Detection of parvovirus B19 DNA by polymerase chain reaction in giant cell arteritis: A case–control study

Abstract

Multiple environmental and genetic factors probably play a role in giant cell arteritis (GCA) susceptibility (1). A close concurrence between the observed incidence peaks of polymyalgia rheumatica (PMR)/GCA and epidemics of Mycoplasma pneumoniae, parvovirus B19, and Chlamydia pneumoniae has been found (2). The involvement of B19 in GCA pathogenesis has been recently supported by the demonstration of a significant association between the histologic evidence of GCA and the presence of parvovirus B19 DNA in temporal artery biopsy (TAB) specimens (3). The aim of this study was to evaluate whether B19 DNA was more likely to be present in the temporal arteries of patients with biopsy-proven GCA or “pure” PMR than in the arteries of control subjects. We prospectively examined fresh TAB specimens from 31 consecutive patients with biopsy-proven GCA, 43 consecutive biopsy-negative PMR patients without clinical manifestations of GCA, and 19 age-matched controls with a diagnosis other than PMR/GCA. Included in the study were all consecutive patients with biopsy-proven GCA seen in the Division of Rheumatology of Reggio Emilia Hospital between January 1998 and January 2001 and all consecutive PMR patients seen in the same division between January 1998 and October 1999. Of the 19 biopsy-negative, age-matched controls, 13 were patients with a suspected diagnosis of GCA (based on the presence of systemic manifestations), but whose final diagnoses after a median followup period of 22 months (range 10–43 months) were sepsis (n = 4), cancer (n = 3), polyarteritis nodosa (n = 2), hypersensitivity vasculitis (n = 2), psoriatic arthritis (n = 1), and adult Still's disease (n = 1). The other 6 controls were patients who had died as a result of neurologic or myocardial acute attacks, and who had had no history of GCA and/or PMR. From these patients, fresh TAB samples were obtained at autopsy. Additional controls consisted of fresh aorta or carotid artery surgical specimens obtained from 32 age-matched patients with atherosclerotic disease who had no history of GCA, PMR, or rheumatoid arthritis, as well as fresh samples of 14 fetal aortas obtained at autopsy. The investigators who performed DNA analyses were blinded to the clinical and histologic diagnoses. The study was approved by the Ethics Committee of Reggio Emilia Hospital, and informed consent was obtained from all patients or their relatives. Fresh samples of arteries were stored at −70°C. Samples were homogenized and incubated in a Tris–EDTA buffer containing proteinase K (20 mg/ml) for 3 hours. After incubation, the DNA was extracted by treatment with a phenol–chloroform–isoamylic alcohol mixture (25:24:1). The extracted DNA was precipitated with absolute ethanol and sodium acetate. After washing with 70% ethanol, the precipitates were dried and redissolved in 100 μl of Tris–EDTA buffer (10 mmoles/liter Tris, 1 mmole/liter EDTA). The DNA solution was stored at −20°C. Plasmid containing B19 sequences was used as a positive control. Sterile water was added to a tube containing polymerase chain reaction (PCR) master mix and used as an amplification negative control. Since high sensitivity of nested PCR is a possible cause of false-positive results, we took extreme precautions in order to avoid any contamination. Samples were handled in laminar-flow hoods, and DNA extraction and reaction mix preparation were performed in separate rooms. All samples were processed using aerosol-resistant pipette tips. Again, to avoid possible cross-contamination during all DNA extraction steps, we used 100 μl of sterile water as a negative control. This negative control was subjected to the extraction procedure and then processed by PCR together with other samples. To check for the presence of PCR inhibitors, we tested all samples performing a PCR reaction to amplify a 250-bp fragment of human β-actin gene. In our study, we used two PCR methods utilizing primers specific for a part of a region of the B19 genome coding for B19 nucleocapsid. All PCRs were performed in a P.E. 9600 thermal cycler (Perkin-Elmer Cetus, Uppsala, Sweden) in a 50-μl reaction volume containing 100 ng total template DNA, 50 mM KCl, 10 mM Tris HCl, 0.1% Triton X-100, 200 μM each of dATP, dCTP, dGTP, and dTTP (Amersham Pharmacia Biotech, Branchburg, NJ), 2.5 mM MgCl2, 0.5 μM of each primer, and 1 unit of Taq DNA polymerase (Perkin-Elmer Cetus). In the first PCR-based method, two primer pairs (4) were used in a nested PCR protocol to identify a 279-bp region of the B19 genome coding for nucleocapsid. Outer primers were 5′-AAGTTTGCCGGAAGTTCCCG-3′ (nucleotide positions 3076–3095) and 5′-AGCATCAGGAGCTATACTTCC-3′ (nucleotide positions 3478–3458); inner primers were 5′-CCCAAGCATGACTTCAG-3′ (nucleotide positions 3118–3134) and 5′-TCTAAATATCTCCATGG-3′ (nucleotide positions 3396–3380). The same PCR profile was used for both amplification rounds. Following an initial denaturation step (2 minutes at 95°C), samples were subjected to 35 cycles of 95°C for 30 seconds, 55°C for 30 seconds, and 72°C for 30 seconds, with a final extension time of 5 minutes at 72°C. In the second PCR round, 2 μl of the first amplification product was added to 48 μl of reaction mixture as template. In the second PCR-based method, the first amplification round (5) was performed using two outer primers located, respectively, at nucleotide positions 2905–2924 (5′-CTTTAGGTATAGCCAACTGG-3′ [forward primer]) and 4016–3997 (5′-ACACTGAGTTTACTAGTGGG-3′ [reverse primer]). Two microliters of amplified product was used as template for the second amplification round. Inner primers for the second PCR reaction were located at nucleotide positions 3187–3206 (5′-CAAAAGCATGTGGAGTGAGG-3′ [forward primer]) and 3290–3271 (5′-CCTTATAATGGTGCTCTGGG-3′ [reverse primer]) and provided a 104-bp PCR product. These primers are also located in a region of the B19 genome coding for B19 nucleocapsid. The PCR profile was 2 minutes of initial denaturation, followed by 35 cycles of 1 minute at 95°C, 1.5 minutes at 55°C, and 1 minute at 72°C for both amplification rounds. The 31 consecutive patients with biopsy-proven GCA, the 43 consecutive patients with “pure” PMR, and the 19 age-matched controls with a diagnosis other than PMR/GCA had similar mean ± SD ages (73.6 ± 6.7 years, 74.6 ± 6.8 years, and 71.3 ± 8.9 years, respectively). The mean ± SD age of the 32 patients with aortic and carotid atherosclerotic disease was similar to those of the 3 other groups (70.1 ± 9.5 years). The two nested PCR methods used to detect the presence of parvovirus B19 in our samples showed a high rate of agreement (97.1%), with 135 of the 139 samples tested giving the same results. In addition, PCR performed on a fragment of β-actin gene to test the DNA extraction produced amplicons in all samples. Table 1, illustrating the frequency with which B19 DNA was detected by PCR in TAB tissue as well as in aorta and carotid artery and fetal aorta specimens, shows that B19 was detected in most of the TAB specimens with a similar frequency in patients with biopsy-proven GCA or “pure” PMR and in age-matched controls, and that there were no differences between men and women. B19 DNA was also detected in most of the specimens of aorta or carotid arteries with atherosclerotic disease. The frequency of B19 DNA detection in this group was similar to that observed in TAB tissue. Epidemiologic data suggest that parvovirus B19 may play a role in the etiology of GCA (1, 2). A longitudinal study of the epidemiology of GCA in Olmsted County, MN, over the last 40 years demonstrated epidemic-like, cyclic fluctuations in the incidence of GCA, occurring every 6–7 years (1). Parvovirus B19 has been shown to occur in epidemic cycles similar to the cycles reported for GCA (6). Therefore, one recent study from the Mayo Clinic investigated whether parvovirus B19 could be found in TAB samples from patients with GCA (3). Investigators in this study reported a highly significant association between the presence of B19 DNA in TAB specimens and histologic evidence of GCA. B19 DNA was detected in 54% of 13 TAB specimens with histologic evidence of GCA and in 10.8% of 37 TAB-negative specimens. These investigators concluded that B19 might have a pathogenetic role in GCA and recommended further research to confirm their findings. Our findings do not confirm the reported association between histologic evidence of GCA and the presence of B19 DNA in TAB tissue. B19 DNA was detected by PCR analysis with similar frequencies in temporal artery tissue obtained from patients with either biopsy-proven GCA or “pure” PMR, as well as from age-matched controls. Furthermore, a similar frequency of B19 DNA localization was found in arterial surgical specimens obtained from age-matched controls with atherosclerotic carotid/aortic disease. To avoid the risk of false-positive DNA amplification results, we followed several precautions against contamination. Our results were further validated by the high rate of agreement observed between the two different PCR tests used to detect B19 DNA. In fact, the use of two amplification systems with different primers specific for the genomic region coding for B19 nucleocapsid helped to rule out false-positive results due to nonspecific amplifications. The reliability of our findings was also confirmed by the absence of B19 DNA from the 14 fetal aorta specimens. The discordant results between our study and that done by investigators at the Mayo Clinic (3) may be attributable to different causes. In the Mayo Clinic study, the controls were patients who had undergone TABs for a suspected diagnosis of GCA. Our control samples not only included TAB specimens from age-matched patients who surely did not have GCA, but they also included carotid and aorta specimens from age-matched patients with atherosclerotic disease as well as fetal aorta specimens obtained at autopsy. Another difference between the two studies is that the Mayo Clinic study used paraffin-embedded samples, while we utilized fresh frozen samples of arteries. Although viral DNA sequences can be reliably detected by PCR in both fresh frozen and paraffin-embedded specimens (7), it is known that nucleic acids from paraffin-embedded tissues are worse templates than those recovered from fresh tissues. The prevalence of IgG antibodies to parvovirus B19 increases with age. In Europe and the US, ∼60–70% of the elderly population has detectable IgG (8). In this report, we describe high frequencies of B19 DNA in the artery specimens of elderly patients that were similar to the reported prevalence of a past B19 infection in the elderly. However, since we did not perform serologic tests, we do not know whether there was a correlation in our patients (as seems likely) between the presence of anti-B19 IgG antibodies and the detection of B19 DNA. B19 DNA seems to be an innocent bystander commonly present in the arteries of elderly people rather than the cause of underlying vasculitis. The expression of the cellular receptor for B19 (erythrocyte P antigen or globoside Gb4) on endothelial cells (9) may explain the B19 localization in the arteries, where it probably persists for years or decades after primary infection. Results similar to ours were recently observed in synovial membranes of young patients with and without chronic arthropathy, where B19 DNA in synovial membrane was identified in 28% of children with chronic arthritis and in 48% of young adult controls without arthropathy (10). The presence of B19 DNA in the synovial membrane was strictly correlated with the presence of serum Ig antibodies to B19. In conclusion, B19 DNA was present in the arteries of a high proportion of elderly patients. No association was found between the presence of B19 DNA in TAB specimens and histologic evidence of GCA. Although the pathophysiologic role of B19 DNA persistence in artery is unclear, B19 seems to be an innocent bystander commonly present in the arteries of elderly people. Future studies will evaluate the cell type in arterial walls that contains B19 DNA and assess whether the virus genome is transcriptionally active.