996

Changes in protein conformation or in electrical chargedue to glycosylation might induce an increase inintemalisation or enhanced passage through intercellularclefts. Glycosylated albumin escape might be due to fluid oradsorptive endocytosis or both; however, from our study, itseems reasonable to exclude a mechanism modulated byreceptor competition, since the native albumin of the

recipient animal could strongly compete with glycosylatedalbumin.

It has been suggested that serum albumin concentrationmodulates vascular permeability, low levels favouringextravascular escape of molecules.12 In our experimentalsetting, an increase of serum albumin 5 or 10 mg above the

physiological level, by earlier injection of native albumin,did not prevent the glycosylated albumin extravasation.Other workers13 have suggested that albumin may modulatethe vascular permeability by narrowing the endothelialpores. However, if this modulation occurs, it does not

prevent glycosylated albumin extravasation.Overglycosylation of albumin occurs in diabetes: our

evidence of the extravascular leakage of glycosylatedalbumin from the venous microvasculature is compatiblewith the hypothesis that albumin overglycation initiatesthe chain of events leading to full-blown diabeticmicroangiopathy.

Correspondence should be addressed to T. S., CNR Institute of ClinicalPhysiology, Via Savi 8, 56100-Pisa, Italy.

REFERENCES

1. Cunha-Vaz JG. Studies on the pathophysiology of diabetic retinopathy. Theblood-retinal barrier in diabetes. Diabetes 1983; 32 (suppl 2): 20-27.

2 Lorenzi M, Healy DP, Hawkins R, Printz JM. Studies on the permeability of theblood-brain barrier in experimental diabetes. Diabetologia 1986, 29: 58-62

3. Kilzer P, Chang K, Marvel J, et al. Albumin permeation of new vessels is increased indiabetic rats. Diabetes 1985; 34: 333-36.

4. Day JF, Thorpe SR, Baynes JW. Nonenzymatically glycosylated albumin. In vitropreparation and isolation from normal human serum. J Biol Chem 1979; 254:595-97.

5. Shaklay N, Garlick RL, Bunn HF. Nonenzymatic glycosylation of human serumalbumin alters its conformation and function. J Biol Chem 1984; 259: 3812-17.

6. Williams SK, Devenny JJ, Bitensky MW. Micropinocytic ingestion of glycosylatedalbumin by isolated microvessels: possible role in pathogenesis of diabeticmicroangiopathy. Proc Natl Acad Sci USA 1981; 78: 2393-97

7. Colantuoni A, Bertuglia S, Berardi PC, Cimini V. Pancreatic islet and microcirculationdamage in normal glycemic alloxan-treated hamsters. Bibliotheca Anat 1981, 20:373.

8. Sampietro T, Lenzi S, Cecchetti P, et al. Nonenzymatic glycation of human plateletmembrane proteins in vitro and in vivo. Clin Chem 1986; 327: 1328-31.

9. Laemmli UK Cleavage of structural proteins during the assembly of the head ofbacteriophage T4 Nature 1970, 227: 680-85.

10. Arfors KE, Rutili G, Svensjo E. Microvascular transport of macromolecules in normaland inflammatory conditions. Acta Physiol Scand 1979; 463: 93-103.

11. Villaschi S, Johns L, Cirigliano M, Pietra GG. Binding and uptake of native andglycosylated albumin-gold complexes in perfused rat lungs. Microvasc Res 1986,32: 190-99.

12. Huxley VH, Curry FE. Albumin modulation of capillary permeability: test of anadsorption mechanism. Am J Physiol 1985; 248: H264-73.

13. Schneeberger EE, Hamelin M. Interaction of serum proteins with lung endothelialglycocalyx: its effect on endothelial permeability. Am J Physiol 1984; 247: H206-17

Hypothesis

IS CYTOMEGALOVIRUS INTERSTITIALPNEUMONITIS IN TRANSPLANT RECIPIENTSAN IMMUNOPATHOLOGICAL CONDITION?

JANE E. GRUNDY1 JOHN D. SHANLEY2PAUL D. GRIFFITHS1

Department of Virology, Royal Free Hospital School of Medicine,London NW3 2QG,1 and Veterans Administration Medical Centre,

Newington, Connecticut 06111 USA2

Summary The conventional explanation for the highfatality rate due to cytomegalovirus (CMV)

pneumonitis among allogeneic transplant recipients is thatimmunosuppression renders the host unable to controlreplication of this opportunistic agent. However, evidencefrom studies in man and the murine model of CMV showthat virus replication in the lung is unrelated to the

development of pathological effects, and that a host immuneresponse is required for the induction of pneumonitis. Thusthe hypothesis is that limited CMV replication in the lungsleads to display of a virus-coded protein, which is recognisedby host T cells, and that the pneumonitis is due to anuncontrolled accumulation and recruitment of such cells inthe lungs. The reason why CMV is found in the lungs ofpatients with the acquired immunodeficiency syndrome(AIDS) without producing pneumonitis is probablybecause these patients cannot mount the pathogenic T-cellresponse. According to the hypothesis stated here, if theimmune capabilities of AIDS patients can be restored,life-threatening CMV pneumonitis may develop.

INTRODUCTION

CYTOMEGALOVIRUS (CMV) infection is a frequent causeof morbidity and mortality in recipients of renal and bone-marrow transplants12 and in patients with the acquired

immunodeficiency syndrome (AIDS). Serious CMV

infection in transplant patients often affects the lungs, toproduce interstitial pneumonitis and death. The pathogen-esis of CMV interstitial pneumonitis is not well understood.However, clinical observations and our laboratory studieshave led us to conclude that in transplant recipients CMVpneumonitis is an immunopathological condition.

CLINICAL OBSERVATIONS

In patients receiving allogeneic bone marrow, graft-versus-host (GvH) disease is often associated with the

development of CMV pneumonitis.1 45 In contrast, despitea similar frequency of CMV infections,’ no cases of CMVpneumonitis were found among 100 recipients of syngeneicbone marrow from identical twins, where no GvH reactionis expected to occur. This discrepancy suggests someinteraction between the immunological reactions involvedin GvH disease and the development of CMV pneumonitis,or in the aggravation GvH disease by CMV infection.

Predisposing factors in renal transplant recipients may bedifferent from that in bone-marrow transplant patientsbecause the procedure does not involve the transfer ofimmunocompetent cells. Moreover, the renal graft is oftenthe vehicle of CMV transmission6-9 irrespective of therecipient’s immune status. CMV pneumonitis is commonerin recipients with primary infection (27% in our centre)than in seropositive recipients (15%), most of whomundergo reinfection,9 and in our centre pneumonitis has notbeen observed in seropositive patients with proven CMVreactivations (unpublished). Thus the finding that primaryinfection or reinfection seems to be associated with CMVpneumonitis more often than in reactivation of pre-existinginfection suggests that prior strain-specific immunity hassome protective effect. The use of antilymphocyte globulinor of high doses of steroids has been reported to be a riskfactor for CMV pneumonitis in patients receiving renaltransplants.10,11

997

EVIDENCE FROM RESPONSES TO TREATMENT OF CMV

PNEUMONITIS

Several antiviral agents have been used without success totreat CMV pneumonitis. Ganciclovir given to bone-marrowtransplant recipients with CMV pneumonitis reduced theCMV titre in the lungs by more than 99-99% but 9 of the10 treated patients died from their pneumonitis, whichsuggests a poor correlation between titres of virus in the lungand the severity of pneumonitis in these patients. In contrastganciclovir therapy seemed to be effective in the treatment ofCMV pneumonitis and retinitis in AIDS patients,13-15although continued therapy was necessary to preventrecurrence.

CMV hyperimmune globulin has been more effective inthe treatment of CMV pneumonitis in transplant recipients(refs 16, 17, and Blacklock HA, Griffiths PD, Prentice HGet al, unpublished)--in an open study of 18 patients withestablished CMV pneumonitis treated with ’Cytotect’hyperimmune globulin, 9 patients survived (Blacklock HA,Griffiths PD, Prentice HG, et al, unpublished). Thebeneficial effects of CMV hyperimmune globulin have beenattributed to its ability to neutralise CMV. However, it isdifficult to understand why the elimination of CMV in vivoby neutralising antibody should be more effective at

inducing recovery from CMV pneumonitis than elim-ination of virus by an antiviral agent. Furthermore, we haveshown that the CMV hyperimmune globulin used to treatthe patients could not neutralise CMV as it is found in

vivo,18 so this mechanism is unlikely. CMV has a strongcapacity to bind the host protein, (32 microglobulin (&bgr;2m).19CMV, when purified directly from body fluids, is "coated"with (32m, which masks the viral antigenic determinantsrecognised by murine monoclonal antibodies182O or humanimmune será,18 thus preventing its neutralisation. Theeffectiveness of CMV hyperimmune globulin in the

treatment of patients with CMV infection is thus due tosome mechanism other than neutralisation of the virus.

LABORATORY STUDIES

We have developed three models of murine CMV(MCMV) pneumonitis which have similar features to CMVpneumonitis in transplant recipients.

GvH Plus MCMV Model

GvH can be induced in unirradiated Fl hybrid mice byinjection of parental spleen cells. When such a GvH

challenge occurs in the presence of acute MCMV infection,the severity of the GvH reaction is augmented.l2 Moreover,a fatal diffuse pneumonitis develops, not seen when eitherMCMV or GvH was given alone.21 Remarkably, virus titresin the lungs of mice with MCMV and GvH were similar tothose in mice given MCMV alone. Thus viral replicationper se is not sufficient for the development of pneumonitis;some facet of the immunological response involved in theGvH reaction is also required. Furthermore, there is a poorcorrelation between the degree of viral replication and thepresence of pneumonitis. In mice given a GvH challengeplus MCMV ganciclovir totally eliminated detectable virusin the lung but had no effect on the development or severityof pneumonitis.22 Thus, the extent of lung virus replication;and lung pathology are not closely correlated in this model.

Cells removed from the lungs of the MCMV-GvH micewith pneumonitis were mainly Thy 1-2 lymphocytes,predominantly of donor origin.23 Thus, much of the

pathological effects is associated with infiltration of donorlymphocytes into the alveolar septae with resultantconsolidation of the lungs.

Cyclophosphamide Model

In BALB/c mice one dose of cyclophosphamide giventogether with MCMV leads to the development of aninterstitial pneumonitis, not seen with MCMV or cyclo-phosphamide alone.24 However, if cyclophosphamide is

given continuously, then pneumonitis does not developdespite an increase in viral titres in the lung.24 Thus somecomponent in the host response which is suppressed bycontinued treatment with cyclophosphamide is importantfor the development of MCMV pneumonitis. Similarly,administration of anti Thy 12 monoclonal antibody todeplete T cells prevents the development of pneumonitis.2$On the other hand, treatment of the animals with ganciclovirfollowing administration of MCMV and one dose ofcyclophosphamide did not prevent the emergence of

pneumonitis despite the reduction of MCMV replication toundetectable levels. Thus as with the GvH model, repli-cation of MCMV in the lungs did not correlate withpneumonitis. Analysis of lung viral replication and

development of pneumonitis using a variety of strains ofmice 26 indicated that the genes controlling susceptibility tolung virus replication segregated in a different way fromthose controlling the development of pneumonitis.

Nude Mouse Model

MCMV given to T-cell deficient athymic nude micereplicates extensively in the lung25 but diffuse pneumonitisdoes not develop until the terminal stages of infection, evenwhen cyclophosphamide is given. When T-cell function isreconstituted with syngeneic cells before the mice are

challenged with MCMV and cyclophosphamide, pneu-monitis develops.25 Thus some component of the T cellimmune response and not viral replication per se is

important in the induction of pneumonitis by MCMV inthis model.

INTERPRETATION OF CLINICAL AND EXPERIMENTAL

FINDINGS

The clinical observations concerning pneumonitis in

bone-marrow transplant recipients and in the three experi-mental models of MCMV-associated pneumonitis have twocommon features: (1) a lack of correlation between lung viralreplication and the development of pneumonitis and (2) theinvolvement of some component of the immune response inthe development of the pneumonitis. In bone-marrow

transplant recipients and in the murine GvH-MCMVmodel the immunological component required seems to bepart of the graft-versus-host reaction. In renal transplantrecipients it is not clear whether the allogenic graft, theimmunosuppression, or both are important. The cyclo-phosphamide model suggests that a slight disturbance of theimmune system associated with the administration of onedose of cyclophosphamide is enough to induce pneumonitisin animals with MCMV infection. However, total

immunosuppression, due to continued cyclophosphamidetreatment or as seen in immunodeficient nude mice, was notassociated with pneumonitis. Kidney transplant recipientsare not as heavily immunosuppressed as are bone-marrowrecipients and may have more resemblance to the cyclo-phosphamide mouse model. It should be noted that CMV

998

pneumonitis develops in bone-marrow recipients only whenthey have a good marrow graft, and this is often many weeksafter they first start to excrete CMV.

PROPOSAL FOR THE MECHANISM OF INDUCTION OF CMV

PNEUMONITIS

On the basis of the two features of CMV pneumonitisdiscussed above and the ineffectiveness of antiviral agents inthe treatment of the condition, we propose that theinterstitial pneumonitis associated with CMV infection intransplant recipients is an immunopathogenetic process dueto a T-cell response to a virally induced antigen expressed oncells in the lung; candidate antigens are a CMV immediateearly (ot) or early (p) gene product.

This hypothesis can account for the lack of efficacy ofantiviral agents since they only block viral DNA repli-cation,27 which occurs after expression of 3( and p geneproducts. It also accounts for the lack of correlation of

pneumonitis with viral titres, since complete viral particleswould not be necessary for the induction of the pneumonitis.The requirement for a component of the immune

response in the development of CMV pneumonitis forms acentral feature of our hypothesis. The target antigen couldbe an a or j3 protein since the main target antigens for T cellson CMV infected cells, at least as far as cytotoxic T cells areconcerned, have been shown to be a gene products withboth the human28 and murine29 viruses. We do not know yetwhether the subset of the T cell thought to be involved in thedevelopment of CMV pneumonitis is a cytotoxic T cell, ahelper T cell, or a delayed type hypersensitivity T cell;however preliminary studies with mice favour delayed typehypersensitivity T cells. It is of course possible thatinteractions between the various subsets are involved.How CMV hyperimmune globulin works can be

explained in two ways. Firstly, the anti-CMV component ofthe hyperimmune globulin could block the CMV antigenson infected cells, towards which T cells are reactive, either atthe induction or effector stage. Thus it is the pathologicaleffect which is inhibited rather than virus replication.Secondly, since the hyperimmune globulin is prepared froma large pool of donors it would also contain anti-HLAantibodies derived from women who have had previouspregnancies, and this anti-HLA component may blockrecognition of viral antigens in association with recipientHLA antigens, again either at the induction or effectorphase. This second possibility should not be overlookedsince in some other virus systems antiviral monoclonalantibodies do not block cytotoxic T-cell recognition of virusinfected cells30 but antibodies directed towards majorhistocompatibility complex (MHC) antigens do .31

The following are possible explanations for the associ-ation between CMV pneumonitis and the GvH reaction.

(1) The GvH reaction is associated with enhanced expressionof class I and II MHC antigens in a variety of tissues. Since classI HLA molecules can function as a receptor for CMV,32 thiswould result in an increased susceptibility of cells to CMVinfection. The enhanced display of class I and II HLA antigensin association with CMV proteins could stimulate an aggressiveT-cell response against infected cells and increase the likelihoodof CMV pneumonitis.

(2) Our in-vitro studies have shown that CMV infection itselfenhances expression of class I HLA antigens" (Grundy JE,Ayles HM, McKeating JA, Butcher RF, Griffiths PD, PoulterLW, unpublished). Thus CMV infection could stimulate the

GvH reaction by enhancing those recipient HLA antigenstowards which donor lymphocytes are reactive.

(3) The GvH reaction is accompanied by polyclonalactivation of both T and B lymphocytes. Furthermore,alloantigens activate T-cells with low-affinity receptors. Thusduring the GvH reaction there may be expansion of clones of T-cells of donor origin which have receptors capable of low-affinity interaction with recipient MHC molecules in

association with CMV antigens.

In renal transplant recipients alloreactive cells are gener-ated by exposure to the foreign graft. Class I and II HLAantigen expression is also enhanced during the host-versus-graft reaction. Thus renal transplant recipients may becomeprone to the development of pneumonitis for reasons similarto those outlined in points (1) and (3) for bone-marrowtransplant recipients.

In AIDS patients the evidence that CMV induces aninterstitial pneumonitis similar in pathology to that seen intransplant recipients is lacking. We believe that in someAIDS patients lung damage may be due directly to thereplication of CMV. This belief is supported by the

observations in the nude mouse model. Furthermore, thisproposal is consistent with the apparent effectiveness ofcontinued ganciclovir therapy in the treatment of pul-monary CMV infection in some AIDS patients, and withthe known lack of immune responsiveness of these patients.In our clinical experience CMV does not seem to be animportant pathogen when found in the lungs of AIDSpatients (unpublished).

IMPLICATIONS FOR THERAPY

Our hypothesis suggests that the search for drugs whichare more effective at reducing CMV replication in vivo maynot help towards improving the treatment of establishedCMV pneumonitis. A combination of immunosuppressivetherapy, combined with an antiviral drug to preventoverwhelming virus dissemination and CMV disease inother organs, may be beneficial. Prophylactic use of antiviraldrugs to prevent the development of lung infection would beanother strategy. The identification of the viral x or )3protein, or peptide fragment thereof, responsible for theT-cell recognition of infected cells could lead to a rationalmeans of screening blood donors for antibodies reactive withthis protein, for use as blocking antibodies in the treatmentof established CMV pneumonitis.

Patients with AIDS often have CMV in the lungs but,presumably because of their inability to mount the requiredT-cell response, do not develop the interstitial pneumonitisseen in transplant recipients. However, if treatment ofhuman immunodeficiency virus infection successfully leadsto correction of this T-cell defect, then according to ourhypothesis, CMV pneumonitis will become a majorproblem in AIDS patients.

Correspondence should be addressed to J. E. G.

REFERENCES

1 Meyers JD, Fournoy N, Thomas ED Risk factors for cytomegalovirus infection afterhuman marrow transplantation J Infect Dis 1986, 153: 478-88

2 Rubin RH, Tolkoff-Rubin NE, Oliv er D, et al Multicenter seroepidemiologic studyof the impact of cytomegalovirus infection on renal transplantationTransplantation 1985; 40: 243-19.

3 Sonnabend J, Witkin SS, Purtilo DT. Acquired immunodeficiency syndrome,opportunistic infections, and malignancies in male homosexuals JAMA 1983, 249:2370-74

999

4. Appelbaum FR, Meyers JD, Fefer A, et al Nonbacterial nonfungal pneumoniafollowing marrow transplantation in 100 identical twins Tr ansplantation 1982, 33:265-68

5 Miller W, Flynn P, McCulbugh J, et al Cytomegalovirus infection after bone marrowtransplantation An association with acute graft-vs-host disease Blood 1986; 67:1162-67.

6 Ho M, Suwansirikul S, Dowling JN, Youngblood LA, Armstrong JA. Thetransplanted kidney as a source of CMV infection N Engl J Med 1975; 293:1109-12

7 Wertheim P, Buurman C, Geelen J, van der Noorda J. Transmission of

cytomegalovirus by renal allograft demonstrated by restriction enzyme analysis.Lancet 1983; i 981-82

8 Grundy JE, Super M, Griffiths PD. Reinfection of a seropositive allograft recipient bycytomegalovirus from donor kidney Lancet 1986; i 159-60

9 Grundy JE, Super M, Lui S, Sweny P, Griffiths PD The source of cytomegalovirusin seropositive renal allograft recipients is frequently the donor kidney TransplantProc 1987; 19: 2136-28

10 Pass RF, Whitley RJ, Diethelm AG, Whelchel JD, Reynolds DW, Alford CA.Cytomegalovirus infection in patients with renal transplants: Potentiation byantithymocyte globulin and an incompatible graft. J Infect Dis 1980; 142: 9-17.

11 Rubin RH, Cosimi AB, Hirsch MS, et al. Effects of antithymocyte globulin oncytomegalovirus infection in renal transplant recipients. Transplantation 1981; 31:143-15.

12 Shepp DH, Dandliker PS, de Miranda P, et al. Activity of 9-[2-hydroxy-1-Hydroxymethyl) ethoxymethyl] guanine in the treatment of cytomegaloviruspneumonia Ann Intern Med 1985; 103: 368-73.

13. Collaborative DHPG Treatment Study Group. Treatment of serious cytomegalovirusinfections with 9-(1,3-dihydroxy-2-propoxymethyl) guanine in patients with

AIDS and other immunodeficiencies. N Engl J Med 1986; 314: 801-05.14 Masur H, Lane HC, Palestine A, et al Effect of 9-(1,3-Dihydroxy-2-Propoxymethyl)

Guanine on serious cytomegalovirus disease in eight immunosuppressedhomosexual men Ann Intern Med 1986; 104: 41-44

15. Laskin OL, Stahl-Bayliss CM, Kalman CM, Rosecan LR. Use of ganciclovir to treatserious cytomegalovirus infections in patients with AIDS J Infect Dis 1987; 155:323-27

16. Blacklock HA, Griffiths PD, Stirk PR, Prentice HG. Specific hyperimmune globulinfor cytomegalovirus pneumonitis. Lancet 1985, ii 152-53.

17. Blacklock HA, Griffiths PD, Stirk PR, Prentice HG. Successful treatment

of cytomegalovirus pneumonitis after allogeneic bone marrow transplantationusing high titre CMV immunoglobulin (Cytotext). Exp Haematol 1985; 13(suppl 17): 76.

18. McKeating JA, Griffiths PD, Grundy JE. Cytomegalovirus in urine specimens has

host &bgr;2-microglobulin bound to the viral envelope A mechanism of evading thehost immune response? J Gen Virol 1987, 68: 785-92.

19 Grundy JE, McKeating JA, Griffiths PD Cytomegalovirus strain AD169 binds &bgr;2microglobulin in vitro after release from cells J Gen Virol 1987; 68: 777-84

20. McKeating JA, Grundy JE, Vargheze Z, Griffiths PD Detection of cytomegalovirusby ELISA in urine samples is inhibited by &bgr;2 microglobulin J Med Virol 1986; 18:341-48

21. Grundy JE, Shanley JD, Shearer GM Augmentation of graft-versus-host reaction bycytomegalovirus infection resulting in interstitial pneumonitis. Transplantation1985; 39: 548-53

22. Shanley JD. The effect of antiviral treatment on interstitial pneumonitis during acutecytomegalovirus infection and graft versus host reaction. Abstracts of the 27thInterscience Conference on Antimicrobial Agent and Chemotherapy Conference,New York, 1987.

23 Shanley JD, Via CS, Sharrow SO, Shearer GM. Interstitial pneumonitis duringcombined murine cytomegalovirus infection and graft-vs-host reaction;characterization of bronchoalveolar cells Transplantation (in press).

24. Shanley JD, Pesanti EL, Nugent KM. The pathogenesis of pneumonitis due tomurine cytomegalovirus. J Infect Dis 1982; 146: 388-96

25 Shanley JD. Murine cytomegalovirus pneumonitis in T-cell deficient nude mice. InAbstracts of the 12th Herpesvirus Workshop, Philadelphia, 1987

26. Shanley JD, Pesanti EL. The relationship of viral replication to interstitial

pneumonitis in murine cytomegalovirus lung infection J Infect Dis 1985; 151:454-58.

27. Tocci MJ, Livelli TJ, Perry HC, Crumpacker CS, Field AK Effect of the nucleosideanolog 2’-nor2’ - deoxyguanosine on human cytomegalovrirus replication.Antimicrob Ag Chemother 1984, 25: 247-52

28 Borysiewicz LK, Morris S, Page JD, Sissons JGP. Human cytomegalovirus-specificcytotoxic T-ly mphocytes requirements for in vitro generation and specificity. EurJ Immunol 1983; 13: 804-09

29 Reddehase MJ, Kosdzinowski UH Significance of herpesvirus immediate early geneexpression in cellular immunity to cytomegalovirus infection Nature (London)1984; 312: 369-71.

30. Dongworth DW, McMichael AJ Inhibition of human T-lymphocyte function withmonoclonal antibodies. Br Med Bull 1984; 40: 254.

31. McMichael AJ, Parham P, Brodsky FM, Pitch JR Influenza specific cytotoxicT-lymphocytes recognise HLA molecules, blocking by anti-HLA antibodies.J Exp Med 1980; 152: 195S-203S.

32. Grundy JE, McKeating JA, Ward PD, Sanderson AR, Griffiths PD &bgr;2 microglobulinenhances the infectivity of cytomegalovirus and when bound to the virus enablesClass I HLA molecules to be used as a virus receptor. J Gen Virol 1987, 68:793-803.

Reviews of Books

Acting as Friends: The Story of the SamaritansMichael De-la-Noy. London: Constable. 1987. Pp 235. 10.95.ISBN 0-094675007.

I

THERE is a thought-provoking ambiguity about the titleof this work. To which school of acting do the Samaritansadhere?, one asks. That of Coquelin, whose coldly calculatedtechnique could spellbind his audiences, invoking theirlaughter, horror, or tears, while he remained emotionallydetached and watchful? Or that of Stanislavsky, whose aimabove all was as complete identification as possible withcharacters in their differing plights? The Samaritans, oneinfers from De-la-Noy’s book, accept both disciplines turnand turn about, acting sometimes in one way and sometimesthe other; and often a bit of both, according to the

personality and the mood of the "befriender", and thedegree of his involvement. It can only be a one-wayfriendship: there is no quid pro quo. The Samaritan "acts"as a friend.This book gives a unique and comprehensive picture of

the Samaritans-warts and all. It traces the movement fromits modest experimental inception in 1935 by a clergyman,the Rev Chad Varah, in the vestry of a London church, to itspresent pride of 182 branches scattered around GreatBritain and Eire, involving a fellowship of more than 20 000unpaid voluntary workers and attracting over two millioncalls a year. The author, who is not a Samaritan himself,marvels how, despite many blunders, great and small, the

I

movement has caught the public’s imagination, has grown,and is still growing, not only in this country but also, as

Befrienders International, throughout the world. He hasinterviewed everyone concerned-not only the generalsecretaries, the chairman of the executive committee,regional representatives, directors of branches, and manyrandom volunteers, but also a great number of anonymous"ex-callers" whose testimony of the saving help they have(or sometimes have not) received-and he has faithfullyrecorded his findings. Indeed, verbatim quotations take upmuch of the text.

Blunders are inevitable when each branch is so largelyautonomous. Branches are not planted out here and there bysome higher authority. They crop up wherever a sufficientlylarge group of public-spirited individuals feel the need for alocal branch and are prepared to shoulder the burden oforganising it. So long as the Seven Principles and SevenPractices are strictly observed (which include 100%

confidentiality), other matters, such as the selection andpreparation of would-be volunteers and the establishment of24-hour telephone manning at the centre, are left to thediscretion of the branch’s directorate. Guidelines are

provided but there is no regimentation. Hence there are notwo branches with exactly the same procedures. And yet,despite this apparent laxity, the movement holds togetherand flourishes. The secret appears to be fellowship.Samaritans meet frequently to discuss their doubts andanxieties, and thus exchange ideas and give each othersupport.The Samaritans make no claim to be reducing the

suicide-rate, which in England and Wales hovers around 8-2per 100 000 of the population. Of anecdotal evidence inindividual cases there is abundance; but the callers’

anonymity precludes any statistical computation. It is

found, however, that about 40% of all first-time callers have