Letters to the Editor |
Author's response to comments by Sigal and Hassett, Phillips et al., and Shapiro et al.
In 1995, Sigal1 wrote ‘We must be aware of the mythology surrounding Lyme disease in our communities and counter with facts and the results of scientific studies’. To do this, we pooled the data from all scientific studies on random selections of patients who had had a diagnosis of Lyme borreliosis (LB) a few years earlier and, for comparison, random selections of subjects without LB from the general population. The prevalence of symptoms in LB patients was over and above the underlying prevalence of symptoms from other diseases such as fibromyalgia in the general population. Our meta-analysis2 shows clearly that a small percentage of patients with LB have symptoms persisting for years. No other data have been provided that contradict this.
Inevitably, diagnoses based on subjective patient reports are prone to error, particularly of post-LB syndrome where the original diagnosis of LB may be in doubt. Steere et al.3 concluded that a large proportion (57%) of patients referred to their Lyme Disease Clinic had not had LB, but these patients would not be representative if many were selected for referral because their diagnosis was uncertain. Fatigue, musculoskeletal problems and neurocognitive difficulties are relatively common in the general population, and, as pointed out by Shapiro et al.,4 some LB patients with symptoms due to other disorders may misattribute them to post-LB syndrome. In their commentary Sigal and Hassett5 report that many of the patients referred to their centre had fibromyalgia, and not post-LB syndrome. The pain following LB seems to be mostly roving, asymmetrical pain in the limbs, which is unlike the pain required for a diagnosis of fibromyalgia.6 So, although fibromyalgia is often accompanied by fatigue and forgetfulness, it should usually be distinguishable from post-LB syndrome. Uncertainty and misdiagnosis in some patients does not mean that these two disorders are not distinct entities. There is clearly a potential for misdiagnosis of post-LB syndrome, and that is an important issue, but it was not the topic of our meta-analysis.
Sigal and Hassett report that many of the patients with post-LB syndrome referred to their centre had positive scores on depression and anxiety scales. Depression and anxiety scales often include symptoms such as fatigue, listlessness, slowed speech, difficulties in working, concentration and memory problems, muscle aches and pain, increased sweating, and weight changes, all of which may be symptoms of post-LB syndrome. And with the distress that often arises from such a chronic illness, it is not surprising if some patients have positive scores on these scales. They also state that this disorder is seen predominantly in women implying this is evidence that it has a psychological basis. Fibromyalgia and many autoimmune diseases are also seen more often in women. It cannot therefore be concluded from their observations that the persistent symptoms following LB are simply due to psychological disturbances.
Shapiro et al. state that the usual course after treatment for LB is a slow resolution of symptoms over weeks to months, and cite three studies. Actually, the results of those three studies are consistent with the results of our meta-analysis. In the first study on patients with early LB,7 20% of patients had not completely responded 12 months after treatment, and 13% had not completely responded after 30 months. In the second study on patients treated for late LB,8 the investigators rated 15% of the patients not cured 12 months after treatment. In the third study on Lyme encephalopathy,9 61% of the patients stated they had not completely recovered after 12–24 months. However, in all three studies, almost all the patients with persistent symptoms had improved.
Shapiro et al. consider the findings of our meta-analysis misleading for a number of reasons, none of which invalidates the main conclusion of our meta-analysis:
- They claim that difficulties in patient recruitment in the clinical trials on long-term treatment point to a low prevalence of the disorder. Even if there are 10 000 or more patients with post-LB syndrome on the US north–east coast, they will be scattered among the many millions of inhabitants. It is therefore not surprising that the clinical trials on treatment of post-LB syndrome had difficulties finding patients. Many clinical trials in even quite common diseases have difficulties with recruitment. This cannot be used as a measure of the prevalence of a disorder, or as evidence that the results of our meta-analysis are incorrect.
- They object to retrospective studies (despite Shapiro using this design himself). This is a valid epidemiological design, and possible biases were discussed in the meta-analysis.
- They object to the lack of timely antimicrobial therapy in some patients and the over-representation of patients with extracutaneous manifestations. The meta-analysis is valid regardless of the composition of the patients. Separate estimates of the proportion with persistent symptoms may be necessary for different groups of patients (e.g. defined by time from infection to start of treatment, stage of LB at start of treatment, age, and sex).
- They object to the fact that some LB patients had not received antibiotic treatment in two of the studies. In the study by Shadick et al.,10 the differences between patients and controls in the proportions with symptoms were well above the proportion with no treatment. Furthermore, Seltzer et al.11 reported that the frequency of symptoms was similar in those who had and those who had not received antibiotics, except for joint or muscle aches. This is not therefore the explanation for the observed differences.
- They object to the use of antimicrobial regimens that are no longer recommended. In three of the studies,10,12,13 some patients had received oral penicillin or intravenous oxacillin, which are no longer recommended. It is not clear whether this had any effect on the results of the meta-analysis.
- They note the lack of 2-stage conditional serologic testing and the use of older serologic methods. This is admittedly a weakness of these studies, particularly the earlier studies. The patients in four of the studies met the CDC case definition for LB, i.e. history of eythema migrans, or clinical manifestations of LB confirmed by positive results on serologic analysis, or both. The possible proportion of patients with a misdiagnosis can be roughly estimated from the number of patients with manifestations of LB but no erythema migrans whose positive serologic result may have been false. The positive predictive value of the test (PPV: percentage of patients with a positive test result who actually have LB) depends on the prevalence of LB among those tested. If 50% of patients with clinical manifestations actually have LB,14 the PPV will be 80% for tests with low test accuracy.14 Assuming this, the differences in the proportions with symptoms between patients and controls were still clearly above the estimated proportion of misdiagnosed LB patients in three studies.10,12,15 The pattern of a higher prevalence of symptoms in LB patients was also seen in the newer studies. This does not therefore appear to be the explanation for the observed differences.
- They note that one study11 included some patients who did not meet the CDC case definition for LB. We had hoped to restrict our analysis to patients meeting the CDC case definition, but unfortunately were informed by Shapiro that the results had been stored and were unavailable. However, some results were provided in the publication by Seltzer et al.11 on the subset of 88 adults meeting the CDC case definition compared with their 88 controls, providing evidence of more ‘role limitations due to physical problems’ in LB patients (P = 0.05).
- They note that patients with chronic arthralgias, fatigue, or cognitive difficulties are now often tested for LB. This may lead to a selection bias whereby more patients are included who have such symptoms due to other disorders. The patients in three of the studies10,12,13 were diagnosed in the 1980s and early 1990s before the extensive media reports, and so this is not likely to have been an issue in these studies. Another study15 only included children with facial palsy attributed to LB. So here too this is not likely to have been an issue.
- They suggest that patients who have had LB may be more likely to notice and then report symptoms. However, the available data from physical examinations and neurocognitive tests confirm the subjective reports (see the meta-analysis discussion). Furthermore, the earlier studies, which also showed a clearly increased prevalence of symptoms in LB patients, were performed before the extensive media reports on this topic. Recall bias therefore does not appear to be the explanation for the consistent findings in these studies.
- They consider it important to assess pre-LB traumatic experiences. Investigation into why some LB patients have persistent symptoms and others do not was not the aim of our meta-analysis.
- They cite other results that, at first, appear inconsistent with the results of the meta-analysis. There were no significant differences in normal tasks of daily living in the study by Seltzer et al.,11 but it appears that only subjects who were attending school or work, doing housework, or attending gym class completed these questions, i.e. subjects unable to perform these activities were excluded. The results are therefore difficult to interpret. The results of neuropsychological tests were in the normal range in LB patients in the study by Vasquez et al.:15 the testing was performed in a small non-random subset of 20 out of the 43 children enrolled who had had LB. That the scores were in the normal range does not mean that none of the 43 patients had had deterioration in cognitive ability. Little can be concluded without a comparison group and inclusion of a large number of randomly selected patients. These results therefore have no impact on the interpretation of the meta-analysis.
Our meta-analysis was performed simply to show that some patients have symptoms persisting for years after treatment for LB, not to determine the cause of the persistent symptoms. Shapiro et al. propose that post-LB syndrome (or chronic LB) may be a ‘functional somatic syndrome’, i.e. a syndrome characterized not by demonstrable abnormalities but by non-specific subjective complaints that are interpreted according to whatever happens to be the latest fashion. This is belied by the specific and consistent pattern of symptoms seen with this disorder and by the results of objective assessments in the studies in the meta-analysis and the brain scans by Fallon et al.16 It is also evidence of a distinct syndrome that post-LB symptoms are associated with multiple erythema migrans skin lesions,17 with neurological symptoms in the acute phase,10,18 and with a delay in treatment.10,12,19 Phillips et al.20 provide an overview of the arguments on why the persistent symptoms may be due to ongoing infection. Some brief counterarguments are provided by Shapiro et al. The controversy whether this disorder is a post-infectious syndrome or due to chronic infection may take some time to resolve.
In the meantime, the main question for doctors and patients is what treatment, if any, should be given. Decisions about antibiotic treatment should be based on evidence from published clinical trials. The data indicate the following:
- Patients with a delay in antibiotic treatment for LB are more likely to have persistent symptoms, so rapid antibiotic treatment in the early stages is critical.
- Patients with confirmed LB who have received inadequate or no antibiotic treatment still appear to benefit from intravenous antibiotic treatment long after being infected (e.g. the patients who had previously only received oral antibiotics in the trial by Krupp et al.21).
- Patients with persistent symptoms after LB do not appear to benefit from further antibiotic treatment beyond the recommended few weeks of intravenous antibiotics.21–23 Even if a larger study with longer treatment duration were to show some improvement in some patients, the risk benefit balance of very long-term treatment seems poor.
 |
References |
|---|
 Top References |
|---|
1 Sigal LH. Anxiety and persistence of Lyme disease. Am J Med 1995; 98 (Suppl. 4A):74S–78S.[CrossRef][Medline]
2 Cairns V, Godwin J. Post-Lyme borreliosis syndrome: a meta-analysis of reported symptoms. Int J Epidemiol 2005;34:1340–45.
3 Steere AC, Taylor E, McHugh GL, Logigian EL. The over diagnosis of Lyme disease. JAMA 1993;269:1812–26.
4 Shapiro ED, Dattwyler R, Nadelman RB, Wormser GP. Response to meta-analysis. Int J Epidemiol 2005; 2005;34:1437–39.
5 Sigal LH, Hassett AL. Commentary: ‘What's in a name? That which we call a rose by any other name would smell as sweet’. Int J Epidemiol 2005; 34:1301–02.
6 Wolfe F, Smythe HA, Yunus MB et al. The American College of Rheumatology 1990 criteria for the classification of fibromyalgia. Report of the Multicenter Criteria Committee. Arthritis Rheum 1990;33:160–72.[Web of Science][Medline]
7 Wormser GP, Ramanathan R, Nowakowski J et al. Duration of antibiotic therapy for early Lyme disease. A randomized, double-blind, placebo-controlled trial. Ann Intern Med 2003;138:697–704.
8 Dattwyler RJ, Wormser GP, Rush TJ et al. A comparison of two treatment regimens of ceftriaxone in late Lyme disease. Wien Klin Wochenschr 2005;117:393–97.[CrossRef][Medline]
9 Logigian EL, Kaplan RF, Steere AC. Successful treatment of Lyme encephalopathy with intravenous ceftriaxone. J Infect Dis 1999;180:377–83.[CrossRef][Web of Science][Medline]
10 Shadick NA, Phillips CB, Sangha O et al. Musculoskeletal and neurologic outcomes in patients with previously treated Lyme disease. Ann Intern Med 1999;131:919–26.
11 Seltzer EG, Gerber MA, Cartter ML, Freudigman K, Shapiro ED. Long-term outcomes of persons with Lyme disease. JAMA 2000;283:609–16.
12 Shadick NA, Phillips CB, Logigian EL et al. The long-term clinical outcomes of Lyme disease. Ann Intern Med 1994;121:560–67.
13 Wang TJ, Sangha O, Phillips CB et al. Outcomes of children treated for Lyme disease. J Rheumatol 1998;25:2249–53.[Web of Science][Medline]
14 Brown SL, Hansen SL, Langone JJ. Role of serology in the diagnosis of Lyme disease. JAMA 1999;282:62–66.
15 Vazquez M, Sparrow SS, Shapiro ED. Long-term neuropsychologic and health outcomes of children with facial nerve palsy attributable to Lyme disease. Pediatrics 2003;112:e93–97.
16 Fallon BA, Keilp J, Prohovnik I, Heertum RV, Mann JJ. Regional blood flow and cognitive deficits in chronic Lyme disease. J Neuropsychiatry Clin Neurosci 2003;15:326–32.
17 Nowakowski J, Nadelman RB, Sell R et al. Long-term follow-up of patients with culture-confirmed Lyme disease. Am J Med 2003;115:91–96.[CrossRef][Web of Science][Medline]
18 Vrethem M, Hellblom L, Widlund M et al. Chronic symptoms are common in patients with neuroborreliosis—a questionnaire follow-up study. Acta Neurol Scand 2002;106:205–08.[CrossRef][Web of Science][Medline]
19 Asch ES, Bujak DI, Weiss M, Peterson MGE, Weinstein A. Lyme disease: an infectious and postinfectious syndrome. J Rheumatol 1994;21:454–61.[Web of Science][Medline]
20 Phillips SE, Burrascano JJ, Harris NS, Johnson L, Smith PV, Stricker RB. Chronic Infection in "Post-Lyme Borreliosis Syndrome". Int J Epidemiol 2005;34;1439–40.
21 Krupp LB, Hyman LG, Grimson R et al. Study and treatment of post Lyme disease (STOP-LD). Neurology 2003;60:1923–30.
22 Klempner MS, Linden TH, Evans J et al. Two controlled trials of antibiotic treatment in patients with persistent symptoms and a history of Lyme disease. N Engl J Med 2001;345:85–92.
23 Kaplan RF, Trevino RP, Johnson GM et al. Cognitive function in post-treatment Lyme disease. Do additional antibiotics help? Neurology 2003;60:1916–22.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||