Embryology of Congenital Malformations
- 221 Downloads
Misconceptions and/or outdated theories concerning normal and abnormal embryology.
A shortage of study material (both normal and abnormal embryos).
A shortage of explanatory images of embryos and developing embryonic organs.
Difficulties in the interpretation of serial sections.
In recent years, a number of animal models had been established which helped to overcome the shortage of both, normal and abnormal embryos. However, a general agreement on when, why and how abnormal development takes place, still does not exist. As a result, many typical malformations are still not explained satisfactorily and pediatric surgeons of all specialties are still confused when they are confronted with the background of normal and abnormal embryologic development.
KeywordsHuman birth defects Animal models Teratology Human embryology
Misconceptions and/or Outdated Theories Concerning Normal and Abnormal Embryology
Our understanding of the normal and abnormal development of embryos is still influenced by two theories:
According to Haeckel’s “biogenetic law,” a human embryo recapitulates in its individual development (ontogeny) the morphology observed in all lifeforms (phylogeny). This means that during its development, an advanced species (a human embryo) seems to pass through stages represented by adult organisms of more primitive species (Gilbert 2003). This theory has been used to “bridge” gaps in the understanding of normal embryonic development and still has an impact on the nomenclature of embryonic organs. This explains why human embryos have “cloacas” like adult birds and “branchial” clefts like adult fish.
A Shortage of Study Material (Both Normal and Abnormal Embryos)
Today, a growing number of animal models exist which allow embryological studies in various embryological fields. This includes studies in normal as well as in abnormal embryos. Especially for the studies of esophageal and anorectal malformations, a number of animal models had been established.
A Shortage of Explanatory Images of Embryos and Developing Embryonic Organs
Advanced technology in a number of fields gives much better insights into human development. This includes ultrasonography of fetuses as well as magnetic resonance imaging (MRI). For detailed embryological studies, scanning electron microscopy is still a very useful tool. Recently, STEDING published a scanning electron microscopic atlas of human embryos which provides detailed insights into normal human embryology (Steding 2009). Scanning electron microscopy is the perfect tool to document embryonic structures:
Serial sectioning of embryos and time-consuming three-dimensional reconstructions are not necessary.
The embryo can be studied in all three dimensions “online.”
- (c)The images and photographs are of superior quality (Fig. 2).
Difficulties in the Interpretation of Serial Sections
Although a number of specific tasks demand the serial section of embryos, the difficulties in the interpretation must not be underestimated. Three-dimensional (3D) reconstructions, although feasible, are tainted with a loss of information, not only caused by the sectioning itself but also by the use of 3D image software.
Animal Models Used for Applied Embryology
Over the last two decades. a number of animal models have been developed with the potential to gain a better understanding of the morphology of not only malformed but also normal embryos. These animal models can be grouped in five subgroups:
Embryos of different species for the study of normal embryology
“Spontaneous” malformations of unclear cause
Embryos of Different Species for the Study of Normal Embryology
Human embryos are rare. Human embryos displaying typical anomalies are extremely rare. Therefore it makes sense to study specific developmental processes in embryos of animals with humanlike abnormalities. However, in all cases of animal models, the detailed study of normal embryos of the same species is mandatory.
We used scanning electron microscopy (SEM) in chicken, rat, and murine embryos in order to study certain embryological processes of the normal embryology of the foregut, the hindgut, the midgut, the testicular descent, and the development of the external genitalia. The advantage of chicken embryos is the high availability at low costs. They are easily accessible in the eggshell, and further breeding is possible when the eggs are treated accordingly. Embryos of rats and mice can be obtained in comparable large numbers; however, local regulations may limit the usage of mammalian embryos.
The chicken embryo was used to study foregut development. The aim was to clarify whether lateral ridges occur in the developing foregut or not and, when present, if they fuse to form the tracheoesophageal septum (Kluth et al. 1987; Metzger et al. 2011a) (Fig. 3).
Rat embryos were used to study, i.e., developmental processes during testicular descent, to clarify if “rotation” takes place during gut development (Fig. 4a), to assess the question if “cloacas” actually exist in rat embryos, and how the differentiation of the developing hindgut takes place (Fiegel et al. 2011; Metzger et al. 2011b; Kluth et al. 2011a) (Fig. 4b).
Apart from these purely embryonic models, a large number of fetal models had been developed in the last 30 years. Although they were mainly created to study the feasibility of fetal interventions (Harrison et al. 1980), they also contributed to our current knowledge of normal and abnormal fetal development and fetal organ systems.
It is well known that a number of chemicals (drugs, chemical fertilizers) can alter normal development of humans and animals alike. Some of these had been used to induce malformations similar to those found in humans. Most important today are:
Models (a–d) have been used to study atresia formation in the esophagus, the midgut, and the anorectum. Model (e) was developed to study malformations of the diaphragm, the lungs, the heart, and kidneys (hydronephrosis). Model (f) was used in chicken embryos to study the formation of cloacal exstrophy.
Many aspects make genetic models the ideal model for the studies of abnormal development. In the past, a number of genetic models have been used for embryological studies of malformations. While older models were mostly the product of spontaneous mutations, newer models are, in most instances, the result of genetic manipulations mainly in mice (transgenic mice). The following models have been used by pediatric surgeons:
Models of spontaneous origin: The SD-mouse model (Dunn et al. 1940; Kluth et al. 1991). In the SD-mouse model, anorectal malformations are combined with anomalies of the kidneys, the spine, and the external genitalia (Fig. 5).
Inheritance models: the pig model of anal atresia (van der Putte and Neeteson 1984; Lambrecht and Lierse 1987) (Fig. 9). In pigs, anorectal malformations are seen quite frequently. One out of 300 newborn piglets present with anorectal malformations without evidence of genetical alterations.
It has been demonstrated, that targeted deletion of Sonic hedgehog resulted in homozygous Shh null mutant mice in the formation of foregut malformations like esophageal atresia/stenosis, tracheoesophageal fistulas, and tracheal/lung anomalies (Litingtung et al. 1998). In the hindgut, the deletion of Sonic hedgehog caused the formation of “cloacas” (Kim et al. 2001), while Gli2 mutant mice presented with the “classic” form of anorectal malformations and Gli3 mutants showed minor forms like anal stenosis (Kim et al. 2001; Mo et al. 2001). Interestingly, the morphology of Gli2 mutant mice embryos resembles that of heterozygous SD-mice embryos, while Shh null mutant mice embryos had morphological similarities with homozygous SD-mice embryos. It is interesting to note that after administration of adriamycin, abnormal pattern of Shh distribution could be demonstrated in the developing foregut (Arsic et al. 2004).
Recently, Botham et al. studied developmental disorders of the duodenum in mutations of the fibroblast growth factor receptor 2 gene (Fgfr2IIIb) (Botham et al. 2012). They noted an increased apoptotic activity in the duodenal epithelium of Fgfr2IIIb -/- embryos at day 10.5, followed by a disappearance of the endoderm at day 11.5. Interestingly, the duodenal mesoderm also disappeared within 2 days, and an atresia was formed. Similar processes had been observed in newborn piglets whose esophageal epithelium was removed via endoscopy (Booß and Okmian 1974; Komfält and Okmian 1973). This procedure resulted in esophageal atresias in these piglets.
In humans, viral infections are known to cause malformations. Animal models that use viral infections important for pediatric surgeons are very rare. One exception is the murine model of extrahepatic biliary atresia (EHBA) (Petersen et al. 1997). In this model, newborn Balb/c mice are infected with rhesus rotavirus group A45. As a result, the full spectrum of EHBA develops as it is seen in newborn with this disease. However, this model is not a model to mimic failed embryology. But it highlights the possibility that malformations are not caused by embryonic disorders but caused by fetal or even postnatal catastrophes.
Spontaneous Malformations Without Genetic Background
This part on embryology and animal models highlights not only the importance to study embryos with experimental malformations but also the study of normal animal embryos. Today, much information in current textbooks on human embryology stems actually from studies done in animals of varies species. Many of these are outdated. The wide use of transgenic mice in order to mimic congenital malformations makes morphological studies of the organ systems in normal mice mandatory. Otherwise the interpretation of the effects by deletion of genetic information can be very difficult or even misleading.
Scanning Electron Microscopic Atlas of Normal and Abnormal Development in Embryos
In this section we want to present examples of normal and abnormal development as we have seen them in our studies in our labs over the past 30 years using scanning electron microscopy (SEM). We use the form of an embryological atlas following the old motto “A picture says more than a thousand words.” We focus on the following developmental processes:
Normal and abnormal foregut development (chicken embryos)
Normal and abnormal development of the diaphragm (rat embryos)
Development of the midgut (rat embryos)
Normal and abnormal development of the hindgut (mice and rats)
The development of the external genitalia and the urethra (rat embryos)
Testicular descent (rat embryos)
Normal Foregut Development
Lateral endodermal ridges appear in the primitive foregut which fuse and form the tracheoesophageal septum.
This solid endodermal septum is partly removed by apoptosis and substituted by mesenchymal cells (Fig. 3).
This theory had been described in detail by Rosenthal (1931) and Smith (1957). However, neither Zaw Tun (1982) nor O’Rahilly and Müller (1984) were able to confirm these sequences of embryological events. According to them, the term “separation” is a misnomer as the formation of the trachea is simply the result of the downgrowth of the respiratory diverticulum (Fig. 3) (Merei and Hutson 2002).
Although a number of models for abnormal foregut development exist, a clear morphological description of the embryological events that finally lead to esophageal atresias is still missing. Based on our observations, the development of the malformation can be explained by disorders either of the formation of the folds or of their developmental movements (Kluth et al. 1987; Metzger et al. 2011a; Kluth and Habenicht 1987):
Atresia of the esophagus with fistula (Fig. 13C1):
The dorsal fold of the foregut bends too far ventrally. As a result the descent of the larynx is blocked. Therefore the common tracheoesophageal space remains partly undivided and lies in a ventral position. Due to this ventral position, the common space differentiates into trachea.
Atresia of the trachea with fistula (Fig. 13C2): The foregut is deformed on its ventral side. The developmental movements of the folds are disturbed, and the tracheoesophageal space is dislocated in a dorsal direction, where it differentiates into esophagus.
Laryngo-tracheo-esophageal clefts (Fig. 13C3): Faulty growth of the folds results in the persistence of the primitive tracheoesophageal space.
In our collection of chicken embryos, we came across an embryo with abnormal foregut features (Fig. 13b). When compared to normal embryos of the same age group (Fig. 13a), the following statements can be made: (I) Obviously, the pharynx ends blindly. (II) The dorsal part of the common foregut space is missing. (III) The ventral part of the common space has the size of a trachea. (IV) This foregut looks like the hypothetical form C1 in Fig. 13.
Development of the Diaphragm
The traditional theories of diaphragmatic development have been summarized by Kluth et al. (1989). Using SEM, we have recently restudied the diaphragmatic development. For practical reasons it is essential to note that the early diaphragm consists of two parts:
The septum transversum which, in young embryos, is identical to the floor of the pericardium
- (b)The structures that surround the pleural cavity. They are:
The posthepatic mesenchymal plate (PHMP) (Irtani 1984), which covers the dorsal aspect of the liver and is in continuity to the septum transversum ventrally and cranially.
The pleuroperitoneal fold (PPF) which separated the pleura from the peritoneal cavity. This fold connects ventrally to the septum transversum and the PHMP and dorsally to the mesonephric ridge (Mayer et al. 2011). This PPF is a structure that is identical to the pleuroperitoneal membrane of the old literature (Kluth et al. 1989).
The dorsal mediastinum which contains the esophagus, the trachea, and the aorta.
In Fig. 15, the final closure of the PPO is shown. In this process the PHMP starts to cover the last free areas of the liver (Fig. 15a). In this process, the pleuroperitoneal fold (PPF) plays only a minor role.
In the literature, the nomenclature of the various parts of the diaphragm is confusing. We use the term PPF for a structure which was formally known as pleuroperitoneal membrane (Kluth et al. 1989; Mayer et al. 2011). The term PPF is used differently by Greer and coworkers (Clugston et al. 2010). Their PPF is very similar to the PHMP as described by IRITANI and us but seems to include the ventral part of our PPF.
In the past, several theories were proposed to explain the appearance of posterolateral diaphragmatic defects:
Pushing of the intestine through posterolateral part (foramen of Bochdalek) of the diaphragm (Bremer 1943)
Abnormal persistence of the lung in the pleuroperitoneal canal, preventing proper closure of the canal (Gattone and Morse 1982)
Abnormal development of the early lung and posthepatic mesenchyme, causing non-closure of pleuroperitoneal canals (Irtani 1984)
Of these theories, failure of the pleuroperitoneal membrane to meet the transverse septum is the most popular hypothesis to explain diaphragmatic herniation. However, using SEM techniques (Kluth et al. 1989; Mayer et al. 2011), we could not demonstrate the importance of the pleuroperitoneal membrane for the closure of the so-called pleuroperitoneal canals (Fig. 15).
It is still speculated that delayed or inhibited closure of the diaphragm will result in a diaphragmatic defect that would allow herniation of the gut into the fetal thoracic cavity. In a series of normal staged embryos, we measured the width of the pleuroperitoneal openings and the transverse diameter of gut loops (Kluth et al. 1995a). On the basis of these measurements, we estimated that a single embryonic gut loop requires at least an opening of 450 μm size to herniate into the fetal pleural cavity. However, in none of our embryos, the observed pleuroperitoneal openings were of appropriate dimensions. This means that delayed or inhibited closure of the pleuroperitoneal canal cannot result in a diaphragmatic defect of sufficient size. Herniation of the gut through these openings is therefore impossible. Thus the proposed theory about the pathogenetic mechanisms of congenital diaphragmatic hernia (CDH) development lacks any embryological evidence. Furthermore the proposed timing of this process is highly questionable (Kluth 1993).
Timing of diaphragmatic defect appearance. Iritani (Irtani 1984) was the first to notice that nitrofen-induced diaphragmatic hernias in mice are not caused by an improper closure of the pleuroperitoneal openings but rather the result of a defective development of the posthepatic mesenchymal plate (PHMP). In our study in rats, clear evidence of disturbed development of the diaphragmatic anlage was seen on ED 13 on the left and ED 14 on the right diaphragm anlage (Fig. 14) (Kluth and Tander 1995; Kluth et al. 1996). In all embryos affected, the PHMP was too short. This age group is equivalent to 4–5 week-old human embryos (Kluth and Tander 1995).
Location of diaphragmatic defect. In our SEM study, the observed defects were localized in the area of the PHMP (Fig. 14). We identified two distinct types of defects (Haeckel 1975): large “dorsal” defects and (Schwalbe 1906) small “central” defects (Kluth and Tander 1995). Large defects extended into the region of the pleuroperitoneal openings. In these cases, the closure of the pleuroperitoneal openings was usually impaired by the massive ingrowth of the liver (Figs. 14 and 15). If the defects were small, they were consistently isolated from the pleuroperitoneal openings which closed normally at the 16th or 17th day of gestation. Thus, in our embryos with CDH, the region of the diaphragmatic defect was a distinct entity and was separated from that part of the diaphragm where the pleuroperitoneal “canals” are localized. We conclude therefore that the pleuroperitoneal openings are not the precursors of the diaphragmatic defect.
Why lungs are hypoplastic. Soon after the onset of the defect in the 14-day-old embryo, the liver grows through the diaphragmatic defect into the thoracic cavity (Fig. 14). This indicates that from this time on, the available thoracic space is reduced for the lung, and further lung growth is hampered. In the following stages, up to two- thirds of the thoracic cavity can be occupied by the liver (Figs. 16 and 17). Herniated gut was found in our embryos and fetuses only in late stages of development (21 days and newborns) (Fig. 17). In all of these, the lungs were already hypoplastic, when the bowel entered the thoracic cavity (Kluth and Tander 1995).
Based on these observations, we conclude that the early ingrowth of the liver through the diaphragmatic defect is the crucial step in the pathogenesis of lung hypoplasia in CDH. This indicates that growth impairment is not the result of lung compression in the fetus but rather the result of growth competition in the embryo: the liver that grows faster than the lung reduces the available thoracic space. If the remaining space is too small, pulmonary hypoplasia will result.
Development of the Hindgut
As in the foregut, a process of septation has been postulated for the proper subdivision of the “cloaca” into the dorsal anorectum and the ventral sinus urogenitalis. Disorders in this process of differentiation are thought to be the cause of cloacal anomalies such as persistent cloaca and anorectal malformations (Stephens 1963).
However, for many years, this process of septation has been under dispute. Some authors (Toumeux 1888; DeVries and Friedland 1974) believe that the descent of a single fold separates the urogenital part from the rectal part by ingrowth of mesenchyme. Others (Retterer 1890) think that lateral ridges appear in the lumen of the cloaca, which progressively fuse along the midline and thus form the septum. Van de Putte (vd Putte 1986) denied the existence of any process of septation.
The “cloaca” is not subdivided into two equal parts (Fig. 19a, b). The much larger ventral part gives rise to the future distal urethra.
The dorsal “cloaca” contains the future anorectal region. The future anal opening is situated in the dorsal part of the “cloacal’ membrane,” close to the tail fold (Fig. 19b).
As already mentioned, a number of animal models exist which allow embryological studies of abnormal hindgut development. In our studies we used embryos obtained from SD-mice. The SD-mouse is a spontaneous mutation of the house mouse, characterized by a short tail (Fig. 5). Homozygous or heterozygous offspring of these mice shows skeletal, urogenital, and anorectal malformations (Kluth et al. 1991). Therefore these animals are ideal for detailed studies of anorectal malformations.
The cloacal membrane was always too short (Fig. 20c, d). In all cases the dorsal part of the cloacal membrane was absent.
The proximal hindgut joined the cloaca at an abnormal position (Fig. 20c, d).
Development of the External Genitalia and the Urethra
Normal Development of the External Genitalia
Rupture of the cloacal membrane (Fig. 23).
At embryonic day 17.5, the dorsal disintegration of the cloacal membrane can be seen (Fig. 23a, b). The ventral (urethral) part of the cloacal membrane remains intact. In Fig. 23c, this process of disintegration is seen in more detail. The ectodermal part of the cloacal membrane shows clear signs of disintegration. The tip of the urorectal fold is seen which later forms the perineum. Ventral to the tip of the urorectal fold, an opening is seen which is in connection to the distal urethra. Dorsally the anal opening is seen. At this time point, the external genitals allow no differentiation between the sexes.
Further development of the external genitalia.
A rupture of the ventral part of the cloacal membrane cannot be seen in older embryos and fetuses. In males, the transient urethral opening disappears. Later a “raphe” is seen in this position (Fig. 24a). In females, this “raphe” is missing (Fig. 24b).
- (c)Special dissections of a rat embryo at embryonic day 18.5 allow the following statements (Fig. 25a–c): The urethra is composed of two parts, the proximal and the distal part. The epithelium of the distal part reaches to the tip of the phallus. It is interesting to see that the urethra is connected to the perineal region by a short canal, the so-called “cloacal canal.” In our opinion this is the future female urethra.
In rats, the urethra is always present as a hollow organ during urethral embryogenesis and that it is always in contact with the tip of the genitals.
Initially a double urethral anlage exists. The differentiation in female and male urethra happens in rats more than 18.5 days old.
We had no evidence for the disintegration of the urogenital cloacal membrane and a fusion of lateral portions within the perineum.
Abnormal Development of the External Genitalia (Hypospadia Formation)
In our opinion, more than one embryological mechanism is at play in the formation of the hypospadia complex (Kluth et al. 1988, 2011a). The moderate degrees, such as the penile and glandular forms, represent a developmental arrest of the genitalia. They take their origin from a situation comparable to the 20-day-old embryo. Consequently the penis, not the urethra, is the primary organ of the malformation. Perineal and scrotal hypospadias are different from the type discussed previously. Pronounced signs of feminization in these forms suggest that we are dealing with a female-type urethra. Origin of this malformation complex is an undifferentiated stage as may be seen in the 18.5-day-old rat embryo.
The Development of the Midgut
In the first phase, the midgut loop develops inside the umbilicus (so-called “physiological herniation of the midgut”). Here, a 90-anti-clockwise rotation around the axis of the mesentery is thought to take place.
After the “return” of the midgut into the abdominal cavity, another anti-clockwise “rotation” of 180° is thought to take place inside the abdominal cavity (second phase). As a result, the region of the cecum moves to the right, thus overcrossing the mesenteric root, while the flexura duodeno-jejunalis is pushed to the left beneath the root of the mesentery (see schematic drawing in Fig. 1b) (Mall 1898; Frazer and Robbins 1915). These two phases sum up 270°. In contrast to this description, Grob (1953) subdivides this intra-abdominal process of rotation into two steps of 90° each.
A central part with the duodenum and the distal colon close to the root of the mesentery.
A ventral part inside the extra embryonic coelom of the umbilicus (so-called “physiological herniation”). Here the cecum and the distal small bowel can be identified.
A middle part which connects the central part with the ventral part inside the umbilicus. Here the umbilical vessel, the small bowel on the right, and the proximal part of the colon on the left can be seen (Figs. 26b and 27d).
In the further development (ED 14–16), growth activities are seen in the area of the duodenum and inside the extraembryonic coelom.
Figure 26 shows the steps important for the duodenal developmental. In Fig. 26b (rat embryonic day 15) the duodenojejunal loop has been formed due to longitudinal growth of the duodenum. Further growth pushes this loop beneath the root of the mesentery (Fig. 26c).
Figure 27 shows the development of the intra-umbilical loops. These loops are the result of longitudinal lengthening of the small gut. Note the absence of any signs of rotation around the axis of the mesentery in Fig. 27d in a phase of active loop development inside the extra embryonic coelom.
Figure 28 shows the “return” of the midgut into the abdominal cavity. The cecum is seen inside the abdominal cavity in a ventral position close to the abdominal wall (embryonic day 17). The colon is entirely to the left in this phase of development. It is a small bowel loop which is still extraembryonic inside the umbilicus. In this phase of small bowel “return,” the bowel loops have already developed locally inside the abdominal cavity (Fig. 26c).
We conclude from our observations that the midgut can be subdivided in three parts, of whom the central and the ventral parts are of major importance. Localized longitudinal growth in the area of the duodenum leads to the formation of the duodenojejunal loop and its final position beneath the root of the mesentery. At the same time, localized growth of the small bowel has led to the formation of bowel loops inside the umbilicus and, later, inside the abdominal cavity. The growth activity of the large bowel is minimal, compared to that of the small bowel. Neither in the phase of loop formation inside the extraembryonic coelom of the umbilicus nor in the phase following the “return” of the gut into the abdominal cavity rotation of the gut around the axis of the mesentery can be observed. All processes in midgut development are the result of longitudinal lengthening of gut.
Since John Hunter in 1762, many researchers studied the embryology of testicular descent. In many of these studies, the importance of the gubernaculum during this process has been highlighted. However, a clear illustration of this rather simple process is still lacking (Heyns and Hutson 1995).
The intra-abdominal descent: In this phase, the testis, which initially lies in close contact to the kidney, moves into the inguinal area.
The inguinal descent: In this phase the testis moves into the area of the scrotum.
The Role of the Gubernaculum
Thus in our findings, we cannot support the opinions about the role of the gubernaculum during the testicular descent. Its main role seems to be its transformation into the PVP. We believe that intra-abdominal pressure probably plays an active role at least in the phase of the inguinal phase of testicular descent.
- Booß D, Okmian L. Ein neues tierexperimentelles Modell für die Ösophagusforschung. Paper presented at the International Symposium on Oesophageal Atresia, Bremen, 31th October and 1st November 1974.Google Scholar
- Bremer JL. The diaphragm and diaphragmatic hernia. Arch Pathol. 1943;36:539–49.Google Scholar
- Broman I. Über die Entwicklung des Zwerchfells beim Menschen. Verh Anat Ges. 1902;16:9–17.Google Scholar
- Felix W. Die Entwicklung der Harn- und Geschlechtsorgane. In: Keibel F, Mall FP, editors. Handbuch der Entwicklungsgeschichte des Menschen, vol. 2. Leipzig: Hirzel; 1911. p. 92–5.Google Scholar
- Frazer TE, Robbins RF. On the factors concerned in causing rotation of the intestine in man. J Anat Physiol. 1915;50:74–100.Google Scholar
- Gilbert SF. Developmental biology 7th edition, chapter 23. Sunderland: Sinauer Associates; 2003.Google Scholar
- Gray SW, Skandalakis JE. Embryology for surgeons. Philadelphia: Saunders; 1972a. p. 359–85.Google Scholar
- Gray SW, Skandalakis JE. Embryology for surgeons. Philadelphia: Saunders; 1972b. p. 595–631.Google Scholar
- Gray SW, Skandalakis JE. Embryology for surgeons. Philadelphia: W. B. Saunders; 1972c. p. 63–100.Google Scholar
- Grob M. Über Lageanomalien des Magen-Darm-Traktes infolge Störungen der fetalen Darmdrehung. Schwabe: Basel; 1953.Google Scholar
- Haeckel E, cited in Starck D: embryologie (ed 3). Stuttgart, Germany, Thieme, 1975.Google Scholar
- Hamilton WJ, Boyd JD, Mossman HW. Human embryology. 3rd ed. Baltimore: Williams & Wilkins; 1962. p. 294–9.Google Scholar
- Holder RM, Ashcraft KW. Congenital diaphragmatic hernia. In: Ravitch MM, Welch KJ, Benson CD, Aberdeen E, Randolph JG, editors. Pediatric surgery, vol. 1. 3rd ed. Chicago: Year Book Medical Publishers; 1979. p. 432–45.Google Scholar
- Jacob HJ. Experimente zur Entstehung entodermaler Organanlagen. Untersuchungen an explantierten Hühnerembryonen. Anat Anz. 1971;128:271–8.Google Scholar
- Kluth D, Habenicht R. The embryology of usual and unusual types of oesophageal atresia. Pediatr Surg Int. 1987;1:223–7.Google Scholar
- Kluth D, Petersen C, Zimmermann HJ, Mühlhaus K. The embryology of congenital diaphragmatic hernia. In: Puri P, editor. Congenital diaphragmatic hernia: modem problems in pediatrics, vol. 24. Basel: Karger; 1989. p. 7–21.Google Scholar
- Kluth D, Fiegel HC, Metzger R. Embryology of the hindgut. Semin Pediatr Surg. 2011b;20152-60Google Scholar
- Komfält SA, Okmian L. JonssonN Healing of circular oesophageal mucosal defect An experimental study in the piglet. Z Kinderchir. 1973;13:184–97.Google Scholar
- Lemez L. Sites for experimental production of tracheal and/or oesophageal malformations in 4-day-old chick embryos. Folia Morphol (Praha). 1980;28:52–5.Google Scholar
- Mall FP. Development of the human intestine and its position in the adult. Bull Johns Hopkins Hosp. 1898;9:197–208.Google Scholar
- Molenaar JC, Tibboel D. The pathogenesis of atresias of the small bowel and colon. S Air J Surg. 1982;20:87–95.Google Scholar
- Retterer E. Sur l’origin et de revolution de la region ano-génitale des mammiferes. J Anat Physiol. 1890;26:126–216.Google Scholar
- Rosenthal AH. Congenital atresia of the esophagus with tracheoesophageal fistula: report of eight cases. Arch Pathol. 1931;12:756–72.Google Scholar
- Schwalbe E. Die Morphologie der Missbildungen des Menschen und der Tiere. 1. Teil Allgemeine Mißbildungslehre (Teratologie). Germany: Jena; 1906. p. 143–4.Google Scholar
- Smith EL. The early development of the trachea and the esophagus in relation to atresia of the esophagus and tracheo-oesophageal fistula. Contrib Embyol Cameg Inst. 1957;36:41–57.Google Scholar
- Spaulding MH. Tire development of the external genitalia in the human embryo. Contrib Embryol Cameg. 1921;13:67–88.Google Scholar
- Steding G. The anatomy of the human embryo – a scanning electron-microscopic atlas. Basel: Karger; 2009.Google Scholar
- Stephens FD. Congenital malformations of the rectum, anus, and genitourinary tract. Edinburgh: Livingstone; 1963.Google Scholar
- Toumeux F. Sur le premiers developpements du cloaque du tubercle genitale et de l’anus chez Fembryon moutons, avec quelques remarques concemant le developpement des glandes prostatiques. J Anat Physiol. 1888;24:503–17.Google Scholar