Br Heart J 1992;68:167-70 167 Monitoring of streptokinase resistance titre in acute myocardial infarction patients up to 30 months after giving streptokinase or anistreplase and related studies to measure specific antistreptokinase IgG

R Fears, H Ferres, E Glasgow, R Standring, K J Hogg, J D Gemmill, J M A Burns, A P Rae, F G Dunn, W S Hillis

Abstract Varying concentrations of antibodies to strep- Objective-To examine the induction tokinase (SK) may be present in patients with of antistreptokinase antibodies after acute myocardial infarction (AMI) as a result giving streptokinase or anistreplase to of previous streptococcal infections."2 The patients with acute myocardial infarc- induction of anti-streptokinase (anti-SK) tion. antibodies after the administration of SK or Design-Patients were randomly anistreplase (Eminase, APSAC) is, therefore, allocated to receive either 15 x 10' IU, an anamnestic response. Previous studies on streptokinase or 30U anistreplase in a SK have measured the antibody response with double blind study. Blood samples were a surrogate endpoint, the functional SK resis- collected immediately before treatment tance titre, which also reflects the influence of and subsequently at intervals up to 30 inhibitors such as a2 antiplasmin and varia- months; plasma samples were assayed tions in other plasma proteins such as fibrin- for streptokinase resistance titre (func- ogen and plasminogen."35 These studies have tional assay) and streptokinase binding shown that a peak value was reached by two by IgG (microradioimmunoassay). weeks after dosing with SK and returned Setting-Cardiology department in a towards pretreatment values by six months. general hospital. The development of neutralising antibodies Patients-128 consecutive eligible may influence the efficacy and risk of an patients. Samples were collected for up adverse reaction to SK or anistreplase on to one year according to a prospective readministration for a subsequent myocardial design: a subsection of 47 patients was infarction, and in the context of the increasing selected for intensive study over the first use of thrombolytic therapy this might 14 days. After one year, all available become an increasingly relevant clinical patients (67) were sampled on one fur- problem. It is therefore important to define the ther occasion. times of induction and decline in a population Results-Antibody responses to strep- large enough to reflect the range of individual tokinase and anistreplase were similar. variations in neutralising antibody concentra- Streptokinase resistance titres exceeded tion. It was the purpose of the present work to pretreatment concentrations five days compare changes in SK resistance titre in after dosing, and values peaked at 14 response to the standard doses of SK and 12 months of SmithKline Beecham days. By after dosing, 92% anistreplase in patients with AMI. The results Pharmaceuticals, resistance titres (n = 84) had returned to from this functional assay were compared Great Burgh, Epsom, within the pretreatment range. Anti- with those from a novel method of assaying Surrey streptokinase concentrations also one class in terms of R Fears IgG antibody (IgG) antigen H Ferres exceeded baseline concentrations within binding.6 E Glasgow five days and peaked at 14 days. Half of R Standring the individual values had returned to University within the pretreatment range by 12 Department of months (n = 84) and 89% 30 months Medicine and by Patients and methods = Therapeutics, Stobhill (n 18). PATIENTS General Hospital, Conclusion-Although we cannot be One Glasgow sure of the clinical significance, because hundred and twenty eight consecutive K J Hogg of the increased likelihood of resistance eligible patients (101 men, 27 women age range J D Gemmill 31-70) were recruited between April 1987 and J M A Burns due to antistreptokinase antibody, strep- 1989. Inclusion and A P Rae tokinase and anistreplase may not be January exclusion criteria F G Dunn if more five and adjunct treatments have been described W S Hillis effective administered than previously.7 Patients were randomised in days after an earlier dose of streptokin- Correspondence to: double blind, double dummy fashion to receive Dr R Fears, SmithKline ase or anistreplase, particularly between either 1-5 x 106IU SK infused Beecham Pharmaceuticals, five days and 12 months, and increased intravenously Great Burgh, Yew Tree over one hour or 30 U anistreplase injected Bottom Road, Epsom, antistreptokinase antibody may increase intravenously overfiveminutes. Initial patency, Surrey KT18 5XQ the risk of allergic-type reactions. Accepted for publication rates of reocclusion, and adverse events are 13 January 1992 (Br Heart J 1992;68: 167-70) described elsewhere.7 168 Fears, Ferres, Glasgow, Standring, Hogg, Gemmill, Burns, Rae, Dunn, Hillis

Evaluation ofvariations in streptokinase resistance titre after dosing with streptokinase or anistreplase

Period offollow up (months with range) Analysis 3 6 12 (11-15) 18 (16-20) 24 (21-27) 30 (28-31) Median evaluation as fold increase over 20 (39)t 15 (38) 4 (60) 5 (22) 2 (20) 2 (18) individual pretreatment value (n) Values (%) returning to: Within pretreatment range 51 72 92 100 100 100 Within 95th percentile pretreatment range 31 44 76 86 77 88 For all patients (n) (45)t (50) (84) (22) (26) (18) Theoretical neutralisation of streptokinase (IU) (combininggroupmedian)* 1-2 x 106 12 x 106 6 x 105 3 x 105 3 x 105 3 x 105 *Theoretical neutralisation by pretreatment median value = 1 5 x IO' (assuming a standard plasma volume of 31). tNumbers of patients in the two types of analysis differ because median evaluation could only be calculated for those patients for whom pre and post treatment pairs of samples were available. At any one sampling time these pairs of samples are fewer than the total number of patients sampled.

Blood samples were collected from patients are normally distributed, and the conclusions before dosing and serially at intervals for up to are similar to those derived from a considera- one year in a prospectively designed study. A tion of median values. The SK resistance and subsection of 47 patients was selected for antiSK IgG results were analysed by the intensive study over the first 14 days although Mann-Whitney U test. not all samples were received from each patient. Figures 1 and 3 show data from this cohort. The remaining figures and tables show Results available data for all patients. After one year all SK RESISTANCE TITRE available patients (67) were sampled on one Early serial changes (one to 14 days) were further occasion and these patients were studied in detail in a subgroup of 47 patients. grouped according to the classification de- The response was similar whether patients scribed in the table. were treated with SK or anistreplase, and baseline (pretreatment) values were consis- LABORATORY ASSESSMENTS tently exceeded by five days after dosing (fig 1). The SK resistance titre and antiSK IgG con- Figure 2 shows results for all patients, centrations were measured in individual monitored up to 30 months. Measurements up plasma samples as described previously.6 Cal- to 12 months are serial analyses in the same culations of the theoretical neutralisation of a patients (although a maximum ofonly 84 out of second dose of SK were performed by multi- a possible 115 samples were collected in the plying the SK resistance titre by the average prospectively designed study). By contrast, plasma volume of 3 litres. each patient was only sampled once over the period 12-30 months; that is, different patients STATISTICAL ANALYSIS appear in the groups collated at 18, 24, and 30 The SK resistance titres and anti-SK IgG months and therefore groups are smaller at concentrations, as analysed by either the Wilk- these times. Sampling times for the points Shapiro or the Kolmogorov-Smirnov test, within the period 12-30 months represent were not normally distributed and so cannot be ranges: the total ranges are specified in the table. represented as arithmetic means. Values for Changes in SK resistance titre were similar SK resistance titres and for comparison IgG for SK and anistreplase throughout the period are shown as median values with ranges. The of follow up (fig 2). Baseline (pretreatment) IgG values have also been analysed as values however were significantly higher in the geometric means as the log transformed values group of patients assigned to treatment with anistreplase (50 v 20 IU/ml, p < 0 05). Peak titres were reached two weeks after Figure 1 Initial changes in SK resistance titre. 'a Values are shown as E median and range for en -~10000 patients who received SK Q (open bars) or anistreplase m (stippled bars). Patient numbers are indicated (n). 0 CL

Das 0 :, (D

Oh 7 14 1 2 3 6 12 18 24 30 1n in - n v-cv cn o a coo am c a) r- ONc Days Months C4 _-_N- -C C C_d4- 4 N _ _ Oh 1 2 3 4 5 6 7 14 Sampling time Days Figure 2 Changes in SK resistance titre up to 30 Sampling time months. See legend tofig 1 for details. Monitoring ofstreptokinase resistance titre in acute myocardial infarction patients 169

Figure 3 Initial changes in anti-SK IgG. See early fall in IgG (fig 3) is not accompanied by a legend tofig 1 for details. E similar decline in SK resistance titre (fig 1). n Figure 4 showed results for all patients Q monitored to a up 30 months: changes in anti- SK IgG were similar for SK and anistreplase 01 throughout the period of follow up. It is ._ noteworthy that baseline values are not higher in the patients assigned to anistreplase, unlike cn the situation with SK resistance titre (fig 2). -0 01 In agreement with the changes in SK resis- cn tance titre, peak values of anti-SK IgG were attained 14 days after dosing (150 fold rise in median). The subsequent decline in anti-SK CD IgG, however, seemed somewhat slower than the decline in SK resistance titre: 48% of IMII IihulIIMInIMIut Isia I Liii I iii loc') Mc"V ON4 ONC, CDC4 N'- (00 r.-co individual anti-SK IgG values returned to n 1-N- N NNN3N4NN_ _ within the normal range by 12 months after Oh 1 2 3 4 5 6 7 14 dosing (14% at six months, 61% at 18 months, Days 65% at 24 months, and 89% at 30 months). Sampling time Discussion dosing (40 fold rise in median). Concentrations We have shown already that the range of declined considerably by three to six months concentrations of anti-SK antibodies found in (fig 2), although the median rise above AMI populations before first treatment- individual pretreatment concentrations was whether measured specifically as IgG or as the still noticeable for those patients in whom a functional SK resistance titre-does not complete set of data was collected (table). usually influence the thrombolytic efficacy of When examining all the data, it was seen that SK and anistreplase or the incidence of side most individual values had returned to within effects.28 The concentration of antibody is the normal (pretreatment) range by six months, usually low by comparison with the concentra- most values returned to within the 95th percen- tion required to bind all SK or anistreplase tile ofpretreatment range by 12 months (table). used,29 0 and the inhibition of lysis is much less When the values for combined median SK than the extent of binding.29 resistance titre (as described in the methods) We have now shown that high values of were used to calculate the theoretical neutral- neutralising antibody are rapidly induced after isation of a second dose of SK (assuming a dosing with anistreplase and SK and the two standard plasma volume of 3 1) it was found agents seem to be indistinguishable. When that most of a standard dose (1-5 x 106 IU) measured in terms of the functional SK resis- might be neutralised up to six months but not tance titre, the present results confirm and thereafter. extend previous findings, showing that most values return to within the normal range by six ANTISK IGG to 12 months. Extrapolating from group SK The early antiSK IgG response was also resistance titres to calculate how much of a similar whether patients were treated with SK second dose of SK might be neutralised, or anistreplase (fig 3). There was an initial fall in opinions on the possibility of readministration antiSK IgG in agreement with previous have varied from three months to greater than results2 because of sequestration as the six months.' 34 5 Our results suggest that much antibody antigen complex but, thereafter, ofa second standard dose (1 5 x 106 IU) of SK values were restored to baseline by day 5. This might be neutralised up to six months after the first dose but not at 12 months. Our assay for specific anti-SK IgG is novel Figure 4 Changes in and these are the first results to define the rate of anti-SK IgG up to 30 E induction and decline after dosing with SK or months after dosing. See co anistreplase. The incremental change from legend tofig I for details. C. co baseline to two months appears similar to the 0 preliminary results on SK published recently C for an enzyme linked immunosorbent assay (ELISA) method." The initial production of 1 IgG, however, appeared to be more rapid when measured by ELISA than by radioimmuno- assay in the present or a previous study.2 At the very high concentrations reached in the first few ._ months, anti-SK IgG is a major contributor to C the functional SK resistance titre. The fall in IgG during the first four days after treatment is < n=(' NN NN NC) eNC')oXa , Oh 7 14 1 2 3 6 not accompanied, however, by a reduction in 12 18 24 30 SK resistance titre presumably because the Days Months fibrinolytic assay reflects the changes in fibrino- Sampling time gen, plasminogen, and ox2-antiplasmin. The 170 Fears, Ferres, Glasgow, Standring, Hogg, Gemmill, Burns, Rae, Dunn, Hillis

significantly higher baseline SK resistance titre centrations may increase the risk of allergic- in anistreplase treated patients and the poor type reactions. correlation between changes in anti-SK IgG and SK resistance from three months onwards We thank Drs E S Johnson, R Cregeen, and Mr S R Hasler for helpful discussions, Nadia James for typing the manuscript, and may also reflect contributions other than anti- Julie Hearn for data entry. SK IgG, possibly a2 antiplasmin. Anti-SK IgG antibodies have occasionally been implicated in adverse reactions after 1 Fletcher AP, Alkjaersig N, Sherry S. The maintenance of a sustained thrombolytic state in man. I. Induction and thrombolytic therapy with SK." In view of the effects. J Clin Invest 1959;39:1096-1 10. very low incidence of severe allergic-type reac- 2 Hoffmann JJML, Fears R, Bonnier JJRM, Standring R, Ferres H, De Swart JBRM. Significance of antibodies to tions in the large mortality trials of anistre- streptokinase in coronary thrombolytic therapy with plase'2 and SK,1314 however, it is not clear streptokinase or APSAC. Fibrinolysis 1988;2:203-10. 3 Kostering H, Barth U, Naidu R. Changes of anti-strepto- whether an appreciable risk is incurred when kinase-titre following long term streptokinase therapy. In: IgG is higher. The concentrations of anti-SK Martin M, ed. New Concepts in Streptokinase Dosimetry. Huber, USA, 1978:110-5. IgG in the few patients in our study who 4 Massel D, Turpie AGG, Gill JB, Cairns JA, Russett J. showed an allergic-like response7 were not Development of neutralizing antibodies after 1-5 million units ofstreptokinase in the treatment ofacute myocardial appreciably different from the concentrations in infarction. Circulation 1989;80(suppl II):350. patients without a response. 5 Jalihal S, Morris GK. Antistreptokinase titres after intravenous streptokinase. Lancet 1990;335:184-5. There are few publications on the size of the 6 Moran DM, Standring R, Lavender EA, Harris GS. Assess- increment in clinically diagnosed allergic-type ment of anti-streptokinase antibody levels in human sera using a microradioimmunoassay procedure. Thromb reactions in response to other antigens when Haemost 1985;52:281-7. IgG is raised. Work on dextrans indicates that 7 Hogg KJ, Gemill JD, Burns JMA, Lifson WK, Rae AP, Dunn FG, Hillis WS. Angiographic patency study of most reactions seen when the IgG antibody anistreplase versus streptokinase in acute myocardial titre increases beyond a range of 10 fold from infarction. Lancet 1990;335:254-8. 8 Gemmill JD, Hogg KJ, Burns JMA, Standring R, Green- the normal median are likely to be mild (grade wood H, Fears R, et al. Pre-treatment streptokinase (SK) I); more severe reactions require a much higher resistance titres and anti-SK IgG levels are not related to hypotensive responses and coronary patency with SK- antibody concentration.'5 Furthermore, containing thrombolytic agents. Br Heart J 1990;64:50-1. although serious allergic type reactions are 9 Fears R, Ferres H, Hibbs M, Standring R. Consequences of antibody binding in vitro on the pharmacological proper- mediated by the complement pathway, com- ties of anisoylated plasminogen streptokinase activator plement is also lysed by plasmin in response to complex. Drugs 1987;33(suppl 3):64-8. 10 Been M, de Bono DP, Muir AL, Boulton FE, Fears R, plasminogen activators such as alteplase.'6 Standring R, Ferres H. Clinical effects and kinetic proper- Therefore, complement activation and any ties of intravenous APSAC-anisoylated plasminogen- streptokinase activator complex (BRL 26921) in acute contingent leucocyte attraction may occur myocardial infarction. Int J Cardiol 1986;11:53-61. regardless of the concentrations of anti-SK 11 Davies KA, Mathieson P, Winearls CG, Rees AJ, Walport MJ. Serum sickness and acute renal failure after strepto- IgG. kinase therapy for myocardial infarction. Clin Exp In conclusion, we have defined the induction Immunol 1990;80:83-8. 12 AIMS Trial Study Group. Long-term effects ofintravenous of neutralising antibodies to SK and anis- anistreplase in acute myocardial infarction: final report of treplase in AMI patients with the standard the AIMS study. Lancet 1990;335:427-31. 3 4 13 GISSI. Effectiveness ofintravenous thrombolytic treatment functional assay of SK resistance titre.' 'The in acute myocardial infarction. Lancet 1986;1:397-402. responses to anistreplase and SK are similar. 14 ISIS-2 Collaborative Group. Randomised trial of intraven- ous streptokinase, oral aspirin, both, or neither among Because of the increased likelihood of resis- 17187 cases of suspected acute myocardial infarction: tance due to anti-SK antibody, SK, and anis- ISIS-2. Lancet 1988;2:349-60. 15 Richter AW, Hedin HI. Dextran hypersensitivity. Immunol treplase may not be effective if administered Today 1982;3:132-8. more than five days after earlier SK or anis- 16 Bennett WR, Yawn DH, Migliore PJ, Young JB, Pratt CM, Raizner AE, et al. Activation ofthe complement system by treplase particularly between five days and 12 recombinant tissue plasminogen activator. J Am Coll months, and increased anti-SK antibody con- Cardiol 1987;10:627-32.