HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH REGISTER		Author Keyword(s) Vol Page [Advanced]

Electronic Letters to:

SUPPLEMENTAL: Potential viral pathogenic mechanism for new variant inflammatory bowel disease [Full text]

eLetters: Submit a response to this article

Electronic letters published:

The value of epidemiology

C Raina MacIntyre   (8 May 2002)

Re: Publication by Uhlmann et al and commentary

John J O'Leary, V Uhlmann, C M Martin, O Sheils, L Pilkington, I Silva, A Killalea, S B Murch, and A J Wakefield   (17 April 2002)

Publication by Uhlmann et al and commentary

Shinji Ijichi   (17 April 2002)

Publication by Uhlmann et al and commentary

Neal A Halsey   (15 March 2002)

The value of epidemiology 8 May 2002
C Raina MacIntyre, Physican National Centre for Immunisation Research, Children's Hospital at Westmead, NSW, Australia Send letter to journal: Re: The value of epidemiology RainaM{at}chw.edu.au C Raina MacIntyre	Dear Editor We write in response to the letter by Professor John O’Leary (April 17th) and the editorial by Morris and Aldulaimi, which both mention “epidemiology” with some disdain. Morris and Aldulaimi’s claim that “epidemiology is a blunt tool”, is sadly misinformed. There are many disciplines within medical research, and each have their uses for differing research purposes. Unfortunately, researchers often fail to understand disciplines outside their own field. Questions of therapy are best answered by randomised, controlled, clinical trials. However, there are some research questions, particularly questions of aetiology and harm, which cannot usually be answered by clinical trials or case series, but require epidemiological studies. A classic example is the research finding of Sir Richard Doll and Sir Austin Bradford Hill that smoking causes lung cancer – this research question was answered by a case control study, which is an observational epidemiological study. [1] Case series analyses rank low on the scale of levels of evidence, but nevertheless, can be useful in describing new diseases or clinical conditions. A good example is the recognition of what we now know as HIV as a new disease, which was first described as a case series of pneumocystis carinii pneumonia in homosexual men. [2] However, the study of Uhlmann and colleagues,[3] by using “controls” and comparing them to “cases” relies (ironically) on epidemiology to make its point. The User’s Guide To The Medical Literature published in JAMA provides useful guidelines for evaluating the validity and of this study.[4] The first question we need to answer is “were there clearly identified groups that were similar with respect to important determinants of outcome other than the one of interest?” [4] The answer is no. Of cases, 85.5% were boys compared to 67% of controls (p<0.01). The age range of cases was 3-14 years, compared to 0-17 years for controls. [3] Although we are given no information on the process of selection of cases and controls, it seems the controls were a convenience sample (a sample of whatever specimens were available at the time) rather than a prospectively collected, matched sample. We are also given no information on the time of specimen collection for cases and controls – for all we know, control specimens may have come from 5 years ago, and cases from the past year. The second question: “were the outcomes and exposures measured in the same way in the groups being compared?” The presence of measles virus detected in tissue may have been measured in the same way, but without knowing the details of when and why specimens were collected for controls, it is uncertain, since the tests may not be as sensitive on older specimens. The diagnosis of developmental disorder is unclear, as we are given no information on the diagnostic process in cases, and whether the diagnosis was excluded in controls. Nor are we given a case definition of “developmental disorder”. The third question: Was follow up sufficiently long and complete? There was no follow up. The fourth question: Is the temporal relationship correct? As there is no information provided in the current study, we can refer back to the information provided by Wakefield et al in his 1998 study, where there was no temporal relationship to support the hypothesis that IBD was followed by autism. In fact, among the 12 cases described in 1998, information on the onset of bowel symptoms was lacking for half the cases, and in 4 of the remaining 6 children, the behavioural symptoms actually preceded the bowel symptoms.[5] The fifth question: “Is there a dose-response gradient?” – none has been shown. The study does not estimate risk, but a crude odds ratio of 60.9 (95% CI 19.4-205) for presence of MV in children with this “syndrome” can be calculated from the data. This is higher than the odds ratio for smoking causing lung cancer, and at such high levels of risk, if MV is, indeed, a causal factor for this yet to be confirmed syndrome, one would expect to see epidemiological evidence for it. Yet there is no such evidence. [6-10] We would do well to remember that before we can even apply the Bradford- Hill criteria for causation for this “new syndrome”, it is a prerequisite that the syndrome itself be given a precise case definition. References (1) Doll R, Hill AB. Lung cancer and other causes of death in relation to smoking: a second report on the mortality of British doctors. Br Med J 1956;2:1071 (2) Anonymous. Pneumocystis pneumonia--Los Angeles. MMWR - Morbidity & Mortality Weekly Report 1981;30:250-2. (3) Uhlmann V, Martin CM, Sheils O, et al. Potential viral pathogenic mechanism for new variant inflammatory bowel disease. Mol Path 2002;55:0-6 (4) Levine M, Walter S, Lee H, Haines T, Holbrook A, Moyer V. Users' guides to the medical literature. IV. How to use an article about harm. Evidence- Based Medicine Working Group. JAMA 1994;271:1615-9. (5) Wakefield AJ, Murch SH, Anthony A, et al. Ileal-lymphoid-nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. Lancet 1998;351:637-41. (6) Taylor B, Miller E, Farrington CP, et al. Autism and measles, mumps, and rubella vaccine: no epidemiological evidence for a causal association. Lancet 1999;353:2026-9. (7) Taylor B, Miller E, Lingam R, Andrews N, Simmons A, Stowe J. Measles, mumps, and rubella vaccination and bowel problems or developmental regression in children with autism: population study. BMJ 2002;324:393-6. (8) DeStefano F, Chen RT. Autism and measles, mumps, and rubella vaccine: No epidemiological evidence for a causal association. J Pediatr 2000;136:125-6. (9) Feeney M, Ciegg A, Winwood P, Snook J. A case-control study of measles vaccination and inflammatory bowel disease. The East Dorset Gastroenterology Group. Lancet 1997;350:764-6. (10) Patja A, Davidkin I, Kurki T, Kallio MJ, Valle M, Peltola H. Serious adverse events after measles-mumps-rubella vaccination during a fourteen-year prospective follow-up. Pediatr Infect Dis J 2000;19:1127-34. Dr C R MacIntyre National Centre for Immunisation Research and Surveillance of Vaccine Preventable Diseases New Children’s Hospital Westmead, NSW Australia
Re: Publication by Uhlmann et al and commentary 17 April 2002
John J O'Leary, MD Coombe Women's Hospital, Dublin, Ireland, V Uhlmann, C M Martin, O Sheils, L Pilkington, I Silva, A Killalea, S B Murch, and A J Wakefield Send letter to journal: Re: Re: Publication by Uhlmann et al and commentary joleary{at}coombe.ie John J O'Leary, et al.	Dear Editor In May 2000, Dr Neal Halsey was the senior author of an extensive review of the proceedings of a meeting convened to examine the very issues to which he alludes to in his opening paragraph [1]. This meeting, to which we refer later in our response, was convened by the American Academy of Pediatrics (AAP) and the Centre for Disease Control and Prevention. Dr Halsey has made certain observations in relation to our manuscript, which are dealt with below [2]. Specimen collection: No differences exist for specimen collection between cases and controls. Frozen biopsies, taken from all cases and controls, having been procured with a sterilised instrument, are placed, individually, in sterile containers. The container is sealed and immersed in liquid nitrogen. It is not opened again until it reaches the laboratory of Professor O'Leary in Dublin. At the Royal Free Hospital the procedure takes place 5 floors from where any laboratory work on measles virus has been carried out. Contamination is not an issue. All paraffin-embedded material from cases and controls had similar anti-contamination procedures applied, and had no contact with any environment where measles is handled. It is notable that measles virus has also been detected in specimens sent directly to Professor O'Leary's lab from other hospitals, including those in the US. Those analysing the biopsies in Professor O'Leary's lab have no knowledge of the clinical details of case or control children. The code of the initial molecular analyses that form part of the study was broken in public. False positive results are highly unlikely given the exactness of the controls used for TaqMan PCR, solution-phase PCR and in cell PCR. Control RNAs consistently yielded negative results, as described in the paper. Extraction controls of duplicate RNAs from cervical, breast and thyroid biopsies etc.were again, consistently negative. PCRs were set up at two geographically distant laboratories in Dublin. In addition there was judicious use of Taq Man PCR controls, including no template (NTC), no amplification controls (NAC), no probe controls (NPC), asymmetric Taq Man PCR controls, irrelevant primers, irrelevant probe etc. Differential detection of virus in inflamed versus normal tissue is highly unlikely since virus is not detected in ileal biopsies from children with established inflammatory diseases. In addition, gene-dosage correction assays were performed which specifically sought to address the issue of PCR inhibition, both in and between samples. Again, Dr Halsey's suggestion that we may be merely looking at persistence of portions of the measles virus in selected individuals with lymphonodular hyperplasia (LNH) is incorrect. As indicated in the paper, we did not identify MV genomes in children with isolated LNH, without co-existent developmental disorder. Genetic sequencing: The assays were designed to detect both wild type and vaccine strains of MV, and not to differentiate between them. Strain specific assignation is currently underway. Cases: Children with developmental disorders were diagnosed on the autistic spectrum. They were investigated for gastrointestinal symptoms. The patients were consecutively investigated in order to avoid selection bias. Controls: Biopsies were mucosal, whereas the seat of the granulomatous inflammation in Crohn's disease (the site of measles virus detection by Wakefield and others) is submucosal/serosal. The reason for including inflammatory controls was to address precisely the question as to whether or not measles virus genomic RNA was sequestered in foci of non-specific inflammation. It did not. Appendicectomies were also included in this study. As pointed out in the paper, the identification of Warthin Finkeldy giant cells indicative of MV infection is described in the appendix. The authors believed that the inclusion of an appendicectomy control cohort was appropriate to define the prevalence of MV infection in a randomly selected population of children. Terminology: Detailed study of the individual children with regressive autism and intestinal symptoms has identified, beyond doubt, a novel syndrome that includes an immune-mediated entero-colitis [3-6]. Epidemiology is wholly inadequate in its ability to assess, let alone refute, the validity of these findings. Please would Dr Halsey identify the scientific literature that demonstrates that chronic ileo-colonic lymphoid nodular hyperplasia - the condition from which these children suffer - is a normal response. Atypical exposure: Dr Halsey himself has confirmed, quite correctly, that if a biological reason exists to account for any adverse reactions due to atypical viral concurrent infections, then large comparative studies should be performed[7]. Clearly, one study of subacute sclerosing panencephalitis showing an association between measles and chicken pox [8] and another one not showing such an association [9] does not necessarily clarify the situation in relation to atypical/concurrent viral infections. Somewhat disingenuously, Dr Halsey raises an issue in relation to communications between the AAP and Professor O'Leary in the fall of 2000. Professor O'Leary wishes to correct, for the record, the erroneous statement made by Dr Halsey. The correspondence between the AAP and Professor O'Leary was in fact in relation to a special meeting, referred to above, that was convened to consider the issue of autism and possible links to MMR vaccine. At that time Professor O'Leary explained verbally and in writing to representatives of the AAP that he was unable to attend for personal reasons. He informed the AAP that the findings were being submitted for peer review and that further communications were inappropriate. John J O'Leary MD, DPhil, MSc, MRCPath, FFPathRCPI. on behalf of the authors References (1) Halsey NA, Hyman SL and the Conference Writing Panel. Measles, mumps, and rubella vaccine and autistic spectrum disorder: report, from the new challenges in childhood immunizations conference convened in Oak Brook, Illinois. Pediatrics; 2000;107:E84 (2) Ulmann V., Martin CM., Shiels O., Pilkington L., Silva I., Lillalea A., Murch SH., Wakefield AJ., O'Leary JJ. Potential viral pathogenic mechanism for new variant inflammatory bowel disease. Molecular Pathology. 2002;55:1-6 (3) Wakefield AJ, Murch SH, Anthony A,, Linnell J, Casson DM, Malik M, et al. Ileal-lymphoid nodular hyperplasia, non-specific colitis, and pervasive developmental disorder in children. Lancet. 1998:351: 637-641. (4) Wakefield AJ., Anthony A., Murch SH., Thomson M., Montgomery SM., Davis S., Walker-Smith JA. Enterocolitis in children with developmental disorders. Am J Gastroenterol. 2000;95:2285-2295 (5) Furlano R., Anthony A., Day R., Brown A., McGavery L., Thomson M., et al. Quantitative immunohistochemistry shows colonic epithelial pathology and gd-T cell infiltration in autistic enterocolitis. J Pediatrics 2001;138:366-372 (6) Torrente F., Machado N., Perez-Machado M., Furlano R., Thomson M., Davies S., et al. Enteropathy with T cell infiltration and epithelial IgG deposition jn autism. Molecular Psychiarty 2002 (in press). (7) F-D-C Reports, Inc., The Pink Sheet, Nov 13th 1995;pp11 (8) Detels R, Brody JA, McNew J, Edgar AH. Further epidemiological studies of subacute sclerosing panencephalitis. Lancet 1973;2:11-14. (9) Halsey NA., Modlin JF., Jabbour JT., Dubey L., Eddins DL., Ludwig DD. Risk factors in subacute sclerosing panencephalitis: a case-control study. Am J Epidemiol. 1980;111:415-424
Publication by Uhlmann et al and commentary 17 April 2002
Shinji Ijichi, M.D. Send letter to journal: Re: Publication by Uhlmann et al and commentary jyajya{at}po.synapse.ne.jp Shinji Ijichi	Dear Editors, In the light of the big impact of this article [1], the authors may have a responsibility to address and explain the following questions and problems. a) The reaction specificity is confirmed in primer/probe sets only for measles virus F and H gene, which are utilized in TaqMan RT-PCR (Fig. 2). The specificity of the set for N gene should also be established to provide a better reliability of RT in situ PCR results. b) The discrepancy between the band density of the PCR amplicons in agarose gel (Fig. 2A) and the luminescent intensity of the hybridized bands (Fig. 2B and 2C) clearly demonstrates that this primer/probe sets are not suitable for quantitative analyses including TaqMan RT-PCR. In F gene PCR amplicons, the agarose gel bands from positive controls are densest (lanes 1 and 2). However, the strongest signal in the Southern blot was obtained from the affected child 2 (lane 4), indicating the affinities between PCR amplicons and the specific probe vary on a case-by- case basis. Furthermore, in H gene PCR amplicons from the affected children, the intensity of hybridized bands (Fig. 2C, lanes 10-13) is not proportional to the density of agarose gel bands (Fig. 2A, lanes 10-13). c) TaqMan RT-PCR standards generated by cloning of PCR products must be sequenced and the cloned inserts should be checked, because amplicons are not necessarily pure. d) For the expanded speculation about virus-reservoirs and virus-host interactions in the DISCUSSION, authors should reveal the number of affected patients whose samples were evaluated by the combination of RT in situ PCR and immunohistochemistry (Fig. 4 maybe show just one patient). e) In the combination of RT in situ PCR and immunohistochemistry, a specific marker for lymphocytes should also be utilized. f) It is described in the figure legend that the samples from four affected children in Fig. 2 are fresh frozen biopsies. However, the number information of sample types (fresh frozen biopsies or fixed and embedded tissues) in both cases and controls is missing from the MATERIALS AND METHODS. g) Some results of control experiments are missing in the RESULTS. h) Authors should explain or address possible reasons of the discrepant results of TaqMan RT-PCR and RT in situ PCR in several cases ("six were positive by TaqMan RT-PCR but negative by in situ PCR, and five were positive by in situ PCR only"). i) Details of the gene dosage correction, reliability of the standard curves, and the criteria for positive results should be addressed in TaqMan RT-PCR methods. j) What is the purpose of the optimization of N2, H2, and F2 primer pairs? k) Case codes or serial case numbers should be used for further constructive discussions. l) The results of TaqMan RT-PCR should be separately demonstrated for F and H genes. Shinji Ijichi, M.D. Institute for EGT (www.iegt.org) Nagahama Shinryojyo 8-3 Nagahama, Shimokoshiki-mura, Satsuma-gun, Kagoshima 896-1411, Japan References (1) Uhlmann V, Martin CM, Sheils O, Pilkington L, Silva I, Killalea A, Murch SB, Wakefield AJ, O'Leary JJ. Potential viral pathogenic mechanism for new variant inflammatory bowel disease. J Clin Pathol: Mol Pathol 2002;55:0-6.
Publication by Uhlmann et al and commentary 15 March 2002
Neal A Halsey Institute for Vaccine Safety, John Hopkins School of Public Health Send letter to journal: Re: Publication by Uhlmann et al and commentary nhalsey{at}jhsph.edu Neal A Halsey	Dear Editors, Given the high level of public concern about MMR, inflammatory bowel disease and autism in recent years, investigators have a special responsibility to use rigorous study methods and to report essential details of their studies so that other scientists can properly interpret the results. Unfortunately, important details regarding several study methods were not provided in the recent study by Uhlmann et al[1]. The journal reviewers, editors, and authors of the accompanying editorial did not fulfill their obligations to identify these shortcomings and set a high standard for the evaluation and publication of research on this topic. In addition to the authors’ interpretation of the results, there are several possible explanations for the study findings: a) Contamination of some specimens during collection or processing. b) False positive results. c) Differences in specimen collection and processing for cases and controls. d) Persistence of measles virus (or portions of measles virus) in lymphoid tissue following measles disease or vaccination with differential detection of virus in inflamed vs. normal tissue. e) Persistence of portions of measles virus in selected individuals who also have lymphonodular hyperplasia. Following Professor John O’Leary’s presentation almost two years ago to a congressional subcommittee in the United States, many questions and concerns were raised about the limited description of the methods provided presented in his online publication. The American Academy of Pediatrics wrote to him in the fall of 2000 requesting more detail on his methods, but he never responded. Since he has chosen to publish these results, I am surprised by the failure to address concerns that were raised earlier. The following points regarding the epidemiologic methods employed in this study should be addressed. Other scientists with more laboratory expertise may address problems with the specific laboratory methods. 1. Specimen collection. The specimens were obtained from the gastroenterology unit at the Royal Free Hospital. Dr. Andrew Wakefield and his colleagues have been studying measles and measles vaccine viruses in this laboratory for several years. Therefore, the possibility of specimen contamination during collection or processing cannot be ruled out. Studies should be designed to collect specimens from patients with inflammatory bowel disease and controls by investigators who are not actively working with these viruses in their clinics or laboratories. The specimens should be processed in laboratories free of any measles viruses and aliquots sent to several different laboratories with the capability of detecting measles viruses using several different techniques. 2. Blinding or masking. No mention is made of masking laboratory investigators as to which specimens come from cases or controls. The code should be broken only after the studies are completed and investigators commit themselves to the results. 3. Concurrent collection of cases and controls. No mention is made as to the timing of the collection of specimens from cases and controls. If the control specimens were obtained after collection of specimens from cases, differences in methods could have been used in the handling of specimens. For example, some laboratories have had problems with inadvertent contamination and then taken steps to eliminate this problem. 4. Genetic sequencing. The authors have not provided key information regarding the viruses detected by RT-PCR. Was there variability in the genetic sequences of these viruses? All viruses mutate and some variability is expected. If there is no heterogeneity in the viruses obtained, this suggests the possibility of specimen contamination. 5. Cases. The 91 'affected patients' are inadequately described. The results section indicates that these patients had developmental disorders, but this is not noted in the methods and the types of disorders are not mentioned. What were the reasons that the children were biopsied? What proportion of all children examined had lymphonodular hyperplasia and how were the patients studied here selected? 6. Controls. The controls appear to have been selected on the basis of being developmentally normal, but they include a variety of conditions including patients with Crohn’s disease and ulcerative colitis. In previous publications, Dr Wakefield has used patients with Crohn’s disease and ulcerative colitis as cases and reported that many of these patients had positive results for tests of measles virus in intestinal tissue.[2] Were the patents studied here a subset of patients previously published selected on the basis of being negative for measles? 7. Terminology. The authors coined the term 'New variant inflammatory bowel disease'. This term appears to have been chosen to attract attention and public concern because of the heightened awareness of new variant Creutzfeld Jacob disease. The evidence presented here and in recent epidemiologic studies does not indicate that there is a new disorder or that there is an increased risk of inflammatory bowel disease in patients with autism or related conditions.[3,4] Lymphonodular hyperplasia may be a normal response to a variety of stimuli. 8. 'Atypical exposure'. The authors cite one small epidemiologic study suggesting a possible increased frequency of chickenpox and measles infections occurring within 6 months in children who subsequently developed subacute sclerosing panencephalitis (SSPE).[5] However, the authors failed to cite a subsequent larger study that found no differences in rates of chickenpox, mumps or rubella occurring within 6 months of measles for children who developed SSPE as compared to matched controls.[6] Measles viruses may persist in persons who do not develop SSPE. However, the data presented here are inadequate to come to such a conclusion. Moreover, if portions of the measles virus or the intact virus are maintained in lymphoid tissue following infection or vaccination, detection of portions of the virus could be enhanced in the presence of inflammatory conditions. I encourage these investigators to participate in coordinated efforts to further investigate the hypothesis and to limit speculation regarding the results of this study. Provision of the missing key information in their recent study would help this process. The recent publication does not change any of the conclusions from our recent review of this topic and I continue to support the use of MMR vaccine for prevention of important childhood diseases. Sincerely, Neal A. Halsey, MD Director, Institute for Vaccine Safety Professor and Director, Division of Disease Control International Health Johns Hopkins Bloomberg School of Public Health Baltimore, Maryland, USA References (1)Uhlmann V, Martin CM, Sheils O, Pilkington L, Silva I, Killalea A, Murch SB, Wakefield AJ, O'Leary JJ. Potential Viral Pathogenic Mechanism For New Variant Inflammatory Bowel Disease. J Clin Pathol: Mol Pathol 2002;55:0-6. (2)Kawashima H, Mori T, Kashiwagi Y, Takekuma K, Hoshika A, Wakefield A. Detection and sequencing of measles virus from peripheral mononuclear cells from patients with inflammatory bowel disease and autism. Dig Dis Sci 2000;45(4):723-9. (3)Halsey NA, Hyman SL. Measles-mumps-rubella vaccine and autistic spectrum disorder: report from the New Challenges in Childhood Immunizations Conference convened in Oak Brook, Illinois, June 12-13, 2000. Pediatrics 2001;107(5):E84. (4)Fombonne E, Chakrabarti S. No evidence for a new variant of measles- mumps-rubella-induced autism. Pediatrics 2001;108(4):E58. (5)Detels R, Brody JA, McNew J, Edgar AH. Further epidemiological studies of subacute sclerosing panencephalitis. Lancet 1973;2(7819):11- 4. (6)Halsey NA, Modlin JF, Jabbour JT, Dubey L, Eddins DL, Ludwig DD. Risk factors in subacute sclerosing panencephalitis: a case-control study. Am J Epidemiol 1980;111(4):415-24.