other low-molecular-weight proteins such as lyso- of melanoma-associated antigens and .vmicroglobulm on cultured human zyme and cytochrome c, which have been shown in animal melanomacells by interferon. J. Nati. Cancer Inst., 68:19-25,1982.

3602 CANCER RESEARCH VOL. 44

14. Louie, A. C., Gallagher, J. G., Sikora, K., Levy, R., Rosenberg, S. A., and 18. Sumpio, B. E. Tubular absorption and catabolism of low-molecular-weight Merigan, T. C. Follow-up observations on the effect of human leukocyte proteins. Ph.D. Dissertation. New York: Cornell University Medical College, interferon in non-Hodgkin's lymphoma. Blood, 58: 712-718,1981. 1981. 15. Maack, T., Johnson, V., Kau, S. T., Figueiredo, J., and Sigutem, D. Renal 19. Sumpio, B. E., and Maack, T. Effect of iodoaoetate and lysine on tubular filtration, transport, and metabolism of low-molecular-weight proteins: a review. absorption of /32-microglobulin and cytochrome C in the isolated perfused rat Kidney Int., 76:251-270, 1979. kidney. Proc. Am. See. Nephrol. 13:152A, 1980. 16. Peterson, P. A., Evrin, P. E., and Berggard, I. Differentiation of gtomerular, 20. Sumpio, B. E., and Maack, T. Kinetics, competition, and selectivity of tubular tubular, and normal proteinuria: determinations of urinary excretion of /3j- absorption of proteins. Am. J. Physiol., 243: F379-F392,1982. microglobulin, albumin, and total protein. J. Clin. Invest., 48:1189-1198,1969. 21. Vignaux, F., andGresser, I. Enhanced expression of histocompatibility antigens 17. Priestman, T. J. Interferon: an anticancer agent? Cancer Treat. Rev., 6: 223- on interferon-treated mouse embryonic fibroblasts. Proc. Soc. Exp. Biol Med 237,1979. 757:456-460,1978.

AUGUST 1984 3603

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1984 American Association for Cancer Research. Urinary Excretion of Interferon, Albumin, and β2-Microglobulin during Interferon Treatment

Bauer E. Sumpio, Marc S. Ernstoff and John M. Kirkwood

Cancer Res 1984;44:3599-3603.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/44/8/3599

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/44/8/3599. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.