Clinical and epidemiological findings in patients with central ray deficiency: Split hand foot malformation (SHFM) in Manitoba, Canada^†

INTRODUCTION

Split hand foot malformation (SHFM) is highly variable in its presentation ranging from a central cleft with no osseous deficiency to monodactyly with only the fifth finger remaining. Various classifications of SHFM have been proposed. According to the International Federation of Surgical Societies of the Hand (IFSSH), SHFM is classified as a Category I malformation failure of formation of parts [Swanson, 1976]. However, syndactyly (Category II) and polydactyly (Category III) are also often present. SHFM has been divided into Typical and Atypical types [Lange, 1937]. Briefly, Typical SHFM is classically characterized by a cone shaped cleft, with or without digital a/hypoplasia. In its most severe form, it is represented by monodactyly. Bilateral hand involvement as well as involvement of the feet can occur and it is often familial. Atypical SHFM often shows unilateral involvement and occurs sporadically. Such defects rarely involve the feet. Clinically, it is represented by a deficiency of the three central rays. The remaining rays are often hypoplastic and a web may exist in place of the deficient digits. It is not uncommon for small nubbins (resembling vestigial digits) to be present in the cleft [Lange, 1937; Barsky, 1964].

We identified individuals born in Manitoba with SHFM and attempted to classify their limb deficiencies into these previously defined categories. In addition, we carried out an epidemiological study evaluating birth prevalence and characteristics of local patients. Findings from the literature were compared to our local population. This paper documents the results of our investigations and the challenges surrounding classifications of SHFM patients.

METHODS

Manitoba is well suited to the epidemiological study of congenital malformations as there is a single tertiary care Children's Hospital and a single program providing medical genetics services, including the only cytogenetics, biochemical genetics, and molecular genetics service laboratories in the province. For most of the years covered by this study, there was an active provincial congenital anomalies surveillance system. Identification of individuals with limb deficiency had been an ongoing interest of our research group. In order to carry out these investigations, we requested and received permission from our Research Ethics Board to analyze data on the individuals documented from these multiple sources as having central axis limb deficiencies.

Patients were included in the study if they displayed any central ray deficiency, which varied from a large gap or cleft without osseous defects to monodactyly. All patients with single digits were carefully evaluated to determine if they indeed represented SHFM patients. Cases of presumed amniotic band disruption sequence were excluded. For each patient, we noted demographic factors including gender and year of birth, and pertinent family and pregnancy history. We recorded the specific limbs involved, and the precise nature of the defects as documented by clinical, radiological and other imaging techniques. Additional congenital malformations were documented along with any syndromal diagnoses and cytogenetic and molecular investigations.

Cases were initially classified in a general anatomic manner in order to differentiate these patients according to the distinctions originally proposed by Lange 1937. However, as several cases could not be easily classified as Typical or Atypical, a third category—“difficult to classify”—was created. Patients in each of these three categories were further subdivided according to whether they had additional malformations. Gender, number and type of limbs involved and other pertinent findings were then compared between cases in these different groups.

Information on provincial birth rates was provided by Statistics Canada (http://www.statcan.ca) for the interval 1998–2003. For earlier years, the rates for Manitoba were provided by The Vital Statistics Agency in Winnipeg.

A literature search was conducted to identify other epidemiological studies of SHFM and our data were compared with their results.

RESULTS

We identified 43 cases with SHFM born between 1957 and 2003. Details are provided in Table I. The total number of births (livebirths and stillbirths) in Manitoba during this time period was 842,124, resulting in a birth prevalence of 0.51/10,000 or 1 in 19,784 births. There were 22 females and 21 males. As seen in Table I, many patients had “mild” limb involvement (e.g., syndactyly, camptodactyly) of a non-SHFM afftected limb. For the following calculations, “affected/involved” refers to limbs with a clinical phenotype that satisfied our inclusion criteria (e.g., having central ray deficiency). Mild findings were not included in these calculations. Four patients (9.3%) had SHFM of all four limbs. The upper limbs were the most frequently affected (39.5% and 46.5% respectively) while 34.8% of patients had right foot involvement and 34.8% had left foot involvement. Five cases had long bone involvement. Associated osseous fusions were present in 13 patients and polydactyly was found in four.

The classification of 32 patients into Typical (Cases 1–29) or Atypical (Cases 30–32) SHFM was straightforward; however, the remaining 11 cases could not be easily categorized in this fashion and thus comprised the “difficult to classify” group (Cases 33–43). Among the Typical cases, 16 (55.1%) had anomalies restricted to the central axis (i.e., were Typical isolated). There were 11 females and 5 males. Among the Typical isolated cases, 7 (43.7%) showed involvement (SHFM) of the upper limbs only, 6 (37.5%) showed involvement of the lower limbs exclusively, and 3 (18.7%) had involvement of hands and feet. Approximately half (56.2%) the cases had a single limb affected and only one case had all four limbs involved.

Thirteen cases had Typical split hand/foot in association with long bone deficiencies or additional malformations outside the musculoskeletal system (i.e., were Typical complex) (Fig. 1). Nine were male and four were female. The frequency of multimelic involvement was higher in this group with four (30.7%) having three or four limbs affected with SHFM clinically. Three cases had only the lower limbs involved, while five had a single hand affected. Several syndromic diagnoses were made in this group including tibial agenesis with split hand/foot (OMIM 119100), fibular aplasia with split hand/foot (OMIM 113310), ectrodactyly-ectodermal dysplasia-clefting (EEC) syndrome (OMIM 129900, 602077, 604292). Acrorenal field defects (Cases 17 and 21) also occurred, while other cases had unrecognized patterns of multiple congenital anomalies. The male with EEC syndrome had a father with facial clefting, but no limb defects; P63 testing has not been carried out in this family. The only two split hand/foot patients with documented karyotypic anomalies fell into the Typical complex group. They were a boy with Gorlin syndrome and a 9q deletion (Case 20) and a girl with de Lange syndrome and a large terminal deletion of 5p (Case 18). No patients had chromosome abnormalities involving the 5 mapped loci for SHFM.

Three patients, two females and one male, were classified as having Atypical isolated split hand/foot. Two had unilateral involvement of the left hand, while the third had an affected left foot. No patients were classified as having atypical split hand foot in association with other defects.

Eleven patients were included in the “difficult to classify” group. Many of these patients had extensive syndactyly; their central axis deficiencies were only apparent after imaging studies had been performed (Figs. 2, 3). Eight (72.7%) had abnormal development restricted to a single hand. Five were male and three were female. In six (75%), it was the left hand that was affected. Three cases had additional anomalies. One female had an affected left hand with a transverse right hemimelia (Case 42). The last two patients had multiple congenital anomalies detected prenatally (Cases 33 and 37). They may represent cases of SCHISIS and VACTERL associations respectively (Fig. 4).

With respect to family and pregnancy histories, two of the cases with Typical isolated defects were sisters; two of their paternal relatives were noted to have syndactyly and polydactyly respectively, but their father's radiographs were normal. Other pertinent findings in the study population as a whole included parental consanguinity in three cases and prenatal exposure to alcohol and cocaine in one.

DISCUSSION

Comparison of our findings with those seen in other populations was hampered by the fact that many studies do not distinguish between different types of limb deficiencies or, when they do evaluate SHFM, they do not necessarily study all four limbs or differentiate between Typical and Atypical cases. Reports describing SHFM specifically in all four limbs yield birth rates ranging from: 0.15/10,000 to 0.98/10,000 total births [Rogala et al., 1974; Aro et al., 1982; Leung et al., 1982; Froster-Iskenius and Baird, 1989; Calzolari et al., 1990; Czeizel et al., 1994; Castilla et al., 1995; Stoll et al., 1996; Galan et al., 2000; McGuirk et al., 2001]. The birth rate, which included both typical and atypical forms of SHFM in the Manitoba population, was calculated to be 0.51/10,000 during the study period. The study with the most similar methodology to our own evaluated Hungarian patients with limb deficiencies born between 1975 and 1984 and, unlike most other studies of SHFM, also recorded split foot. Limb deficiencies, in general, were found in 5.53/10,000 total births, while SHFM including both Typical and Atypical occurred in 93 individuals or 0.60/10,000 total births [Czeizel et al., 1994]. The Manitoba study results also resembled the Hungarian study results with respect to the proportion of complex cases at 34.8% and 41.2% respectively. Two studies evaluating livebirths, stillbirths, and pregnancy terminations also found similar results for the birth frequency of SHFM: 0.54/10,000 [Stoll et al., 1996] and 0.60/10,000 [McGuirk et al., 2001].

Potentially, the highest birth rate for SHFM including monodactyly (0.98/10,000) was derived from a study of consecutive singleton livebirths from upstate New York [Lin et al., 1993]. It is of interest that this rate is so high, considering that stillbirths and terminations of pregnancy were not included. The overall prevalence of limb deficiency defects in this study was not elevated (3.76/10,000); however SHFM cases represented 26.2% of their all limb deficiencies, which is much greater than the proportion seen in other limb defect studies. The Hungarian rate, for example was 10.7% [Czeizel et al., 1994].

In contrast, the study with the lowest birth frequency of SHFM was from South America (0.15/10,000) [Castilla et al., 1995]. Interestingly, the total limb deficiency rate was comparable to that of other studies at 5.88/10,000; however, their SHFM group comprised only 2.8% of the total limb deficiency patients.

Evaluating specific epidemiological and morphological characteristics, we noted that in our Typical isolated group, there was a non-significant excess of females and, when the feet were involved, the defect was usually bilateral. Various authors have proposed preferential involvement of the feet in SHFM, particularly in certain familial cases [Ozen et al., 1999; Tackels-Horne et al., 2001]. These published cases represent isolated typical SHFM, or typical SHFM associated with deafness. In this group in the Manitoba population, the right-sided limbs were affected somewhat more often than the left in the patients with unilateral involvement (60% vs. 40%, P = 0.6531). This preferential involvement of the right side in Typical isolated cases has been found in other epidemiological SHFM studies. Birch-Jensen saw a preference for right-sided involvement, as was also the case in the Hungarian population [Birch-Jensen, 1949; Czeizel et al., 1994]. In his classic Danish study, Birch-Jensen identified 36 proposita(i) with Typical split hand. Bilateral defects of the hands were more common than unilateral defects (58.3% vs. 41.7%) and, of the unilateral defects, 73.4% were right-sided.

For SHFM as a whole, upper limb involvement was more common than lower limb involvement in the Manitoba population (72% vs. 44%, respectively, P = 0.2536). Atypical SHFM is generally unilateral and involves the upper limbs. Thus populations with a high frequency of this or related forms will tend to show a higher proportion of upper limb involvement. The Hungarian study, for example, had 54 isolated cases; 33 (61.1%) were atypical and 48 (88.9%) had affected hands [Czeizel et al., 1994]. In another Canadian population study from British Columbia, central ray defects of the hand were also found to be more frequent than those involving the feet [Froster and Baird, 1992, 1993]. For the hand, the birth rate for defects of the middle digits was 0.61/10,000 (for 1952–1984) whereas, central ray defects of the feet were 0.098/10,000. In these two studies of the same population group (1,213,913 consecutive livebirths), the authors excluded cases of amniotic band disruption and diagnoses of a hereditary or syndromic nature. However, other studies have found preferential involvement of the lower limbs, especially among patients with Typical isolated SHFM. In the South American population studied [Castilla et al., 1995], 72.4% had their feet affected compared to 41.3% with split hand, 10.1% having both. However, in our population there was a higher incidence of both upper and lower involvement (30.7%) when other malformations were present.

The apparently conflicting patterns between these studies may well be due to variations in inclusion/exclusion criteria, the relative frequency of the underlying etiological factors and, in particular, the single gene disorders that predispose to involvement of several limbs, and the feet in particular.

An unanticipated result of our study was that differentiation of cases into the previously described Typical and Atypical forms was far from straightforward. Many patients that we placed in the “difficult to classify” group had extensive syndactyly such that their central axis deficiencies were only apparent after imaging studies had been performed. Lower limb involvement was very rare within this group and none of the patients had bilateral lower limb involvement. Most patients had unilateral involvement of the left upper limb. The monomelic involvement of one upper limb is suggestive of atypical SHFM; however these patients were not classified as such due to their phenotype.

This group emphasizes the importance of radiographic evaluations to further characterize the limb defect, thereby influencing management and counseling. Follow-up clinical and radiographic evaluations of very young patients are critical, particularly to confirm that the defect is indeed central. Examination of both parents—both clinically and radiographically—is also important, as subtle findings in a parent would confirm an inherited malformation. Although the relationship between osseous syndactyly, central ray polydactyly and split hand has been well covered in the literature by the Japanese clinicians [Miura, 1976; Ogino, 1979, 1990], the relationship of extensive cutaneous syndactyly and central deficiencies is less frequently discussed. The degree of syndactyly in these patients resulted in a more “syndactylous” phenotype, rather than cleft phenotype, since the syndactyly extends so far distally that a true “cleft” may not be apparent. However, radiographic evaluation confirms that these patients have a true central ray deficiency, potentially distinct from the vascular insults thought to be the cause of at least some Atypical cases [Hoyme et al., 1983; Graham, 1986].

Another patient in the hard to classify group had an interesting phenotype with the left hand showing a classical split hand with 3 rays present and the right hand showed a mixed deficiency involving a transverse deficiency at the wrist and a missing ulna. Transverse defects have occasionally been reported in patients with SHFM and comprise another interesting phenotypic association [Kelikian, 1974; Spranger and Schapera, 1988; Morey and Higgins, 1990].

Lack of fine morphological data in most of the other epidemiological studies makes it difficult to assess whether such difficult to classify cases have been overlooked as having central axis deficiencies or have been included as Atypical even though their phenotype is different from that initially described by Lange and others [Lange, 1937; Barsky, 1964; Swanson, 1976].

In conclusion, we found a birth rate of SHFM of 0.51/10 000 in the Manitoba population, which was comparable to similar epidemiological studies. We did not find the excess of males or right-sided defects that has been reported by some researchers. The upper limbs tended to be preferentially involved in Manitoba patients. This, and the high frequency of monomelic involvement, potentially indicates a relatively low frequency of the classic inherited forms of SHFM in our population. We identified a group of patients who were difficult to classify, most of whom showed involvement of one upper extremity and had cutaneous syndactyly. In total, 65.2% of our patients had isolated SHFM and 34.8% had SHFM in addition to involvement of another system. With respect to our complex cases, finding conditions such as EEC syndrome and acrorenal field defects was not surprising; however, we identified two patients with chromosomal deletions that did not involve any of the mapped loci for SHFM.