Dumonde and Glynn (1962) have demonstrated that intra-articular injection of autologous fibrin into an animal sensitised against this immunogen produced a.
Br. J. exp. Path. (13970) 51, 81.
THE EFFECT OF ARVIN ON EXPERIMENTAL IMMUNE ARTHRITIS IN RABBITS P. M. FORD, W. R. BELL, R. BLUESTONE, J. M. GUMPEL AND F. W. S. WEBB From the Department of Medicine, Royal Postgraduate Medical School, London, W. 12 Received for
p)ublication
September 12, 1969
SUMMARY.-The production of near total defibrination in rabbits by the use of Arvin, an extract of the venom of the Malayan pit viper (Agkistrodon rhodostoma), failed to influence the development of an experimentally induced chronic immune arthritis in these animals.
THE chronicity of the inflammatory lesion in rheumatoid arthritis in man and in experimental immune arthritis in animals is well recognised. This chronicity may result from autoimmunisation to one or more of the components of the inflammatory process (Glynn 1968). That a chronic arthritis may be induced in this way has been demonstrated experimentally by Phillips, Kaklamanis and Glynn (1966). They immunised rabbits with the exudate of artificially induced inflammatory lesions from the same animal, and were subsequently able to induce a chronic arthritis by intra-articular injection of the same immunogen. Dumonde and Glynn (1962) have demonstrated that intra-articular injection of autologous fibrin into an animal sensitised against this immunogen produced a similar chronic arthritis, although heterologous fibrin produced a more severe response in animals in which it had been used as the immunising agent. Fibrin or one of its degradation products may thus be the autoimmunogen present from the start of the inflammatory process and the subsequent autoimmunisation may then be responsible for the chronicity of the lesion in both experimental immune arthritis in animals and rheumatoid arthritis in man. We sought therefore, to determine whether reduction of the amount of fibrin available to partake in the initial inflammatory reaction would modify in any way either the severity or the chronicity of the subsequent lesion. The animal used was the rabbit and almost total defibrination was attained by i.v. administration of Arvin. Arvin, an extract of the venom of the Malayan pit viper (Agkistrodon rhodostoma), acts by converting the fibrinogen in the blood to fibrin microclots (Regoeczi, Gergely and McFarlane, 1966). These microclots are rapidly phagocytosed by the reticuloendothelial system resulting in a rapid fall in circulating fibrinogen. At the same time the presence of these microclots in the circulation promotes an increase in local fibrinolytic activity in the blood (Regoeczi et al., 1966). If sufficient Arvin is administered almost total defibrination may be achieved in the normal animal and this effect may be maintained at least for a period by subsequent repeated dosage.
82
P. M. FORD, W. R. BELL, R. BLUESTONE, J. M. GUMPEL AND F. W. S. WEBB MATERIALS AND METHODS
Twelve pure strain New, Zealand white rabbits initially weighing between 3-4 kg. were immunised after the method of Dumonde and Glynn (1962) using ovalbumin (5 x crystallised, Koch-light Laboratories Ltd.) instead of the fibrin used by these workers. Freund's complete adjuvant (1 ml.) containing 10 mg. of ovalbumin was injected i.d. in multiple sites between the shoulder blades of each animal. Each animal was skin tested 10 days after immunisation using an i.d. injection of 0.1 ml. of an aqueous solution of ovalbumin containing 100 ug./ml. These tests were read at 24 hr, an area of palpable induration greater than 10 mm. in diameter was taken as being positive. Intra-articular injection was carried out the day after the skin test became positive. Under anaesthesia 1 ml. of a saline solution of ovalbumin containing 10 mg./ml., which had previously been passed through a 0-22 millipore filter (Millipore (U.K.) Ltd.), was then injected into the right knee of each animal, under sterile conditions. The animals were divided into 3 groups of 4; groups I, II and III. The animals of group III in addition to having ovalbumin injected into the right knee had 1 ml. of sterile normal saline injected into the left knee. Group I served as control animals and received no further treatment being killed 28 days after injection of protein into the knee joint. Group II animals were treated for 10 days with Arvin after which time treatment was terminated and the survivors were killed after a further 18 days. Group III animals were treated with Arvin for 28 days, the survivors being killed on the 28th day. Defibrination was induced 12 hr before injection of antigen into the knee joints of animals in groups II and III. This was effected by slow i.v. injection over 20 min. of a saline solution of Arvin (Twyford Laboratories Ltd.) containing 1 unit/ml. The starting dose was 1 unit/kg. body weight. Subsequent injections were given i.v. at 12 hr intervals, the dose being adjusted to maintain satisfactory defibrination. Daily blood samples were taken 12 hr after the preceding Arvin injection and the level of defibrination estimated by the tube titre method of Sharpe, Howie, Biggs and Methuen (1958), aiming to keep the titre at or below 1: 8. In addition plasma fibrinogen was estimated twice weekly using the method of Ratnoff and Menzie (1950) on blood samples from groups II and III. Daily measurements in millimetres of the transverse joint diameters of both knees were made (Fig. 6). Animals dying during the course of the experiment were examined immediately after death. All animals surviving to 4 weeks were killed on the 28th day. The knee joints were opened under sterile conditions and immediately cultured both aerobically and anaerobically. The joints were then opened fully and a sample of synovium was taken from each and fixed in formol saline. The remainder of the joint was also fixed in formol saline and later decalcified. Sections were stained with haematoxylin and eosin and also with Perl's stain for iron. RESULTS
All 4 group I animals survived to 28 days. Of the 4 group II animals one died during the induction of defibrination and a second animal died at 4 days. Of the 4 group III animals one died suddenly at 8 days and a second died at 16 days, 4 days after the onset of hemiplegia and nystagmus. The remaining 2 animals survived to 28 days. Defibrination in the animals of group II was satisfactorily maintained, the tube titre never rising above 1: 8 and normally being 1: 4 or less (tube titres in the control ran at 1: 256 during the first 10 days period). The graph (Fig. 1) shows the fibrinogen values for the group I controls and group III animals over the period of 28 days, and demonstrates a significant degree of hypofibrinogenaemia. Serial joint measurements (Fig. 6) show that the most severe joint swelling occurred in the animals of group III, although a marked increase in joint diameter was noted in all groups. Microscopic examination of the protein injected knee joints of group I at 28 days showed a diffuse plasma cell infiltration of the synovium with lining cell
ARVIN AND EXPERIMENTAL IMMUNE ARTHRITIS
800 I
700
\
I I'
600
I
\
2
-J
E
500
c.'
\NControl group I
400 -i
_
La.'
z
0
z
200
U-
100
co
_
_
300
__-
_
Group III
---
:-I0
0
4
8
12
16
20
24
28
DAYS FIG. 1.-Showing fibrinogen levels in groups I and Ill. 141')
11 I
1: 11
I-
0
z
I,, z
11
I
Iu
4
8
12
16
20
24
28
DAYS
FIG. 6.
Change in knee joint diameter (mm.), showing the
mean
change for each group.
83
84
P. M. FORD, W. R. BELL, R. BLUESTONE, J. M. GUMPEL AND F. W. S. WEBB
hyperplasia and aggregations of lymphocytes (Fig. 2 and 3). The non-injected left knees of this group showed a normal synovium of which Fig. 4 is an example. The injected joints of the animals of group II surviving to 28 days had a similar appearance to those of group I again with lining cell hyperplasia, numerous plasma cells and lymphocytic aggregations (Fig. 5). The non-injected joints of this group showed a normal synovium. The group II animal dying at 4 days showed extensive muscle necrosis in the left quadriceps above the patella indicating that the protein injection had failed to enter the joint; this might possibly have contributed to the early death of this animal. The protein injected knees of all animals in group III including those dying at 8 and 16 days also showed a severe inflammatory reaction with marked lining cell hyperplasia, numerous lymphocytic aggregations and a diffuse infiltration with plasma cells (Fig. 7 and 8). Erosions of the marginal articular cartilage were also noted in the joints of the 2 animals surviving to 28 days (Fig. 9 and 10). A mild but definite synovitis with some lining cell hyperplasia and occasional plasma cells was also noted in the saline injected knees of this group. Sections stained for iron from all groups showed no excess of stainable iron suggesting that excessive haemorrhage had not occurred. Joint cultures both aerobic and anaerobic showed no bacterial growth at 48 hr. DISCUSSION
The results show that almost complete defibrination with Arvin exerts no inhibitory effect on either the acute phase of experimental immune arthritis as assessed by measurements of joint swelling, nor upon the establishment of chronicity as assessed by histological examination. It has been suggested (Barnhart, 1968) that fibrin and its immediate immunological relatives play a central role in the acute inflammatory process. The work of Phillips et al. (1966), who produced an immune arthritis in rabbits by intraarticular injection of autologous inflammatory exudate in animals previously immunised to that exudate, led them to suggest that the chronicity of the inflammatory reaction in human rheumatoid arthritis might be due to a similar process of autoimmunisation. Fearnley, Chakrabarti and Hocking (1965) and Fearnley and Chakrabarti (1966) have demonstrated a significant improvement in the clinical condition of patients with rheumatoid arthritis, following a course of fibrinolysis-inducing therapy with phenformin.
EXPLANATION OF PLATES FIG. 2.- Group I rabbit: lining cell hyperplasia, a lymphocytic aggregation and diffuse plasma cell infiltrate. x 150. FiG. 3. Group I rabbit: plasma cell infiltration of the synovial membrane. X 430. FIG. 4.-Group I rabbit: normal synovial membrane of non-injected knee joint. x 265. FIG. 5. Group II rabbit: dense plasma cell infiltration around a lymphocytic aggregation, with an area of fibrin in the membrane. x 265. FlG. 7.-Group III rabbit: severe plasma cell infiltration with lymphocytic aggregations. x 65. FIG. 8. Group III rabbit: synovial membrane hyperplasia with plasma cell infiltrate and young fibroblasts. x 265. FIG_. 9. Group III rabbit: marginal erosion showing an osteoclast and the edge of the plasma cell infiltrate. x 265. Fi(n. 10. Group III rabbit: erosion of cartilage margin by inflammatory infiltrate. x 65.
BRITISH JOURNAL
OF
EXPERIMENTAL PATHOLOGY.
.w:lp'
_
Vol. LI, No. 1.
5
I.
P,
J,I ...4
i
t. 4t
f; ~~~~~~~~~~~*.
f.
.W
.fs
,l.~~N
**
.
,*5
6-
I
Ford, Bell, Bluestone, Gumpol and Webb.
BRITISH JOURNAL
OF
EXPERIMENTAL PATHOLOGY.
Vol. LI, No. 1.
.
Do..;. *
:. Ri
A
I
ws.q A
:s
:1 .
I;
Ford, Bell, Bluestone, Gumpel and Webb.
ARVIN AND EXPERIMENTAL IMMUNE ARTHRITIS
85
The possibility arose therefore, that the significant autoimmunogen might be the animal's own fibrin. If this were so, removal of fibrin available to take part in the acute process might be expected to modify the process in both the acute and chronic stages. Arvin, which has been reported as being used in the treatment of rheumatoid arthritis (Pitney, Bray, Holt and Bolton (1969)), acts in 2 ways which might mediate an anti-inflammatory role. Firstly it precipitates fibrinogen in the form of fibrin microclots which are then rapidly removed by the reticuloendothelial system, and secondly it promotes increased local blood fibrinolytic activity (Regoeezi et al., 1966). The histological lesion in experimental immune arthritis resembles closely in its main features that of rheumatoid arthritis in humans (Dumonde and Glynn, 1962) and it would seem reasonable to suppose that failure of Arvin to influence the histological picture in the animal model, might indicate that a similar failure could be expected if used in human rheumatoid arthritis. The occurrence of a mild synovitis in the saline injected knees of the group III animals was an unexpected finding. It has been postulated however (Glynn: personal communication) that this might be caused by small quantities of antigen from the protein injected knee being released into the circulation and taken up by the synovium of the saline injected knee joint during the period of transitory synovial inflammation which is known to occur following injection of saline into knee joints (Hollingsworth, 1965). The absence of any reaction in the noninjected knees of the animals in group II would argue against this being a direct effect of the Arvin. We were unable to demonstrate the presence of excessive amounts of iron pigment which would be expected to be present if extensive bleeding were responsible for these changes (Roy and Ghadially, 1969). It is probably not possible to attain total defibrination in vivo using Arvin, but as may be seen from Graph I not only was the plasma fibrinogen level markedly reduced at the onset of the acute inflammatory response but the rise which occurred in the controls following injection of antigen was completely prevented in the treated group III. In the animals of group III the amount of Arvin required to maintain a satisfactory level of defibrination (i.e. a thrombin tube titre of 1: 8 or less) rose during the course of treatment and in the 2 animals surviving for 28 days was 31 times the starting dose during the last 7 days. Arvin is a protein, the molecular weight of the monomer form being 30,000 (Esnouf and Tunnah, 1967) and might be expected to be antigenic. Resistance to Arvin is known to develop in man (Pitney et al., 1969) and a possible explanation of the increased dose requirements in this study is the formation of Arvin antibodies, but such antibodies were not looked for. In conclusion, therefore, we were unable to demonstrate any inhibitory effect of reduction of available fibrin on either the externally observable features of the acute inflammatory response or upon the ultimate histological picture in experimental immune arthritis in rabbits. This suggests that treatment with Arvin of rheumatoid arthritis in humans is unlikely to be of benefit. We can not, however, completely exclude the possibility that autoimmunisation to fibrin in the inflammatory process may play a part in the establishment of a chronic inflammatory reaction as only minute quantities of fibrin may be required and it was not possible to obtain total defibrination in the animals used. It is also possible that autoimmunisation to some product of the inflammatory process does not occur for some months, the earlier reaction being maintained by persisting traces of the 8
86
P. M. FORD, W. R. BELL, R. BLUESTONE, J. M. GUMPEL AND F. W. S. WEBB
initiating foreign antigen. In which case treatment with Arvin, to be effective, would have to be continued for several months. We are grateful to Dr. L. E. Glynn and Dr. W. R. Pitney for their helpful advice, to Twyford Laboratories Ltd., for supplies of Arvin, and also to Professor E. G. L. Bywaters for his interest. REFERENCES BARNHART, J. I.-(1968) Biochem. Pharmacol., 205 (March special supplement). DUMONDE, D. C. AND GLYNN, L. E.-(1962) Br. J. exp. Path., 43, 373. ESNOUF, M. P. AND TUNNAH, G. W.-(1967) Br. J. Haemrt., 13, 581. FEARNLEY, G. R. AND CHAKRABARTI, R.-(1966) Lancet, ii, 757. FEARNLEY, G. R., CHAKRABARTI, R. AND HOCKING, E. D.-(1965) Lancet, i, 9. GLYNN, L. E.-(1968) Ann. rheum. Dis., 27, 105. HOLLINGSWORTH, J. W.-(1965) Yale J. biol. Med., 37, 300. PHILIPS, J. M., KARLAMANIS, P. AND GLYNN, L. E.-(1966) Ann. rheum. Dis., 25, 165. PITNEY, W. R., BRAY, C., HOLT, P. J. L. AND BOLTON, G.-(1969) Lancet, i, 79. RATNOFF, 0. D. AND MENZIE, C.-(1951) J. Lab. clin. med., 37, 316. REGOECZI, E., GERGELY, J. AND McFARLANE, A. S.-(1966) J. clin. Invest., 45, 1202. Roy, S. AND GHADIALLY, F. N.-(1969) Ann. rheum. Dis., 28, 402. SHARPE, A. A., HOwIE, B., BIGGS, R. AND METHUEN, D. J.-(1958) Lancet, ii, 1309.
