A collaborative study to establish a U.S. reference for tissue plasminogen activator (t-PA)

Biologic&

(1991)

l&229-232

COLCABORATWE STUDY A CoMxwatlve Study to EstaMsh a U.S. lW#wwwe for Tissue Plasminogen Actiwator (t-PA) Deborah P. Beebe* and Laura L. Wood Blood Products Division, Center for Bioiogics Evaluation and Research, Bethesda, Maryland, U.S. A.

8800 Rockviile

Pike,

Abstract. In 1987 the Second International Standard for tissue plasminogen activator (t-PA) was established by the World Health Organization following an international collaborative study. At that time, the Center for Biologics Evaluation and Research (CBER) decided to establish a national reference t-PA to be used in lot release potency testing of Alteplase, a licensed t-PA biological or of other t-PAS in development. A candidate recombinant t-PA (rt-PA) preparation was donated by Genentech, Inc. (South San Francisco, California) for this purpose and a collaborative study was launched to calibrate this material against the 2nd I.S. Four laboratories (including the Center for Biologics Evaluation and Research (CBER) and three manufacturers) participated in the study to establish the potency of the rt-PA preparation using a clot lysis assay. The results indicate that the potency of the U.S. reference for t-PA is 2900 international units (IU) per vial. Introduction Tissue-type plasminogen activator (t-PA) is an endogenous activator of plasminogen.’ In 1987, Alteplase, a recombinant t-PA, manufactured by Genentech, was licensed in the U.S. for use in acute myocardial infarction.2 Standardization of t-PA has been conducted by the National Institute for Biological Standards and Control (NIBS0 in the U.K.3 The First International Standard for tissuetype plasminogen activator (t-PA) was established in 1984 and was derived from melanoma culture supernatants. The standard was intended for use in quantitating t-PA activity in plasma and biological fluids, as well as in production batches of t-PA products. For the Second International Standard, two candidate preparations were examined.6 One was purified from a cultured melanoma cell supernatant; the other was a recombinant t-PA preparation from Genentech, Inc. (South San Francisco, California). Both preparations compared favorably with the First International Standard for t-PA and both appeared to be equally stable to degradative conditions. However, on the basis of continuity of source material for replacing the first standard, the melanoma-derived material (coded 86/670) was chosen as the Second International Standard t-PA.5 * To whom correspondence should be addressed. 1045-1056/91/030229+04

$03.00/O

The Center for Biologics Evaluation and Research (CBER) routinely performs confirmatory assays as part of a lot release system. Although manufacturers may choose to calibrate internal working standards against the International Standard for each particular product, U.S. reference preparations are generally used for lot release testing both by the companies and by CBER. In 1984 the Center for Biologics (CBER) requested bids from several firms for the manufacture of material to be used as a U.S. t-PA reference for lot release potency assays and/or for research purposes. Genentech offered to donate recombinant t-PA produced in Chinese hamster ovary (CHO) cells and 3000 vials were sent to CBER in 1987. Testing for standardization of the material was begun after the Second International Standard for t-PA was established (1987). This report summarizes the results of a collaborative study involving four laboratories to establish the potency of the U.S. Reference for t-PA. Materials

and methods

t-PA preparation The reference preparation contains purified recombinant tissue plasminogen activator (&PA) which was manufactured by Genentech, Inc. (South San Francisco, California) and formulated with 5 0 1991 The International

Association

of Biological

Standardization

230

D. P. Beebe and L. L. Wood

mg/ml human albumin, 0.1 M arginine and 0.01% Tween 80. By comparison, routine production lots (50 or 20 mg vial sizes) of Alteplase, generic name for product &PA, are formulated without albumin. Alteplase has been shown to have an estimated specific activity of 500 000 IU per mg protein and > 60% single chain.6 For the present studies each vial of the proposed reference material was reconstituted with 1 ml of water and dilutions were made in assay buffer. Second International Standard

Samples of the Second International Standard for t-PA were generously donated by Dr Patrick Gaffney (National Institute for Biological Standards and Control, South Mimms, Herts, U.K.). The standard contains in addition to t-PA O-05 M phosphate buffer, pH 7.4 and 0.5% human serum albumin (free of peptidase activity).5 Each vial was reconstituted in 1 ml of assay buffer and assayed immediately. Clot lysis assay

The WHO recommended procedure for the clot lysis assay incorporated the following buffers and solutions: 0.06 M phosphate, pH 7.4, O-15 M sodium chloride, 0.05% bovine albumin, 0.01% Tween 80 and sodium azide as a preservative; 0.1 M borate buffer, pH 7.4; human thrombin, 30 III/ml dissolved in phosphate buffer; human Glu-plasminogen, 0.5 mg/ml dissolved in phosphate buffer; human fibrinogen, 3 mg/ml clottable protein dissolved in borate buffer. Plasminogen solution (200 ml) is mixed with 1 ml of thrombin solution and 120 ul of the resulting plasminogen/thrombin solution is added to each clean glass test tube (50 mm X 7 mm) and incubated in a water bath at 37°C. Aliquots of t-PA dilutions are added (6-50 ul), followed immediately by 1.0 ml of fibrinogen solution. A stopwatch is started at this point. Ninety seconds later, a 6 mm diameter glass bead is placed on top of the clot and the time to lysis of the clot (time at which bead falls to the bottom of the tube) is recorded visually. Other clot lysis assays which were used included a kinetic fibrinolytic assay7 in microtiter plates or in a microcentrifugal analyser.6 These methods use reagents that are the same as or similar to those used in the WHO recommended procedure and lysis is quantitated by turbidometric analyses. Study design

Twenty vials of the proposed reference t-PA were sent to each laboratory along with four ampoules of the Second International Standard for t-PA. The lab-

oratories are listed in the Acknowledgements. Each laboratory was asked to perform at least three independent assays with a fresh vial of both samples each time. Statistical analysis

Each laboratory submitted raw data which were entered into a parallel line bioassay statistical program relating the log of the lysis time to the log of the dilution. If the dose response curves were linear and parallel, potency estimates were derived from the horizontal distance between log dose response lines of the International Standard and the sample. A separate program to determine the arithmetic mean of all estimated potency determinations was used to calculate the overall value for the reference standard preparation. Results The values obtained from three of the laboratories (CBER, Wellcome Biotech and Knoll Pharmaceutical) which ran the proposed reference against the Second International Standard are shown in Table 1. The mean was 2849 III/ml (SD = 213 and SEM = 64) with 95% confidence limits of 2430-3267. The coefficient of variation was 7.5%. In addition, Genentech ran samples of the proposed reference 1. Potency values for the U.S. reference for t-PA assayed against the Second International Standard for t-PA

Table

Laboratory 1 1 1 2 2 2 3 3 3 3 3 Mean (confidence SD SEM

Potency value (IU Vial)

limits)

3125 3044 2849 2663 2708 2528 2869 3104 2559 2857 3034 2849 (2430-3267) 213 64

Method Plate Plate Plate Tube Tube Tube Tube Tube Tube Tube Tube

Test samples (rt-PA) were dissolved in 1 ml of water and the Second International Standard was dissolved in 1 ml of assay buffer. Dilutions of each preparation were made in assay buffer and tested in a clot lysis assay (plate or tube -..method, see Methods). Data were analysed with a parallel line assay.

231

U.S. reference for t-PA

Table 2. Potency values for the U.S. reference for t-PA assayed against Genentech internal standard IU vial Study 1

Study 2

3004 2991 3063 2897 2874 2968 2926

2773 2916 2938 2834 2917 2866 2893 2951

Confidence

Mean 2960 limits (2831-3090) SD 66 SEM 25

3008 2899 (2764-3034) 69 23

The U.S. Reference for t-PA was assayedby Genentech on two separate occasions against the manufacturer’s internal standard which had previously been calibrated against the First International Standard for t-PA. Samples were handled as described in Table 1 and assayedby the turbidometric assay.7

against their current working in-house standard which had been standardized against the First International Standard for t-PA. This was performed in two separate studies. In the first study, the mean value obtained for the proposed standard was 2960 (SD = 66 and SEM = 25) with 95% confidence limits of 2831-3090 (Table 2). In the second study, the proposed standard was 2899 (SD = 69 and SE = 23) with 95% confidence limits of 2764-3034. The consensus from these results indicates a potency of 2900 IU per vial. As shown by Genentech in previous work,6 the specific activity of Alteplase is estimated to be 500 000 IU per mg. When routine production lots of Alteplase are assayed against the U.S. reference using the assigned potency of 2900 IU/ml, all lots meet the potency specifications indicating no change in the specific activity (data on file at CBER and Genentech). Therefore, potency determinations of Alteplase using the Second International Standard, a Genentech internal standard for the U.S. reference remain consistent. Discussion The purpose of a U.S. reference for t-PA is for lot release potency assays of U.S. licensed t-PA products. Since Alteplase W-PA) is the only t-PA licensed in the U.S., a recombinant t-PA reference appeared

to be appropriate. Although the Second International Standard for t-PA is derived from melanoma cells, results from the collaborative study to determine the Second International Standard for t-PA indicated that both candidate preparations which included a melanoma-derived t-PA and a recombinant t-PA (Genentech, South San Francisco, California) were suitable to replace the First International Standard. From these results, there is no reason to believe that different standards are required for naturally-derived or recombinantderived t-PA. Thus, the U.S. reference, even though it is a recombinant protein, could be used to measure the potency of melanoma-derived or other naturally derived t-PA products as well as recombinant t-PA. The present study was actually performed twice, since variable results were obtained when the Second International Standard for t-PA was not diluted in assay buffer and when it was not assayed immediately. A recent publication8 states that, depending on the buffers and assay conditions, the potency of the Second International Standard varies with time. Once the parameters of buffer reconstitution and time were carefully controlled, there was excellent agreement between laboratories. Hence, the recommended way for using this standard is to resuspend the material in assay buffer and use it immediately. Stability studies on the U.S. reference t-PA were not incorporated in the design of this study. However, samples of the material have been periodically assayed for potency against an internal standard by Genentech. As shown in Table 3, the potency of the U.S. reference has remained stable for over 3 years. The potency of the material will continue to be monitored at intervals by both CBER and Genentech. Table

3. U.S. reference for t-PA stability

summary

Per cent activity Days

Remaining

0 16 602 1210

100.0 104.8 103.4 101.7

n 150 20 50 45

Samples of the U.S. Reference for t-PA were held by Genentech for various time periods (at -70°C) at which point they were assayed by the clot lysis assay. The precision of the clot lysis assay (Genentech) is +6% RSD. Therefore, all values were the samewithin experimental error. n refers to the number of vials assayed.

232

D. P. Beebe

The establishment of a U.S. reference for t-PA will facilitate lot release testing for licensed t-PA (Alteplase) and be useful for potency testing of t-PA products in research and development. Vials of the U.S. reference for t-PA are available upon written request to the Center for Biologics Evaluation and Research. Acknowledgements We would like to thank the following laboratories for participating in this study: Genentech Inc., 460 Point San Bruno Blvd, South San Francisco, California 94080, U.S.A. Wellcome Biotech, Langley Court, Beckenham, Kent BR3 3BS, U.K. and Knoll Pharmaceutical Co., 30 North Jefferson Road. Whippany, New Jersey 07981, U.S.A. References 1. Fears R. Biochemical pharmacology and therapeutic aspects of thrombolytic agents. Pharmacol Rev 1990; 42: 201-222.

and

L. L. Wood

2. Collen D, Stump DC, Gold HK. Thrombolytic therapy. Ann Rev Med 1988; 39: 405-423. 3. Campell PJ. International biological standards and reference preparations. I: preparation and presentation of materials to serve as standards and reference preparations. J Biol Stand 1974; 2: 249-258. 4. Gaffney PJ, Curtis AD. A collaborative study of a proposed international standard for tissue plasminogen activator (t-PA). Thromb Haemost 1985; 53: 134-136. 5. Gaffney PJ, Curtis AD. A collaborative study to establish the Second International Standard for tissue plasminogen activator (t-PA). Thromb Haemost 1987; 59: 1085-1087. 6. Carlson RH, Garnick RL, Jones AJS, Meunier AM. The determination of recombinant human tissue-type plasminogen activator activity by turbidimetry using a microcentrifugal analyzer. Anal Biochem 1988; 168: 428-435. 7. Beebe DP, Aronson DL. An automated fibrinolytic assay performed in microtiter plates. Thromb Res 1987; 47: 123-128. 8. Ranby M. On the properties of the international standards for tissue plasminogen activator. Thromb Haemostas 1990; 63: 139.

Received for publication 3 September accepted 10 April 1991

1990;