Immunological Standardization
The first international standard for serum amyloid A protein (SAA). Evaluation in an international collaborative study

https://doi.org/10.1016/S0022-1759(98)00057-XGet rights and content

Abstract

An ampouled preparation of acute phase serum rich in serum amyloid A protein (SAA) was evaluated in seven laboratories in six countries for its suitability to serve as the international standard for immunoassay of SAA. A variety of different immunoassays were used. On the basis of the results reported here and with the authorization of the Expert Committee on Biological Standardization of the World Health Organization (WHO) this preparation (coded 92/680) was established as the first international standard of SAA.

Introduction

SAA is an apolipoprotein of high density lipoprotein (HDL) particles and is the polymorphic product of a set of genes located on the short arm of chromosome 11. SAA is highly conserved in evolution and is a major acute phase reactant in all species in which it has been studied. Most of the SAA in plasma is produced by hepatocytes in which the synthesis is under transcriptional regulation by cytokines, especially IL-1, IL-6 and TNF, acting via NFκB, CEβP and possibly other transcription factors (Edbrooke et al., 1991; Betts et al., 1993). After secretion it rapidly associates with HDL from which it displaces apoAI. The circulating concentration can rise from normal levels of about 3 mg/l to over 1000 mg/l within 24–48 h of an acute stimulus, whilst with ongoing chronic inflammation the level may remain persistently high (de Beer et al., 1982; Pepys and Baltz, 1983; Wilkins et al., 1994). Certain isoforms of SAA, the products of different genes, are predominantly synthesized by macrophages, adipocytes and other cells. Although these isoforms also associate with HDL, their acute phase synthesis is stimulated differently and they presumably have different functions (Meek et al., 1992). There is also a closely related family of trace apoproteins of HDL, the so-called constitutive SAAs, which are not acute phase reactants (Whitehead et al., 1992).

In most circumstances the serum concentration of SAA correlates with that of the classical, and also highly sensitive, acute phase reactant, C-reactive protein (CRP) (Pepys and Baltz, 1983), but SAA values reach higher levels and may respond more rapidly. The cytokine regulation of SAA production is also different from that of CRP, and it has been suggested that there may be clinically important differential acute phase responses of SAA and CRP to different stimuli (Maury, 1985; Hachulla et al., 1991; Smith et al., 1992; Miwata et al., 1993; Nakayama et al., 1993). Furthermore, although more sensitive procedures are now becoming available, most of the routine commercial assays for CRP only measure reliably levels above 5 mg/l, whereas 50% of healthy subjects have CRP concentrations in the range 0.07–0.8 mg/l, and 90% have less than 3 mg/l (Shine et al., 1981). There may thus be increases of up to 100 fold in CRP concentration before this analyte is detected, let alone recorded as abnormal (Wasunna et al., 1990). Immunoassays for SAA, covering the whole range from normal to peak acute phase levels, are therefore useful in clinical monitoring of the acute phase response.

In addition, circulating SAA is the precursor of amyloid A protein (AA), which forms the fibrils deposited in the tissues in reactive amyloidosis complicating chronic inflammatory and infective conditions (Pepys, 1994). Amyloidosis is a serious, and usually fatal condition for which no specific therapy exists. Sustained increased production with a high circulating concentration of SAA is a necessary, though not sufficient, condition for development of AA amyloidosis. In patients with established AA amyloid, persistently high values of SAA are associated with progression of amyloid deposition whilst regression can occur when SAA production is reduced to normal by spontaneous or therapeutically induced remission of the primary disease (Falck et al., 1983; Hawkins et al., 1993). The capacity to monitor serum SAA levels frequently and precisely can thus make a significant contribution to the management of patients with AA amyloid.

Effective management of patients receiving renal allografts requires early recognition and adequate treatment of rejection episodes. Apart from direct monitoring of renal function, a variety of putative markers of rejection have been evaluated in serum and urine, but none are both highly specific and highly sensitive. Nevertheless there is substantial evidence that the serum concentrations of CRP (White et al., 1981; Freed et al., 1984; Groth et al., 1984; Lempert et al., 1985; Van Lente et al., 1986; Kaden et al., 1987; Lempert et al., 1987; Bruzzone et al., 1987; Dyck, 1988; Cohen et al., 1988; Lalla et al., 1988; Fischbach et al., 1994) and SAA (Maury, 1985; Maury et al., 1983b, Maury et al., 1983a; Maury and Teppo, 1984; Maury et al., 1985; Hocke et al., 1989; Müller et al., 1992; Casl et al., 1995), can provide useful information in this context. These observations were first made with CRP before the widespread introduction of cyclosporin. Subsequently it has become clear that standard immunosuppressive therapy with cyclosporin and prednisolone markedly suppresses the acute phase response of this protein to transplantation surgery and to acute rejection (Cohen et al., 1988, Cohen et al., 1986), although not the response to intercurrent infection. In contrast the SAA response to transplant surgery still occurs during cyclosporin and steroid treatment, and the responses to rejection and infection are apparently unimpaired (Maury, 1985; Maury et al., 1983b, Maury et al., 1983a; Maury and Teppo, 1984; Maury et al., 1985; Hocke et al., 1989; Müller et al., 1992; Casl et al., 1995).

The acute phase response is a non-specific phenomenon induced by almost all forms of tissue damage, inflammation and infection, and although both CRP and SAA are exquisitely sensitive and rapid reactants covering extremely wide dynamic ranges, their serum levels can never be diagnostic on their own (Pepys and Baltz, 1983; Pepys, 1995). They must be interpreted in the context of the full clinical picture and can then provide invaluable information for both diagnosis and management. In a recent study a new, rapid, automated immunoassay (Wilkins et al., 1994) was used to confirm and extend earlier observations of the sensitive and frequently predictive SAA response to episodes of renal allograft rejection (Hartmann et al., 1997). The contrast between specific failure of the CRP response to rejection in these cyclosporin treated patients and preservation of the response of both proteins to infection enhances the value of frequent routine measurement of both SAA and CRP.

In order to facilitate development and promote the availability of properly standardised robust routine clinical assays for SAA, the National Institute for Biological Standards and Control has distributed into ampoules a candidate international standard preparation of SAA for evaluation by collaborative study. The preparation of SAA, consisting of pooled acute phase serum diluted in pooled normal serum, was generated in the Immunological Medicine Unit, Department of Medicine, Royal Postgraduate Medical School, Hammersmith Hospital, London. Although there are several isoforms of acute phase SAA, and these might conceivably vary in relative abundance during the acute phase response or in different conditions, the assay methods for SAA in whole serum do not distinguish between them. Furthermore, the candidate standard was prepared from pooled material of 400 individual donors and is therefore likely to be generally representative.

Section snippets

Aims of the study

The aims of the study were to: (1) assess the suitability of the candidate preparation to serve as the international standard (IS) for the assay of SAA by comparing, in a variety of assay systems, the candidate standard with other samples containing different amounts of SAA; (2) assay the SAA content of the candidate standard; (3) compare the candidate standard with local standards; (4) compare the various assay systems for SAA.

Preparations for the study

Ampouled preparations of SAA were prepared under the supervision of Dr. P. Dawson in the Standards Processing Division of NIBSC, according to the procedures used for international biological standards (Annex 4, 40th ECBS Report, 1990). These included the candidate standard, samples of serum containing different amounts of SAA and a sample of purified SAA1 (Patel et al., 1996). The candidate standard for SAA, consisting of acute phase serum diluted in 2 l of normal human serum was passed first

Design of the study

Participants received two sets of ampoules coded by each letter from A to F (see Table 1). Also, some participants received ampoules coded by each letter from V to Z. Each participant was asked to carry out at least two independent assays by each method used in their laboratory (see Table 2) and including as far as possible all the preparations to be tested, each to be tested at several dilutions in the linear part of the response range. Participants were also asked to include in the assay any

Participants

Seven laboratories participated in the study. Participants were:

  • Professor M.B. Pepys, Immunological Medicine Unit, Department of Medicine, Hammersmith Hospital, Du Cane Road, London W12 ONN

  • Professor Ruth Shainkin-Kestenbaum, Sapir Medical Center, Meir Hospital, 44281, Kfar Saba, Israel

  • Professor Jean D. Sipe, Boston University School of Medicine, Department of Biochemistry, 80 East Concord Street, Boston, MA 02118-2394, USA

  • Dr. Peter Merle, Behringwerke, Postfach 11 40, D-35001 Marburg, Germany

Statistical analysis

All assays were analysed as multiple parallel line bioassays (Finney, 1978), comparing assay response with log concentration. When plotted against log concentration, linear response lines which are parallel are essential for this analysis. The statistical validity of the assays was assessed by analysis of variance tests for linearity and parallelism.

For laboratories 01 and 03, responses at the extremes of the response range which showed no change with further increase, or decrease, of dose were

Dose–response relationships

Log dose–transformed response lines for the ampouled SAA preparations were generally linear and in the majority of cases parallel to one another. Exceptions were those for (preparation) A in laboratory 05 and to a lesser extent in laboratory 07 and for E in laboratory 05, which tended to be steeper than those for B, C, D and F, and for B in laboratories 05 and 06B and to a lesser extent C in laboratory 07 and F in laboratory 06B which tended to be flatter than those for A, D and E. For the

Discussion

Estimates for the ampoule content of the candidate standard (92/680) in terms of the various local (in-house) standards were between 169 μg/ampoule (95% confidence limits: 128–224) if all estimates were included and 154 μg/ampoule (131–182) if outlying values were excluded. However the mean value of 154 μg/ampoule is in good agreement with the nominal ampoule content prior to lyophilisation of 156 μg. The value of 156 μg had been determined by comparison with a highly purified preparation of

Acknowledgements

We thank Professor Patricia Woo of the Department of Molecular Pathology, University College London Medical School, UK, for providing SAA1 protein produced in-vitro, and the participants who gave their time and resources to carry out this study.

References (42)

  • D.J. Cohen et al.

    Serum C-reactive protein concentrations in cyclosporine-treated renal allograft recipients

    Transplantation

    (1988)
  • H.M. Falck et al.

    Correlation of persistently high serum amyloid A protein and C-reactive protein concentrations with rapid progression of secondary amyloidosis

    B.M.J.

    (1983)
  • Finney, D.J., 1978. In: Statistical methods in biological assay, 3rd edn. Charles Griffin,...
  • M. Fischbach et al.

    Evaluation of combination of biologic markers (C-reactive protein, neopterin, and microalbuminuria) in the follow-up of kidney transplantation in children: four years of clinical experience

    Transplant. Proc.

    (1994)
  • B. Freed et al.

    Early detection of renal allograft rejection by serial monitoring of serum C-reactive protein

    Transplantation

    (1984)
  • R. Gaines Das et al.

    Iterative weighted regression analysis of logit responses. A computer program for the analysis of bioassays and immunoassays

    Comput. Prog. Biomed.

    (1980)
  • J. Groth et al.

    C-reactive protein and clinical kidney transplantation

    Nephron

    (1984)
  • E. Hachulla et al.

    Serum amyloid A concentrations in giant-cell arteritis and polymyalgia rheumatica: a useful test in the management of the disease

    Clin. Exp. Rheumatol.

    (1991)
  • A. Hartmann et al.

    Serum amyloid A protein is a clinically useful indicator of acute renal allograft rejection

    Nephrol. Dial. Transplant.

    (1997)
  • P.N. Hawkins et al.

    Serum amyloid P component scintigraphy and turnover studies for diagnosis and quantitative monitoring of AA amyloidosis in juvenile rheumatoid arthritis

    Arthritis Rheum.

    (1993)
  • G. Hocke et al.

    A rapid laser immunonephelometric assay for serum amyloid A (SAA) and its application to the diagnosis of kidney allograft rejection

    Klin. Wochenschr.

    (1989)
  • Cited by (0)

    View full text