Viral aetiology of Paget's disease of bone: a review'
6
Journal of the Royal Society of Medicine Volume 77 November 1984 943 Viral aetiology of Paget's disease of bone: a review' L Harvey MA MRCPath Department of Pathology, Queen's Medical Centre, Nottingham NG7 2UH Introduction Sir James Paget published the first comprehensive description of an unusual bony deformity, illustrated by its clinicopathological features in 5 patients, in 1877. His personal term for the condition, based on macro- and microscopic examination of the affected bones, was 'osteitis deformans'. It has subsequently become universally known as Paget's disease of bone. The aetiology of the disease, however, remains an enigma and its treatment, discussed by Russell (1979), is symptomatic or palliative rather than curative (although some drug regimens may reverse or even normalize many of the observed anomalies). Several theories for its pathogenesis (Table 1) have arisen from a consideration and integration of the known clinical, epidemiological, radiological, biochemical and pathological features of the disorder, including Paget's own theory of a chronic inflammation. The observed radiological progression of disease in single (long) bones coupled with results from research into Pagetic bone cell activity indicate that, locally at least, disease begins with a 'wave' of non-physiological osteoclast-mediated osteolysis. It is now widely accepted that Paget's disease is, pathogenetically, an example of a primary osteoclast dysfunction. Consequently research has focused on the control mechanisms involved in normal or abnormal activity of this cell in an attempt to elucidate the pathological trigger factors. An unexpected observation was of an ultrastructural anomaly which has led to the most recent proposal of a viral aetiology for Paget's disease of bone. Osteoclast abnormality Rebel et al. (1974) were the first to publish, in French, observations on the atypical ultrastructure of certain bone cells derived from Pagetic biopsies. These workers demonstrated the presence of tubulofilamentous paracrystalline particles in the osteoclast nuclei and cytoplasm. The individual filaments had a measured diameter of 15 nm with a central lucent zone approximately 7 nm in diameter, seen both transversely and longitudinally. Goniometrically, the particle filaments retained these same measurements (Malkani et al. 1976), which in conjunction with other control observations confirmed their 3-dimensional nature and that they were not processing or technical artefacts. Such inclusions were noted only in osteoclasts and were absent from other bone, connective tissue or haemopoietic cells present in the biopsies. They were noted in every case of Pagetic bone examined. Examples of these structures (from the author's series) are shown in Figure 1. These findings were published in English only in 1976 (Rebel et al. 1976). By this time they had been confirmed by Mills & Singer (1976) in the USA and subsequently by Schulz et al. (1977) in Germany, Gherardi et al. (1980) in Italy and Harvey et al. (1982) in the UK. Reports from such geographically disparate groups mirrored and emphasized known epidemiological aspects of the disease (Table 2). Similarly, all agreed that only the Pagetic osteoclasts contained the observed inclusions and that every single case of Paget's disease was positive. More recently, Mills et al. (1979) demonstrated the presence of these inclusions in bone tissue culture, and Harvey et al. (1982) showed in a preliminary survey that the biopsy inclusion-body load may be related to disease severity as manifested locally. The latter observation is presently being reassessed using more strictly defined 'Based on paper read to Section of Pathology, 15 November 1983. Accepted 9 August 1984 0 141-0768/84/110943-06/$O 1.00/0 1984 The Royal Society of Medicine
Transcript of Viral aetiology of Paget's disease of bone: a review'
Journal of the Royal Society of Medicine Volume 77 November 1984 943
Viral aetiology of Paget's disease of bone: a review'
L Harvey MA MRCPathDepartment of Pathology, Queen's Medical Centre, Nottingham NG7 2UH
IntroductionSir James Paget published the first comprehensive description of an unusual bony deformity,illustrated by its clinicopathological features in 5 patients, in 1877. His personal term for thecondition, based on macro- and microscopic examination of the affected bones, was 'osteitisdeformans'. It has subsequently become universally known as Paget's disease of bone.The aetiology of the disease, however, remains an enigma and its treatment, discussed by
Russell (1979), is symptomatic or palliative rather than curative (although some drugregimens may reverse or even normalize many of the observed anomalies). Several theoriesfor its pathogenesis (Table 1) have arisen from a consideration and integration of the knownclinical, epidemiological, radiological, biochemical and pathological features of the disorder,including Paget's own theory of a chronic inflammation.The observed radiological progression of disease in single (long) bones coupled with
results from research into Pagetic bone cell activity indicate that, locally at least, diseasebegins with a 'wave' of non-physiological osteoclast-mediated osteolysis. It is now widelyaccepted that Paget's disease is, pathogenetically, an example of a primary osteoclastdysfunction. Consequently research has focused on the control mechanisms involved innormal or abnormal activity of this cell in an attempt to elucidate the pathological triggerfactors. An unexpected observation was of an ultrastructural anomaly which has led to themost recent proposal of a viral aetiology for Paget's disease of bone.
Osteoclast abnormalityRebel et al. (1974) were the first to publish, in French, observations on the atypicalultrastructure of certain bone cells derived from Pagetic biopsies. These workersdemonstrated the presence of tubulofilamentous paracrystalline particles in the osteoclastnuclei and cytoplasm. The individual filaments had a measured diameter of 15 nm with acentral lucent zone approximately 7 nm in diameter, seen both transversely andlongitudinally. Goniometrically, the particle filaments retained these same measurements(Malkani et al. 1976), which in conjunction with other control observations confirmed their3-dimensional nature and that they were not processing or technical artefacts. Suchinclusions were noted only in osteoclasts and were absent from other bone, connective tissueor haemopoietic cells present in the biopsies. They were noted in every case of Pagetic boneexamined. Examples of these structures (from the author's series) are shown in Figure 1.
These findings were published in English only in 1976 (Rebel et al. 1976). By this timethey had been confirmed by Mills & Singer (1976) in the USA and subsequently by Schulz etal. (1977) in Germany, Gherardi et al. (1980) in Italy and Harvey et al. (1982) in the UK.Reports from such geographically disparate groups mirrored and emphasized knownepidemiological aspects of the disease (Table 2). Similarly, all agreed that only the Pageticosteoclasts contained the observed inclusions and that every single case of Paget's diseasewas positive. More recently, Mills et al. (1979) demonstrated the presence of these inclusionsin bone tissue culture, and Harvey et al. (1982) showed in a preliminary survey that thebiopsy inclusion-body load may be related to disease severity as manifested locally. Thelatter observation is presently being reassessed using more strictly defined
'Based on paper read to Section of Pathology, 15 November 1983. Accepted 9 August 1984
0 141-0768/84/110943-06/$O 1.00/0 1984 The Royal Society of Medicine
944 Journal of the Royal Society of Medicine Volume 77 November 1984
Table 1. Paget's pathogenesis: theories Table 2. Epidemiologicalfeatures of Paget's disease relevant toconsidered prior to 1974 the proposed viral aetiology
(1) Chronic progressive multifocalinfilmmation
(2) Collagen tissue disorder(3) Autoimmune disease(4) Slowly progressive (multifocal)
neoplasm(5) Arising as a result of endocrine
imbalance(6) Unusual vascular anomaly
(1) The disease is rare outside mainland UK, parts of (mainlyNorthern) Europe, parts of the USA and Australasia
(2) High incidence foci within geographically-affected zones(3) Variable male: female ratio but always male> female(4) An apparent 'selection' of bones by the disease process i.e.
Skull > Vertebrae > Pelvis > Long Bones> Small Bones(5) Possible familial/genetic predisposition(6) Apparent periodicity of disease with the year of birth
histomorphometric parameters for comparison and a greater number of proven Pageticbiopsies. Two groups have quantitated the osteoclast inclusion-body load: Mills & Singer(1976) found that 20-40% of all sampled clasts and up to 25% of their individual nucleiwere positive, whereas in the author's series (Harvey et al. 1982) the figures were 60-100%and 34% (average) respectively. This apparent difference may be due to differences in cellselection for electronmicroscopic examination or sampling at different developmental ortherapeutic stages of the disease. It may also depend upon distinct aetiopathogenetic triggersin different parts of the world, or local variation in disease expression as a result of variable'virulence' of any putative (possibly identical) trigger agent.The described inclusions are morphologically analogous to those seen in proven
paramyxovirus (PMV) infections described by Nakai et al. (1969) in vitro and Kallman et al.(1959) in vivo. The possibility that they are nonspecific condensations of cell protein orprocessing artefacts has been effectively dismissed by the results of several independentcontrol studies, as discussed in detail by Rebel et al. (1976) and illustrated by Harvey et al.(1982). A further control series to establish the overall prevalence of these inclusions inosteoarticular lesions (Harvey, in preparation) - which involved the examination ofcongenital, inherited, traumatic and tumorous bony lesions as well as a similar spectrum ofmetabolic conditions previously examined by other workers - has merely served toemphasize the consistent and absolute presence of such inclusions in Pagetic osteoclasts;their occurrence in nucleus and cytoplasm (but mainly the former), together with theiroverall appearance, simulated a PMV infection - measles or respiratory syncytial virus(RSV). PMV inclusions were, however, detected in giant cells in 2 out of 10 cases of non-Paget's associated giant cell tumours (GCT) of bone examined (Harvey, in preparation), a
A~~~~~~~~~~~~~mFigure 1. Osteoclast inclusion bodies. A: Oil immersion light microscopy showing 2 of 4 positive nuclei in a cellcontaining 16 nuclei (Toluidine blue x 800). B: Lower power electronmicrograph showing 3 inclusion-body-bearingnuclei (one arrowed) highlighting their distinct appearance compared to nucleoli (Uranyl acetate, lead citratex 7500). c: High power electronmicrograph showing their tubulofilamentous structure (Uranyl acetate, lead citratex 30 000)
Journal of the Royal Society of Medicine Volume 77 November 1984 945
finding already documented by Welsh & Meyer (1970) and Le Charpentier et al. (1977).Mirra et al. (1981) have also recently commented on the inclusion-body positivity of someGCTs complicating Paget's disease. Two further examples of non-Pagetic osteoclastinclusion-body positivity have been identified: the first a peripheral reparative giant cellgranuloma from a 28-year-old male and the other a 2-month old Asian suffering fromAlbers-Sch6nberg disease (osteopetrosis). In all these non-Pagetic lesions, the cellularinclusion body incidence was extremely low (< 10% of sampled cells) compared to Pageticclasts. The significance of these occasional non-Pagetic inclusion-body-positive lesions will bediscussed elsewhere (Harvey et al., in preparation).The inclusions are generally accepted as morphological evidence for a viral presence or
involvement in the aetiology of Paget's disease. Corroborative evidence has been sought in avariety of ways. The possibility that Paget's disease, like scrapie, kuru, Jakob-Creutzfelddisease, etc., is a chronic, slow virus disorder, has now been added to the theories listed inTable 1.
Support for a viral aetiologyHard scientific evidence or proof of a viral aetiology has been considerably hampered by thelack of a suitable animal or laboratory model for the disease. Too close a comparison withsubacute sclerosing panencephalitis (SSPE) - itself proven to be of measles aetiology byisolation, propagation and comparative studies (Horta-Barbosa et al. 1969, Chen et al. 1969and Oyanagi et al. 1971 respectively) - has confused the issue. For reasons related toattenuation of the viral particle and/or its pathogenetic process, SSPE may not in fact be aparticularly relevant model for study. Despite these two problems, there are several areas ofsupport for the viral hypothesis.
Clinicopathological aspects of Paget's disease have several features in common with otherproven slow-virus disorders (Table 3). In addition, the inclusion-body-bearing osteoclastshave been shown to contain some PMV-type antigens. Rebel et al. (1980) have demonstratedsuch antigens from measles virus using both immunofluorescence and immunoperoxidase(Heyderman 1979) techniques. Using similar methodology, Mills et al. (1980) havedemonstrated similar antigens from RSV and, more recently, elements of both viruses (Millset al. 1984)! Basle et al. (1983) have also identified 4 of 6 structural measles antigens inPagetic osteoclasts, using a monoclonal antibody technique, and emphasized the absence oftwo antigens associated with viral replication; the significance of this is discussed later. Otherworkers have had less successful immunohistochemical results, however, suggesting thatstrict adherence to a specific method may be critical (Woods 1984).
Attempts to retrieve or grow infectious virions by tissue culture have had, at best,equivocal results (Mills et al. 1979). More recent animal experiments, however, do suggestthe possibility that a transmissible agent may ultimately give rise to osteosarcoma in therecipient (Singer 1984). This is an extremely significant observation with connotations forthe aetiopathogenesis of tumorous complications of Paget's disease in the human. Thepossible interrelationships of these neoplastic lesions have already been alluded to (Harveyet al. 1982) and are being further analysed histomorphometrically and immunocyto-chemically in order to identify common features.
Several investigators have tried to confirm a viral 'infection' in Paget's disease by means ofa serological screen for raised antiviral antibody titres. These have focused mainly onstraightforward complement fixation (CF) techniques. All those reported have beennegative, including the author's own series comparing 19 Pagetic sera to 30 age- and sex-matched controls. Pringle (1984), however, has implied relative persistence of RSVantigenaemia (? in Paget's particles) on the basis of statistically significant abnormalities ofthe anti-RSV titres. Again, this is not seen in the author's personal series where RSV titresare universally low.
Perhaps the most relevant technique to apply to this particular situation is viralDNA/RNA probe hybridization, as successfully used in studies of herpes simplex II andcarcinoma of the cervix. So far, there have been remarkably few reports of its use and those
946 Journal of the Royal Society of Medicine Volume 77 November 1984
Table 3. Features of Paget's disease incommon with other proven slow-virus disorders
(1) Long latent period(2) Chronic recrudescent disease activity(3) Single target organ system(4) Presence of atypical giant cells(5) Detectable intracellular inclusions(6) Minimal inflammatory response(7) Possible genetic/familial predisposition
publicized have been criticized on the grounds of poor control or methodology.On the basis of certain epidemiological features (Table 3), the well documented and
reproducible morphological studies and the albeit capricious and often unreproducibleimmunohistochemical data, several authors have strongly supported the proposed viralaetiology, especially Rebel et al. (1981), though it has not been simply or uncautiouslyaccepted as fact (Lancet 1982).
Criticisms of a viral aetiopathogenesis for Paget's diseaseThe main problems/criticisms of the theory encountered to date are the following:(1) The absence of a demonstrable immune reaction to the proposed viral antigens.(2) The inability to recover infectious virions or demonstrate, convincingly, the presence ofthe viral genome in Pagetic cells.(3) The possibility that activated giant cells (i.e. the osteoclasts) may ingest nonspecificallyany virus present at the time.(4) The apparent inconsistent immunological results, namely the demonstration of RSV andmeasles virus alone, or together, in different studies.(5) Finally, the necessity to invoke co-factors or secondary trigger factors in the eventualexpression of the disease.
These and other topics were discussed by workers in the field, independent researchers andinvited observers at a recent international symposium held by the Medical Research Councilin Southampton. The following assessment is based upon the discussions at that meeting,with additional comments from the author of this paper.
Criticisms assessedA detectable immune reaction in the form of a simple elevation of CF titres (the techniquemost often quoted) need not necessarily be observed or expected. Integration of virus-codedhereditary nuclear proteins into the osteoclast (or osteoclast precursor) at a timeconsiderably predating their pathogenetic effect could adequately explain the absence of arising or raised CF titre, normally seen only in acute infections. Similarly, the occult natureof these viral genes, even following their morphological expression as an inclusion body,would prevent extracellular exposure to the immune system. The body defences would thenbe alerted only if there was a cell surface antigenic mutation, but not by the simple increasein clast lytic activity actually observed with the onset of the disease. The possibility that thevirus is, anyway, atypical or attenuated and consequently of altered immunogenicity issupported by the observation by Rebel et al. (1981) of absent 'replication' associatedantigens (L+P). These arguments, however, do not all pertain in SSPE, where blood andcerebrospinal fluid contain an IgM anti-measles antibody (Kiessling et al. 1977).The known difficulties of culture-retrieval techniques, as experienced in SSPE, lessen the
validity of this criticism. The incomplete or attenuated nature of the proposed Paget's virusas outlined above would also lessen the likelihood of successful culture.The negative hybridization studies should probably also be ignored as they seem to have
relied on incomplete viral genomes with a consequently low potential for recombination inthe test situation. -o
Journal of the Royal Society of Medicine Volume 77 November 1984 947
The argument that activated giant cells nonspecifically take up any free virus with themorphological expression of entrapped virion can be effectively countered in several ways. Ifthis were the case, why are similar findings not seen in a higher incidence and a greaterspectrum of other diseases characterized by similar giant osteoclasts, e.g.hyperparathyroidism? Secondly, it would be unrealistic in this case for all the entrappedvirus to be PMV-like; surely other viruses such as herpes, cytomegalovirus, polio, etc.,would also be encountered? The known fact that viruses, especially PMV, stimulate giantcell production, and that osteoclast lytic activity may be proportional to the nuclear contentof the cell, seem more relevant observations and support the viral story!The paradoxical RSV/measles immunohistochemical results do not themselves invalidate
this concept. As referred to earlier, they may simply represent two alternative local triggerviruses. These viruses also share many virion components with significant overlap ofantigenic determinants. The results may be due to the inability of our relatively crudeimmunohistochemical preparations and techniques to separate them rather than indicating atrue involvement of separate viruses. The results of Mills et al. (1984) showing an apparentdual presence emphasizes this point. Maybe one or more simultaneous trigger viruses areactually involved!The disparity of the known prevalence of measles infection and the geographical
distribution of Paget's disease, documented histologically (Collins 1956) andepidemiologically (Detheridge et al. 1982, Barker et al. 1980), has resulted in the invocationof co-factors or secondary triggers in the pathogenesis of the disease. Its prevalence inemigrants from high to lower incidence areas (Gardner et al. 1978) further suggests thatthese secondary agents are themselves acting on a predisposed population. Sofaer (1984) hasalso highlighted the possibility of a birth-cohort-related cyclical disease incidence. This is,arguably, the hardest criticism to contend with and it remains, as yet, an unexplainedproblem. Nevertheless, the concept of co-factors has been used effectively to explain othersystems as diverse as malaria and Burkitt's tumour, Hodgkin's disease, primary biliarycirrhosis and the development of many malignancies.ConclusionsPaget's disease is widely accepted as arising from a primary osteoclastic dysfunction.Relative ignorance of the interaction of bone cells at the biochemical level, theircommunications in maintaining normal bone turnover, the effects of acute virus infectionson the former, and the utilization of inadequate virological techniques combined withinconclusive epidemiological/family studies, may have obscured many features ofaetiological significance.
Nevertheless, some of Koch's (modified) criteria for establishing a lesion as infective inaetiology have been met. Juvenile exposure to a PMV with integration of virally-codedDNA into the host genome of an osteoclast progenitor population, with clinical expressionof Paget's disease at a later date following the effect of one (or more) exogenous orendogenous co-factors, seems the most exciting and tenable hypothesis to date. It does notnecessarily exclude previously suggested theories (Table 1) and may indeed integrate withsome, e.g. a slowly progressive neoplasm and Paget's own theory.
Increasing sophistication in the application of techniques already in use should help toconfirm or refute the viral aetiopathogenetic theory. Its connotations for juveniletumorigenesis and our understanding of certain giant cell lesions of adults are enormous(Harvey et al., in preparation). Likewise its results may significantly influence the directionof research into anti-Pagetic therapy.
Finally, as the incidence of Paget's disease is apparently declining (Barker 1984) letus hope that our knowledge of the various pathogenetic mechanisms involved, possiblyrelevant to other diseases, become known before it eventually disappears!
Acknowledgments: I would like to thank Professor D R Turner for his advice and criticismof the manuscript. The author's research was supported by the Yorkshire Cancer ResearchCampaign.
948 Journal of the Royal Society of Medicine Volume 77 November 1984
ReferencesBarker D J P (1984) Proceedings MRC Symposium on Paget's disease, Southampton, 1983 (Scientific Reports No.
5). Ed. D J P Barker. Medical Research Council, London; p 1Barker D J P, Chamberlain A T, Guyer P B & Gardner M J (1980) British Medical Journal 280, 1105-1107Basle M, Rebel A, Filmon R & Pilet P (1983) 17th European Symposium on Calcified Tissue, Davos, Switzerland;
Abstract 43Chen T T, Watanabe I, Zaman W & Mealey J (1969) Science 163, 1193-1194Coffins D H (1956) Lancet ii, 51-57Detheridge F M, Guyer P B & Barker D J P (1982) British Medical Journal 285, 1005-1008Gardner M J, Guyer P B & Barker D J P (1978) British Medical Journal 276, 1655-1657Gherardi G, Lo Cascio V & Bonucci E (1980) Histopathology 4, 63-74Harvey L, Gray T, Beneton M N C, Douglas D L, Kanis J A & Russell R G G (1982) Journal of Clinical Pathology
35, 771-779Heyderman E (1979) Journal of Clinical Pathology 32, 971-978Horta-Barbosa L, Fuccillo D L & Sever J L (1969) Nature 221, 974-975Kallman F, Adams J M, Williams R L & Imagawa D T (1959) Journal of Biophysical and Biochemical Cytology 6,
379-392Kiessling W R, Hall W W, Yung L L & Ter Meulen V (1977) Lancet i, 324-327Lancet (1982) ii, 1198-1199Le Charpentier Y, Le Charpentier M, Forest M et al. (1977) La Nouvelle Presse Medicale 6, 259-262Malkani K, Basle M & Rebel A (1976) Journal of Submicroscopic Cytology 8, 229-236Mills B G & Singer F R (1976) Science 194, 201-202Mills B G, Singer F R, Weiner L P & Holst P A (1979) Calcified Tissue International 29, 79-87Mills B G, Singer F R, Weiner L P & Holst P A (1980) Arthritis and Rheumatism 23, 1115-1120Mills B G, Singer F R, Weiner L P, Suffin S C, Stabile E & Hoist P (1984) Clinical Orthopaedics and Related
Research 183, 303-311Mirra J M, Henrik Bauer F C & Grant T T (1981) Clinical Orthopaedics and Related Research 158, 243-251Nakai T, Shand F L & Howatson A F (1969) Virology 38, 50-67Oyanagi O, Ter Meulen V, Katz M & Koprowski H (1971) Journal of Virology 7, 176-187Paget J (1877) Medico-Chirurgical Transactions 60, 37-63Pringle C R (1984) Proceedings MRC Symposium on Paget's disease, Southampton, 1983 (Scientific Reports(No. 5). Ed. D J P Barker. Medical Research Council, London; p 29
Rebel A, Basle M, Pouplard A, Malkani K, Filmon R & Le Patezour A (1981) Metabolic Bone Disease and RelatedResearch 3, 235-238
Rebel A, Malkani K, Basle M & Bregeon Ch (1976) Calcified Tissue Research 20, 187-199Rebel A, Malkani K, Basle M, Bregeon Ch, Le Patezour A & Filmon R (1974) Revue de Rhumatisme 41, 767-771Rebel A, Pouplard A, Filmon R, Basle M, Kouyoumdjian S & Le Patezour A (1980) Lancet ii, 344-346Russell R G (1979) Clinics in Rheumatic Diseases 5, 673-695Schulz A, Delling G, Ruige J D & Ziegler R (1977) Virchows Archives (Pathological Anatomy and Histology) 376,
309-328Singer F R (1984) Proceedings MRC Symposium on Paget's disease, Southampton, 1983 (Scientific Reports No. 5).
Ed. D J P Barker. Medical Research Council, London; p 22Sofaer J A (1984) Proceedings MRC Symposium on Paget's disease, Southampton, 1983 (Scientific Reports No. 5).
Ed. D J P Barker. Medical Research Council, London; p 7Welsh R A & Meyer A T (1970) Laboratory Investigation 22, 63-72Woods C G (1984) Proceedings MRC Symposium on Paget's disease, Southampton, 1983 (Scientific ReportsNo. 5). Ed. D J P Barker. Medical Research Council, London
