Statistics from Altmetric.com
In the Western world approximately 2–3% of all pregnancies result in an infant with a major malformation. Inevitably the parents of such children undergo a period of grieving for “the loss of their normal child” and as a part of that grief they seek to apportion blame. If ocular involvement is a feature of the malformation then it is during the period of their grief that we as ophthalmologists are likely to come into contact with the family. We need to understand this process and during our early contact with the family avoid the twin problems of either being dismissive of their search for a causative agent or, alternatively, being too eager to support their proposal that a particular action or event has been responsible for the outcome of the pregnancy. We are rarely aided in this by the attentions of the “popular press” where the demands of investigative journalism are incompatible with a studied evaluation of possible cause and effect. When trying to identify the effect of a possible teratogen on a pregnancy, there are well established criteria which help in the attribution of blame.
The possible association is often first recognised through case reporting (and that remains a major value of the isolated case report), but these can be misleading. The picture of affected pregnancies in cocaine addicted mothers is a case in point, where it begins to seem possible that some of the described abnormalities are a consequence of other features of the addict’s lifestyle, such as poor nutrition and alcohol abuse, rather than the cocaine itself. It is important, therefore, that epidemiological studies should be undertaken to demonstrate a statistical association between exposure to the “teratogen” and an abnormality in exposed children (ideally in different population groups). The thalidomide tragedy of the late 1950s and early ’60s demonstrated this mechanism, but thalidomide was remarkable in that as far as can be determined 20% to 25% of all exposed pregnancies resulted in a child with major malformations, while exposure between the 20th and 40th day after conception always resulted in some limb shortening.1 Few if any other teratogens have had such a “strike rate” (which allowed relatively rapid identification with epidemiological studies in Scandinavia particularly), so for other agents identifying an association may be more delayed.
Corroborating evidence can be derived from longitudinal studies if they are able to show an increase in a particular defect associated with exposure to the teratogen, accompanied by a reduction when the teratogen is eliminated. Again, this approach has been valuable with the rubella virus, identifying first a high incidence of the profound embryopathy with first trimester exposure, and then subsequently a reduction in frequency following the introduction of effective immunisation.
Animal studies showing a teratogenic effect from the agent being investigated are useful though subject to the vagaries of species variation, and it is also helpful to look at the dynamics of an agent under suspicion to see if its mode of action (in the case of a chemical) shows the potential to interfere with organogenesis (as might be expected if the pregnant mother is using antimetabolites). These criteria have been enunciated by Brent2 and serve as guidelines when we as clinicians evaluate reports of induced ocular malformations.
The danger of accepting unsubstantiated associations was demonstrated in 1980 when major defects, including cyclopia, were attributed to salicylate ingestion during the first trimester.3 Only a properly conducted, prospective, collaborative study was able to allay the anxiety induced by these reports. However, we must at the same time guard against an over-resistance to information coming to us from non-medical sources, since we then run the risk of failing to identify a teratogen and expose further pregnancies unnecessarily.
This dilemma has recently been highlighted by the publicity surrounding the benzimidazole antifungal agent Benomyl. In January 1993 under the headline “Mystery of babies with no eyes” an article appeared inThe Observer detailing the birth of nine children in an area of Lincolnshire within a 40 mile radius of Louth village, born with microphthalmia or anophthalmia. The report went on to identify other clusters in the UK, some, though by no means all, occurring in areas where agriculture (and thus the possible use of a popular fungicide) represented a major feature of the economy. The report cited evidence from animal experiments carried out in 1991 in which craniocerebral defects, including ocular malformation with retinal dysplasia and microphthalmos/anophthalmos, occurred in the offspring of pregnant rats fed a dose of Benomyl of 62.4 mg/kg between the 7th and 21st days of the pregnancy.4 The researchers speculated that the antitubulin action of Benomyl could conceivably impair microtubule formation and thus disrupt neuronal proliferation and migration.
The leap from this preliminary animal work (which could not be reproduced in a rabbit model) to ascribing isolated microphthalmos with no associated malformation to possible exposure to the fungicide being used in an agricultural or market garden setting is enormous. Nevertheless, in June 1996 the American courts awarded US$4 million to the Castillo family on the basis that the microphthalmos affecting their child might be caused by exposure to Benomyl. The child had no other neurological deficit and was not growth retarded.
Subsequently, the support group in the United Kingdom have, naturally, seized upon this judgment, and many of us have met parents who now confidently accept the association. What then are we to do? To dismiss this legal decision would be naive, even though the evidence presented in court was heavily criticised. What seems most important is to have a mechanism which can rapidly and effectively investigate this and any future worries, using the principles which Brent has laid down. Yet again this episode indicates the pressing need for a continuation of our move towards “evidence based” medicine, but in addition it suggests a role for a rapid response mechanism when a possible toxic effect is suggested, so that the popular press with its different priorities, is not left to take the lead. The current reporting system should work when medicines come under the spotlight, but is inappropriate for environmental agents. Perhaps the Royal College of Ophthalmologists, in conjunction with its sister colleges, might consider the establishment of a body to scan the scientific literature, and ensure that the clinicians in daily contact with patients are kept as informed as possible.
For the time being the case against Benomyl continues to be investigated. Since Australia banned its use several years ago it might well be time for the Australian College of Ophthalmologists to assess whether the incidence of this particular malformation has fallen in their community. In the United Kingdom isolated reports continue to appear, sometimes associated with very high dose application to nearby crop growing areas, and coordination of this reporting is available.