Syndactyly is a condition well documented both in textbooks and current literature mainly due to it being the most common congenital hand defect.

Coming from the Greek syn (“συν” meaning together) and dactyly (“δακτυλος” meaning digits) it describes an embryological failure of finger separation. Common in some species, including birds and kangaroos, it has a large aesthetic and functional significance for humans born with the condition.

It has long been believed to have a genetic component, and this has been researched greatly over the last decade. Operative management of this condition has historically been, aside a conservative approach, the only definitive care for patients, but are new options becoming available as our understanding of the condition develops?

This article aims to discuss and summarise the current perspective of syndactyly in terms of epidemiology and genetics, as well as considering any new ideas in management of this condition.

Syndactyly is the fusion of adjacent digits. It is the most common of all congenital hand deformities with an incidence of around 1 in 2000 live births and is twice as common in males, as well as in the Caucasian population [1-4].

Syndactyly may involve fusion of the soft tissues with or without bony fusion. It mainly occurs due to the failure of differentiation between adjacent digits caused by the absence of apoptosis in the interdigital mesenchyme during the seventh and eighth weeks of gestation [1, 5]. The third, fourth, second and first web spaces are affected in decreasing frequency with around 57% of cases occurring in the third web space [2, 3, 6]. The condition is bilateral in half of cases [2, 6].

It is mostly seen in a sporadic appearance, but there is a family history in 10-40% of cases [2, 7]. Inheritance is thought to be autosomal dominant with variable penetrance and expressivity, and this is thought to possibly explain the male predominance [1, 2].

Syndactyly can be an isolated finding or seen with other anomalies such as acrosyndactyly, clindodactyly, synostosis, cleft hand and polydactyly or as a feature in several syndromes including Apert, Poland’s, Pfeiffer, Jackson-Weiss and Holt-Oram.

Syndactyly can be classified in several ways. Anatomically the syndactyly is either simple or complex, and complete or incomplete. Simple syndactyly involves only soft tissues whereas complex includes side-to-side bony fusion. When the adjacent digits are fused all to the finger tip it is described as complete syndactyly, whilst in incomplete refers to only partial union.

The most severe presentation is complex-complicated syndactyly where there is skeletal deformity accompanied by tendon and neurovascular abnormalities; the incidence of which rises as the complexity of the syndactyly increases [2].

Since its first description in the literature, syndactyly has also been classified by its phenotype. The simple and complex, and complete and incomplete are simple anatomical classifications which help to communicating the malformation. The phenotypical classification is more specific for the digits involved, and has led to syndromic and non syndromic syndactylies being described.

Temtamy and McKusick [8] concluded in 1978, from information gathered from both the literature and their own experience, that there were at least 5 phenotypically different types of syndactyly involving the hands, with or without foot involvement. The majority of these are thought to be inherited as autosomal dominant traits. Within each pedigree there is uniformity of the type of syndactyly, allowing for the variation characteristic seen in dominant traits. These genetic forms of syndactyly are required to be analysed separate to syndactyly with congenital amniotic bands for which currently, there is little or no evidence of a genetic basis. This paper will focus on syndactylies involving the hand, and will not discuss lower limb only malformations, or the amniotic band abnormality.

These non syndromic syndactylies appear to only involve digit and appendage malformation, and have since been expanded to nine phenotypes, named syndactyly I to IX, although some are known by their synonyms [9, 10]. See Table 1 for an overview of the nine non-syndromic syndactyly phenotypes.

The syndromic syndactylies, as the name suggests, are linked to other abnormalities in the body and these appear to occur alongside the digit anomalies during foetal development. The list of syndromic syndactylies is extensive and appears to be constantly growing as syndactyly is discovered with additional malformations. In this paper, we will briefly note the genes currently assigned to be causative in the better known and commonly presenting syndromes.

Before discussing the genetics aspects behind each syndactyly, it is useful to understand how the malformation is believed to develop. This paper does not aim to explore the molecular biology of limb formation in detail, but aims to summarise the current findings in a simple, clear, and systematic approach.

The formation of both the upper and lower limbs in the vertebrate appears to be linked to multiple genes as well as a vast number of encoding proteins.

The limb buds and consequent upper and lower limb are formed between the 4^th and 8^th weeks of gestation and arise from the main trunk, or body. The limb bud is initially directed along three axes of asymmetry, along which the mesodermal cells grow and later become fixed. These axes involve the proximal-distal axis from shoulder to finger, the dorsal ventral axis, from the back to the palm of the hand and the anterior-posterior axis from thumb to little finger. It is this latter axis which appears most important in digit formation. The final and specific limb architecture resulting in the aesthetic limb, normally involves cell proliferation, cell fate determination, cell differentiation and also apoptosis [11, 12].

Two signal centres have been identified to be in control of the human limb structure and positional identity [13-15]. These are the apical ectodermal ridge (AER) which is key for limb growth, and the zone of polarizing activity (ZPA). This latter centre specifies the position and overall patterning in relation to the anterior-posterior axis, although each is dependant on the other [16]. By day 44 the ZPA begins to regress, with metacarpophalangeal joints and proximal phalanges beginning to form. At day 48 the middle phalanges chondrify, followed by the distal bones by day 51 and digit separation by day 54.

This whole process is under the influence of several encoding proteins as illustrated in Fig. (1). In particular the hedgehog pathways, fibroblast growth factors, bone morphogenetic proteins, cartilage derived morphogenetic protein and WNTs have been most discussed in the literature in relation to limb formation [17].

Fig. (1). Diagram representing Shh/Gli pathways invloved in limb formation.

Both polydactyly and syndactyly appear to have a relationship with the Hedgehog family of intercellular signalling proteins. These have a majority of functions in cell fate, and most research has involved the Hedgehog (Hh) and Sonic hedgehog (Shh) pathways, specifically in Drosophilia and mice studies [18]. These have particular relevance as Shh is expressed in the ZPA where it controls anterior-posterior limb patterning [19]. In Mice, Shh appears to be a secreted molecule, related to the Drosophilia Hh, which regulates the balance of Gli3 repressor and activator and through these its target genes.

Indian hedgehog (Ihh) is biologically similar to Shh and appears to play a key role in a pathway which is involved in the rate of chondrocyte differentiation regulation [20] and appears to be repressed by fibroblast growth factor receptor 3[21, 22]. Ihh also appears to have a part to play in bone ossification [23]. Multiple papers have suggested a role for the Ihh pathway specifically, in the later development of syndactyly as well as in other congenital abnormalities [24, 25].

The ZPA positioning, and its involvement with Shh, is mainly determined by the transcription factors dHand, Gli3, Alx4 and several Bmp antagonists (Formin and Gremlin). Changes in any of these pathways have been found to lead to the –dactyly malformations (brachy-, syn-, and poly-) [26-30].

Wingless-type MMTV integration site family, members 6 and 10B (WNT6, WNT10B) have both been described as possible genes which require further research due to their involvement in cell apoptosis, and expression in the developing limb bud of the mouse. There has also been noted a relationship to the region of 2q35, a loci hypothesized as the source of syndactyly type 1[31-33].

The WNT family appear to be expressed at a similar time to the fibroblast growth factors (Fgf, in particular Fgf8) in chicken studies [34]. and their pathway is likely to be involved in the final stages of mesenchymal ossification. The Fgf group have been suggested as causative in some of the syndactyly syndromes, and are referenced later in the paper.

The last stage of digit formation involves cell apoptosis, where cell death along anterior, posterior and finally interdigital necrotic zones leads to the familiar profile of the hand [35-37]. This appears to coincide with restriction of Fgf8 expression and down-regulation of Gremlin in these regions [35, 36, 38-40].

Signalling molecules, including the Bone morphogenetic proteins (Bmp) and their antagonist Noggin (Nog) have recently been shown to influence the number of phalanges, possibly having a role in apoptosis [41-47] and blocking their signalling pathway has been shown to result in syndactyly [48-50].

Final digit identity appears dependant on the interdigital mesenchyme. Dahn and Fallon [51] found removal of this in chickens resulted in loss of digit identity, and it appears this is related to Shh and Gli3 pathways [26, 27]. The metalloproteases are also under scrutiny for their involvement in the formation of normal hand architecture, and appear to have a role independent of the Bmp for interdigital web regression [52].

N-Myc and several zinc-finger transcription factors need further research too, as defects in these molecules appear to induce soft tissue syndactyly in mice [53, 54]. Body patterning also appears to be governed by a family of transcription factors which originate from HOX gene encoding. There are 39 HOX genes in the human, as in most vertebrates, organised into 4 clusters and these play an important role in the development of the axial skeleton, central nervous system as well as the urogenital and gastrointestinal tracts, and our main interest, the limbs. There have been links to limb abnormalities with both deletions of some of these HOX clusters (-A and –D) and mutations affecting one or more HOX genes [55].

The specific HOX genes involved in syndactyly will be discussed later in the paper.

It should be noted that sporadic syndactyly with no familial history has been documented and environmental factors in-utero that predispose the foetus to syndactyly, as well as other congenital hand abnormalities, have been evaluated. Man conducted a study which reports a probable association with these conditions and maternal smoking [56] and there are suggestions that syndactyly occurrence is associated with lower nutritional and economic status, including increased meat and egg intake whilst pregnant, although more research is required before suggesting these are causative factors [57].

This paper will specifically look at the genetic attributes to syndactyly and we will now discuss the familial form of this anomaly.

It has been well documented that syndactyly appears to have an autosomal dominant transmission with variable expression and penetrance [1-4]. This is best represented with the increased prevalence in male offspring, possibly due to reduced penetrance in females. It has also been seen to skip generations and present in a reduced form showing variable phenotype.

This autosomal dominant transmission can be associated either in a simple fashion involving only the distal limbs or as part of a syndrome, some of which will be listed below.

The non syndromic forms of syndactyly are most easily analysed in their subgroups, rather than as one entity and are shown in Figs. (2-10).

Fig. (2). Syndactyly I: Zagodactyly.

Fig. (3). Syndactyly II: Synpolydactyly.

Fig. (4). Syndactyly III: Spectrum of oculodentodigital dysplasia.

Fig. (5). Syndactyly IV: Haas type.

Fig. (6). Syndactyly V.

Fig. (7). Syndactyly VI: Mitten hand.

Fig. (8). Syndactyly VII: Cenani lenz.

Fig. (9). Syndactyly VIII.

Fig. (10). Syndactyly IX: Mesoaxial synostotic.

Other presentations of syndactyly are as part of a syndrome, usually with other congenital abnormalities. A brief overview of the key genetic areas in the more common syndromes follows.

Acrosyndactyly, a term for syndactyly associated with congenital constriction bands, appears to lack a genetic basis, with Tentamy and McKusick [8] being first to find little or no evidence of a clear or simple genetic link.

Poland Syndrome (MIM 173800) presents with unilateral hypoplasia or absence of pectoralis muscle with ipsilateral hand and digit anomalies. As of yet no gene or loci have been implemented in its origin.

Acrocephalosyndactyly; there are 5 types of acrocephalosyndactyly, a condition involving syndactyly and craniosynostosis.

Type I, synonymous with the term acrocephalosyndactyly, is Apert syndrome (MIM101200). Associated with the FGFR2 gene, and the loci 10q26, includes mid face hypoplasia, foot and hand syndactyly with a trend for distal bony fusion [126].

Type II, Carpenter syndrome (MIM201000) has been linked to RAB23 gene originating from 6p11, with malformations including foot and hand syndactyly/brachydactyly, and acrocephaly [127].

Saethre-Chotzen syndrome, type III (MIM101400), involves syndactyly of the second and third fingers, as well as the third and fourth toes, as well as eyelid anomalies and cranial abnormalities. It has been linked to the loci 7p21.2 and 10q26 involving the TWIST 1 and FGFR 2 genes respectively [128, 129].

Type IV, was known as Goodman syndrome (MIM2010 20) but is thought now to be a variant of type II [130].

Type V, also known as Pfeiffer syndrome (MIM101600) has been linked to the FGFR 1 and 2 genes [131,132].

Other syndromes and chromosomal location include Acropectorevertebral dysplasia (MIM102510) and 2q36 and Fraser syndrome (MIM 219000) associated with both the sites 4q21 and 13q13, involving the FRAS1 and FREM2 basement membrane genes respectively [133,134], which have also been shown to be linked to fin deformity in zebrafish [135].

Greig cephalopolysyndactyly (MIM 175700) is an autosomal dominant disorder associated with haploin sufficiency of GLI3. This appears to be caused by deletions, truncations or point mutations of the associated Gli3 gene. Similarly the zinc finger domain of Gli3 has been found to be causative in Pallister Hall syndrome whose phenotype includes central nervous system and craniofacial deformities, as well as anal defects [136].

A recent study states a wide range of phenotypes can occur with only a Gli3 mutation, ranging from non-syndromic to syndromic syndactyly [137].

Other genes and loci that have been linked to digital anomalies include the Xq25 loci, with associated developmental delay [138]. Cholesterol metabolism [139], ROR2 [140], nidogen [141], GAS [142] and MBOAT [143] genes have been shown to be related to limb and digit formation in animal and patient groups, likewise mutations in Jagged [144], Serrate [145] and MSX [146] genes appear to cause syndactyly among other congenital abnormalities.

The current mainstay for the treatment and management of syndactyly is operative. The indications in general for hand anomalies run true with syndactyly and include;

Functional needs: The degree of function required by the patient; including whether delaying surgery may alter hand function and grip development.

Aesthetic consideration: The appearance of the deformity and its subsequent effects on psychological and social aspects of the patient’s life.

It must be remembered that surgical intervention is not urgent. Despite this there are age related targets for reconstruction and dependency on the type of anomaly present.

Border digit syndactyly involving the thumb, index and ring fingers is felt to benefit from earlier release, usually between the ages of 6 to 12 months, and should always be released first in multiple digit involvement. The result in delayed surgery can involve deformity of the digit relating to forced flexion, angulation and/or rotation.

Long finger syndactyly usually involves similarly lengthened digits and for this reason is often delayed, and KettleKamp and Flatt [147] found results to be better in those aged greater than 18 months, particularly with long term final appearance of the commissure.

Multiple digit involvement, as stated above, should involve in the first instance release of the border digits, with subsequent digit release after a 6-month waiting period. These subsequent releases involve either radial or ulnar sides of a finger as releasing both sides may endanger the digit viability.

It is generally regarded that operative therapy should be done pre-school age. The main reason for this is that interaction with other children whilst having the anomaly is thought to have potentially detrimental affects on the child’s social, functional and psychological development.

The main concern with surgical therapy is the fact that combined circumference of the two normal digits is approximately 1.4 times the circumference of the fused digits. Pre reconstruction it is felt that massaging the skin where the new web space will be constructed facilitates skin loosening and maximises potential soft tissue needed during the procedure.

Full thickness skin grafts are used for soft-tissue coverage in the majority of cases at present, with altering levels of flap use. The choice of flap, and consequent extra need for full thickness graft coverage, appears to be dependant on the surgeon’s preference, with multiple techniques and variances described in the literature. Several other methods have been used, including split thickness grafts [148] and tissue expanders [149]. Both have suffered with less favourable outcomes and increased complications compared to the use of full thickness grafts and have hence failed to establish themselves as alternative techniques. The use of no graft is more common, but this appears only to be of use in mild cases of simple syndactyly.

Follow up should be until skeletal maturity, mainly due to the prevalence of web creep until this age. Complications of all the surgical techniques include web creep, finger deviation (particularly in complex syndactyly) as well as those complications associated with any surgical procedure. The greater the degree of syndactyly, the more need there is for the surgeon to be aware of neurovascular variance and this takes on greater importance in graft and flap survival.

In light of genetic and peri-natal investigations, both these fields have potential for future management input. Ultrasound has allowed earlier diagnosis and this can now be supplemented by genetic review of likely carriers. Constriction bands have been released in-utero allowing a more normal limb development and with minimal scarring [150]. However, in utero surgery at present is high risk and currently the benefit of operating does not outweigh the risks and associated complications for non-fatal conditions.

The continuing development in both the genetic basis of syndactyly, and the improvement in tissue bioengineering, bodes that in the future both these avenues will expand management options.

In conclusion, there are currently 9 types of phenotypically diverse syndactyly. This number has increased since the original classification in 1978 by Tentamy and McKusick [8]. The non-syndromic syndactyly are inherited as an autosomal dominant trait, although the more severe presenting types (SD7, SD9 and to a lesser extent SD8) and sub types appear to have autosomal recessive and in some cases X-linked hereditary.

Genetic analysis has found that these phenotypes appear to not be one gene specific, although each phenotype does appear to have its own localised loci, but dependant on a wide range of genes, and subsequent proteins and transcription factors involved in limb formation. The principal genes so far defined in their involvement in congenital syndactyly mainly concern the pathways connected to the Zone of Polarizing Activity and it’s linked Shh pathway.

New syndromes, genes and causative loci are being found as research continues into congenital hand defects and it appears that each new finding gives as many further avenues for further investigation as answers in this field.

Research into the individual phenotypes appears to complicating phenotypical classification as new genes are found both linked, and not linked, to each malformation.

This has been noted by several researchers [151, 152] and attempts have been made to simplify the current classifications although these are yet to be recognised across all specialities.

There is still plenty of research required in the origins and causes behind syndactyly, both as a non-syndromic and syndromic entity, and until this is completed we feel the phenotypical and anatomical description of syndactyly is easiest to learn and use day to day when in clinical practice.

Management is still by surgical correction, and requires optimal planning in terms of patient age and degree of malformation present. This is a non-urgent procedure and it must take into account the psychological and social impact on the child.

Despite increasing knowledge of the causes of syndactyly, management has not changed greatly over the years, although the future is likely to see new techniques involving gene manipulation and tissue engineering at least to be created and studied, although current management is low risk.

None declared.