| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology and Radiological Science, Johns Hopkins Hospital,
Phipps Building, Room B112, 600 N. Wolfe St., Baltimore, MD 21287.
2 Department of Environmental Health Sciences, Johns Hopkins Bloomberg School of
Public Health, Baltimore, MD 21205.
3 Geisinger Clinic, Center for Health Research and Rural Advocacy, Danville, PA
17822.
Received March 10, 2005;
accepted after revision August 22, 2005.
This research was supported by grant RO1 AG10785 from the National
Institutes on Aging (NIA). Its content is solely the responsibility of the
authors and does not necessarily represent the official views of the NIA.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Structural MR images obtained for the research study were reviewed by a neuroradiologist on an ongoing basis for findings of clinical concern. The subjects were recruited as part of a longitudinal research study designed to examine the effects of past lead exposure on CNS structures. The cohort examined consisted of 589 men who previously worked at an organolead manufacturing facility and had a wide range of lead exposure durations and intensities. MR images were also reviewed from 67 population-based control subjects from the same study who had no history of occupational lead exposure.
RESULTS. Incidental findings were detected in 84% (548/656) of research subjects. Of the 548 subjects with abnormalities, 30% (167) required no referral, 51% (280) required routine referral, 17% (93) required urgent referral, and 1.5% (8) required immediate referral. Incidental findings of all categories were observed in a larger percentage of older subjects, aged 60 years and above. Using multivariable logistic regression analysis, we found that age (p < 0.0001), but not lead history, was associated with an increased risk of incidental neuroradiologic abnormalities.
CONCLUSION. In the population examined, there was a high prevalence of incidental brain and head and neck abnormality. Of particular concern were the serious health problems observed requiring urgent or immediate referral in 18.5% of the subjects. These findings underscore the need for radiologists to evaluate the anatomic images generated by research studies, particularly those with an older population base.
Keywords: lead • medicolegal risk • MRI • neuroimaging • neurotoxicants • occupational exposure
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
All study subjects were men. The mean (SD) ages of former lead workers and control subjects who completed a research MRI were 60.1 (8.1) years (age range, 35-82 years) and 66.7 (6.3) years (age range, 53-78 years), respectively. In addition to dividing the scan pool into former lead workers and control clusters, subjects were separated by age into two groups (above 60 years and below 60 years) for the analysis. The mean (SD) ages of these two groups were 53.4 (4.9) years and 67.0 (4.3) years, for the younger (n = 299) and older groups (n = 357), respectively. The protocol for recruitment of former lead workers and control subjects was approved by the Johns Hopkins Bloomberg School of Public Health Committee on Human Research. Written informed consent was obtained from all participants.
MRI Parameters
Sagittal T1-weighted (TR/TE, 500/20), axial proton density weighted
(2,500/20), axial T2-weighted (2,500/80), and coronal spoiled gradientecho
scans (45/8) were performed through the brain as part of the research
protocol. Vascular stenoses and occlusions were assessed with a side-by-side
comparison of luminal diameter, loss of flow void on axial scans with
suppression pulse applied, loss of flow-ghosting artifacts, or demonstration
of intraluminal clot on all pulse sequences or a combination of these methods.
Corroboration was made with follow-up MR angiographic sequences in selected
individuals on clinical studies performed outside the defined research
protocol.
Classification of Incidental Findings
Incidental findings were classified into four categories according to a
scheme adopted in previous studies
[2-4,
7]. Abnormalities were rated as
either no referral necessary (normal or findings common in asymptomatic
subjects, e.g., minimal paranasal sinus disease); routine referral (e.g.,
acute sinusitis, small vessel ischemic disease); urgent referral required
within 1 week of study (e.g., nonacute parenchymal stroke, neoplasm); or
immediate referral required (e.g., definite aneurysm).
Statistical Analysis
Chi-square tests for contingency tables were used to compute p
values for crude (unadjusted) analysis. In addition, multivariable logistic
regression was used to evaluate associations of several potential risk factors
for incidental neuroradiologic abnormalities including age (as both a
continuous and binary variable, divided as described); smoking (current or
previous smoking history vs never); alcohol (current or previous alcohol
history vs never); hypertension (yes, defined as systolic blood pressure >
140, diastolic blood pressure > 90, or self-reported use of
anti-hypertensive medications vs no); apolipoprotein E genotype
(APOE) (one or two 
4 alleles vs none); and history of lead
exposure (former workers vs controls). APOE 4 is a gene
hypothesized to exacerbate the toxic effects of lead on cognitive function
[11]. Two separate models were
used to evaluate risk factors for two levels of severity of the study findings
(routine abnormalities vs normal and urgent or immediate abnormalities vs
normal). Odds ratios and 95% confidence intervals are presented from the
logistic regression models to identify factors that significantly affected the
presence and severity of incidental findings. Information on APOE
genotype was missing in 48 cases, and these subjects were thus excluded from
the regression analysis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
4 allele
(2.0% homozygous). In contrast, the 67 control subjects were 93% white, 7.5%
current smokers, 60% previous smokers, 61% current drinkers, 31% previous
drinkers, 28% hypertensive, and 29% carriers of at least one APOE
4 allele (1.6% homozygous). (See Materials and Methods section for
additional demographic data of each population.) Of the total 656 scans
examined, abnormalities were detected in 548 (84%) and abnormalities requiring
clinical follow-up evaluation were detected in 381 (58%). Of the 548
incidental findings, 167 cases required no referral (25% of the total
population, 30% of abnormalities), 280 cases required routine referral (43% of
the total population, 51% of abnormalities), 93 cases required urgent referral
(14% of the total population, 17% of abnormalities), and eight cases required
immediate referral (1.2% of the total population, 1.5% of abnormalities)
(Table 1).
|
The classification and distribution of all detected abnormalities are presented in Table 2. Ten research subjects had multiple lesions; eight subjects in the routine referral category presented with both acute sinusitis and small-vessel ischemic disease; two subjects in the urgent referral category presented with multiple findings: one with hydrocephalus plus hemorrhagic stroke and another with encephalomalacia plus neoplasm.
|
Although sinusitis was the primary reason for referral in only 18 cases (2.7% of the total population, 3.3% of abnormalities, and 6.4% of routine referrals), we found a 56% prevalence (367/656) of sinus inflammatory disease in the population of research subjects observed. The next most common abnormality noted was small-vessel ischemic changes in the white matter or deep gray matter structures, present in 247 (38%) of the 656 studies and representing 45% of the 548 reported abnormalities. This was the cause for referral in 88.2% (247/280) of the routine classifications.
Incidental findings were identified in a larger percentage of the older group (89%, 319/357) than in the younger group (77%, 229/299, p < 0.0001). In addition, radiologic abnormalities requiring follow-up (routine, urgent, and severe) were observed more frequently in the older group (71%, 253/358) than in the younger group (43%, 128/299, p < 0.0001). Similarly, severe (urgent or immediate referral required) abnormalities were observed in 12% (35/299) of the younger versus 18% (66/357) of the older group (p < 0.0001).
Incidental abnormalities were detected in 83% (486/589) of MRI studies of lead-exposed subjects and in 93% (62/67) of MRI studies of control subjects. Of the 486 incidental findings in former lead workers, 155 cases required no referral (26% of the 589 former lead workers, 32% of the 486 former lead workers with abnormalities), 243 cases required routine referral (41% of the 589 former lead workers, 50% of the 486 former lead workers with abnormalities), 81 cases required urgent referral (14% of the 589 former lead workers, 17% of the 486 former lead workers with abnormalities), and seven cases required immediate referral (1.2% of the 589 former lead workers, 1.4% of the 486 former workers with abnormalities). Of the 62 incidental findings in control subjects, 12 cases required no referral (18% of the 67 control subjects, 19% of the 62 control subjects with abnormalities), 37 cases required routine referral (55% of the 67 control subjects, 60% of the 62 control subjects with abnormalities), 12 cases required urgent referral (18% of the 67 control subjects, 19% of the 62 control subjects with abnormalities), and one case required immediate referral (1.5% of the 67 control subjects, 1.6% of the 62 control subjects with abnormalities) (Table 3).
|
Although the demographics of the lead-exposed and control groups were similar, the prevalence of incidental findings in each group was compared to assess the validity of presenting the data together as a whole. Interestingly, incidental findings were identified in a larger percentage of the control group (93%, 62/67) than in the former lead workers (83%, 486/589, p < 0.05). Radiologic abnormalities requiring follow-up (routine, urgent, and severe) were also observed more frequently in the control group (75%, 50/67) than in the former lead workers (56%, 331/589, p < 0.01). And similarly, severe (urgent or immediate referral required) abnormalities were observed in 19% (13/67) of the control versus 15% (88/589) of the former lead workers (p < 0.05).
Given these findings, logistic regression was next used to identify risk factors for incidental neuroradiologic findings in models that evaluated age, smoking history, alcohol history, hypertension, APOE genotype, and former occupational lead exposure. These adjusted results showed that age was a strong predictor of mild (routine referral required) and moderate to severe (urgent or immediate referral required) versus no incidental findings when modeled as a continuous variable, with an odds ratio (95% confidence interval, p) of 1.11 (1.08-1.14, p < 0.0001) increased odds per year of age for mild abnormalities and 1.12 (1.08-1.16, p < 0.0001) increased odds per year of age for moderate to severe abnormalities.
When age was modeled as a binary variable, older (> 60 years old) adults were found to have an odds ratio (95% confidence interval, p) of 3.23 (2.20-4.72, p < 0.0001) for mild disease and 3.36 (1.97-5.74, p < 0.0001) for moderate to severe disease when compared with younger (< 60 years old) adults. We found that APOE genotype was predictive of mild abnormalities in all ages with an odds ratio (95% confidence interval, p) of 1.73 (1.13-2.63, p = 0.011). However, our analysis revealed that no other variables tested, including lead-exposure history, hypertension, and smoking and alcohol history, were associated with risk of either mild or moderate to severe incidental neuroradiologic findings.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our population of research subjects— men with an average age falling within the sixth to seventh decade of life and a history of past occupational exposure to organolead—we found a high prevalence of severe (urgent or immediate referral required) incidental brain abnormalities (15%, 88/589) and paranasal sinus disease (55%, 326/589) compared with findings cited in previous studies. Multivariable analysis showed that age was a strong predictor of mild (routine referral required) and moderate to severe (urgent or immediate referral required) versus no incidental findings when modeled as a continuous or binary variable. Given that the mean ages of the populations examined in the studies reviewed above were close to, if not more than, half that of the populations examined in this study, it is likely that the high prevalence of severe incidental brain and head and neck abnormalities may be due to older age alone. Furthermore, in the one study examining incidental findings in an elderly population [6], the authors comment that their findings are likely to be underestimated because their population of community-dwelling people (65 years old and older) who participated in the Cardiovascular Health Study tended to be healthier than those who did not participate. The high prevalence of incidental brain and head and neck abnormality reported here highlights the need for radiologists to evaluate the anatomic images generated by research studies, particularly those with an older population base.
To maintain consistency with earlier literature reporting incidental findings, abnormalities were classified into four categories according to a scheme adopted in previous studies [2-4, 7]. Thus, our classification system was limited by broad categories such as "neoplasm," which encompassed lesions eliciting varying degrees of concern. Neoplastic lesions identified included those suspicious for meningioma, pituitary adenoma, schwannoma, parotid pleomorphic adenoma, and so on. In addition, because radiologic abnormalities often require clinicopathologic correlation for definitive diagnosis, our study was further limited by the lack of formal follow-up to confirm the presence of findings in subjects with suspected incidental neuroradiologic abnormalities. Thus, our findings may represent an overestimate of actual disease in the population in question. Finally, our study was limited by potential risk factors not evaluated in the logistic regression analysis, such as the white, male, blue-collar demographic, which may reduce the external validity of our findings.
In large-scale studies, even relatively uncommon but serious abnormalities will be found in some subjects. We have recently reported that cumulative lead dose, as estimated by measurement of tibia lead, was associated with an increased prevalence and severity of white matter lesions in the study population reported here [8]. Tibia lead was also associated with longitudinal declines in neurobehavioral test scores in a wide range of cognitive domains [10]. However, using multivariable logistic regression to compare controls to lead-exposed subjects, we found that age and APOE genotype, but not lead exposure history, were associated with the presence and severity of incidental neuroradiologic abnormalities. Here, we establish serious concerns for the appropriate management of incidental neuroradiologic findings generated by imaging research.
The obvious concern of incidental findings, especially serious ones, is ethical and involves the well-being of the research subject. However, there are additional more practical concerns surrounding medicolegal issues in the interpretation of research MRI studies. Illes et al. [2] highlight matters of protocol that need to be addressed, such as the appropriate operation of MRI scanners to ensure patient safety. Furthermore, lawsuits are an imminent threat when informed consent is ambiguous and a patient's health is at stake. In the field of genetics, preventive ethics has been used to address both ethical and legal issues engendered by incidental information revealed by genetic testing [17, 18]. This approach requires that genetic clinics identify policies for dealing with such findings before the implementation of any testing as a means to minimize potentially problematic issues.
Why not advocate having a trained radiologist review all neuroradiologic research studies that have structural images? To some this may seem like a natural solution to the problem of incidental findings. However, the advocates of a "don't ask; don't tell" policy cite the following rationale: having a neuroradiologist review films will incur a cost that would adversely affect funding sources; patients' expectations for a timely review that may be of clinical benefit will be falsely raised; patients may be motivated to participate if they secretly believe they harbor intracranial disease; detection of disease may adversely affect insurability if the patient is uninsured and rates if the patient is insured; the principal investigators will be unduly burdened with clinical care issues; some research centers, including those unaffiliated with hospitals, may not have access to radiologists; the cost of workup for false-positive or equivocal findings because the research scans are not definitive or are suboptimal may lead to increased health care expense; medicolegal implications could result regarding incorrect diagnoses, delayed diagnoses, and increased patient anxiety; and findings may adversely affect the collection of research data.
On the other side of the issue are advocates of patients' and research subjects' rights. They cite the importance of treating human subjects as individuals, not laboratory mice or pieces of data; the medicolegal risk of missing a pertinent abnormality that could lead to disabilities; the influence of incidental findings on the quality of the data; the benefit of earlier diagnosis of abnormalities in the course of many diseases; the improved environment for recruitment if subjects believed their images would be screened for abnormalities; and the overarching ethics of consistently providing optimal treatment.
The literature examining incidental neuroradiologic findings from brain imaging research suggests that serious clinical abnormalities will be detected in 1-2% of healthy cases. With those numbers, it has been previously suggested that the neuroscience community take a preventive approach and develop consistent, universal protocols that attend to the appropriate handling of incidental findings [1, 2, 5]. Here, we showed a high prevalence of both routine and serious clinical abnormalities in a population of older men with past occupational exposure to a known neurotoxicant who were enrolled in a longitudinal study designed to examine the effects of cumulative lead dose on CNS function. This work highlights and reinforces previous recommendations for the need for trained physicians, namely radiologists, in the mandatory, consistent, and accurate evaluation of MRI scans in anatomic brain research.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
