- Email: [email protected]
Diagnosis of vascular graft infections with antibodies against staphylococcal slime antigens
RESEARCH LETTERS
pregnancies complicated by a CHM who wish to continue their pregnancy have a 40% chance of a successful outcome. Moreover, this choice is associated with a fairly small chance of serious obstetric complications and does not seem to increase the risk of development of pGTD.
77 women with CHM and healthy co-twin pregnancies
19 elective termination of pregnancy before week 14
58 remained pregnant until at least week 14
3 persistent GTD
12 persistent GTD
3 treated with single agent
8 treated with single agent
4 treated with multi agent
Contributors N J Sebire helped design the study, obtained and analysed data, and wrote the report; M Foskett helped with data collection and analysis; F J Paradinas contributed to pathological review and data collection; R A Fisher contributed to data collection and analysis, genetic studies, and writing of the report; R J Francis and D Short contributed to the data collection; E S Newlands contributed to project design, data analysis, and writing of the report; M J Seckl conceived and helped design the study, and contributed to data analysis and writing of the report.
Conflict of interest statement None declared.
Acknowledgments Figure 2: Frequency of persistent gestational trophoblastic disease (GTD) in a series of 77 twin pregnancies complicated by the co-existence of a complete hydatidiform mole (CHM) and a healthy fetus
MJS is supported by the Department of Health, Cancer Research UK, and the Cancer Treatment, Research Trust. The Department of Health funds the Gestational Trophoblastic Disease Service. Cancer Research UK part funds a fluorescence activated cell sorting facility used to distinguish PHM from CHM. Cancer Treatment and Research Trust funds GTD research. 1
a spontaneous vaginal delivery but did not develop pGTD. 15 women developed pGTD (19%, 95% CI 11–30), including three of 19 (16%, 3–39%) who had a termination during the first trimester and 12 of 58 (21%, 11–33%) who progressed into the second trimester (figure 2). There was no significant difference in the risk of developing pGTD between these two groups (p=0·54, odds ratio 1·4, 95% CI 0·3–8·6). Single-agent chemotherapy cured all three patients whose pregnancy was terminated and seven of 12 of the women who developed pGTD after continuing their pregnancy. Only four patients required multi-agent chemotherapy. There were no maternal deaths and all women were well and free of pGTD at a median follow-up of 12 years (range 2–31). Our data indicate that 60% of CHM and healthy co-twin pregnancies, which are not electively terminated, will result in either intrauterine death of the co-twin or spontaneous pregnancy loss. Most of these events happen in the second trimester, before fetal viability. Nevertheless, nearly 40% of women who choose to continue their pregnancies have live babies, most of which are delivered beyond 32 weeks’ gestation, avoiding the risks of prematurity. CHM and healthy co-twin pregnancies can be associated with potentially lifethreatening maternal complications such as thromboembolic disease and severe pre-eclampsia. However, only one patient had pre-eclampsia that might not have been trophoblast related, and the 6% risk of developing severe pre-eclampsia, although higher than in normal twin pregnancies, is no greater than that of continuing singleton CHM pregnancies.3 The development of pGTD is another potential life threatening risk that can arise after continuation of CHM and healthy co-twin pregnancies. Crucially, our data show that the risk of pGTD does not differ between women undergoing elective first-trimester termination of pregnancy and those who continue their pregnancies. Furthermore, the risk is similar to that after a singleton CHM (16%).1 Moreover, all women with pGTD were successfully treated with chemotherapy. By contrast with our results, those of a previous smaller series of twin pregnancies with CHM indicated a much higher overall risk of pGTD (50%).4 This discrepancy might result from ascertainment bias, since the previous study identified cases via mailed questionaires sent to 87 hospital departments nationally. However, our larger study is still underpowered to exclude small but significant increases in the risk of pGTD. Furthermore, although there were no maternal deaths in our series, we cannot exclude the possiblity of a death due to a complication of such pregnancies in future. Our findings, therefore, suggest that women with twin
2166
2 3
4
5
Seckl MJ, Fisher RA, Salerno G, et al. Choriocarcinoma and partial hydatidiform moles. Lancet 2000; 356: 36–39. Goldstein DP, Berkowitz RS. Current management of complete and partial molar pregnancy. J Reprod Med 1994; 39: 139–46. Newlands ES. Presentation and management of persistent gestational trophoblastic disease and gestational trophoblastic tumors in the UK. In: Hancock BW, Newlands ES, Berkowitz RS, eds. Gestational trophoblastic disease. London: Chapman and Hall, 1997. Matsui H, Sekiya S, Hando T, Wake N, Tomoda Y. Hydatidiform mole coexistent with a twin live fetus: a national collaborative study in Japan. Hum Reprod 2000; 15: 608–11. Paradinas FJ. The histological diagnosis of hydatidiform moles. Curr Diagnostic Pathology 1994; 1: 24–31.
Trophoblastic Disease Unit, Departments of Histopathology (N J Sebire MD, Prof F J Paradinas FRCPath) and Cancer Medicine (M Foskett MA, R A Fisher PhD, D A Short, Prof E S Newlands FRCP, M J Seckl FRCP), Imperial College School of Medicine at Charing Cross Hospital, London W6 8RF, UK Correspondence to: Dr Michael J Seckl (e-mail: [email protected])
Diagnosis of vascular graft infections with antibodies against staphylococcal slime antigens Laura Selan, Claudio Passariello, Luigi Rizzo, Paola Varesi, Francesco Speziale, Giulio Renzini, Maria Cristina Thaller, Paolo Fiorani, Gian Maria Rossolini Late-onset infections of synthetic vascular grafts (LO-SVGIs) are generally caused by staphylococci that produce a slime polysaccharide and grow as a biofilm on the graft surface. We developed an ELISA to detect serum antibodies against staphylococcal slime polysaccharide antigens (SSPA). Patients with an ongoing staphylococcal LO-SVGI had greater titres of IgM antibodies against SSPA than did patients in other groups. Antibody titres of 0·40 ELISA units (EU) or more, or 0·35 EU or more detected 97% and 100% of staphylococcal LO-SVGIs, respectively, 0% and 2% titre/unit false-positive results. Our findings suggest that such an ELISA represents a sensitive, specific, and non-invasive diagnostic test for staphylococcal LO-SVGIs.
Lancet 2002; 359: 2166–68
Infections can cause severe complications soon after implantation of synthetic vascular grafts or at a later stage. Late-onset infections, can arise months or years after
THE LANCET • Vol 359 • June 22, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
RESEARCH LETTERS
Group A (n=38) IgM titre (mean, SD) (EU) IgG titre (mean, SD) (EU) Positive tests for an IgM titre (number, %) (EU) ⭓0·35 ⭓0·40
0·69 (0·37) 1·13 (0·36) 38 (100%) 37 (97%)
B (n=10)
C (n=16)
D (n=17)
E (n=58)
B+C+D+E (n=101)
0·28 (0·07) 0·86 (0·46)
0·26 (0·06) 0·82 (0·57)
0·18 (0·05) 0·46 (0·27)
0·17 (0·06) 0·55 (0·26)
0·19 (0·08) 0·60 (0·37)
2 (20%) 0
0 0
0 0
0 0
2 (2%) 0
Group A=patients with an ongoing LO-SVGI caused by Staphylococcus epidermidis (n=6), S aureus (n=2), coagulase-negative staphylococci other than S epidermidis (n=24), and mixed infection by one more staphylococcal species plus enterococci, Pseudomonas aeruginosa, or Escherichia coli (n=6); group B=patients with LO-SVGI caused by bacteria other than staphylococcus—P aeruginosa (n=2), mixed infections by gram-negative bacilli (Pseudomonas spp or Enterobacteriaceae; n=5), mixed infection by Enterococcus spp and Enterobacter cloacae (n=1), mixed fungal-bacterial infections (Candida albicans plus Enterococcus spp or plus P aeruginosa) (n=2); group C=healthy patients with a history of staphylococcal LO-SVGI, followed by successful graft replacement; group D=healthy patients with a synthetic vascular graft, with no previous history of graft infections; group E=patients with no prosthetic device.
Comparison of titres of IgM and IgG antibodies against staphylococcal slime polysaccharide antigens (SSPA) (ELISA units [EU]) in sera from patients with an ongoing staphylococcal late-onset infection of synthetic vascular graft (LO-SVGI) and in controls
THE LANCET • Vol 359 • June 22, 2002 • www.thelancet.com
attending the hospital for non-infectious reasons (group E). Each serum sample was tested in triplicate in two different assays (SDs were always ⬍10%), and we recorded the mean value (table). We judged infections as LO-SVGIs if they arose after the graft had been implanted for at least 2 years. All infections (ongoing and past) had been microbiologically confirmed on the explanted graft. We obtained sera from infected patients shortly before graft replacement. Grafts were always in the aorto-femoral tract. The local ethics committee approved the study, and all patients provided written informed consent. Titres of IgM antibodies against SSPA were higher in patients with an ongoing staphylococcal LO-SVGI than in controls, with an almost complete absence of overlap between the values of the former patients and those of other controls (table and figure). IgM antibody titres against SSPA of 0·40 EU or more indicated an ongoing 2·0 1·8
ELISA units
implantation, have an incidence rate of up to 8% per year, and are mostly caused by coagulase-negative staphylococci.1,2 In late-onset infections of synthetic vascular grafts (LO-SVGIs), the graft is colonised by a microbial biofilm, which has great resistance to antimicrobial agents and host defences. Production of slime polysaccharide, which is crucial to biofilm generation and maintenance, plays a major part as a virulence factor.2 LO-SVGIs are usually characterised by a long, clinically silent period, and diagnosis is often made only at an advanced stage once severe complications, such as suture erosion or formation of abscesses or fistulae, have taken place.1 Specific and non-invasive laboratory tests to diagnose these infections are not available. Here, we describe an ELISA to diagnose serum antibodies to staphylococcal slime polysaccharide antigens (SSPA), which could be useful in the diagnosis of staphylococcal LO-SVGI. We prepared SSPA from Staphylococcus epidermidis SA1545 (a slime-producing clinical isolate) grown for 6 days at 37°C in Hussain, Hastings, and White chemically defined medium.3 We extracted SSPA, by homogenisation, from cells suspended in ice-cold normal saline, after a freeze-thaw step. The clarified homogenate was extensively dialysed against distilled water, freeze-dried, and rehydrated in 5% trichloroacetic acid to precipitate protein contaminants. We collected the SSPA from the supernatant by precipitation with cold ethanol, vacuum-dried the antigens, and resolubilised them in distilled water. We analysed the purified preparations by polyacrylamide gelelectrophoresis and quantified them by densitometric scanning of gels stained with 5% Alcian blue and destained with 1% acetic acid—45% methanol with known amounts of colominic acid (Sigma, St Louis, MO, USA)—as a standard. We did ELISAs in polystyrene microtitre wells (NUNC Maxisorp, Life Technologies, Milan, Italy) sensitised with 0·5 g of SSPA, and blocked with phosphate buffered saline (PBS) pH 7·2, containing 10% calcium-free soy milk. Serum dilutions (1/160 for IgM and 1/320 for IgG) were left to react for 1 h at 37°C. After washing with PBS, containing 0·05% Tween 20, antigen-bound immunoglobulins were detected with peroxidase-conjugated antibodies against human IgG or IgM (Dako, Glostrup, Denmark). Antibody titres are expressed as units of optical density (ELISA units [EU]). We measured titres of antibodies (IgG and IgM) against SSPA in sera of 38 patients with a staphylococcal LO-SVGI (group A), and in various controls: ten patients with a nonstaphylococcal LO-SVGI (group B); 16 patients with a previous history of staphylococcal LO-SVGI with successful graft replacement done 6–72 months before (group C); 17 patients with a synthetic vascular graft implanted 14–78 months before study entry and no previous history of graft infections (group D); and 58 individuals free of prosthetic devices, randomly selected from patients
IgM
1·6 1·4 1·2 1·0 0·8 0·6 0·4 0·2 0 2·0 IgG 1·8 1·6 1·4 1·2 1·0 0·8 0·6 0·4 0·2 0
A
B
C
D
E
Group IgM and IgG antibody titres against SSPA Group A=individuals with staphylococcal LO-SVGI; group B=individuals with non-staphylococcal LO-SVGI; group C=individuals with previous history of staphylococcal LO-SVGI, with successful graft replacement; group D=individuals with synthetic vascular graft implanted 14–78 months before study entry and no previous history of graft infections; group E=individuals free of prosthetic devices.
2167
For personal use. Only reproduce with permission from The Lancet Publishing Group.
RESEARCH LETTERS
staphylococcal LO-SVGI, with detection rates of 97% (95% CI 84·6–99·9) and a false-positive result rate of 0% (0·0–4·6). This detection rate was increased to 100% (88·6–100·0) and the false-positive result rate to 2% (0·3–7·7) when antibody titres were 0·35 EU or more. The false-positives detected by a threshold titre set at 0·35 EU arose in patients with an LO-SVGI caused by bacteria other than S epidermidis, and were possibly related to a polymicrobic nature of the infection, including some staphylococcal component eventually replaced by other bacteria or missed by microbiological testing. Although the highest IgG antibody titres to SSPA were noted in individuals with an ongoing staphylococcal LO-SVGI, high titres were also seen in controls, which precluded the use of IgG titres for diagnostic purposes. Results of other studies have shown the potential of serological assays for diagnosis of coagulase-negative staphylococcal infections (not including LO-SVGIs).4,5 Measurement of serum IgM antibodies against SSPA could be a simple, sensitive, and non-invasive laboratory test useful in the diagnosis of staphylococcal LO-SVGIs. A cohort of patients, presenting with a suspected LO-SVGI, however, needs to be assessed before the true clinical predictive value of this test is known. The lower titres observed in patients who reported a previous staphylococcal LO-SVGI suggest that IgM antibody titres against SSPA decrease rapidly after successful graft substitution. This finding was confirmed in three individuals who had serological follow-up. In these patients, successful substitution of an infected graft was always associated with a reduction in IgM antibody titres (reduction to below 0·35 EU usually arose within 3 months), whereas recurrence of graft infection was associated with a renewed increase in IgM antibody titre. Testing for IgM antibodies against SSPA could, therefore, prove useful in the follow-up of patients with vascular grafts. The SSPA preparation described here could detect antibodies in LO-SVGIs caused by different staphylococcal species. Identification of conserved antigens produced in biofilms by the various causative agents could be used to develop a panel of immunoassays useful for diagnosis of all types of LO-SVGIs. Contributors C Passariello, L Selan, M C Thaller, and G M Rossolini developed the diagnostic technique; C Passariello, L Selan, and P Varesi did in-vitro experiments; L Rizzo, P Varesi, F Speziale, and P Fiorani were responsible for clinical sampling and collection of clinical data; and C Passariello, L Rizzo, and G M Rossolini did data analysis and wrote the report. The work was supervised by G Renzini, P Fiorani, M C Thaller, and G M Rossolini. L Selan, C Passariello, and L Rizzo contributed equally to this research.
Conflict of interest statement This work was funded, in part, by a research grant received in 1997 from Bracco SpA, Milan, Italy. The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. On the basis of the research results, a patent application for the diagnostic test was filed in 1998 (MI98A000197 and WO 99/40440 for international extension) by the sponsor, with L Selan, M C Thaller, C Passariello, and G M Rossolini as investors. After submission of the patent application, the sponsor granted the authors permission to submit the paper for publication. None of the authors have ever been employed by the sponsor. They have no stock ownership or patent applications related to the funding source, nor have they received consultancies, honoraria, paid expert testimony, and travel grants from the funding source within 3 years since the beginning of the study. 1 2
3
Calligaro KD, Veith FJ. Diagnosis and management of infected prosthetic aortic graft. Surgery 1991; 110: 805–13. Henke PK, Bergamini TM, Rose SM, Richardson JD. Current options in prosthetic vascular graft infection. Am Surg 1998; 64: 39–45. Hussain M, Hastings JG, White PJ. A chemically defined medium for slime production by coagulase-negative staphylococci. J Med Microbiol 1991; 34: 143–47.
2168
4
5
Karamanos NK, Syrokou A, Panagiotopoulou HS, Anastassiou ED, Dimitracopoulos G. The major 20-kDa polysaccharide of Staphylococcus epidermidis extracellular slime and its antibodies as powerful agents for detecting antibodies in blood serum and differentiating among slime-positive and -negative S epidermidis and other staphylococci species. Arch Biochem Biophys 1997; 342: 389–95. Lambert PA, Worthington T, Tebbs SE, Elliott TSJ. Lipid S, a novel Staphylococcus epidermidis exocellular antigen with potential for the serodiagnosis of infections. FEMS Immunol Med Microbiol 2000; 29: 195–202.
Dipartimento di Scienze di Sanità Pubblica (C Passariello MD, L Selan BS, G Renzini PhD) and Dipartimento di Chirurgia Vascolare (L Rizzo MD, F Speziale MD, Paolo Fiorani MD), Università di Roma “La Sapienza”, Roma; Dipartimento di Biologia, Università di Roma “Tor Vergata”, Roma (M C Thaller PhD); and Dipartimento di Biologia Molecolare, Sezione di Microbiologia, Università di Siena, Siena, Italy (P Varesi MD, G M Rossolini MD) Correspondence to: Dr Gian Maria Rossolini (e-mail: [email protected])
Association between aplastic anaemia and mutations in telomerase RNA Tom Vulliamy, Anna Marrone, Inderjeet Dokal, Philip J Mason The main cause of aplastic anaemia remains elusive. Germline mutations in the gene encoding the RNA component of telomerase (hTR) have been seen in the autosomal dominant form of dyskeratosis congenita—an inherited syndrome characterised by aplastic anaemia. By screening the hTR gene, we identified mutations in two of 17 patients with idiopathic aplastic anaemia, three of 27 patients with constitutional aplastic anaemia, but in none of 214 normal controls (p<0·0001). Furthermore, patients with hTR mutations had significantly shorter telomeres than age-matched controls (p=0·027). These data indicate that, in a subset of patients with aplastic anaemia, the disorder might be associated with a genetic lesion in the telomere maintenance pathway.
Lancet 2002; 359: 2168–70
Aplastic anaemia is a serious disorder of the bone marrow that affects between two and five people per million per year. About 75% of these cases are classified as idiopathic, and the main cause remains unexplained.1 In about 15% of cases, a drug or infection can be identified that precipitates the aplasia, although why only some individuals are susceptible is unclear. In about 5–10% of patients, the disease is constitutional—ie, is familial or presents with one or more other somatic abnormalities. Bone-marrow failure in X-linked dyskeratosis congenita, in which the initial presentation can occasionally be aplastic anaemia, is caused by mutations in the nucleolar protein, dyskerin. This protein binds to a particular class of small nucleolar RNAs called H/ACA as well as to telomerase RNA. The very rare autosomal dominant form of dyskeratosis congenita is caused by mutations in the RNA component of telomerase, and patients consequently have very short telomeres.2 Patients with idiopathic aplastic anaemia also have shorter telomeres than normal controls.3 We aimed to find out whether mutations in the telomerase RNA gene (hTR) can also be found in patients with aplastic anaemia. We screened the hTR gene of 44 patients with aplastic anaemia (17 patients with idiopathic disease and 27 patients with constitutional disease) and 214 normal controls for mutations after amplification by PCR of a 653 bp
THE LANCET • Vol 359 • June 22, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
pregnancies complicated by a CHM who wish to continue their pregnancy have a 40% chance of a successful outcome. Moreover, this choice is associated with a fairly small chance of serious obstetric complications and does not seem to increase the risk of development of pGTD.
77 women with CHM and healthy co-twin pregnancies
19 elective termination of pregnancy before week 14
58 remained pregnant until at least week 14
3 persistent GTD
12 persistent GTD
3 treated with single agent
8 treated with single agent
4 treated with multi agent
Contributors N J Sebire helped design the study, obtained and analysed data, and wrote the report; M Foskett helped with data collection and analysis; F J Paradinas contributed to pathological review and data collection; R A Fisher contributed to data collection and analysis, genetic studies, and writing of the report; R J Francis and D Short contributed to the data collection; E S Newlands contributed to project design, data analysis, and writing of the report; M J Seckl conceived and helped design the study, and contributed to data analysis and writing of the report.
Conflict of interest statement None declared.
Acknowledgments Figure 2: Frequency of persistent gestational trophoblastic disease (GTD) in a series of 77 twin pregnancies complicated by the co-existence of a complete hydatidiform mole (CHM) and a healthy fetus
MJS is supported by the Department of Health, Cancer Research UK, and the Cancer Treatment, Research Trust. The Department of Health funds the Gestational Trophoblastic Disease Service. Cancer Research UK part funds a fluorescence activated cell sorting facility used to distinguish PHM from CHM. Cancer Treatment and Research Trust funds GTD research. 1
a spontaneous vaginal delivery but did not develop pGTD. 15 women developed pGTD (19%, 95% CI 11–30), including three of 19 (16%, 3–39%) who had a termination during the first trimester and 12 of 58 (21%, 11–33%) who progressed into the second trimester (figure 2). There was no significant difference in the risk of developing pGTD between these two groups (p=0·54, odds ratio 1·4, 95% CI 0·3–8·6). Single-agent chemotherapy cured all three patients whose pregnancy was terminated and seven of 12 of the women who developed pGTD after continuing their pregnancy. Only four patients required multi-agent chemotherapy. There were no maternal deaths and all women were well and free of pGTD at a median follow-up of 12 years (range 2–31). Our data indicate that 60% of CHM and healthy co-twin pregnancies, which are not electively terminated, will result in either intrauterine death of the co-twin or spontaneous pregnancy loss. Most of these events happen in the second trimester, before fetal viability. Nevertheless, nearly 40% of women who choose to continue their pregnancies have live babies, most of which are delivered beyond 32 weeks’ gestation, avoiding the risks of prematurity. CHM and healthy co-twin pregnancies can be associated with potentially lifethreatening maternal complications such as thromboembolic disease and severe pre-eclampsia. However, only one patient had pre-eclampsia that might not have been trophoblast related, and the 6% risk of developing severe pre-eclampsia, although higher than in normal twin pregnancies, is no greater than that of continuing singleton CHM pregnancies.3 The development of pGTD is another potential life threatening risk that can arise after continuation of CHM and healthy co-twin pregnancies. Crucially, our data show that the risk of pGTD does not differ between women undergoing elective first-trimester termination of pregnancy and those who continue their pregnancies. Furthermore, the risk is similar to that after a singleton CHM (16%).1 Moreover, all women with pGTD were successfully treated with chemotherapy. By contrast with our results, those of a previous smaller series of twin pregnancies with CHM indicated a much higher overall risk of pGTD (50%).4 This discrepancy might result from ascertainment bias, since the previous study identified cases via mailed questionaires sent to 87 hospital departments nationally. However, our larger study is still underpowered to exclude small but significant increases in the risk of pGTD. Furthermore, although there were no maternal deaths in our series, we cannot exclude the possiblity of a death due to a complication of such pregnancies in future. Our findings, therefore, suggest that women with twin
2166
2 3
4
5
Seckl MJ, Fisher RA, Salerno G, et al. Choriocarcinoma and partial hydatidiform moles. Lancet 2000; 356: 36–39. Goldstein DP, Berkowitz RS. Current management of complete and partial molar pregnancy. J Reprod Med 1994; 39: 139–46. Newlands ES. Presentation and management of persistent gestational trophoblastic disease and gestational trophoblastic tumors in the UK. In: Hancock BW, Newlands ES, Berkowitz RS, eds. Gestational trophoblastic disease. London: Chapman and Hall, 1997. Matsui H, Sekiya S, Hando T, Wake N, Tomoda Y. Hydatidiform mole coexistent with a twin live fetus: a national collaborative study in Japan. Hum Reprod 2000; 15: 608–11. Paradinas FJ. The histological diagnosis of hydatidiform moles. Curr Diagnostic Pathology 1994; 1: 24–31.
Trophoblastic Disease Unit, Departments of Histopathology (N J Sebire MD, Prof F J Paradinas FRCPath) and Cancer Medicine (M Foskett MA, R A Fisher PhD, D A Short, Prof E S Newlands FRCP, M J Seckl FRCP), Imperial College School of Medicine at Charing Cross Hospital, London W6 8RF, UK Correspondence to: Dr Michael J Seckl (e-mail: [email protected])
Diagnosis of vascular graft infections with antibodies against staphylococcal slime antigens Laura Selan, Claudio Passariello, Luigi Rizzo, Paola Varesi, Francesco Speziale, Giulio Renzini, Maria Cristina Thaller, Paolo Fiorani, Gian Maria Rossolini Late-onset infections of synthetic vascular grafts (LO-SVGIs) are generally caused by staphylococci that produce a slime polysaccharide and grow as a biofilm on the graft surface. We developed an ELISA to detect serum antibodies against staphylococcal slime polysaccharide antigens (SSPA). Patients with an ongoing staphylococcal LO-SVGI had greater titres of IgM antibodies against SSPA than did patients in other groups. Antibody titres of 0·40 ELISA units (EU) or more, or 0·35 EU or more detected 97% and 100% of staphylococcal LO-SVGIs, respectively, 0% and 2% titre/unit false-positive results. Our findings suggest that such an ELISA represents a sensitive, specific, and non-invasive diagnostic test for staphylococcal LO-SVGIs.
Lancet 2002; 359: 2166–68
Infections can cause severe complications soon after implantation of synthetic vascular grafts or at a later stage. Late-onset infections, can arise months or years after
THE LANCET • Vol 359 • June 22, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.
RESEARCH LETTERS
Group A (n=38) IgM titre (mean, SD) (EU) IgG titre (mean, SD) (EU) Positive tests for an IgM titre (number, %) (EU) ⭓0·35 ⭓0·40
0·69 (0·37) 1·13 (0·36) 38 (100%) 37 (97%)
B (n=10)
C (n=16)
D (n=17)
E (n=58)
B+C+D+E (n=101)
0·28 (0·07) 0·86 (0·46)
0·26 (0·06) 0·82 (0·57)
0·18 (0·05) 0·46 (0·27)
0·17 (0·06) 0·55 (0·26)
0·19 (0·08) 0·60 (0·37)
2 (20%) 0
0 0
0 0
0 0
2 (2%) 0
Group A=patients with an ongoing LO-SVGI caused by Staphylococcus epidermidis (n=6), S aureus (n=2), coagulase-negative staphylococci other than S epidermidis (n=24), and mixed infection by one more staphylococcal species plus enterococci, Pseudomonas aeruginosa, or Escherichia coli (n=6); group B=patients with LO-SVGI caused by bacteria other than staphylococcus—P aeruginosa (n=2), mixed infections by gram-negative bacilli (Pseudomonas spp or Enterobacteriaceae; n=5), mixed infection by Enterococcus spp and Enterobacter cloacae (n=1), mixed fungal-bacterial infections (Candida albicans plus Enterococcus spp or plus P aeruginosa) (n=2); group C=healthy patients with a history of staphylococcal LO-SVGI, followed by successful graft replacement; group D=healthy patients with a synthetic vascular graft, with no previous history of graft infections; group E=patients with no prosthetic device.
Comparison of titres of IgM and IgG antibodies against staphylococcal slime polysaccharide antigens (SSPA) (ELISA units [EU]) in sera from patients with an ongoing staphylococcal late-onset infection of synthetic vascular graft (LO-SVGI) and in controls
THE LANCET • Vol 359 • June 22, 2002 • www.thelancet.com
attending the hospital for non-infectious reasons (group E). Each serum sample was tested in triplicate in two different assays (SDs were always ⬍10%), and we recorded the mean value (table). We judged infections as LO-SVGIs if they arose after the graft had been implanted for at least 2 years. All infections (ongoing and past) had been microbiologically confirmed on the explanted graft. We obtained sera from infected patients shortly before graft replacement. Grafts were always in the aorto-femoral tract. The local ethics committee approved the study, and all patients provided written informed consent. Titres of IgM antibodies against SSPA were higher in patients with an ongoing staphylococcal LO-SVGI than in controls, with an almost complete absence of overlap between the values of the former patients and those of other controls (table and figure). IgM antibody titres against SSPA of 0·40 EU or more indicated an ongoing 2·0 1·8
ELISA units
implantation, have an incidence rate of up to 8% per year, and are mostly caused by coagulase-negative staphylococci.1,2 In late-onset infections of synthetic vascular grafts (LO-SVGIs), the graft is colonised by a microbial biofilm, which has great resistance to antimicrobial agents and host defences. Production of slime polysaccharide, which is crucial to biofilm generation and maintenance, plays a major part as a virulence factor.2 LO-SVGIs are usually characterised by a long, clinically silent period, and diagnosis is often made only at an advanced stage once severe complications, such as suture erosion or formation of abscesses or fistulae, have taken place.1 Specific and non-invasive laboratory tests to diagnose these infections are not available. Here, we describe an ELISA to diagnose serum antibodies to staphylococcal slime polysaccharide antigens (SSPA), which could be useful in the diagnosis of staphylococcal LO-SVGI. We prepared SSPA from Staphylococcus epidermidis SA1545 (a slime-producing clinical isolate) grown for 6 days at 37°C in Hussain, Hastings, and White chemically defined medium.3 We extracted SSPA, by homogenisation, from cells suspended in ice-cold normal saline, after a freeze-thaw step. The clarified homogenate was extensively dialysed against distilled water, freeze-dried, and rehydrated in 5% trichloroacetic acid to precipitate protein contaminants. We collected the SSPA from the supernatant by precipitation with cold ethanol, vacuum-dried the antigens, and resolubilised them in distilled water. We analysed the purified preparations by polyacrylamide gelelectrophoresis and quantified them by densitometric scanning of gels stained with 5% Alcian blue and destained with 1% acetic acid—45% methanol with known amounts of colominic acid (Sigma, St Louis, MO, USA)—as a standard. We did ELISAs in polystyrene microtitre wells (NUNC Maxisorp, Life Technologies, Milan, Italy) sensitised with 0·5 g of SSPA, and blocked with phosphate buffered saline (PBS) pH 7·2, containing 10% calcium-free soy milk. Serum dilutions (1/160 for IgM and 1/320 for IgG) were left to react for 1 h at 37°C. After washing with PBS, containing 0·05% Tween 20, antigen-bound immunoglobulins were detected with peroxidase-conjugated antibodies against human IgG or IgM (Dako, Glostrup, Denmark). Antibody titres are expressed as units of optical density (ELISA units [EU]). We measured titres of antibodies (IgG and IgM) against SSPA in sera of 38 patients with a staphylococcal LO-SVGI (group A), and in various controls: ten patients with a nonstaphylococcal LO-SVGI (group B); 16 patients with a previous history of staphylococcal LO-SVGI with successful graft replacement done 6–72 months before (group C); 17 patients with a synthetic vascular graft implanted 14–78 months before study entry and no previous history of graft infections (group D); and 58 individuals free of prosthetic devices, randomly selected from patients
IgM
1·6 1·4 1·2 1·0 0·8 0·6 0·4 0·2 0 2·0 IgG 1·8 1·6 1·4 1·2 1·0 0·8 0·6 0·4 0·2 0
A
B
C
D
E
Group IgM and IgG antibody titres against SSPA Group A=individuals with staphylococcal LO-SVGI; group B=individuals with non-staphylococcal LO-SVGI; group C=individuals with previous history of staphylococcal LO-SVGI, with successful graft replacement; group D=individuals with synthetic vascular graft implanted 14–78 months before study entry and no previous history of graft infections; group E=individuals free of prosthetic devices.
2167
For personal use. Only reproduce with permission from The Lancet Publishing Group.
RESEARCH LETTERS
staphylococcal LO-SVGI, with detection rates of 97% (95% CI 84·6–99·9) and a false-positive result rate of 0% (0·0–4·6). This detection rate was increased to 100% (88·6–100·0) and the false-positive result rate to 2% (0·3–7·7) when antibody titres were 0·35 EU or more. The false-positives detected by a threshold titre set at 0·35 EU arose in patients with an LO-SVGI caused by bacteria other than S epidermidis, and were possibly related to a polymicrobic nature of the infection, including some staphylococcal component eventually replaced by other bacteria or missed by microbiological testing. Although the highest IgG antibody titres to SSPA were noted in individuals with an ongoing staphylococcal LO-SVGI, high titres were also seen in controls, which precluded the use of IgG titres for diagnostic purposes. Results of other studies have shown the potential of serological assays for diagnosis of coagulase-negative staphylococcal infections (not including LO-SVGIs).4,5 Measurement of serum IgM antibodies against SSPA could be a simple, sensitive, and non-invasive laboratory test useful in the diagnosis of staphylococcal LO-SVGIs. A cohort of patients, presenting with a suspected LO-SVGI, however, needs to be assessed before the true clinical predictive value of this test is known. The lower titres observed in patients who reported a previous staphylococcal LO-SVGI suggest that IgM antibody titres against SSPA decrease rapidly after successful graft substitution. This finding was confirmed in three individuals who had serological follow-up. In these patients, successful substitution of an infected graft was always associated with a reduction in IgM antibody titres (reduction to below 0·35 EU usually arose within 3 months), whereas recurrence of graft infection was associated with a renewed increase in IgM antibody titre. Testing for IgM antibodies against SSPA could, therefore, prove useful in the follow-up of patients with vascular grafts. The SSPA preparation described here could detect antibodies in LO-SVGIs caused by different staphylococcal species. Identification of conserved antigens produced in biofilms by the various causative agents could be used to develop a panel of immunoassays useful for diagnosis of all types of LO-SVGIs. Contributors C Passariello, L Selan, M C Thaller, and G M Rossolini developed the diagnostic technique; C Passariello, L Selan, and P Varesi did in-vitro experiments; L Rizzo, P Varesi, F Speziale, and P Fiorani were responsible for clinical sampling and collection of clinical data; and C Passariello, L Rizzo, and G M Rossolini did data analysis and wrote the report. The work was supervised by G Renzini, P Fiorani, M C Thaller, and G M Rossolini. L Selan, C Passariello, and L Rizzo contributed equally to this research.
Conflict of interest statement This work was funded, in part, by a research grant received in 1997 from Bracco SpA, Milan, Italy. The sponsors of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report. On the basis of the research results, a patent application for the diagnostic test was filed in 1998 (MI98A000197 and WO 99/40440 for international extension) by the sponsor, with L Selan, M C Thaller, C Passariello, and G M Rossolini as investors. After submission of the patent application, the sponsor granted the authors permission to submit the paper for publication. None of the authors have ever been employed by the sponsor. They have no stock ownership or patent applications related to the funding source, nor have they received consultancies, honoraria, paid expert testimony, and travel grants from the funding source within 3 years since the beginning of the study. 1 2
3
Calligaro KD, Veith FJ. Diagnosis and management of infected prosthetic aortic graft. Surgery 1991; 110: 805–13. Henke PK, Bergamini TM, Rose SM, Richardson JD. Current options in prosthetic vascular graft infection. Am Surg 1998; 64: 39–45. Hussain M, Hastings JG, White PJ. A chemically defined medium for slime production by coagulase-negative staphylococci. J Med Microbiol 1991; 34: 143–47.
2168
4
5
Karamanos NK, Syrokou A, Panagiotopoulou HS, Anastassiou ED, Dimitracopoulos G. The major 20-kDa polysaccharide of Staphylococcus epidermidis extracellular slime and its antibodies as powerful agents for detecting antibodies in blood serum and differentiating among slime-positive and -negative S epidermidis and other staphylococci species. Arch Biochem Biophys 1997; 342: 389–95. Lambert PA, Worthington T, Tebbs SE, Elliott TSJ. Lipid S, a novel Staphylococcus epidermidis exocellular antigen with potential for the serodiagnosis of infections. FEMS Immunol Med Microbiol 2000; 29: 195–202.
Dipartimento di Scienze di Sanità Pubblica (C Passariello MD, L Selan BS, G Renzini PhD) and Dipartimento di Chirurgia Vascolare (L Rizzo MD, F Speziale MD, Paolo Fiorani MD), Università di Roma “La Sapienza”, Roma; Dipartimento di Biologia, Università di Roma “Tor Vergata”, Roma (M C Thaller PhD); and Dipartimento di Biologia Molecolare, Sezione di Microbiologia, Università di Siena, Siena, Italy (P Varesi MD, G M Rossolini MD) Correspondence to: Dr Gian Maria Rossolini (e-mail: [email protected])
Association between aplastic anaemia and mutations in telomerase RNA Tom Vulliamy, Anna Marrone, Inderjeet Dokal, Philip J Mason The main cause of aplastic anaemia remains elusive. Germline mutations in the gene encoding the RNA component of telomerase (hTR) have been seen in the autosomal dominant form of dyskeratosis congenita—an inherited syndrome characterised by aplastic anaemia. By screening the hTR gene, we identified mutations in two of 17 patients with idiopathic aplastic anaemia, three of 27 patients with constitutional aplastic anaemia, but in none of 214 normal controls (p<0·0001). Furthermore, patients with hTR mutations had significantly shorter telomeres than age-matched controls (p=0·027). These data indicate that, in a subset of patients with aplastic anaemia, the disorder might be associated with a genetic lesion in the telomere maintenance pathway.
Lancet 2002; 359: 2168–70
Aplastic anaemia is a serious disorder of the bone marrow that affects between two and five people per million per year. About 75% of these cases are classified as idiopathic, and the main cause remains unexplained.1 In about 15% of cases, a drug or infection can be identified that precipitates the aplasia, although why only some individuals are susceptible is unclear. In about 5–10% of patients, the disease is constitutional—ie, is familial or presents with one or more other somatic abnormalities. Bone-marrow failure in X-linked dyskeratosis congenita, in which the initial presentation can occasionally be aplastic anaemia, is caused by mutations in the nucleolar protein, dyskerin. This protein binds to a particular class of small nucleolar RNAs called H/ACA as well as to telomerase RNA. The very rare autosomal dominant form of dyskeratosis congenita is caused by mutations in the RNA component of telomerase, and patients consequently have very short telomeres.2 Patients with idiopathic aplastic anaemia also have shorter telomeres than normal controls.3 We aimed to find out whether mutations in the telomerase RNA gene (hTR) can also be found in patients with aplastic anaemia. We screened the hTR gene of 44 patients with aplastic anaemia (17 patients with idiopathic disease and 27 patients with constitutional disease) and 214 normal controls for mutations after amplification by PCR of a 653 bp
THE LANCET • Vol 359 • June 22, 2002 • www.thelancet.com
For personal use. Only reproduce with permission from The Lancet Publishing Group.