Complement I.‘ 147—159 ([984)
`
`01984 S. Karger AG. Basel
`0253-5076/84/00I3-0l4782.7$/0
`
`Quantification of Cl -Inhibitor Functional Activities by
`Immunodifl'usion Assay in Plasma of Patients with Hereditary
`Angioedema 9 Evidence of a Functionally Critical Level of
`Cl-Inhibitor Concentration
`
`P.J. Spd'th 3, B. Witt/trich b, R. Bt't‘tlera
`
`“Blood Transfusion Service SRC, Central Laboratory, Beme;
`b Dermatological Clinic, Allergy Unit, University Hospital, Zurich, Switzerland
`
`Key Words. Attenuated androgcns - Hereditary angioedema ~ Cl-inhibitor - C4 -
`Immunodiffusion assay
`
`CSL V. Shire
`
`Abstract. The relationship of Cl-inhibitor (Cl-INH) concentration and apparent func-
`tional activity was investigated in 11] plasma samples from 21 patients with the common
`form of hereditary angioedema (HAE). Functional C l -INH was analyzed by means ofa modi-
`fied version of immunodilfusion assay. Down to a Cl-lNH level of approximately 0.075 g/l
`(38% of normal) apparent Cl-INl-l functions were fOund within the normal range, while
`below this level functional adequacy of Cl-INH could no longer be ascertained. When C4
`concentrations, considered to reflect approximately functional Cl-INH, were related to Cl-
`INH antigen levels of individual samples, a relationship emerged which was identical to that
`between Cl-INH concentration and apparent function. No attacks of edema could be associ-
`ated with Cl-INH concentrations above 0.075 g/l, while it was possible to associate attacks
`with concentrations below this level. In experiments where patient plasma and normal plasma
`were mixed in various ratios or where HAE plasma was replaced by purified Cl-INH, an
`increase in Cl-INH antigen to concentrations of 0.06—0.08 g/I was followed by a sharp rise in
`apparent functions to normal values. The rise of functional Cl-INH became moderate when
`Cl-INH antigen further increased. The results supported the idea of a functionally critical
`level of Cl-INl-l in the common form of HAE.
`
`Introduction
`
`All these inhibitors become modified by the
`
`serine proteases which they inactivate [l].
`is a regulatory The serine proteases inhibited by Cl-INH
`Cl-inhibitor (Cl—INH)
`plasma protein. This az-glycoprotein belongs
`are the active forms of the two enzymatic
`to a class of inhibitors such as ol-proteinase
`subunits of the first component of comple-
`inhibitor, antithrombin III and antiplasmin. ment, Clr and C15, and also plasmin, kalli-
`
`This material was copied
`nthe "LM and may be
`Subject USCop-ltight lam
`
`CSL EXHIBIT 1083
`
`Page 1 of 13
`
`CSL EXHIBIT 1083
`CSL v. Shire
`
`Page 1 of 13
`
`

`

`148
`
`
`
`Spéith/Wiithrich/Biitler
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`krein and coagulation factors Xla, XIIa and
`
`
`
`
`
`
`XIIf[2—8]. In addition, Cl-INH controls the
`
`
`
`
`autoactivation of zymogen Clr [9].
`
`
`
`
`
`
`
`A variety of clinical symptoms may be
`
`
`
`
`
`
`associated with reduced functional activity of
`
`
`
`
`
`Cl-INH. In hereditary angioedema (HAE)
`
`
`
`
`
`the reduced Cl-INH functional activity in-
`duces recurrent attacks of edema of the ex-
`
`
`
`
`
`
`
`
`
`
`
`
`
`tremities, the gastrointestinal tract, the face
`
`
`
`
`
`
`
`
`or the larynx, and death may result from air-
`
`
`
`
`
`way obstruction [10]. Affected persons are
`
`
`
`
`
`
`
`heterozygous for a defect of the Cl -INH pro-
`
`
`
`
`
`
`
`
`
`
`tein gene [1 1]. In the common form of the
`
`
`
`
`
`
`
`disease, the reduction in functional activity is
`
`
`
`
`
`
`due to Cl-INH concentrations being only 5—
`
`
`
`
`
`
`
`
`
`30% of normal. In the variant form of the
`
`
`
`disease, immunochemically detectable con-
`
`
`
`
`
`
`centrations of Cl-INH are normal or ele-
`
`
`
`
`
`
`vated, but a dysfunctional mutant protein is
`
`
`
`
`
`
`synthesized [12]. A reduction of functional
`
`
`
`
`
`
`Cl-INH may also be acquired; such condi-
`
`
`
`
`
`
`
`
`tions may or may not be associated with
`
`
`
`
`
`
`reduced antigen levels [13, 14]. Furthermore,
`
`
`
`
`
`
`functional
`inadequacy of Cl-INH can be
`
`
`
`
`
`
`
`
`transient without any decrease in the level of
`
`
`
`
`
`Cl-INH [15]. Thus the unequivocal diagno-
`
`
`
`
`
`
`
`
`sis of all Cl-INH deficiency stages is based
`
`
`
`
`
`on the demonstration of a functionally inade-
`
`
`quate Cl-INH.
`
`
`
`
`
`
`Recently, Ziccara’i and Cooper [16] de-
`
`
`
`
`
`
`scribed a simple and easily performable assay
`
`
`
`
`
`system for assessment of Cl-INH functional
`
`
`
`
`
`activity. Unlike other methods, no special-
`
`
`
`
`
`
`
`ized techniques are needed and all reagents
`
`
`
`
`are commercially available. Furthermore,
`
`
`
`
`
`
`
`
`
`
`this assay system is the only one where one of
`
`
`
`
`
`
`
`
`the natural substrates of Cl-INH, Clr, in its
`
`
`
`
`
`physiologic concentration serves directly as
`
`
`
`
`
`
`
`an indicator of functional activity of the in-
`
`
`
`
`
`
`
`
`hibitor. So far a drawback of this assay sys-
`
`
`
`
`
`
`
`
`tem was the lack of a valid method to quan-
`
`
`
`tify functional Cl-INH.
`
`
`
`
`
`
`
`
`
`
`A modification of the immunodiffusion
`
`
`
`
`
`
`
`
`
`
`
`
`
`assay is proposed in this study which allows
`
`
`
`
`
`
`quantification of functional Cl-INH. It was
`
`
`
`
`
`
`
`used to investigate routinely the plasma of a
`
`
`
`
`
`
`large number of patients for functional C1-
`
`
`
`
`
`
`INH and to select those patients suffering
`from the common form of HAE. A further
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`aim of this study was to reinvestigate the
`
`
`
`
`
`
`effect of attenuated androgens in the com—
`
`
`
`
`
`
`mon form of HAE. Indeed,
`the literature
`
`
`
`
`
`reports that therapy with impeded androgens
`
`
`
`
`
`prevents attacks of angioedema and normal-
`
`
`
`
`
`
`izes the concentration of the fourth compo—
`
`
`
`
`
`
`nent of complement. However, there are con—
`troversies about the effect on the level of Cl-
`
`
`
`
`
`
`
`
`
`
`
`
`
`INH antigen [17—23]. The results presented
`
`
`
`
`
`
`
`
`in this study suggest that there is a function—
`
`
`
`
`
`
`
`ally critical level of Cl-INH concentration of
`
`
`
`
`
`
`
`approximately 0.075 g/l (38% of normal),
`
`
`
`
`
`
`above which apparently normal function of
`Cl-INH and thus normal concentrations of
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`C4 seem to be assured in the plasma ofpatients
`
`
`
`
`
`
`
`affected by the common form of HAE.
`
`
`
`Patients and Methods
`
`
`
`
`
`
`
`
`
`
`Subjects
`
`21 patients were studied, 15 women and 6 men, all
`
`
`
`
`
`
`
`
`
`suffering from the common form of HAE. In each
`
`
`
`
`
`
`
`
`case, diagnosis was made on the basis of clinical and
`
`
`
`
`
`
`
`
`
`family history as well as on serological findings. Of the
`
`
`
`
`
`
`
`
`
`111 plasma samples analyzed, 22 were from 13 pa-
`
`
`
`
`
`
`
`
`tients receiving no therapy. 2 of these 13 patients were
`
`
`
`
`
`
`
`
`
`
`eventually treated with attenuated androgens and 1
`
`
`
`
`
`
`
`with tranexamic acid. 5 samples were collected from 3
`
`
`
`
`
`
`
`
`
`patients receiving therapy with tranexamic acid.
`I of
`
`
`
`
`
`
`
`
`these patients was later treated with danazol. 7 pa-
`
`
`
`
`
`
`
`
`tients were managed by danazol or stanozolol (Win-
`
`
`
`
`
`
`
`throp, Basel, Switzerland). 84 samples originated from
`
`
`
`
`
`
`
`this group of patients.
`
`
`
`
`
`Laboratory Methods
`
`
`Heat-aggregated human lgG (aIgG) was prepared
`
`
`
`
`
`
`by heating a solution of 16 mg/ml for 20 min to 63 °C.
`
`
`
`
`
`
`
`
`
`
`
`
`After cooling and centrifugation for 15 min at 1,500 g
`
`
`
`
`
`
`
`
`
`
`
`This mate rial was copied
`
`
`
`
`
`
`aims NLM and may be
`
`
`
`
`SL5 Eject U5 Earp-yright Laws
`
`
`
`
`
`
`Page 2 of 13
`
`Page 2 of 13
`
`

`

`Functionally Critical Level of C l-INH Concentration
`
`
`
`
`
`
`
`149
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`INH. The method proposed estimates func-
`
`
`
`
`
`
`
`
`tional Cl-INH by calculating the slope of the
`
`
`
`
`
`
`
`regression line formed when the added aIgG
`
`
`
`
`
`
`concentrations are plotted against ratios of
`
`
`
`
`
`
`immunologically detectable Clr in the 5 ali-
`
`
`
`
`
`
`
`
`quots of the recalcified plasma sample to be
`
`
`
`
`
`
`analyzed. To obtain a significant coefficient
`
`
`
`
`
`
`
`of regression (r2), it proved useful first to cal-
`
`
`
`
`culate the percentages of immunologically
`
`
`
`
`
`
`
`
`detectable Clr in the 5 aliquots. The Clr con-
`
`
`
`
`
`
`
`tent of plasma pooled from 50 healthy blood
`
`
`
`
`
`
`donors served as reference (=100%). Accu-
`
`
`
`
`
`
`
`
`racy of Clr quantification from day to day
`
`
`
`
`
`
`
`
`
`
`
`was i 5 % (n = 45). To avoid a dependency of
`
`
`
`
`
`
`apparent functional activity of Cl-INH on
`
`
`
`
`
`
`
`
`levels of Clr, calculation of the ratios of
`
`
`
`
`
`immunologically detectable Clr in the ali-
`
`
`
`
`
`
`quots A0_4 was a prerequisite. Calculation in
`
`
`
`
`
`
`parallel of the regression lines in individual
`
`
`
`
`
`serum and recalcified citrate plasma samples
`
`
`
`
`
`
`from 50 healthy blood donors generally re-
`
`
`
`
`
`
`
`
`
`sulted in a higher r2 value for the recalcified
`
`
`
`
`
`plasma samples. Quantification of functional
`
`
`
`
`
`
`
`
`Cl-INH in EDTA plasma proved not to be
`
`
`
`
`
`possible, even though Ca2+ and Mg“ were
`
`
`added in a 10-fold excess.
`
`
`
`
`
`
`
`
`
`
`
`
`
`In each run of analyses the slope of the
`
`
`
`
`
`regression line of recalcified plasma pooled
`
`
`
`
`
`
`
`from 50 healthy blood donors was also esti-
`
`
`
`
`
`
`
`
`
`
`
`mated and was set at 100% (in fig. 1, b =
`
`
`
`
`
`—0.577). The percentage of Cl-INH func-
`
`
`
`
`
`
`
`tional activity in plasma from a HAE pa-
`
`
`
`
`
`
`
`
`
`
`tient is reflected by the value of the slope of
`
`
`
`
`
`
`
`
`
`
`the regression line (in fig. 1, b = -—0.196) and
`
`
`
`
`
`
`
`
`can be calculated as given in this example
`
`
`
`
`
`
`
`(fig. 1):
`(—0.196/—0.577) X 100% = 34%.
`
`
`
`
`
`
`
`The accuracy of the determination within 1
`
`
`
`
`
`
`
`
`
`
`
`day was found to be i 10% (n = 15) and
`
`
`
`
`
`
`
`
`
`
`
`
`
`from day to day to be i 13% (n = 45). The
`
`
`
`
`
`
`normal range of functional Cl-INH levels
`
`
`
`
`
`
`varied between 70 and 135% (95% confi-
`
`
`dence interval).
`
`
`
`the protein concentration was adjusted to exactly
`
`
`
`
`
`
`15 mg/ml.
`
`
`Concentrations of C4 and Cl-INH were deter-
`
`
`
`
`
`
`mined nephelometrically (Hyland) and by single radial
`
`
`
`
`
`
`
`immunodiffusion (RID), respectively. Standards from
`
`
`
`
`
`Atlantic Antibodies (Merz + Dade, Diidingen, Swit-
`
`
`
`
`
`
`zerland) and from Hyland (Travenol, Switzerland)
`
`
`
`
`
`served as references. Purified, functionally active C1-
`
`
`
`
`
`
`INH was kindly donated by II.P. Kocher, University
`
`
`
`
`
`
`
`of Geneva, Switzerland.
`
`
`
`
`
`
`
`
`
`
`Quantification of Functional CI—INH by the
`
`
`
`
`
`Immunodiffusion Method
`
`
`9 volumes of blood were drawn into siliconized
`
`
`
`
`
`
`
`
`glass tubes containing 1 volume of0.11 mol/l trisodium
`
`
`
`
`
`
`
`citrate. After adequate mixing and centrifugation,
`
`
`
`
`
`
`plasma was stored in plastic tubes at —70 °C. To 495 pl
`
`
`
`
`
`
`
`
`
`
`
`plasma 5 pl of 2 mol/l CaClz solution was added. The
`
`
`
`
`
`
`
`
`
`
`formed clot was removed and the supernatant was
`
`
`
`
`
`
`
`
`divided into 5 aliquots of 80 ul (A04); aIgG solution
`
`
`
`
`
`
`
`
`
`
`was added to 4 of the aliquots to obtain final concen-
`
`
`
`
`
`
`
`
`
`
`tration of 0.25 (A1), 0.5 (A2), 1.0 (A3) and 1.5 mg
`
`
`
`
`
`
`
`
`
`
`
`aIgG/ml (A4). The volumes of the aliquots were ad-
`
`
`
`
`
`
`
`
`justed to 100 pl with isotonic saline. To the 5th of the
`
`
`
`
`
`
`
`
`
`
`
`aliquots (A0) 20 ul isotonic saline was added. The mix-
`
`
`
`
`
`
`
`
`
`tures were incubated for 60 min at 37 °C. Incubation
`
`
`
`
`
`
`
`
`
`
`was stopped by the addition of6 ul of0.2 mol/l EDTA,
`
`
`
`
`
`
`
`
`
`
`pH 7.4. In each run, pooled plasma from 50 healthy
`
`
`
`
`
`
`
`
`
`
`blood donors was treated the same way. Clr levels in
`
`
`
`
`
`
`
`
`
`
`A04 were assessed by RID in 1 % (w/v) agarose, pH 8.0,
`
`
`
`
`
`
`
`
`
`
`
`containing 3.25% (v/v) ofan IgG fraction ofgoat anti-
`
`
`
`
`
`
`
`
`
`
`human Clr (Atlantic Antibodies). The Clr contents of
`
`
`
`
`
`
`
`
`various dilutions of the aliquot of the pooled plasma
`
`
`
`
`
`
`
`
`
`that did not receive aIgG served as a calibration curve.
`
`
`
`
`
`
`
`
`
`Ratios ofClr were formed by dividing the percentage of
`
`
`
`
`
`
`
`
`immunologically detectable Clr in A04 by the C 1 r level
`
`
`
`
`
`
`
`
`
`in A0. In order to quantifiy Cl—INH functional activity,
`
`
`
`
`
`
`
`
`
`the logarithms of the added aIgG concentrations were
`
`
`
`
`
`
`
`
`plotted against ratios ofClr in the aliquots and the slope
`
`
`
`
`
`
`
`
`
`
`
`of the regression line was calculated.
`
`
`
`
`
`
`
`
`
`Results
`
`
`
`
`
`
`
`Quantification of Functional CI-INH by
`
`
`Immunodiffitsion Assay
`
`
`
`
`
`The immunodiffusion assay of Ziccardi
`
`
`
`
`
`
`
`
`
`and Cooper [16] was modified in such a way
`
`
`
`
`
`
`
`as to allow quantification of functional C1-
`
`
`
`Th is mate rial was copied
`
`
`
`
`
`
`aims NLM and may DE
`
`
`
`
`EL: Eject U5 Copyright Laws
`
`
`
`
`
`
`Page 3 of 13
`
`
`
`
`
`
`Page 3 of 13
`
`

`

`ISO
`Spéith/Wiithrich/Biitler
`—__—____________—_____———————
`
`oa:
`
`“0
`I.CtrA)
`'1.Ctr Da:
`
`,‘c
`
`Page 4 of 13
`
`Thus, the most exteme C l-lNH concentra-
`tion and related function (0.02 g/l and 85%,
`respectively; one of the open triangles in
`fig. 3a) were seen in the plasma from a pa-
`tient with an abscess on his neck. All other
`markedly divergent concentrations and func-
`tions were observed only in recalcified
`plasma from members of a particular family.
`All these patients were heterozygous for the
`Cl-lNl-l and the factor I (C3b/C4b-INA)
`protein gene, and some of them even showed
`
`Functional Adequacy of CI-INH in the
`Common Form of HAE
`Figure 2 gives an overview of the relation-
`ship between C ] -lNH concentrations and ap-
`parent functions in clinical conditions. These
`include the common and the variant form of
`
`HAE (sample of the variant form kindly pro-
`vided by C. Alper, Boston, USA), acquired
`Cl-lNl-l deficiency with complete remission
`and normalization of various complement
`parameters for 6 months after surgical re-
`moval of a giant myoma of the fundus uteri,
`deficiencies and dysfunctions of components
`of complement other than Cl-INH, hypo- or
`agammaglobulinemia, immune complex dis-
`eases, acute and chronic idiopathic thrombo-
`
`cytopenic purpura, malignancies, bacterial
`and viral infections and urticaria. The results
`
`indicate a correct quantification of functional
`
`Cl-lNH in cases of the acquired and variant
`forms of angioedema. However, in II] recal-
`cified plasma samples from 21 patients with
`the common form of HAE, unexpectedly
`high functional activities associated with
`subnormal levels of C l-lNH were observed.
`
`A closer view of Cl-lNH antigen levels and
`related functions in these samples is given in
`figure 3a. The association of apparently nor-
`mal function and abnormally low Cl-lNH
`level was found almost exclusively in recal-
`cilied plasma from patients treated with at-
`tenuated androgens. Among the 23 samples
`with Cl-lNH levels above 0.075 g/l (38% of
`normal) only 1 sample (4%) displayed an
`apparently inadequate Cl-lNH function. In
`
`contrast, among the 88 samples with Cl-INH
`concentrations below 0.075 g/l 65 (74%)
`showed functionally inadequate Cl-lNH.
`In the range of 0.035-0.075 g/l Cl-lNl-l
`antigen the functional activities could vary
`from 10 to 135% ofnormal, and in 1 sample
`an apparent Cl-lNH function as high as
`
`Rah:ofdetectableClr\ oa
`
`_ON
`
`0
`
`.
`
`A,
`A.J
`‘__,
`I
`025
`0
`.3qu added, mglml
`
`Fig. I. Quantification of Cl-INH functional activ-
`ity in HAE using the immunoditTusion method. The
`regression lines between the concentrations of added
`algG and the ratios of immunologically detectable Clr
`in aliquots (A0...) of recalcified plasma pooled from
`healthy blood donors (El) and in aliquots of rccalcificd
`plasma from a patient (0) are shown. The coefficients
`of correlation (r1) in the examples given were 0.999 (D)
`and 0.99l (0), respectively.
`
`156% emerged (open circle in fig. 3a). This
`and all other samples with extremely diver-
`gent Cl-lNI-l concentrations and apparent
`functions were collected from patients with
`manifest infections or from patients where a
`subclinical infection was strongly suspected.
`
`l‘hlsmalerill wascopied
`um NLM Ind maybe
`Subject US Copyright Laws
`
`Page 4 of 13
`
`

`

`Functionally Critical Level of C l-lNH Concentration
`
`15 1
`
`Fig. 2. C l-l NH concentrations and
`apparent function s in 364 recalcified
`plasma samples analyzed routinely. T he
`symbols indicate samples from patients
`with the common form of HAE under
`various therapeutic regimens (•), ac(cid:173)
`quired angioedema before and afier sur(cid:173)
`gical intervention (o), the variant form
`of HAE(•), the common form of HAE
`a tier infusion of C 1-INH concentrate
`(•), HAE under discussion (o) and
`plasma sa mples from members of HAE
`families (L\. ), or from subjects with var(cid:173)
`ious other diseases (see text) ( + ). Nor(cid:173)
`mal 95 % range of immunologica lly de(cid:173)
`tectable C l-INH is given by the vertical
`and of functional C l-INH by the hori(cid:173)
`zontal , partially dotted , lines.
`
`. +
`
`•
`
`I
`
`• +
`
`I
`
`150
`
`100
`
`~ 50 -
`I
`
`G
`'ii
`c
`
`2 u c
`
`:J
`lL
`
`+
`
`+
`
`0.1
`0
`Cl-INH ant igen. g ll
`
`0.2
`
`0.3
`
`depressed plasma concentrations of C4-bind(cid:173)
`ing protein. Levels of C3 were usually found
`below the lower limit of the normal range and
`the patients had prolonged phases of infec(cid:173)
`tions.
`Figure 3a does not clearly indicate the re(cid:173)
`lationship of C l-INH concentration and
`function in the sam ples analyzed . However,
`when C l-INH antigen was ranged in classes
`of 0.02 g/I, as indicated by bars in figure 3b,
`the means of C I-INH concentrations and
`apparent fun ctions of the individual classes
`clearly indicated a nonlinear relationship.
`Two approaches calculating the relationship
`between concentration and function were
`made. In a first attempt, the regression lines
`of concentrations and functions in samples
`with C 1-INH levels in the ranges 0.012-0.08
`and 0.08-0.13 g/I, respectively, were calcu(cid:173)
`lated. For this purpose, individual C I-INH
`concentrations and functions of plasma sam-
`
`pies as given in figure 3a were considered.
`The two regression lines are shown in figure
`3b. In a second attempt the regression line
`was calculated after logit transformation of
`apparent C I-INH function. Such transforma(cid:173)
`tion requires values below 100%. Thus, it
`was reasonable to do a calculation using the
`mean values of functional Cl -INH as given
`in figure 3b. The use of the mean values
`ensured that the corresponding functional ac(cid:173)
`tivities of an equivalent of C l-INH antigen
`up to a concentration of 0.09 g/I were taken
`into consideration. The retransformed values
`of the calculated regression line are given in
`figure 3b. Both calculations indicated that a
`concentration of Cl-INH antigen of 0.07-
`0.08 g/I represents a threshold level for C 1-
`INH function.
`The concept of a functionally critical level
`of C l-INH concentration was further sup(cid:173)
`ported by relating C l-INH concentration to
`
`Thi.s m at,e·ri al w a scopie-d
`atthe-N LM and maybe
`Subj ect US Copyright Law s
`
`Page 5 of 13
`
`

`

`152
`
`Spiith/ Wlithrich/ Blitler
`
`150
`
`i'- 100
`I '
`z
`I
`G 50
`;;;
`c
`0
`·.;=
`u
`c
`:J
`u_
`
`0
`
`04
`
`0.3
`
`02
`
`:::
`Ol
`c
`Q)
`Ol
`~
`"'
`"
`u
`
`0.1
`
`0
`
`0
`
`0
`
`..
`"
`•
`"
`• •
`•• •• • • •
`Ill •
`• •
`• •
`. -...~··
`0 ..... •
`• •
`•
`. ~· . •
`.,~:
`., .
`~ .. \ •
`d .
`
`6
`
`. . . .
`
`0
`
`'
`
`• ~~! '
`'
`"
`
`a
`
`,,..
`
`c
`
`d
`
`b
`
`e
`
`,
`
`I
`
`I
`
`, _______
`
`.. " • • •
`..
`•
`9\. • •
`..
`•• .z,. .
`. ,..,, ••
`"'·:.···
`~ ...
`
`t ~
`17~ 0
`•" oli!ib
`" "
`
`0.05
`
`0.15
`0.1
`C1- INH antigen. g I I
`
`005
`
`0.1
`
`0.15
`
`Fig.3. Q uantification of C l -INH and C4 antigen
`and functiona l C 1-INH in 111 indi vidual plasma sam(cid:173)
`ples from 2 1 patients with the common form of HAE.
`T he vertical lines indicate the lower limit of the nor(cid:173)
`mal range (-2 SD) of C l-INH concentra tion. The
`horizontal lines represent the lower and the upper li m(cid:173)
`its of the normal range( ± 2 SD) of apparent C l -IN H
`function, respecti vely. a Plot of indi vidual C l-INH
`antigen levels against apparent functional acti vities.
`The therapeutic management of the patients at the
`time when the samples were collected is ind icated: no
`therapy (o), therapy with tranexamic acid (Y'), therapy
`with danazol (•) or stanozolol (&). 6 or o = Samples
`from patients with present or recent bacterial or viral
`infections and therapy with danazol or stanozolol; IZI =
`C l-INH concent rations and functions in samples
`from an asymptomatic woman, a member of the
`famil y with concomitant heterozygoty fo r the C 1-INH
`and the factor I protein gene, herself heterozygous onl y
`fo r C l-INH . b Means( ± I SD) of C l -IN H concent ra(cid:173)
`tions and functions as can be calculated when ranges
`
`of C l-I NH antigen levels are fo rmed . Li mits o f the ·
`ranges we re 6-1 5, 16-25 , ... , 56- 67% of normal co n(cid:173)
`centration ( J 00 % = O. I 96 g/I). o = Calculated curve
`of mean C l-I N H concent rations versus logit tra ns(cid:173)
`fo rmed and retransformed mean functi onal acti vi(cid:173)
`ties; - - - = regression lines between indi vidual C 1-
`IN H concentra tions in the ra nges 6-40 and 41-67 %
`fun ctions, respecti vely.
`of normal and associated
`c C J-INH antigen levels and mani fes tation of attacks
`of edema.• = C l-INH concent ra tions in plasma col(cid:173)
`lected after an attack; o = mean concentrations of
`two difTerent quantifications. During the peri od o f
`quantifi cation an attack was observed, but the time
`between the plasma collections was too long to per(cid:173)
`mit direct correlation of C l-INH concent ration and
`attack of edema. d Indi vidual C l-I NH levels, plotted
`against C4 concentrations. T he sym bols used are the
`same as in a. e Means(± I SD) o f C l -IN H and C4
`concentrations as can be calculated when ranges o f
`C l-INH antigen levels are fo rmed . Symbols and ex(cid:173)
`planations as in b.
`
`This. mate ri al w asc.o pie<I
`at t he NLM and may be
`Subject US Copyright Law5
`
`Page 6 of 13
`
`

`

`Functionally Critical Level of C 1-I NH Concentration
`
`153
`
`manifestations ofangioedema. C l-INH con(cid:173)
`centrations above 0.07 5 g/l (38 % of normal)
`could not be related to an attack of edema,
`while in all samples collected after an attack
`C 1-INH concentrations were below 0.07 5 g/l
`(fig. 3c). Attacks were most frequent when
`C l-INH levels were below 0.035 g/l (18% of
`normal).
`
`Relationship between Cl-IN/-/ and C4
`Concentrations in the Cornman Form of
`HAE
`In HAE a close relationship between func(cid:173)
`tional Cl -I NH and levels of C4 antigen is
`assumed. The assumption of such a close
`relationship is widely accepted and serves
`worldwide for routine laboratory approxima(cid:173)
`tion of functional Cl-INH in HAE. Thus,
`quantification ofC4 levels by routine labora(cid:173)
`tory technique in parallel to C l-INH func(cid:173)
`tion offered a chance to verify the correctness
`of the method proposed for quantification of
`C 1-INH function. Figure 3d depicts the rela(cid:173)
`tionship between Cl-INH and C4 concentra(cid:173)
`tion of individual samples. Analogous to the
`results as given in figure 3a, C l -INH concen(cid:173)
`trations above 0.075 g/l were associated with
`C4 levels within the normal range. Further(cid:173)
`more, Cl-INH antigen within a range of
`0.035-0.07 5 g/l could be associated with C4
`concentrations of 0.03-0.25 g/l. Finally, be(cid:173)
`low a C l-INH concentration of 0.035 g/l C4
`levels were greatly reduced. When the same
`procedure was applied to results presented in
`figure 3d as was applied to parameters in fig(cid:173)
`ure 3a to obtain the results in figure 3b, a
`nonlinear relationship between C l-INH and
`C4 concentrations became apparen t (fig. 3e).
`Again a C l -INH threshold level of0.07-0.08
`g/l was found, above which normal C4 con(cid:173)
`centrations seemed to be assured in patients
`suffering from the com mon form of HAE.
`
`Cl -INH Functions in Mixtures of HAE
`and Normal Plasma and in HAE Plasma
`Replaced by Purified CJ-INH
`The results in figure 3a reflect the fluctua(cid:173)
`tion in C l-I NH concentration and apparent
`function in the plasma of various HAE indi(cid:173)
`viduals receiving various therapeutic regi (cid:173)
`mens. The results imply, but do not conclu(cid:173)
`sively prove, a nonlinear relationship be(cid:173)
`tween C l-INH concentration and apparent
`function. To obtain more evidence of the
`existence ofa functionally critical level of C 1-
`INH concentration, recalcified plasma from
`2 individuals with the common form of HAE
`was used for furth er in vitro studies. At the
`time of plasma collection both patients were
`receiving tranexamic acid and thus had low
`C 1-I NH concentrations and functions. The
`concentration of factor B in the plasma from
`1 of these patients was elevated, indicating
`that there was, or recently had been , an
`inflammatory
`process. T he
`recalcified
`plasma of the patient with a normal factor B
`concentration was mixed in various ratios
`with recalcified plasma from a healthy per(cid:173)
`son. Cl -INH concentration and apparent
`function of the mixtures were quantified and
`the results are shown in figure 4. To demon(cid:173)
`strate the relationship between Cl -INH con(cid:173)
`centration and apparent function, the regres(cid:173)
`sion lines of these parameters within C l-INH
`concentration ranges of 0.02-0.06 and 0.06-
`0.18 g/l, respectively, were calculated. Again
`a disproportionately rapid rise of apparent
`function over immunologically detectable
`C l-INH emerged in the C l-INH concentra(cid:173)
`tion range of 0.02-0.06 g/l. T his dispropor(cid:173)
`tion was reversed as soon as C 1-INH concen(cid:173)
`tration exceeded a val ue of 0.06 g/l.
`The study described above was repeated
`with plasma containing elevated concentra(cid:173)
`tions of factor B. A rise in C 1-INH concen-
`
`This. material w a.s.copiedl
`at t he NLM and may be,
`Subject USCopyright Law>
`
`Page 7 of 13
`
`

`

`154
`
`Spath / Wiithrich/ Biitle r
`
`tration from 0.035 to 0.065 g/I was associ(cid:173)
`ated with an overshoot in apparent func(cid:173)
`tional activ ity reaching values of 180%. A
`furth er rise in C l-I NH antigen as a result o f
`the addition of normal recalci lied plasma re(cid:173)
`sulted in a constant decrease of apparent
`functional activity to the values o f the recal(cid:173)
`cified normal plasma (fig. 5). Replacement of
`th e HAE plasma containing elevated concen(cid:173)
`trations of factor B by increasing amounts of
`purified C 1-INH led again to an initially dis(cid:173)
`proportionate ri se in apparent function s; at a
`concentration of0.075 g/I the apparent fun c(cid:173)
`tion of C l-I NH was approximately 100%
`and subnormal concentrations of C l-INH
`around 0 .1 g/I were associated with apparent
`fun ctions above the upper limit of normal. A
`ri se in C l-I NH concentration from 0.1 to
`0.35 g/I showed li ttle effect on fun ctional C l(cid:173)
`IN H (fig. 5).
`
`150 -
`
`100 -
`
`I
`I
`•I
`
`I'
`z
`' 50
`u
`'ii c
`Q
`u c
`t.i'
`
`0 -i---'-r----.--r-.--1r-.----r--Y-~
`0. 15
`0.10
`0
`0.05
`C l-INH an tigen. gi t
`
`Fig. 4. Concentrations and functions o f C l - lN H in
`recalcified mixtures with va rying ratios of HAE
`plasma and normal plasma. Regression lines o f C I -
`!NH functions and concentrations in the ra nges 0.02-
`0.06 and 0.06-0. 18 g/l, respecti vely, were calculated.
`Vertical and horizontal lines indicate normal ranges of
`C l-I NH concentration and apparent function.
`
`Discussion
`
`The results of this study suggest that there
`is a fun cti onally criti cal level of Cl-I NH in
`recalcified plasma. T his evidence was ob(cid:173)
`tained by analysis of 111 recalcified plasma
`samples from 21 patients with the common
`form o f HAE under various types of therapy,
`and also from recalcified HAE plasma sup(cid:173)
`plemented by recalcified normal plasma or
`replaced by puri lied C I -I N H. The existence
`o fa functiona lly critica l level ofCl-INH was
`found to be rooted in a nonlinear relationship
`between C l-INH concentration and apparent
`function. The value of the fun ctionally criti(cid:173)
`cal level was found to be approximately
`0 .075 g/I (38% of normal); indeed, the fun c(cid:173)
`tional adequacy of C l-IN H, the concentra(cid:173)
`ti ons of C4 within the normal range and the
`absence of attacks of edema were guaranteed
`
`as long as th e antigen concentration did
`not fall below thi s threshold level. Thus,
`thi s is the first study to explain the obser(cid:173)
`vation of this and other reports [1 8- 23) of
`normal C4 concentrations in association
`with subnorm al levels of C l-INH in HAE
`(fig. 3).
`At the time when th e preli m inary results
`of this 4-year stud y were presented [24], no
`other reports existed which could support the
`idea of a nonlinear relationship between C I -
`!NH concentration and fun ction; indeed all
`reports analyzing concentrations and fun c(cid:173)
`tions of C 1-INH by va rious methods fa(cid:173)
`voured a linear relationsh ip [25-31 ]. How(cid:173)
`ever, in the meantime four publications pre(cid:173)
`sented data which are compatible with a
`functiona lly critical level of C 1-I NH concen(cid:173)
`tration . Very recentl y, Ghebrehiwet et al. [32)
`studied the spontaneous breakdown of he-
`
`Th i.s m ateri a I w as co pi e.c:I
`at the NLM and may b-e,
`Subjert US Copyright Law.s.
`
`Page 8 of 13
`
`

`

`Functionally Critical Level of C l -INH Concentration
`
`155
`
`Fig. 5. Concentration s and func(cid:173)
`tions of C 1-INH in a recalcified
`mixture of HA E plasma with ele(cid:173)
`vated factor B level and normal
`plasma(•) or purified C l-JNH (.6.).
`The indicated relationships of C 1-
`INH concentrations and function s
`as they arc given by the drawn
`curves are approximations and were
`not ca lculated. The normal range
`(RID) is indicated by vertical lines,
`the apparent normal range of fimc(cid:173)
`tional Cl-INH by horizontal lines.
`
`150 -
`
`.~ 100 -
`I.
`z
`
`I u 50.
`
`;;;
`c
`Q
`u c
`~
`LL
`
`~,---,---,--,-
`
`0 1
`005
`0
`C HNH antigen, g /l
`
`0 15
`
`0 2
`
`0.25
`
`0 3
`
`0.35
`
`molytic complement in diluted sera obtained
`from patients with HAE. They found that the
`concentration ofCl-INH in NHS is in excess
`of what is required to prevent the breakdown
`of hemolytic complement in the diluted sam (cid:173)
`ples. In 1 HAE serum sample with no detect(cid:173)
`able functional C l-I NH but substantial he(cid:173)
`molytic complement prior to dilution, the
`replacement of highly purified and function(cid:173)
`all y active C l-IN H resulted in a dose-depen(cid:173)
`dent prevention of spontaneous complement
`reduction which reached its maximum effect
`at concentrations around 50 % of normal. In
`another publication, Yelvington et al. [33]
`showed C l-INH concentrations of approxi(cid:173)
`mately 0 .08 g/I in sera of patients with HAE
`in association with normal and abnormal
`results of the immunodiffusion assay. Below
`C l-INH concentrations of approximately
`0 .075 g/I the immunodiffusion assay indi(cid:173)
`cated a functionally diminished C l-I NH in
`all samples, whilst above a concentration of
`approximately 0 .095 g/l functional C l-I NH
`was apparently normal. Further ev idence ofa
`nonlinear relationship between C l-INH con-
`
`centration and function was reported by
`Doekes et al. [34]. They analyzed in vitro the
`inhibition of the C4-converting activity after
`activation of Cl by algG of definite size in
`the presence of various amounts ofCl-INH.
`The maximum C4 consumption that could
`be reached at high concentrations of algG
`could be significantly reduced at a Cl-INH:
`C4 ratio of only 0.6 although in vivo the
`value of normal Cl-INH:C4 ratio is 5- 10.
`The fourth publication which supports the
`findings of th is study is by Ziccardi [9]. He
`ana lyzed the concentration dependence of in(cid:173)
`hibition by Cl -I NH of spontaneous Cl acti(cid:173)
`vation . Although a different method was
`used to quantify functional Cl-INH, the re(cid:173)
`sults obtained were similar to those of this
`study. A rise in Cl-INH concentration from
`25 to 35 % of normal was associated with a
`rise in apparent inactivating capacity of C 1-
`INH from 20 to 70% of normal. Variations
`in Cl -I NH concentration above 35% of nor(cid:173)
`mal resulted only in minor changes of appar(cid:173)
`ent functional activity ofC l-INH. Although
`Ziccardi 's [9] results and those of this study
`
`This. mat erial w as. copie-d
`at t he NLM and ma y Oe
`Su boject US Copyright La w s
`
`Page 9 of 13
`
`

`

`156
`
`Spath/Wiithrich/Biitler
`
`accord, it must be borne in mind that two
`different functions of Cl-INH were ana(cid:173)
`lyze