 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
To the Editor.—
Rintoul et al1 reviewed the records of 297 patients born after 1983 with an open meningomyelocele whose lesion was closed at their hospital and who were subsequently followed in their spina bifida clinic, to describe clinical outcomes. The authors are to be commended for documenting the results of conscientious and comprehensive care. I have, however, reservations about using their cross-sectional survey for their stated purpose "to serve as a comparison group ... to assess the efficacy" of in utero (fetal) surgery for meningomyelocele.
First, I do not believe that their rate of placement of a ventriculoperitoneal (VP) shunt for treatment of hydrocephalus (81%) or that from any cross-sectional survey can serve as a "gold standard" against which the rate of developing hydrocephalus after fetal surgery can be compared. The decision about who among infants with spina bifida and hydrocephalus requires placement of a VP shunt is arbitrary and the prevalence of shunting varies from clinic to clinic. For instance, in a cohort study performed by our clinic some years ago, we found that 95% of patients had been shunted.2 Likewise, there is a significant difference in the rate of VP shunt placement in the 2 series of patients who underwent fetal surgery for meningomyelocele, as acknowledged by the authors. In the series of patients who had fetal surgery reported from the authors’ institution, reported by Sutton et al,3 only 1 of 9 (11%) had VP shunt placement, but Bruner et al,4 doing fetal surgery on patients at a slightly more advanced gestational age, shunted 59%. We now follow a patient who had fetal surgery at another hospital and was not shunted there but who had marked ventricular dilatation when seen by us and was shunted subsequently. Because the decision about shunt placement is arbitrary, it would be helpful if objective measures of ventricular dilatation could be used to compare the rate of developing hydrocephalus after fetal surgery to that of patients who had surgery postnatally. Objective measures of ventricular dilatation might include the ratio of the average of the frontal and occipital horn widths to the interparietal distance5 or, potentially better yet, the ratio of ventricular volume to intracranial volume, determined by volumetric magnetic resonance imaging, at an age equivalent to 40 weeks’ gestation and/or prior to VP shunt placement.
Second, the data presented are not complete enough to use for comparison to determine if fetal surgery reduces the occurrence of Chiari II-related symptoms. The prevalence of a history of symptoms attributable to the Chiari II malformation was not recorded in the authors’ cross-sectional survey, although they state that 9.4% had undergone posterior fossa decompression and laminectomy. Six of the 297 patients (2%) died from Chiari II-related symptoms in their cross-sectional survey. These are lower prevalences of symptoms and mortality attributable to Chiari II-related symptoms than reported by us2 and by McLone et al.6 The authors’ better outcome can be substantiated convincingly only by presenting data from a cohort study of consecutively referred patients, all followed forward over time to a preestablished endpoint, including those patients whose care was transferred to another hospital.
Third, motor function was not accurately enough determined to use for comparison. McDonald et al,7 in a carefully performed study, demonstrated that neurosegmental level was not uniform in individual patients with meningomyelocele and that patients should be "grouped according to the strength of specific muscles rather than by neuro-segmental level." Further, the authors presented the "best motor level" recorded in the chart, rather than the best (most accurately) determined motor level. In our own clinic, there is sometimes a difference between the 2.
Third, their data are insufficient to control for the confounding effect of method of delivery on motor outcome. Luthy et al8 demonstrated a substantially better motor outcome in patients who were delivered by elective cesarean section before a trial of labor than in those who were delivered vaginally or by cesarean section after a trial of labor. As the authors state, they do not have information about whether cesarean section was performed for their patients before or after a trial of labor. To determine if motor outcome is better after fetal surgery, it is necessary to compare motor outcome to that of infants delivered by elective cesarean section before a trial of labor, since this is "current best practice."
Fourth, the occurrence and age of occurrence of symptomatic tethered cord requiring surgery were not reported by the authors. It is possible that symptomatic tethered cord might be more common in patients who have had fetal surgery than in those whose surgery was postnatal.
Rather than determine the efficacy of fetal surgery by comparing outcomes to a cross-sectional survey of patients who had surgery postnatally, it would be far better if a randomized trial was undertaken. New surgical procedures ought to be held to the same standards of proof of efficacy as are required for new medical therapies. A randomized trial for this experimental surgery is especially necessary from an ethical standpoint because of the risks to both the fetus and mother, because of the potential risk to future pregnancies (however small), and because of the option for therapeutic abortion. Ideally, outcome should not be determined by the surgeons who performed the fetal surgery, since it is now recognized that surgeons are "stakeholders" in operations that they develop or promulgate; although "the driving force [in developing new surgical procedures is to save lives, there is also the appeal of technical challenge and competition to meet demand."9
It is likely that the number of centers performing this as of yet unproven surgery will soon increase because all too often in medicine, invention is the mother of necessity. Now, therefore, is the time to organize, seek funding for, and conduct a randomized trial of fetal surgery for meningomyelocele, while it is still done at only a few centers. Comparison of outcomes of fetal surgery for meningomyelocele to any cross-sectional survey will not do to establish efficacy.
REFERENCES
In Reply.—
We appreciate the comments of Dr Worley regarding our article. We acknowledge the limitations of a historical cross-sectional survey. In our discussion, we made note of the limitations of comparing the incidence of shunting in patients with myelomeningocele repaired prenatally with those repaired postnatally. Although not entirely "arbitrary," there is certainly some subjectivity as to when a shunt is placed.
The primary purpose of this study was to gather data that could serve to estimate the possible benefits of fetal intervention that could serve in the design of a prospective clinical trial. Such a trial has now been funded by the National Institutes of Health, and should open shortly. It is an unblinded randomized clinical trial of fetuses with open myelomeningoceles. Initial intake information will go through a central referral at the Biostatistics Center at Georgetown University. If initial entry criteria are met, patients will be assigned to undergo additional evaluation at 1 of the 3 study centers: Children’s Hospital of Philadelphia, Vanderbilt University, or University of California at San Francisco. If all of the detailed entry criteria are met, patients will be randomly assigned to fetal closure or to standard postnatal closure at the study center. Detailed follow-up evaluation will be performed according to a standard battery of tests by individuals not directly involved with the surgical procedures. The criteria for shunt placement have been defined, although the study will rely on the cooperation of all neurosurgeons, because these patients will require care in their home communities. It is estimated by the power calculations based on our historical control series and the early results of the fetal surgery experience that 200 patients will be required to test the hypothesis that fetal surgery significantly decreases the incidence of shunting in this population. Appropriate interim analysis and stop rules are in place. The results of the trial should be available in 2006.
The success of this trial will depend on the cooperation of the entire medical community. New fetal centers must not allow a "back door" to randomization. It is hoped that pediatricians and perinatologists will be supportive of our efforts.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
||
|
||
Home | My Pediatrics | Journal Information | Current Issue | Past Issues | Subscriptions & Services | Contact Us Click here for faster international access © 2003 American Academy of Pediatrics. All rights reserved. |
||
|
||
