In North America, twins occur in every 80 to 90 pregnancies. Twins may arise either from 2 zygotes resulting in dizygotic twins or from 1 zygote resulting in monozygotic twins. Approximately two-thirds of twins are dizygotic and one-third are monozygotic. Approximately 1 in every 400 monozygotic twin pregnancies have incomplete twinning and result in conjoined twins. The incidence of conjoined twinning varies in different parts of the world. In the United States, the incidence ranges from 1:100,000 to 1:250,000, while in sub-Saharan Africa, it may be as frequent as 1:50,000 with isolated incidences of 1:14,000 in some areas. The higher estimates of incidence are the result of episodes of several births in a circumscribed geographic area over a short period of time. The incidence for live-born conjoined twins is significantly different. Conjoined twins are stillborn up to 60% of the time while another 35% die in the first 24 hours of life. This makes the incidence of live-born conjoined twins closer to 1 in 200,000 births. The frequency of monozygotic twins is not affected by racial differences or maternal age. It is reasonable to assume that the frequency of conjoined twins is similarly unaffected. However, autopsy evaluation reveals a predominance of females being affected by conjoined twinning approximately 3 times as frequently as males.
Conjoined twins have been described in many species in both the plant and animal kingdoms. The exact embryology remains theoretical and controversial. There are 2 different theories as to how these twins develop. The first suggests that conjoined twinning results from failure of separation of the embryonic disc late in the blastomere stage at 15 to 17 days gestation, also termed incomplete fission. The second theory suggests that the twins result from partial fusion of 2 embryonic plates at 3 to 4 weeks of gestation. Each theory has strengths and weaknesses and neither has been proved.
Conjoined twins are always monozygotic and monochorionic, though 1 set of diamniotic omphalopagus twins has been reported. Twins are always of the same sex, which would seem to lend more weight to the fission theory as it would seem just as likely that a male and female embryonic plate would fuse; however, this has never been observed. Several cases of pseudohermaphrodites, however, have been documented. Conjoined twins have been reported as part of a set of triplets or quadruplets with the conjoined twins always being isosexual while the other infant or infants may be of the opposite sex. There are also extremely rare sets of conjoined tripling and quadrupling that have been reported as well.
Conjoined twins are always joined in a homologous position, cranially, caudally, ventral to ventral, or dorsal to dorsal. If one accepts fusion as the most likely cause of conjoined twinning, the type of twin that develops depends on where the embryonic discs come into contact. The theory of fusion, which may also be termed the spherical theory, was developed by Dr. Rowena Spencer after analysis of many hundreds of conjoined twins cases. This theory postulates that 2 monovulvar embryonic discs are present floating on either the outer surface (yolk sac) or inner surface (amniotic cavity) of a sphere. These discs may come into contact with each other and fuse either ventrally over a single yolk sac (cephalopagus, thoracopagus, omphalopagus, ischiopagus, and parapagus twins) or dorsally (craniopagus, rachipagus, and pygopagus twins). The orientation of the embryonic disc determines the type of conjoined twinning that develops. The union is not random, however, and may only take place where surface ectoderm is absent or is genetically programmed to break down such as at the oropharyngeal and cloacal membranes or along the neural tube. These unions are always homologous in that the oropharyngeal and cloacal membranes never fuse. Ventral union in general leads to twins sharing 1 abdomen and 1 umbilicus, while dorsal union of the neural tube leads to infants with 2 separate abdomens and 2 umbilical cords.
While the spherical theory of conjoined twin development is appealing in that it can explain almost all types of twinning, additions to the theory must be considered to further explain development. Dr Spencer calls these additions, “adjustments to conjunction,” which include aplasia of contiguous lateral anlagen, and division and diversion of midsagittal structures. Division and diversion are best illustrated by cephalopagus twins. With this type of twin, as the oropharyngeal membrane fuses, it splits sagittally, diverts laterally, and fuses with the laterally diverted membrane of the other disc. This results in the Janus type of cephalopagus twin in which there appears to be a face on either side of the body, one half of the face from each twin. Brain imaging demonstrates that in these twins, one half of the frontal lobes derives from each twin as the result of the same process. Aplasia of contiguous lateral anlagen is demonstrated in parapagus twins where the structures in the area of union fail to develop. Using parapagus twins as an example the medial leg for each child is absent because of this process.
There is one other type of twinning that may occur that is not well explained by the above theories: the development of a parasitic or atypical twin. These twins have only a partially formed twin extending from the body of the better developed twin. These parasitic twins cannot survive independently. As many as 20% of conjoined twins have developmental malformations in multiple organ systems. These anomalies are not random, and are typically found in the right-sided twin, and in male sets of twins. The risk of recurrence for conjoined twins in essentially negligible.
To classify conjoined twins appropriately requires a common terminology that is all-too-often lacking when these children are described, which thus gives rise to many different terms in the literature. The classification of the type of twin depends on the primary area of union. Since conjoined twins represent a continuum of development in the manner in which the 2 embryologic discs fused or split, it may at times be difficult to classify 1 type of twin based on external appearance only. Picturing a continuum of development is a helpful means of visualizing conjoined twin development. Ventrally united twins may be described by 2 fetuses joined ventrally and then rotated in both the x–y axis and the y–z axis. By rotating the infants in this manner, one can produce every type of ventrally united conjoined twins. This challenge in classifying twins and the reason there are so many terms in the literature may be due to some overlap between 1 type of twin's external appearance and that of another type.
When referring to a particular infant in a set of conjoined twins it is important to use uniform terminology in describing both the twins and their anatomy. The terms anterior and posterior are not applied to individual twins, but rather refer to the twins as a whole. When the twins are lying in a position such that the abdomen is most fully exposed, right and left are oriented from the twins' perspective and the infants are termed “the twin on the right” and “the twin on the left.” Structures within individual twins should be referenced as dorsal and ventral. Using these reference points works well for describing the anatomic relationships of all twins except craniopagus and rachipagus twins.
Conjoined twins are classified into 8 main categories: omphalopagus (umbilicus), thoracopagus (chest), ischiopagus (hip), craniopagus (helmet), and pygopagus (rump), parapagus (side), cephalopagus (head), and rachipagus (spine). Additionally, parasites, teratomas, and fetus-in-fetu may also be considered part of this process. The Greek term “pagus” means to be fixed or joined, and is preceded by a prefix that describes where the twins are joined. Suffixes are used to describe other aspects of the twins. For example: prosopus refers to the face; brachius to the upper limb; and pus to the lower limb. Other prefixes include di-, tri-, and tetra-, used to describe the number of limbs or heads. The most recent and most logical classification of conjoined twins is based on physical examination with the exception of the important distinction between omphalopagus and thoracopagus twins (Fig. 60-1). Externally, omphalopagus and thoracopagus twins may be indistinguishable. However, thoracopagus twins always have some degree of cardiac union, whereas omphalopagus twins never have a cardiac connection.
Classification of conjoined twins based on the anatomy of the shared area. A. Cephalopagus. B. Thoracopagus. C. Omphalopagus. D. Ischiopagus. E. Parapagus. F. Craniopagus. G. Pygopagus. H. Rachipagus.
The 8 types of twins can be organized into 2 primary groups based on the orientation of the union, either ventral to ventral or dorsal to dorsal. This ventral union includes those twins joined at the rostral and mid-ventral areas, such as cephalopagus infants, to those fused more caudally, such as ischiopagus and parapagus twins. Infants in this group share some aspect of the gastrointestinal tract and generally have 1 umbilicus. The ventral union of twins often is more typically a ventrolateral orientation as the twins may not directly face each other. Three types of twins in the ventral union group have all 4 extremities present and no pelvic union. These include cephalopagus, thoracopagus, and omphalopagus. The cephalopagus twins are united from the vernix of the head to the umbilicus. These twins share the forebrain, optic chiasm, and the upper aerodigestive tract. Two faces are present, although 1 of the faces may be rudimentary. All of these twins are either stillborn or die shortly after birth. Thoracopagus twins are the most common type of twin comprising 40% of conjoined twins. Historically, they have often been classified with omphalopagus twins because of their common external appearance. Thoracopagus twins are united from the upper thorax to the umbilicus and always have a cardiac connection, even if only by a single vessel. This distinction from omphalopagus twins is important because thoracopagus twins often cannot be separated while omphalopagus twins can. Thoracopagus twins typically share a compound heart of varying degrees of complexity. The upper gastrointestinal tract is shared in 40% and the biliary tree in 22%. The gastrointestinal tract union, if present, usually starts at the level of the mid-duodenum and may extend into the distal small bowel. A portion of the liver is shared and biliary union should be suspected if the twins have a common duodenum. Successful separation of thoracopagus twins with a shared heart is unusual. There are recorded cases of atrial conjunction that have survived separation; however, when there is complex ventricular union, there are no cases described where both twins have survived separation and survival of even 1 twin is rare. All attempts at separation with complex cardiac union, have either attempted to salvage the heart for 1 child with planned sacrifice of the other or have attempted to create 2 cardiac units, 1 for each child. Neither approach has been particularly successful.
Omphalopagus twins account for 35% of conjoined twins and are joined from the lower thorax to the umbilicus. There is never any type of cardiac connection. The union usually extends from the xiphoid to the umbilicus. There is a common bridge of liver tissue in almost all such twins and a part of the diaphragm may be shared. In some cases this may be the only connection. The digestive tract is often shared with one sixth of cases joining in the upper tract at the mid-duodenum. The bile ducts and portal drainage may also be anomalous particularly in cases where the duodenum is fused. The lower aspect of the sternum may be joined and some of these twins will also have an omphalocele.
Ischiopagus twins account for 6% of conjoined twins. Depending on the number of lower extremities these twins may be united from the lower chest through the pelvis. Usually, there are 2 sacrums and 2 symphyses pubis. There may be 2 (dipus), 3 (tripus), or 4 (tetrapus) lower extremities, and external genitalia and anus are always involved. Ischiopagus dipus and parapagus twins may be confused with each other because ischiopagus twins and parapagus twins represent a continuum of rotation along the x, y, and z axes. Parapagus twins are united laterally and may share parts of the upper gastrointestinal system, while ischiopagus twins will not. In parapagus infants, the developing fetus succumbs to aplasia of contiguous lateral anlagen, such that in the true parapagus twins, the infants are united laterally with no internal rotation along the z axis, and all medial structures from each twin such as the leg or arm, depending on the degree of union, are absent. Parapagus twins may be only united in the lower body and in this case will have 4 arms, while other sets may be more completely united and have only 2 or 3 arms. Parapagus twins always have only 2 legs.
With ischiopagus twins, internal rotation along the z axis occurs such that the twins start to face each other; aplasia of contiguous lateral anlagen lessens and anomalies increase as the various structures of 2 separate infants become more apparent. As the twins rotate internally to face each other, duplicate sets of structures begin to emerge dorsally. The dorsal structures are often somewhat rudimentary. In these cases, the medial buttock or leg of each twin may become apparent. The fused leg may first appear as an undeveloped appendage and, as internal rotation continues, becomes apparent as the tripus limb, comprised of the fused medial limb of each child. Internal rotation also causes the increasing presence of diminutive dorsal genetalia. In the continuum of development, the spine of each twin may not only be more internally rotated toward the opposite twin, but also be oriented farther apart along the x and y axis, such that the twins may face each other but are joined only at the pelvis with 4 legs, as in ischiopagus tetrapus infants. In these infants, the spine is oriented 180° from the other child.
The internal anatomy of ischiopagus twins may be quite complex but usually will allow separation. The upper gastrointestinal tract is separate but the more distal small bowel may be joined at the level of a Meckels diverticulum, leading to a shared distal ileum and colon with a dual mesentery, half from each child. This colon may end as an imperforate anus, cloaca, urogenital sinus, or perineal fistula. One or two anal sphincter complexes may be present depending on the degree of development of the dorsal complex. The liver is usually fused in the ischiopagus tripus and parapagus twins. The biliary tree in ischiopagus twins is separate, while for parapagus twins, it may be shared. The genitourinary system will likely present a difficult reconstructive challenge as the union may be very complex. The number of kidneys varies from 1 to 4, and ureters may cross the midline and empty into the opposite twin. With ischiopagus twins, there may be 2 uteri and 4 ovaries in various degrees of development. The external genitalia in ischiopagus tetrapus twins are rotated 90° from the normal orientation, and are shared by each twin. Between each set of legs there exist a normal labia, clitoris, urethra, and vaginal orifice. With ischiopagus tripus twins, the external genitalia may open on either side of a common anus or may be part of a cloacal malformation or urogenital sinus. In male ischiopagus tetrapus twins, a penis and scrotum are present between each set of legs, but, as in girls, half of the genitalia derives from either twin. As union moves toward ischiopagus tripus, boys may have 2 scrotums and 1 penis or other various anomalies.
The second basic group of conjoined twins includes those united dorsally: twins joined at the neural tube and the cranio-vertebral axis. These infants face away from each other and have 2 umbilical cords, and either separate or nearly separate gastrointestinal tracts. There are 3 types of twins in this group: craniopagus, pygopagus, and rachipagus. All types in this group have 4 arms and 4 legs. The craniopagus twins account for 2% of conjoined twins and may be united at any portion of the skull except the foramen magnum and the face. They are distinctly different from the cephalopagus twins that are ventrally united, with further union in the trunk. Craniopagus union may not be symmetric but may involve the bony skull, meninges, and even the surface of the brain. They can be subclassified in terms of the site of union: frontotemporal/frontoparietal 25%; parietal 45%; and occipital/occipitoparietal 30%. Further categorization of these twins describes either partial or total conjunction. Total conjunction refers to the brains being connected or separated by the arachnoid only. In the partial form, the brain is separated by the bone and dura and each has a separate leptomeninges. The ability to separate these types of twins centers on the degree to which the dural venous sinuses are connected and the subsequent risk of hemorrhage or severe neurologic deficit.
Pygopagus (rump) twins are united at the sacrococcygeal and perineal areas. They account for 20% of conjoined twins. Seven percent of these cases are boys. The twins face away from each other and are joined at the sacrum, buttocks, and perineum. The spinal cord may sometimes be part of the union. Forty percent of these twins will have a common dural tube. Close to 50% of these children have some anomaly of the central nervous system (CNS) and/or vertebral column, which may include hydrocephalus, myelomeningocele and spina bifida, and rachischisis. The pelvis is separate except for osseous fusion in the lateral portion of the sacrum and coccyx. There are usually 2 sets of external genitalia with fusion in the perineum at the area of the posterior forchette in girls. There is only 1 anus, which is posterior to the urogenital fusion. The lower gastrointestinal tracts are separate but the rectum usually unites several centimeters proximal to the anus. Twenty five percent will have renal anomalies including renal agenesis, ectopia, and fusion. Usually these children can be separated, but the perineal reconstruction may be challenging.
Rachipagus (spine) twins are extremely rare and can be categorized as dorsal union from above the sacrum cephalad. The union can include the occiput and segments of the vertebral column. Table 60-1 summarizes the possible unions of conjoined twins.
Table 60-1Possible Unions of Conjoined Twins |Favorite Table|Download (.pdf) Table 60-1 Possible Unions of Conjoined Twins
|Group ||Type ||Cardiac ||Liver ||Upper Gastrointestinal |
|Ventral ||Cephalopagus ||+ ||+ ||+ |
| ||Thoracopagus ||+ ||+ ||+ |
| ||Omphalopagus || ||+ ||+ |
| ||Ischiopagus || ||+ || |
| ||Parapagus ||+ ||+ ||+ |
|Dorsal ||Craniopagus || || || |
| ||Pygopagus || || || |
| ||Rachipagus || || || |
|Group ||Type ||Lower Gastrointestinal ||Genitourinary || |
|Ventral ||Cephalopagus || || || |
| ||Thoracopagus || || || |
| ||Omphalopagus || || || |
| ||Ischiopagus ||+ ||+ || |
| ||Parapagus ||+ ||+ || |
|Dorsal ||Craniopagus || || || |
| ||Pygopagus ||+ ||+ || |
| ||Rachipagus || || || |
|Group ||Type ||Nervous System ||Bony Structures || |
|Ventral ||Cephalopagus ||+ ||+ || |
| ||Thoracopagus || ||+ || |
| ||Omphalopagus || || || |
| ||Ischiopagus || ||+ || |
| ||Parapagus ||+ ||+ || |
|Dorsal ||Craniopagus ||+ ||+ || |
| ||Pygopagus ||+ ||+ || |
| ||Rachipagus ||+ ||+ || |
The diagnosis of conjoined twins may be emotionally traumatic for the family and an ethical challenge for everyone involved. Many fetuses are likely terminated at the time of diagnosis. For those carried to term and who then survive, decisions must be made regarding their subsequent care. If the diagnosis is known prenatally, it is most helpful to have discussions with the family before birth so that many of the issues regarding care will have already been decided. In many cases, prenatal evaluation of the children will provide enough information to accurately predict the types of problems the twins are likely to encounter, and the parents may be able to make informed decisions about their children's care. The decision whether to separate the twins or not may be fairly clear if 1 twin has obvious anomalies not compatible with life. If both children can survive, however, the decision to separate the infants is not necessarily straightforward, as some would argue that conjoined twins left together may live long and fulfilling lives. There are numerous historical examples of this. Cultural experience and societal norms all play a role in the decision. How the Western world views conjoined twins may be very different from the viewpoint of other societies, and the parent's background must be taken into consideration. Even if separated, the individual twin's quality of life must be considered as each may have significant medical issues as a result of the separation surgery. In some cases it may well be that if left alone, the children will require little medical care outside of routine health maintenance.
While certain types of conjoined twins are quite likely to survive separation surgery with 2 surviving infants, not all separation surgeries may be so successful. Perhaps the most ethically challenging situation is that in which both twins are born alive but it becomes apparent that both cannot survive in the long term and consideration is given to sacrifice 1 twin to save the other. Numerous ethical studies have addressed this situation. This brings to the forefront ethical arguments regarding what constitutes life, the individual, societal norms, and the law. Christian, Jewish, and Islamic scholars have all written opinions that separation surgeries performed where 1 child is likely to die as a result of the surgery are acceptable if each child has the opportunity to live; in other words, if the surgery is not a planned sacrifice of 1 twin. Arguments for separation in these cases include the doctrine of double effect, 1 twin as an “unjust aggressor,” appendage, or surrogacy and the view that 1 twin has been marked for death by nature. Some ethics scholars have written that separations performed to save 1 twin and cause the demise of the other may be accepted by Western society if the decision was made fairly, for example, by means of anatomy. Each argument has both strengths and weaknesses but none can fully withstand in-depth scrutiny.
Physicians must pay attention to their own moral compass and come to terms with the Hippocratic Oath when dealing with some of these cases. Surgeons need to examine their own motives for offering the surgery and ensure that they are motivated to determine and provide what is best for the children and family rather than for any personal gain through notoriety or acclaim. It is crucial in these cases to partner not only with medical colleagues but also with the hospital staff and administration. Not all hospitals may be willing to take on such cases for any of a variety of reasons, including religious beliefs, media exposure, and financial concerns. We have found that for particularly complex cases where ethical concerns may be more prominent, presentation at hospital or university ethics boards and hospital forums directed toward and limited to the nurses, physicians, and other health professionals who will be involved with the care of the twins does a great deal to alleviate concern and allows the development of a common purpose when caring for these children. When the health care team understands up front why certain decisions were made, problems and concerns can be addressed and resolved.
The most important part in the decision-making process is to provide the parents with enough information to allow an informed decision. While entire teams of health care professionals may be involved with these children, the physician–patient relationship with 1 surgeon who is leading the care may allow many of these ethical issues to be resolved. We have often found it useful to have 1 of our neonatologists who is a member of the hospital ethics board involved in these discussions as this allows the parents another perspective and helps ensure that a realistic set of expectations are created when all of the options are discussed. A practical reference that may be useful to help direct decision-making is the Great Ormand Street Ethical Guidelines for Conjoined Twin Separation. These guidelines state that when separation is feasible with a reasonable chance of success with 2 live infants, surgery should be offered and carried out. If surgery is not possible, custodial care should be offered and nature allowed to take its course. If 1 twin is dead or has a lethal anomaly and cannot survive independently from its normal twin, and both would die if not separated, separation to save the healthy twin should be attempted.
Conjoined twins may be born either by C-Section or transvaginally. Elective C-Section at 36 to 38 weeks after the lungs are determined to be mature may be preferable, though premature labor is common. At the Red Cross War Memorial Children's Hospital in Cape Town, South Africa, vaginal deliveries are allowed if the combined weight of the infants is predicted to be less than 3.9 kg. Transvaginal deliveries of larger babies may suffer a higher rate of problems with dystocia and stillbirth, while smaller infants including thoracopagus and ischiopagus twins were born transvaginally without injury to the connecting bridge with 1 exception of a ruptured omphalocele.
Postnatal Management and Timing of Surgery
After delivery, the twins need to undergo immediate resuscitation and evaluation for the need for emergent separation. Stable twins are a clear indication for systematic and well-organized evaluation and preparation. In cases requiring emergent surgery, infants should undergo echocardiography and plain skeletal films. Based on these studies and the physical examination much of the expected anatomy may be predicted and specific studies may be obtained as time and circumstances allow. While some have recommended separation in the first 4 to 6 weeks of life, O'Neill notes that emergent separation or separation within the first 4 months of life carries a much higher mortality rate, as high as 70% compared to elective separation. Emergent surgery is needed if there is injury to the connecting bridge of tissue, the status of 1 infant threatens the life of 1 or both twins, or there is a condition such as a bowel obstruction or a cardiac defect that requires surgical intervention (Fig. 60-2). Procedures needed should be planned with the eventual separation surgery in mind. Infants with imperforate anus, for example, will need a diverting ostomy and this should be placed with separation planes and eventual tissue flaps and blood supply given due consideration. Placement of an ostomy through the umbilicus or in the midline, for example, may preserve future flap blood supply.
Pygopagus twins. The left twin has a life-threatening cardiac anomaly and severe brain anomalies noted on prenatal MRI. This knowledge predelivery allowed the operating team to be ready for emergent separation at birth to save the right twin.
If the twins are stable, a more thorough evaluation is necessary. Elective separation is generally performed 4 to 12 months after birth. In some cases, staged operations may be necessary to achieve the desired results. Delay of the separation surgery allows for growth and for more information on the anatomic union to be obtained. Anesthetic management becomes easier with time and the infants are better able to tolerate blood loss and physiologic derangement. Vascular access becomes easier, though, for some types of twin separations, adequate central venous access will still be needed at the time of separation. It is important to consider the psychologic aspects of the twins as well. There is evidence that twins separated after 1 year of age may have a more difficult time developing a sense of their own individual identity as previously they have identified self as the conjoined state. Not surprisingly, there is frequently a more dominant twin based at least partly on size differences. Sometimes this dominance recedes when the twins are separated.
Attention to nutrition and growth is essential in forming the timeline for surgery. Differential growth rates of the infants are not uncommon despite what would seem to be normal caloric intake for both children. Feeding tubes for either or both infants may at some point be necessary. Differences in portal blood flow and cross-over circulation may account for some of the differences in growth between infants. If there is a distinct difference in growth between the two, separation may be needed in order for the smaller infant to gain weight and grow adequately.
The team described earlier should meet to plan the operation. The operating room nurses and technicians should be chosen and included in the planning sessions. A diagram of the operating room and of the location of necessary equipment should be made. This should include provisions for arranging the operating tables and personnel after the infants have been separated but also prior to completion of the operation. Sometimes it is also helpful for all the personnel to actually have a dress rehearsal using toy dolls. Surgical skin preparation, draping, placement of electrocautery pads, and monitoring devices may be challenging and should be considered and rehearsed prior to the actual separation surgery.
Conjoined twins often attract media attention. In addition to the considerations and planning involved in designing the actual operation, the team leader must also keep in mind the publicity and security issues. Discussions with the family should be held early to determine what if any media will be involved. If the media is to be involved, it is probably best for the medical team to take charge up front and control the flow of information. The communications department of the hospital can provide help in arranging news conferences and in directing the access of the press to the family and medical staff involved.
Security is related to the decision of how to handle the press and public. If access is limited, which is usually necessary at least for the safety of the infants and personnel, the security staff will need to devise a plan to secure parking access and hospital entrances. In addition, the security staff will need to plan for controlling access to the operating room. Special identification of the medical staff involved should be required. If the media is involved, plans should be made for a postoperative press conference and for periodic conferences as the infants recover. In addition to questions regarding the medical condition of the conjoined twins, the medical staff and hospital representatives should be prepared to answer questions regarding the financial considerations of the care of the infants. The staff may also need to explain difficult considerations regarding the survival of one or both infants.
Plans may also be made for recording the events related to the conjoined twins and the operation to separate them. When the hospital plans to record the operation, the hospital and medical staff can decide what to release to the media. Furthermore, a closed-circuit monitor can be established in a separate but limited-access room for interested medical personnel to view the operation. This limits the number of people in the operating room to only those needed.