| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters to the Editor |
Fetal Neurology Unit, Department of Obstetrics and Gynecology, Genetics Institute and Pediatric Neurology Unit, Edith Wolfson Medical Center and Sackler School of Medicine, Tel-Aviv University, PO Box 5, Holon, Israel. e-mail: malinger@inter.net.il
Editor:
In the October 2003 issue of Radiology, Dr Levine and colleagues (1) attempted to study the effect of magnetic resonance (MR) imaging on changes in diagnosis, patient counseling, and case management regarding fetuses suspected of having central nervous system (CNS) anomalies.
We find that despite the large number of patients, this study does not overcome the biases that we have commented on in a recent editorial in Ultrasound in Obstetrics and Gynecology (2). We would like to comment on these aspects.
In the introduction of their article, Dr Levine and colleagues (1) claim that their group and others have demonstrated that MR imaging improves diagnosis of CNS anomalies. However, this affirmation has not been proved scientifically. The only article in which statistical calculations have been performed showed no significant difference between ultrasonography (US) and MR imaging in the diagnosis of CNS anomalies (3). It is a pity that in the present study (1), they again failed to statistically analyze the effect of MR imaging versus US in the diagnosis and management of fetuses suspected of having CNS anomalies.
The introduction of new technologies should be based on prospective randomized studies in which the results of the different techniques are blinded to avoid interpretation bias. The authors seem to be, a priori, convinced of the superiority of MR imaging, so it is possible that they have invested less time and effort in the US examination. This seems obvious when the quality of US and MR images are compared and when the poor performance of US is evaluated in the diagnosis of easily diagnosed abnormalities, such as agenesis of the corpus callosum, septum pellucidum, and hemorrhage (4–6).
It is not clear from the text and tables how MR imaging findings led to modification of patient counseling or case management. We also do not see how some of the figures demonstrate this; in figure 3, the US findings of ventriculomegaly, Dandy-Walker malformation, and suspected intraventricular hemorrhage were sufficient to recommend termination of pregnancy. The MR imaging study was not indicated on the basis of the fact that the parents were not considering termination of pregnancy. It is not clear why the authors failed to observe the porencephalic changes by using coronal and/or transverse US planes. In figure 4, a head circumference that corresponds to 13 weeks of pregnancy in a 16-week-old fetus without visualization of normal intracranial structures is enough to recommend termination of pregnancy. It is worth noting that MR imaging did not resolve the diagnostic issue. In figure 5, the gestational age is not specified, but the US finding of severe ventriculomegaly with associated cerebellar hypoplasia is again enough to recommend termination of pregnancy based on an almost 100% chance of neurodevelopmental retardation.
Since MR imaging is expensive and not readily available in most centers worldwide, until large prospective studies prove the superiority of MR imaging in a statistically significant number of patients, we believe that it should remain an investigational tool, restricted to selected clinical situations in which the results are expected to modify case management.
REFERENCES
Lucille Salter Packard Children’s Hospital at Stanford, Palo Alto, Calif. e-mail: dlevine@caregroup.harvard.edu
We appreciate the comments of Drs Malinger, Lev, and Lerman-Sagie but feel that their conclusions are not valid.
They claim that there is no scientific proof that MR imaging improves diagnosis of CNS anomalies. There are multiple studies and case reports in the literature that demonstrate the ability of MR imaging to depict anomalies that are not apparent at US. They quote an article (1) in which receiver operating characteristic analysis was used for findings of 30 MR examinations performed for either fetal or uterine anomalies, for which no statistical improvement was shown for MR imaging versus US in the entire patient population. The small patient number is one reason why statistical significance was not achieved in that article, since MR imaging findings changed the US diagnosis in three of 16 (19%) fetuses with CNS anomalies. In our study (2), we assessed 145 fetuses with US abnormalities of the CNS and found additional MR imaging findings in 31.7%. Calculation of an increased sensitivity, specificity, and accuracy of MR imaging in this population would be misleading (and unnecessary), given the referral bias in patients in our study.
Dr Malinger and colleagues state that new technologies should be assessed in prospective trials in which the readers of each modality are blinded to avoid interpretation bias. Since our contention is that MR imaging is an adjunct to US and should not be performed without good-quality US, it is not realistic to assess MR imaging diagnosis in a vacuum, since it should only be performed after US. The authors of the letter imply that we invested less time and effort on our US examinations, since we were convinced of the superiority of MR imaging. To the contrary, all patients in our series underwent a full fetal US survey with additional targeted views of the brain to demonstrate the abnormality. As we stated in our materials and methods section, when the fetus was in cephalic presentation, we also performed vaginal US.
We disagree that the poor performance of US was caused by poor-quality examinations. There were many cases of partial agenesis of the corpus callosum, which can be a very difficult US diagnosis. Likewise, the cases of hemorrhage that were missed were located in regions that were difficult to visualize with US.
Most patients in our study were referred from either perinatologists or radiologists experienced in high-risk obstetric US. The anomalies missed during our US examinations were missed in their US examinations, as well. We acknowledge that high-quality US is crucial to accurately characterize the incremental improvement of MR imaging over US. For this reason, we did not compare the MR imaging results with referral diagnoses but instead compared them with the results of confirmatory US.
MR imaging findings led to modification of patient counseling whenever a more specific diagnosis was rendered and to modification of patient care when the obstetrician made a clear change in management based on the MR findings. Perhaps there are regional biases in how decisions are made regarding termination of pregnancy, because in our population, the types of findings we described had a clear effect on management. When patients are referred for MR imaging at our institution for evaluation of a US abnormality of the CNS, they are frequently undecided as to how to proceed with their pregnancy, and they desire additional information. The additional findings of porencephaly, hemorrhage, and severely disorganized brain are helpful to many patients and their clinicians in understanding the severity of the abnormality, reinforcing the feeling that they were able to come to an informed decision regarding pregnancy continuation or termination.
We agree that US remains the procedure of choice for fetal screening. However, if a CNS anomaly is visualized and more information is desired regarding the anomaly, MR imaging commonly adds additional information beyond that available with US. We acknowledge that MR imaging is not available worldwide. However, it is available to many patients in many parts of the world. Where it is available, and there are imagers with sufficient expertise to perform the study and interpret the images, then the information provided by MR imaging can be useful in patient understanding of the anomaly and decisions regarding the pregnancy.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  P. D Griffiths, M. N J Paley, E. Widjaja, C. Taylor, and E. H Whitby In utero magnetic resonance imaging for brain and spinal abnormalities in fetuses BMJ, September 10, 2005; 331(7516): 562 - 565. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
