CHDD researchers develop tool to aid in diagnosis of Fetal Alcohol Syndrome

Efforts to prevent FAS hinge on effective surveillance that provides an accurate gauge of its prevalence, so that cost-effective programs can be implemented for treating women whose use of alcohol places them at risk for giving birth to children with the syndrome. Preventing or reducing secondary disabilities such as low self esteem, depression and school failure in individuals who already have FAS depends on early diagnosis leading to appropriate interventions.

Studies conducted by CHDD research affiliates Drs. Susan Astley and Sterling Clarren aim to bolster FAS prevention efforts by laying the groundwork for improved surveillance and diagnostic tools.

"FAS is characterized by central nervous system dysfunction, a cluster of minor facial anomalies that is unique to FAS, and growth deficiency," explains Astley, assistant professor of epidemiology. "Because central nervous system dysfunction and growth deficiencies are also characteristic of other developmentally related conditions, the ideal case definition for screening and surveillance of FAS would focus on the unique cluster of facial features." To derive a more specific and objective definition of the facial phenotype of FAS, Astley and Clarren, professor of pediatrics, studied facial measurements recorded directly from children evaluated in the FAS Clinic at CHDD. They also studied facial measurements recorded from photographs of individuals diagnosed with FAS.

The cluster of facial anomalies that characterizes the FAS facial phenotype includes small palpebral fissures (eye openings), a smooth philtrum (an indistinct or missing groove between the upper lip and nose), and a thin upper lip. However, criteria have never been established as to how small, how smooth, or how thin these features must be, explains Astley. Nor, she notes, have criteria been established as to how many of these features must be present. Are two of the three sufficient, must all three be present or does it depend on the magnitude of expression of each? "In our studies we were looking for the minimum number of facial features in the cluster that could be measured accurately, precisely and efficiently to define the facial phenotype," says Astley.

"There are several advantages to using photographs," Astley notes. "Since a computer is used to capture the measurements from the photographs, the data can be collected with greater accuracy and precision. We also frequently have to rely on a childhood photograph when performing a diagnostic evaluation on an adult, for the facial features of FAS often diminish with adulthood.

"In both the direct and photographic studies, we collected key facial measurements from individuals with and without a diagnosis of FAS and applied a technique known as discriminant analysis to identify which cluster of facial features best differentiated the two groups."

Both analyses identified the same cluster of three facial features - a thin upper lip, a smooth philtrum and small eye openings. When the measures were collected directly by the clinician, the clinician ranked the thinness of the upper lip on a three-point scale (very thin, somewhat thin, not thin at all) using a photographic example of each as a guide. The clinician used the same technique to rank the smoothness of the philtrum. The size of the eye opening was measured with a ruler and compared to a norm based on age.

In contrast, the photographic method relied almost entirely on a computer to measure the level of expression of each feature. The thinness of the lip was measured by outlining the upper lip with a computer drawing tool and having the computer compute a thinness score. The higher the score, the thinner the upper lip. Philtrum smoothness was ranked as by the clinician, but on a finer five-point scale. The size of the eye opening was measured indirectly by having the computer measure the ratio of the eye size to the distance between the eyes. "Although a reduced ratio is not always indicative of small eyes, this reduced ratio in combination with the other two features was highly specific to FAS," explains Astley.

Both analyses generated formulas which could be used to compute an overall score for each individual face. "We can take the facial measurements, run them through the formula, and if the score is greater than or equal to 1.5 from the direct measures, or greater than or equal to 0.7 from the photographic measures, the facial phenotype is consistent with that of FAS," explains Astley.

These formulas performed with very high accuracy. The formula derived from the direct facial measures performed with 100 percent sensitivity (identified all 39 patients with FAS correctly) and 89 percent specificity (misclassified 17 of the 155 patients without FAS). The formula derived from the photographic facial measures performed with 100 percent sensitivity and specificity. It differentiated the 42 patients with FAS from the 84 patients without FAS, without error. The accuracy of both methods was unaffected by race, gender and age.

Although the two methods were highly effective in differentiating individuals with and without FAS, further studies are necessary to validate the results in larger, more diverse populations.

Like positive results from a routine Pap smear or tuberculin test, which indicate that it is worthwhile to proceed with further testing and evaluation, a positive result from either of these facial measurement methods is only a first step in the diagnosis of FAS. "Whether these tools are being used as diagnostic aids or for screening, final diagnosis must remain in the hands of a skilled clinician who can consider the case as a whole," says Astley.

Top of Page | Next Story | CHDD Home