Smith-Lemli-Opitz syndrome : Clinical and biochemical findings in Brazilian patients

Smith-Lemli-Opitz syndrome (SLOS) or RSH syndrome comprises multiple congenital anomalies and mental retardation. The underlying defect is a deficiency in the activity of -sterol reductase, which decreases cholesterol and increases 7-dehydrocholesterol (7-DHC) levels. Our aim was to identify and evaluate the frequency of SLOS manifestations in a group of Brazilian patients. Based on our own data and those reported previously, we present a simple method that allows the estimation of probabilities favoring the diagnosis of SLOS. We evaluated 30 patients clinically and determined their plasma levels of cholesterol and 7-dehydrocholesterol. In 11 patients, the diagnosis was confirmed by ultraviolet spectrophotometry (UV). Of 19 patients with normal laboratory results, 17 showed a high probability favoring the diagnosis of SLOS. The most significant signs and symptoms observed in over 2/3 of the biochemically confirmed cases were mental retardation (10/11), delayed neuropsychomotor development (10/11), syndactyly of 2/3 toes (10/11), and craniofacial anomalies including microcephaly (11/11), incompletely rotated ears (8/11), palpebral ptosis (10/11), anteverted nostrils (10/11), and micrognathia (9/11). Genital anomalies were found in all male patients (6/6).

The incidence of SLOS is not known accurately; estimates range from 1:10,000 to 1:60,000 (Lowry and Yong, 1980;Ryan et al., 1998;Kelley, 2000;Kelley and Hennekan, 2000;Nowaczyck et al., 2001).This variation may be due to different criteria used in the evaluation of patients.
The clinical manifestations most frequently reported in SLOS include craniofacial anomalies, mental retardation, syndactyly of the second and third toes, and genital malformations in males.Facial anomalies are typical and comprise microcephaly, narrow frontal area, epicanthic folds, ptosis, short nose, broad nasal bridge, anteverted nostrils, micrognathia, and posteriorly angulated or low-set ears.Strabismus, cataract, high-arched or cleft palate, and short neck are also common.Limb defects include postaxial polydactyly of feet and syndactyly of the second and third toes.Genital anomalies (cryptorchidism, hypospadias) are reported in males.Other findings include: delayed global developmental progress with hypotonia and neuropsychomotor disturbances; pyloric stenosis; infection susceptibility during infancy; and various cardiac, renal, and pulmonary anomalies (Curry et al., 1987;Guzzeta et al., 1996;Cunnif et al., 1997;Ryan et al., 1998;Krajewska-Walasek et al., 1999;Nowaczyk and Waye, 2001;Haas and Kelley, 2001).
The clinical diagnosis of SLOS remains problematic because signs and symptoms presented by affected individuals range from minor anomalies associated with behavioral and learning disabilities to severe malformations and mental retardation, or even neonatal death (Porter, 2000).Biochemical analyses of cholesterol and 7-dehydrocholesterol levels are fundamental to establish a diagnosis of SLOS.
The purpose of the present work was to identify the most common and important findings in SLOS, to compare our data with those of the literature and thus further to delineate the classical SLOS phenotype.In addition, the frequency of the diverse signs and symptoms was used to determine probability figures favoring the diagnosis of SLOS in patients exhibiting negative biochemical results, in order to help the differential diagnosis of SLOS.

Patients
The study group included 30 patients (21M, 9F) with findings consistent with SLOS, who were clinically and biochemically evaluated from 2000 through 2002.
The study protocol was approved by both local and national committees of ethics, and informed parental consent was obtained in each case.To standardize clinical ascertainment and anamnestic procedures, the patients were evaluated according to the same protocol.
Figure 1 shows the facial appearance of all 11 patients whose diagnosis was confirmed biochemically.

Cholesterol analysis
Total plasma cholesterol was determined by an enzymatic colorimetric method (Kit Sera-Pak ® Cholesterol, fast color, Bayer ® ) and auto-analyser RAX Bayer, as described by Allain et al. (1974).
7-DHC analysis 7-DHC was determined by UV spectrophotometry.For this purpose, 200 mL of plasma were transferred to glass tubes and vortexed for 10 s after addition of 200 mL of ethanol.Subsequently, 1 mL of n-hexane was added and the mixture vortexed for another 20 s.After centrifugation at 400 g for 1 min, the clear n-hexane layer was collected in a 1 mL quartz cuvet and used for UV measurement.The samples were scanned using a n-hexane blank in the reference beam, and absorption maxima at 271, 282 and 234 nm were used for detection of 7-DHC in plasma samples; for quantitative analysis, absorption maxima at 234 and 282 nm were used, as proposed by Honda et al. (1997).

Statistical analysis
Table 1 shows the frequencies of SLOS findings in our group of 11 confirmed patients and in cases from the lit-  erature.The frequencies of signs/symptoms pertinent to the diagnosis of SLOS were calculated from literature reports, using two different estimation methods (Pardono et al., 2003).Estimate 1 was obtained using the formula P/(P+A+N), where P is the number of cases in which the finding was present, A is the number of cases in which it was absent, and N is the number of cases in which it was not mentioned.In estimate 1, it is hypothesized that, whenever the anomaly is not mentioned in a reference paper, it is absent.Estimate 2 was obtained using the formula P/(P+A); when the finding is not mentioned, it is assumed that is was not searched for by the authors.The true value of the estimate is obviously an unknown quantity, but its upper and lower limits are given by estimates 1 and 2, respectively.A  (Grier et al., 1983;Cappa et al., 1987;Fryns et al., 1992;Guion-De-Almeida and Richieri Costa, 1992;Mc Donald et al., 1993;Teebi et al., 1993;Fernandez et al., 1994;Gibbons et al., 1995;Reardon et al., 1995).The estimates thus obtained are shown in

Results
The clinical diagnosis of SLOS was confirmed biochemically by abnormally high levels of 7-dehydrocholesterol and low plasma cholesterol levels in 11 of the 30 patients studied (Table 3).In these 11 patients, cholesterol and 7-dehydrocholesterol levels ranged from 50 to 143 mg/dL and 1.56 to 28.81 mg/dL, respectively.The remaining 19 patients had 7-dehydrocholesterol levels that could not be quantified and normal cholesterol levels (94-221 mg/dL).
The clinical traits observed in our sample of 30 patients are summarized in Table 4.Among the eleven patients with positive biochemical results, only one patient presented apparently normal development (case 11).Of the remaining 10, seven could not walk or talk (cases 1, 2, 4, 5,  7, 9 and 10), and three started walking and talking after 3 years of age (cases 3, 6 and 8).All the 11 patients with positive biochemical results and 16 of the remaining 19 patients were cytogenetically evaluated and no chromosomal anomalies were detected in any of them.
Table 3 shows the estimated frequencies of the traits found in patients with SLOS (cases from the literature and from the present study).Combining the data of our 11 patients with biochemically confirmed SLOS with those from the literature, at least 2/3 of them have microcephaly, epi-Smith-Lemli-Opitz syndrome: Brazilian study 433 Discriminant analysis showed that only 3/30 of our patients presented a low probability favoring the diagnosis of SLOS in relation to the other three conditions, Opitz (G/BBB) syndrome, Aarskog syndrome and alpha-thalassemia/mental retardation (Table 5).

Discussion
Eleven of the 30 patients suspected of having SLOS exhibited increased levels of 7-dehydrocholesterol.In the literature, studies including controls and SLOS carriers have shown that 7-DHC and cholesterol levels were normal in all controls, whereas in patients with SLOS the 7-DHC levels were always increased; cholesterol levels were normal only in 10% of patients (Axelson, 1991;Nowaczyk et al., 1999).These and other similar studies in the literature, however, do not clearly state the clinical inclusion criteria used in the diagnosis of SLOS in cases which are not completely typical.
The plasma 7-DHC levels found in our 11 patients by UV spectrophotometry were higher than 1mg/dL of plasma, a result consistent with that found by Honda et al. (1997).
The clinical diagnosis of SLOS is hindered by the fact that none of its signs or symptoms is pathognomonic.Despite this difficulty, the association of mental retardation with genital anomalies in males and a typical facial appearance seems to be the most important clinical indicator of SLOS (Guzzeta et al., 1996;Cuniff et al., 1997;Ryan et al., 1998;Krajewska-Wallasek et al., 1999).That is exactly what we observed among our biochemically confirmed patients.
Thirty-four cases of the literature were reviewed and their clinical findings were compared with those of our biochemically confirmed SLOS patients.The most frequent findings in our sample were: mental retardation (10/11), microcephaly (11/11), delayed neuropsychomotor development (10/11), and syndactyly of the 2 nd and 3 rd toes (11/11).The six male patients had genital anomalies.A typical face (with incompletely rotated ears, palpebral ptosis, anteverted nostrils, flattened nasal base, and micrognathia) is easily recognized in all 11 patients.According to the review by Kelley and Hennekan (2000) including 164 individuals with SLOS, the anomalies present in over 50% of the biochemically confirmed cases are: syndactyly of the 2 nd /3 rd toes (97%), mental retardation (95%), microcephaly (84%), anteverted nostrils (78%), and palpebral ptosis (70%).The corresponding frequencies observed in the present study, therefore, do not differ from those observed in a larger number of affected individuals.
Another facial anomaly commonly found among our SLOS patients, but not frequently reported in the literature (though usually observed in photos), was a tented upper lip (10/10), a characteristic described by Neri and Opitz (1999) as "elongated upper lip" and by Kelley (2000) as "large mouth, which, combined with frequent micrognathia, gives a distinctive appearance".Another facial trait that seems not to have been described in the literature so far, but was frequently observed in our patients (8/11), is the presence of medial hypoplastic eyebrows.
The comparison between the frequencies of the most significant SLOS signs and symptoms observed in our pa-  tients and those observed in three other entities, by using a simplified discriminant analysis technique, led us to conclude that all 11 patients with positive biochemical results and 16 of the 19 with negative results had a high probability of having SLOS.This fact suggests that the latter group of patients should be investigated molecularly.
Recent studies have indicated that the malformations associated with SLOS can be traced to the role cholesterol plays in normal embryonic development.Most malformations seen in SLOS are likely to be related with a role of cholesterol in the hedgehog signal transduction pathway (Roux et al., 2000;Waterham and Wanders, 2000).Hedgehog (Hh) proteins form a family of embryonic signaling molecules, which are important for patterning in embryos.In humans, these proteins are called Sonic hedgehog (SHH), Desert hedgehog (DHH), and Indian hedgehog (IHH).The covalent addition of cholesterol to SHH is essential for its autoprocessing.SHH participates in the development patterning of the ventral forebrain, neural tube and limbs, among other structures.A similar process may be responsible for abnormal responses to DHH and IHH in the development of the genital tract and skeleton, respectively (Porter et al., 1996;Hammerschmidt et al., 1997;Tabin and McMahon, 1997).
Based on the discovery that the impaired function of development proteins (Hh) may explain a number of malformations present in SLOS, our hypothesis, to be tested in future studies, is that some of the patients who show typical malformations of SLOS but negative biochemical results have a defect, not in the DHCR7 gene (that codifies D 7 -sterol reductase), but in a developmental gene that codifies for Hh proteins.

Table 1 -
Estimated frequencies of the signs/symptoms found in SLOS patients (literature and present study).

Table 2 .
These values allow the calculation of the probabilities favoring the diagnosis of SLOS in relation to each one of the other three conditions.If no information about a given trait is available, the probability rate (likelihood) takes the indifferent value 1.To exemplify the use of the table data, let us suppose that in a given patient suspected of having SLOS or Opitz syndrome, only the first four features shown in the table were searched and that the first three features were present and the last one was absent.In this case, the likelihood favoring the diagnosis of SLOS is the product: 95/5 x 66/30 x 54/80 x 78/5 = 440.754,that is, given this set of anomalies, the probabilities favoring both diagnoses are in the ratio of 440.754: 1; therefore, the probability of SLOS diagnosis is obtained by normalizing 440.754 and 1: P1 = 440.754/(440.754+1) = 440.754/441.754= 0.9977.

Table 2 -
Reported frequencies of signs and symptoms in SLOS, Opitz syndrome, Aarskog syndrome and alpha-thalassemia/mental retardation syndrome.Whenever the frequency of a given feature (with the exception of polydactyly) was not increased in a given syndrome, a frequency of 5% was assigned to it.

Table 4 -
Clinical findings in 30 patients suspected of having the Smith-Lemli-Opitz syndrome.

Table 5 -
Probability figures favoring the hypothesis of SLOS in each of our 30 patients.