The occasional presence of sternal muscle in humans has been of great interest to anatomists. Since the sternal muscle was firstly described by Cunningham (Cunningham, 1884), a high number of anatomical publications report the presence of the sternalis muscle, which have a prevalence of about 3–8% and a variable presentation (Arraez-Aybar et al., 2003; Orts Llorca, 1970; Snosek et al., 2014). There are differences by gender: the sternal muscle is slightly more frequent in women (8.7%) than in men (6.4%) (Scott-Conner and Al-Jurf, 2002) and differences by ethnic groups: in white population its incidence is approximately 4–7%, while in the black population 8.4%, and in the Asian population 11.5% (Bergman et al., 1988). The cases described in this work correspond to white male, meaning they belong to the less probable group of occurrence.
Embryology of the sternal muscle
The embryological origin of the sternalis muscle is still on discussion. It has been postulated that it comes from adjacent muscles, which may be the panniculus carnosus, sternocleidomastoid, pectoralis major or rectus abdominis muscles (Jelev et al., 2001; Kida et al., 2000; Orts Llorca, 1970; Raikos et al., 2011). Turner (1867) considered it as an atavic form of the pectoralis cutaneous of lower animals (Turner, 1867). It is described that the somatic origin of the sternalis muscle, is a part of the ventral longitudinal muscular column that arises from the ventral portion of the thoracic hypomeres, whose equivalent in the abdomen gives rise to the rectus abdominis muscle while its persistence in the thorax gives rise to the sternal muscle (Kumar et al., 2003; Saeed et al., 2002). Other work indicate that is derived from the pectoralis muscle group including the subcutaneous trunci muscle (Kida and Kudoh, 1991). The innervation of the muscle give us some hints about its origin: most of the sternalis muscles are innervated by branches of the internal or external thoracic nerves (55%), or by branches of the intercostal nerves (43%), or both (2%) (O'Neill and Folan-Curran, 1998). Some cases describe innervation from the pectoral nerves, even if the muscle is in direct contact with the thoracic wall (Kida and Kudoh, 1991), however it is also described that the innervation comes from the intercostal or extramural nerves (Yamada & Mannen, 1985; Kodama et al., 1986). It is also suggested that the sternalis muscle could be arising from pectoralis major with innervations from pectoral nerve or from rectus abdominis with innervations from intercostal nerves (Saeed et al., 2002; Vaithianathan et al., 2011).
A high correlation between anencephaly and the presence of the sternal muscle is described together with some variations and origins in the pectoralis major muscle, implicating similar developmental pathways (Abraham, 1883; Shepherd, 1885). These results motivated us to examine an anencephalic infant, finding it bilaterally and corroborating the variations of the major pectoral muscles mentioned by Abraham (Abraham, 1883): both sternal muscles of the infant were originated from the larger pectoralis on the opposite side, a unique characteristic that was not seen in the case of the adult.
The relationship between anencephaly and the presence of alteration on the chest muscles is not clear. At the light of the current knowledge in developmental biology, this can be caused by the effect of a common molecular control, in which a common signal can be involved in the generation of both conditions. Probably, if this signal is altered at early stages. it can affect to the neural tube and somites. However, if the signal is altered later in time and in a downstream component, it can affect just the muscles, as we saw in the case of the adult. Other possibility is that two separate aetiologies are acting, however, same molecular control can be shared in same phenotype. Several genes are involved in the differentiation of the neck and chest muscles, such as Mef2c-AHF, Islet1, Mesp1and Pax3. Chest muscles, such as pectoralis, are controlled by the most posterior somite myogenic program, which depends on Pax3 (Heude et al., 2018). Importantly, mice homozygous for Pax3 mutation develop significantly higher cranial neural tube defects (Burren et al., 2008; Epstein et al., 1991), making Pax3 a candidate gene to be analysed for this muscular variation. Relation with anencephaly can be also linked with a positional effect, however the anencephaly is mainly caused in anterior regions in which paraxial mesoderm and the muscles of the chest are related to the more posterior somites, making this unlikely.
Classification and innervation
Given the different forms of presentation of the sternal muscle, Jelev developed a classification (Jelev et al., 2001) that was later modified by Snosek (Snosek et al., 2014) trying to accommodate varieties not previously considered. The former classification of Jelev (Jelev et al., 2001) established four morphological criteria to denominate a muscle as "sternal muscle" and propose a numerical nomenclature to define the different types. The modified classification of Snosek (Snosek et al., 2014) includes the four criteria indicated, but redefines the types and the subtypes of the sternal muscle. The sternal muscles of the adult and the infant presented in this study meet all the four above mentioned criteria to be considered as a sternal muscle: 1) to be placed on the fascia of the pectoralis major muscle; 2) to be originated from the sternum or from the infraclavicular region; 3) to be inserted in the lower ribs, costal cartilages, aponeurosis of the external oblique muscle of the abdomen or the sheath of the rectus abdominis muscle; and 4) to be innervated by the medial or intercostal pectoral nerves (Jelev et al., 2001). Applying the classification modified by Snosek (Snosek et al., 2014), the variety found in the adult cadaver seems to be similar to the type "others" subtype "cross-linked", although with some differences with respect to the muscle described in this work. The "cross-linked" variety as schematically represented in this classification, corresponds to a muscular arrangement in an "X", similar to the observed in this study (Fig. 1); but with two muscular bellies fused in the midline in front of the sternum, in contrast to the case we are describing here, in which the muscular bellies are not fused, rather they share a common superior tendon (Fig. 2). The "X" aspect in our case is given by the extension of their superior tendons with the sternal head of the sternocleidomastoid muscle. Therefore, the muscle described in this work does not match the classification of Snosek (Snosek et al., 2014).
In relation to the classification of the sternal muscles of the infant, the right sternal muscle could be close to a muscle of a simple type and of a cross-linked subtype, while the muscle of the left side could be similar to a mixed type and divergent bicipital subtype crisscrossed (Snosek et al., 2014), nevertheless, they do not match exactly the current classification due to its origin is in the larger pectoralis muscles.
Regarding the innervation of the sternal muscle, Shepherd (Shepherd, 1885) indicated that it was given by the medial pectoral nerve, similar to what was found by Kida (Kida et al., 2000). Other cadaveric and surgical explorations have reported that the sternal muscle is innervated by the pectoral nerves or anterior branches of the intercostal nerves (Snosek et al., 2014), or a combination of both (Hung et al., 2012). In the adult case described here, the nervous branches from the anterior cutaneous nerves of the neighbouring intercostal nerves reached the sternalis muscle (Fig. 2B), but no innervation of the medial pectoral nerve was observed in none of the two sternalis bellies. In our current report was not possible to identify the vascularization and innervation in the sternal muscles of the infant, probably due to its small size.
The clinical interest of the sternalis muscle relates to its imagenological and surgical implications: it is necessary for professionals to become familiar with this muscle to improve their medical practice. Here we confirmed the high variability in the presentation of this muscle that can be found and how numerous classifications are still not covering all types of variants that can be founded. The sternal muscle presence and variability is extremely important in the field of imagenology, in which this muscle has been studied due to its accidental finding during mammograms. To reduce the possibility of omitting a neoplasic condition in the image, it has been emphasized the correct positioning of the patient to cover the largest amount of breast tissue on the detector. Nowadays this, together with the improvements of the technique, results in a greater detection of the sternal muscle. In the imagenological studies, the sternalis muscle is observed as an irregular mass of medial situation in the craniocaudal projections of the mammary gland, and its shape varies according to the position of the patient, resembling a band when in supine position and making a bulge in a prone position (Nuthakki et al., 2007). Reports of the sternal muscle in imagenological tests increased in the last decades. The study by Bradley et al., (Bradley et al., 1996), reported finding 4 cases with sternal muscles in the review of mammograms corresponding to more than 32,000 women. In contrast, the prevalence reported in more recent studies that have used multidetector computed tomography is similar to the values of the cadaveric reports: reaching 5.8% according to the study by Ge et al., (Ge et al., 2014), which involved the revision of 6000 images of Chinese adults; 6.2% according to the study by Young et al., (Young Lee et al., 2006) that involved the review of 1,387 images of Korean patients; and 10.5% according to the study by Shiotani et al., (Shiotani et al., 2012), that considered 948 exams taken consecutively in Japan. Undoubtedly, the soft tissue image quality that this test provides, the intention to spot the muscle, and the study of the Asian population, explain the higher prevalence on these studies compared with the older ones.
The prevalence of the sternalis muscle reported in surgical studies is also low. The study by Bailey & Tzarnas (Bailey and Tzarnas, 1999) identified the sternal muscle in the mastectomies of only 3 patients in a period of 15 years, whereas in the study by Harish & Gopinath (Harish and Gopinath, 2003), which reviewed 1151 operative mastectomy records, the prevalence was 0.7%. According to Snosek et al. (Snosek et al., 2014), this low prevalence can be related to the lack of awareness of the surgeon regarding the existence of the sternal muscle, together with its high variability. It can also be unnoticed during mastectomies or breast implant surgeries (Salval et al., 2012), despite of breast implant is the most frequent cosmetic surgery in the United States (Alderman et al., 2014).
Numerous contributions from the surgical field take into account the sternal muscle. For example, Schulman & Chun (Schulman and Chun, 2005) reported a modified technique of tissue expander placement in breast reconstructive surgery in the presence of a sternal muscle. Kabay et al., (Kabay et al., 2005) reported that they include the sternalis muscle when removing a breast during a radical mastectomy in cancer surgery and according to Khan (Khan, 2008) the sternal muscle can be used to give more coverage of the breast implant. Other authors also suggest that it could be used as a flap in reconstructive surgery of the thorax or neck (Raikos et al., 2011; Salval et al., 2012).