Introduction

Paget’s disease of bone is a common skeletal disorder characterised by focal abnormalities of increased and disorganised bone turnover at one or more skeletal sites [1]. Many patients are asymptomatic but those that do come to clinical attention can develop various complications including bone pain, deformity, pathological fractures, osteoarthritis, deafness and rarely osteosarcoma [2, 3]. Paget’s disease is a complex disorder. Current evidence suggests that genetic factors play a key role in susceptibility [4], but there is also a strong environmental component evidenced by the fact that the disease has become less common and less severe in several countries over recent decades [5–7]. Several potential environmental triggers for Paget’s disease of bone (PDB) have been suggested including dietary calcium intake [8], vitamin D deficiency [9] and excessive biomechanical loading of affected bones [10], but the most widely studied is paramyxovirus infection [11, 12]. Paramyxoviruses were first implicated as a potential trigger of PDB by the morphological finding of nuclear inclusion bodies in osteoclast nuclei which were thought to resemble measles nucleocapsids [13, 14]. Despite extensive research over the past 30 years, the role of paramyxovirus infection in the pathogenesis of PDB remains controversial [15, 16]. Some researchers have reported finding evidence of measles virus (MV) nucleic acids or antigens in cells and tissues from patients with the disease [17–20]. Respiratory syncytial virus antigens were also detected in one study [21], and in another evidence was presented to suggest that antigens for both MV and respiratory syncytial virus (RSV) may be present [22]. Canine distemper virus (CDV) nucleic acids have also been reported by one group to be present in Pagetic bone cells [23, 24], whereas other researchers failed to find evidence of paramyxovirus nucleic acids or protein in bone tissue or peripheral blood from PDB patients [15, 25–29]. A characteristic feature of the persistent measles infection, subacute sclerosing panencephalitis, is the presence of a marked increase in antiviral antibodies both in the serum and cerebrospinal fluid [30, 31]. Based on this observation, several researchers have measured antibodies to paramyxoviruses in patients with PDB [32–35]. These studies yielded negative results, but the sample sizes were small and had limited power to detect possible differences between PDB cases and controls. Here, we have evaluated the antibody response to paramyxoviruses in PDB in the largest study performed to date using a case control design and determined if antibody concentrations were associated with disease extent or complications. We also evaluated the antibody response to other common viral infections not previously implicated in the pathogenesis of PDB including mumps, rubella and varicella zoster virus.

Patients and Methods

Patients

The study cohort comprised 463 patients who participated in the Paget’s disease, Randomised Trial of Intensive versus Symptomatic Management (PRISM) study (ISRCTN12989577) [36, 37]. Participants were included in this study if sufficient serum was available from the baseline study visit for analysis of the antibodies of interest. One subject included in this study had a history of osteosarcoma. The controls comprised 202 age and gender match subjects not known to have PDB who were predominantly spouses of PRISM participants.

Clinical Assessments

The clinical assessments performed within the PRISM study have been previously described [36]. Health-related quality of life was assessed by the SF-36 questionnaire. Deformity was assessed by the attending physicians who were asked to assess whether the patient had clinical evidence of bone deformity using a three-point scale as follow: 0 = no deformity; 1 = mild or moderate deformity and 2 = severe deformity. The presence of bone pain was recorded and physicians were asked to assess if they thought the pain was caused by PDB. Information was collected on previous fractures and whether they had occurred in affected bone; on orthopaedic surgical procedures; on the use of a hearing aid for deafness in those with skull disease; on age at first diagnosis of PDB; and family history of PDB. The extent of PDB was recorded based on involvement on bone scan examination. Information was recorded on previous bisphosphonate treatment, and the number of treatment courses given. Based on the clinical data, we devised a composite scoring system for disease severity as previously described [38] taking into account the number of affected bones, age at first diagnosis, family history, presence of bone pain, pathological fracture, orthopaedic surgery for PDB, bone deformity, deafness associated with skull involvement and previous bisphosphonate treatment for PDB.

Genotyping

Genotyping was conducted for SQSTM1 mutations by DNA sequencing on DNA extracted from peripheral blood using standard techniques as previously described [38].

Detection of Antiviral Antibodies

Enzyme-linked immunosorbent assays (ELISA) were used for the detection and quantitative determination of IgG antibodies to MV, Rubella, Mumps, Varicella Zoster Virus (VZV), RSV and CDV. The ELISA were performed on serum samples that had been collected at the baseline visit of the PRISM study between 2001 and 2004 and stored frozen at −80oc until analysis in the present study. We used the Trinity Biotech Captia assays TM for MV (catalogue number 2326000), mumps (catalogue number 2325900) rubella (catalogue number 2325300) and VZV (catalogue number 2325600). We used the MP Biomedical IgG ELISA (catalogue number 071-516002) for RSV. For CDV, we used the ImmunoComb® antibody test kit. All assays were performed according to the manufacturer’s instructions. The results of the ELISA assays were read on a Synergy HT Multi-Mode Micro plate reader (Bio-Tek), 450 nm filter. Dual wavelength was used and the reference filter set to 600–650 nm. All samples were tested in duplicate and the mean value of the two samples calculated. Duplicate samples that yielded values that differed by more than 20% were repeated. With the exception of CDV, data were expressed as the immune status ratio (ISR). This was calculated by dividing the optical density value of the patient sample by a calibrator value provided with the kit. These results were converted to International Units (IU) using the natural exponential function f(x) = ex. For CDV, semi-quantitative analysis was performed based on a colour scale provided with the kit using Comb Scan software. The reference ranges are shown in Table 1.

Table 1 Reference ranges for viral antibodies
Full size table

Statistical Analysis

Differences in viral antibody concentrations in PDB cases and controls were assessed quantitatively using Student’s t-test and semi-quantitatively by χ2 test in which patients were categorised into two groups if antibody concentrations suggested there had been previous infection. The relation between viral antibody titres and disease extent was evaluated by dividing PDB subjects into three groups based on the severity score and comparing viral antibody titres and other clinical characteristics between the groups by ANOVA or χ2 test as appropriate. The significance level was set at 0.008 to account for the fact that antibodies against six different viruses were tested.

Results

Antiviral Antibodies in PDB Cases and Controls

Demographic characteristics and antiviral antibody concentrations in PDB cases and controls are shown in Table 2. A high proportion of PDB cases and controls tested positive for measles virus, rubella, VZV, mumps and RSV indicating previous infection, with no significant difference between the groups. All cases and controls tested positive for at least one virus and more than 90% tested positive for four of the viruses. Antibodies were also detected that cross-reacted with distemper virus in about 45% of PDB cases and controls with no difference between the groups. Quantitative analysis showed that antibody concentrations did not differ between cases and controls for MV, CDV, RSV and VZV (Table 2). However, antibody concentrations for mumps virus were significantly higher in PDB cases as compared with controls (3.00 ± 0.85 vs. 2.56 ± 0.89; p < 0.001). For rubella virus, antibody concentrations in PDB cases were also higher, but this was not significant taking multiple testing into account (29.1 ± 11.4 vs. 27.30 ± 9.82, p = 0.039).

Table 2 Clinical characteristics and antibody concentrations in PDB cases and controls
Full size table

Relation Between Severity of PDB and Antiviral Antibodies

To determine if circulating antibody concentrations were related to the severity of PDB, patients were grouped into three categories based on the disease severity score as previously described [38]. Mild disease was defined as a score of 3 or less, moderate disease as a score of 4 to 6 and severe disease as a score of 7 or more. The result is shown in Table 3. There was no association between disease severity and circulating antibody concentrations to any of the viruses tested. We repeated the analysis for number of affected bones but no association was observed (data not shown). The presence of SQSTM1 mutations was associated with greater severity. Not surprisingly, patients with more severe disease had a significantly lower physical functioning as assessed by the SF36 instrument. Interestingly, there was difference between the severity groups in SF36 bodily pain or mental functioning. The patient with a history of osteosarcoma tested positive for all the viruses studied, except CDV where the results were negative. The circulating antibody concentrations were unremarkable in this patient.

Table 3 Antiviral antibodies and other variables in relation to severity of PDB
Full size table

Antiviral Antibodies and SQSTM1 Mutations

It has previously been suggested that measles virus infection may interact with SQSTM1 mutations to act as a trigger for PDB [39]. To determine whether antibody concentrations were associated with SQSTM1 mutation status, we performed a subgroup analysis in which we analysed circulating antibody concentrations in SQSTM1 mutation positive and negative PDB cases as compared with controls. The results are summarised in Table 4. There was no difference in circulating antibody concentration to any of the viruses tested according to SQSTM1 mutation status, although concentration of mumps virus antibodies were higher in SQSTM1 negative and SQSTM1 positive PDB cases as compared with SQSTM1 negative controls.

Table 4 Antiviral antibodies in PDB cases and controls according to SQSTM1 status
Full size table

Discussion

The aim of this study was to determine if antibody response to measles virus, RSV or distemper virus was associated with the occurrence or severity of PDB. In addition to testing for antibodies to the paramyxoviruses which have previously been implicated in PDB [11, 12, 22], we also measured antibody concentrations to other common viruses such as mumps, rubella and varicella zoster virus.

The results showed that a high proportion of patients in both the PDB group and the control group had circulating antibodies to the viruses tested indicating that previous infection had occurred. For most viruses, the prevalence of those that tested positive was greater than 95% with no difference between cases and controls. These results confirm the findings of previous small scale studies which failed to find a difference in antiviral antibody concentrations between PDB cases and controls [32–35]. For CDV, only about 45% of cases and controls tested positive. Since there is extensive homology within the paramyxovirus family [40–44], it is unclear if these antibodies truly represented previous distemper virus infection or were cross reacting antibodies due to exposure to other paramyxoviruses such as measles [45]. Further studies will be required to investigate this fully. On quantitative testing, there was no significant difference in circulating antibody concentrations between cases and controls except for mumps virus where concentrations were significantly higher in cases. This raises the possibility that the immune response to mumps may be altered in PDB and suggests that the role of mumps in PDB might be worth further study.

We also sought to determine if the levels of virus antibody were associated with the severity of PDB. For this analysis, patients were divided into three categories of mild moderate and severe PDB based on disease extent and complications as previously described [38]. No significant association between circulating antibody levels and disease severity or extent was observed. Given that persistent virus infections such as SSPE are associated with high titres of antibody production [30, 31], this observation argues strongly against the possibility that there is a state of persistent paramyxovirus viral infection in PDB contrary to what has been suggested [46]. While the present study provides no support for the hypothesis that patients with PDB have persistent paramyxovirus infection, the study design cannot completely rule out the possibility that previous exposure to a viral illness early in life might play a role in PDB. However, as previously reported, there was a significant association between disease severity and the presence of SQSTM1 mutations [38, 47] emphasising the importance of genetic factors in the pathogenesis of PDB and its severity. In summary, the present study does not provide support for the notion that persistent paramyxovirus infection is involved in the pathogenesis of PDB.