Pure Red Cell Aplasia: Studies on an IgG Serum Inhibitor Neutralizing Erythropoie tin CESARE PESCHLE, ALBERTO M. MARMONT,* GIANNI MARONE, ~ MARIOCONDORELLI ARTURO GENOVESE, GUIDOF. S A S S OAND

Institirte of Medical Pathology, II Faci4lty of Medicine and Surgery, University of Naples, Naples, It a1y (Reccived 16 Deccniber 1974; accLptcdfor pirblication 4 February 1975) SUMMARY. A new type of IgG serum inhibitor in adult pure red cell aplasia (PRCA) has been investigated. This inhibitor is directed against circulating erythropoietin (Ep) (PRCA type B), rather than thc erythroid marrow (PRCA type A), Thus, the IgG inhibitor, after interaction with Ep in solution, is precipitated together with Ep by addition of goat anti-human gamma-globulins. Pre-therapy PRCA serum, although apparently devoid of Ep, shows considerable Ep activity following acidification and boiling. The inhibitor is absent from post-therapy scrum, while Ep levels are elevated. An experimental model for PRCA type B has been established in normal mice after proloiigcd administration of pre-remission serum IgG. Pure red cell aplasia (PRCA) in the adult is presently considered an autoimmune condition (Krantz, 1973 ; Peschle, 1973), characterized by both a favourable response to non-steroid iiiiniunosuppressive therapy (Krantz & Kao, 1969; Krantz, 1972; Marmont, 1973) and the presence of an IgG serum inhibitor of erythropoiesis (Krantz & Kao, 1969; Krantz, 1973 ; Peschle, 1973 ; Zalusky et al, 1973). Extensive evidence (Krantz & Kao, 1969; Krantz, 1973 ; Peschle, 1973 ; Zalusky et al, 1973) indicates that the inhibitor acts at the level of erythroblasts and/or an carly, unidentifiable erythroid precursor, which has been differentiated from the erythropoietin-responsive-cell(s) by the Ep stimulus (PRCA type A). However, in one out of nine patients examined so far by our group, the presence of an IgG serum inhibitor neutralizing circulating erythropoietin (Ep) has been demonstrated (PRCA type B). The physiopathology of this type of PRCA is presented here. CASE REPORT A 70-year-old woman was admitted to hospital on 12January 1973 because of severe anaemia, which was first recognized in 1966. O n admission the patient was extremely pale, but otherwise her physical examination was negative. Fluoroscopic chest examination was negative; electrocardiographic tracings were normal. Laboratory examination showed : H b 8 g/dl,

* Division of Haetnatology, Ospedali Civili di Genova, Genova, Italy. t Institute of Medical Semeiology (V), University of Rome, Rome, Italy. Correspondence: Dr Cesare Peschle, Istituto Patologia Medica, Nuovo Policlinico, Via S. Pamini, 8013 I Napoli, Italy. 411

412

Cesare Peschle et a1

Hct 0.25; RBC 2.27 x I O ” / ~ . ; peripheral reticulocytes werevirtually absent. WBC9.o x 10~11.; the differential count was : neutrophils 70%, lymphocytes 20%, monocytcs 10%. One myelocyte and three metamyelocytes were also observed. The platelet count was 85 x 109/l. The total serum protein concentration was 59 g/l. (37 g albumin and 22 g globulins); immunoglobulin values by radical immunodiffusion were: IgG 10.8, IgA, 2.65, IgM 1.48 g/l. No paraprotein peaks could be detected. The LE test was negative; anti-nuclear and anti-IgG antibodies were not detected. Both direct and indirect erythrocyte antiglobulin tests, Ham’s and siicrose tests were negative. Osmotic fragility was in the normal range. Serum iron was 39 pmoI/I.; all other biochemical data were within normal limits. A series of sternal and iliac crest examinations showed a hypercellular marrow with a hyperplastic granulocytic component and near-normal megakaryocytosis;erythroblasts were not observed on any smears. Focal lymphoid infiltration was present. Remission of the PRCA condition was observed following an initial cycle of combined therapy with dexamethasone (8 mg/day) and cyclophosphamide (200 mg/day up to a total dosage of 5.2 g of the latter agent) and a subsequent cycle of equine anti-lymphocytic globulin (ALG, Behringwerke, Marburg, West Germany) up to a total dosage of 145 ml over 9 days. The reticulocyte count progressed up to a peak of 10% and the packed cell volume to 0.37 although no blood transfusions had been administered. MATERIALS AND METHODS

Collection of Serum Heparinized or non-heparinized sterile blood was collected both prior to and after treatment. Serum and plasma were maintained frozen at - 30°C until used.

Preparation of Rabbit Anti-human Ep Serum IgG Fraction from PRCA Sera Rabbit anti-human Ep serum (anti-Ep) was obtained by a modification (Peschle et al, 1971) of the method reported by Schooley & Garcia (1965). One ml of anti-Ep employed here neutralizes at least 125 i.u. of human standard B Ep (National Institute for Medical Research, London) or 12.5 i.u. of mouse or sheep Ep (Peschle et al, 1974). All PRCA sera were subjected to DEAE-cellulose column chromatography (Sober et al, 1956) and the IgG fraction thereby separated. Standard immunodiffusion and immunoelectrophoresis techniques were employed respectively to quantitate the IgG fractions and establish their purity. The amount of IgG is always expressed in terms of the equivalent volume of original serum. Preparation of Test Materialsfor Assay in Ex-hypoxic Polycythaemic Mice (A)Incubation o f E p with serum P R C A IgG and goat anti-human gamma-globulin ( G A H G G ) (Peschle ct al, 1972). The IgG serum fraction prior to therapy was incubated with Ep (Step I sheep plasma Ep, with a specific activity of 0.5 i.u./mg of protein; Connaught Medical Research Lab., Toronto) in a water bath incubator with constant shaking at 37’C for 30 min (amount of IgG equivalent to that in 0.1 ml of original serum/o.2 i.u. Ep). Thereafter, GAHGG (Antibodies Inc., Davis, California) was added to the mixture for an additional 15 min undcr the above conditions. The precipitate was thereafter discarded by centrifugation and the supernatant was tested in polycythaemic mice. In control vessels, the pre-remission

Pure Red Cell Aplasia 413 IgG serum fraction from another patient with PRCA type A (case No. I : Marmont, 1973 ; Peschle, 1973) was similarly tested. A further control involved incubation of Ep with rabbit anti-human Ep (0.I in1 anti-Ep/o.z i.u.) and then addition of goat anti-rabbit gamma-globulin (GARGG) as described above. The appropriate amount of GARGG or GAHGG was previously ascertained by a two-step incubation: ( I ) known quantities of anti-Ep or IgG were incubated with graded amounts of GARGG or GAHGG respectively; (2) after centrifugation thc supernatant was incubated with Ep and then injected in polycytliaeiiiic mice. Thcsc incubations were performed as described above. (B) Acidijication a d boiling ofwhole PRCA S ~ Y M W Z .Sera . obtained before and after treatment were acidified and boiled as described by Garcia & Scliooley (1971). Thus, samples were acidificd to pH 5.0 with 0.1 N HC1 aiid placed in a boiling bath for 5 niin. They were then cooled, centrifuged for 10 min, and the supernatants were decanted. The precipitates were washed once with an equivalent volume of saline and the supernatants were combined and tested in polycythaemic mice. Control sera wcrc diluted I :I with physiological saline and assayed in polycythaemic mice.

Assny

of Test Materials in Norniaf or Ex-hypoxic

Polycythncmic Mice CF I female mice weighing 20-25 g were employed. The animals were maintained 011 a diet of laboratory pellets and tap water ad libittinr. (A) Norm1 nrice received test materials either intraperitoneally or subcutaneously. Radioiron (0.5 PCi 59Fecitrate in sterile physiological saline) was injected intravenously or intraperitoneally 24 h after the last test injection aiid 24 h per cent RBC-radioiron incorporation values were determined. Blood volume was assumed to be 5 % ofthc body weight (Camiscoli & Gordon, 1970). Mice with a final packed cell volume of less than 0.40 were discarded. (B) Ex-hypoxic polycythaevnic mouse assay (Peschle et al, 1973). The animals were rendered polycythaemic by exposure to intermittent hypoxia (0.42 atmospheres of air) for 18 hiday up to a total of approximately 220 h. Test materials, either incubated or not as described above, were injected either intraperitoneally or subcutaneously starting from day 5 posthypoxia. Radioiron was administered as in normal mice on day 7. 48 h per cent RBC-”Fc incorporation values were determined. The blood volume was assumed to be 7%. of body weight values (Camiscoli & Gordon, 1970). Mice with a final packed cell volume of less than 0.56 were discarded. RESULTS A N D DISCUSSION As indicated in Table I, injection of pre-treatment serum IgG in normal mice induced a significant inhibitory effect on the erythropoietic rate. Similar results have been previously reported (Peschle, 1973; Zalusky et al, 1973). O n the other hand, however, pre- or posttreatment serum showed respectively either absent or elevated Ep activity, in spite of low packed cell volumes at the time of collection (respectively 0.21 and 0.26). This pattern is in sharp contrast to the standard phenomenon observed in PRCA patients, i.c. elevated Ep levels prior to but not after successful treatment (Peschle, 1973; Zalusky et al, 1973). These findings may suggest that in this case the IgG inhibitor functioned as an antibody to circulating Ep. Although the possibility of an anti-Ep inhibitor in PRCA has been previously postulated by Jepson & Lowenstein (1966), no evidence has been provided yet to substantiate

Cesare Peschle et al

414

this hypothesis. Furthermore, the majority of patients examined so far clearly showed an IgG inhibitor directed against the erythroid marrow rather than Ep (Peschle, 1973; Kraiitz & Kao, 1969; Krantz et al, 1973 ;Zalusky et a], 1973). hi this regard, incubation ofPRCA IgG with Ep and then GAHGG led to restoration of the Ep activity, as evaluated in polycythaemic inice (Peschle, 1973; Zalusky ct a!, 1973). This clearly indicated that the Ep molecule did not interact with IgG, which did not function therefore as an antibody to Ep. This finding was TABLE I. Effect of the pre- and post-treatment serum IgG on the erythropoietic rate of normal mice, as evaluated on the basis of 24 h per cent 59Fe incorporation values (mean k SEM) ~

Crottp No.

1

1

Mean % 59Fe incorporation valrres SEM Saline Normal seruin IgG (0.1niliday x 3) Pre-treatment PRCA IgG (0.1 ml/day x 3)

23.14k 1.71 20.59f 1.36 7.11 +o.go*

Saline Post-treatment PRCA IgG (0.1ml/day x 3) Post-treatment PRCA IgG (0.3 ml/day x 3)

25.09 k 2.98 24.74f 1.25 21.87+ 1.37

A minimum of six mice per group. Each mouse received test materials on day and 2, radioiron on day 3. IgG: amount in thc volume of original serum.

* P I >I

I

Standards 6 Saliuc 7 0.05 i.u. Ep 8 0.20 i.u. Ep ~~~~

0.86 _+ 0.03 2.01 0.37 7.53 k 1.04 ~

~

~~

~

~

~

~

~

~

~

~

~

A minimum of six mice per group. N.D. : no detectable activity.

* The scruin w a s collected in the course of the reticulocytosis

following inimuiiosuppressive therapy.

Cesare Peeschk et aE

416

in contrast with the response observed when injecting serum IgG from other PRCA patients (Peschle, 1973 ;Zalusky et a2,1g73), the decrease in packed cell volume was not associated with an inverse rise of Ep titres in serum. At the end of the injection period, an experimental model for PRCA type B was therefore established. Thus, Witebsky's (1958) criteria for the autoimmune pathogenesis of a disease (demonstration of the autoantibody, identification of the autoantigen, experimental model of the disease following injection of the autoantibody) are fully satisfied.

s.

20L

I

0

I

I

I

I

I

~

I

~

I

I

I

10

5

'

I

I

I

~

15

Days

FIG I . Effect in normal mice of prolonged administration of serum IgG (equivalent to 0.05 nil of original serumlday), from either a normal subject (. . .) or the PRCA patient (-) on mean values of (from the top): (I) 24 h per cent 59Feincorporation, (2) Hct and (3) serum Ep levels (evaluated on thc basis of 48 hr per cent 59Fe incorporation values in recipient ex-hypoxic polycythemic mice receiving a total amount of I ml of donor mouse serum). * P