Surgical nuances in the separation of craniopagus twins - Our experience and a follow up of 15 years
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.227289
Source of Support: None, Conflict of Interest: None
Keywords: Craniopagus, conjoined twins, preoperative evaluation, surgery
Craniopagus twins are conjoined twins fused at the cranium. This is one of the rarest of the anomalies prevalent among the conjoined twins and accounts for 2 – 6% of conjoined twins. Conjoined twins, on the other hand, are seen in about 10–20 children per billion births.Craniopagus twins are twins of the same gender, and conjoined twins have a female preponderance in the ratio of 3:1. Most of the craniopagus twins die in utero, or present as stillborn. Those craniopagus twins who survive present with unique challenges to the neurosurgeons even in today's era consisting of markedly advanced diagnostic and therapeutic armamentarium. We present a case of a partially successful elective separation of a pair of partial angular craniopagus twins performed in 2002 and highlight the management nuances.
The craniopagus twins were partial angular joining the frontotemporal region to the frontal regions. The children were named Sou and Af (names changed). We received them when they were 4 months old. The children were born of non-consanguineous parents and delivered by a Caesarian section. Both the children were healthy except that Sou had a telangiectasia in the lumbar region [Figure 1].
Their blood parameters, chest radiograph and abdominal ultrasound were normal. Their computed tomographic (CT) scan revealed that the skull defect was 13 cm in circumference and the widest diameter of the defect was approximately 8 cm. The brain parenchyma of both the twins appeared normal and separate from each other (except in a narrow region). The magnetic resonance imaging (MRI) showed evidence of a good arachnoidal plane between the brain parenchyma of both the children. However, in the center of the defect, there were areas of merging of the two brains. The superior sagittal sinus was complete in the case of Sou, and the anterior half of sinus was incomplete in the case of Af. There was an intercommunication between the superior sagittal sinuses of the two children at one place [Figure 2].
On cerebral angiogram, the injection of the dye through Sou's left internal carotid artery showed a pial supply to the brain parenchyma of Af. During the venous phase of the same injection, the dye was seen draining towards the jugular vein of Sou. The right carotid injection showed a normal flow.
The dye injection through Af's right internal carotid artery showed a normal right carotid angiogram without any evidence of sharing of blood with the other child. The venous phase revealed that the dye was draining towards the veins of Sou. The left carotid injection showed a normal flow. When the dye was simultaneously injected into the left internal carotid artery of Sou and the right internal carotid artery of Af, there was a significant flow of the dye from the circulation of Af to Sou, and also the draining of blood to the ipsilateral jugular veins of both the children. The deep venous system was found to be normal in both the children [Figure 3].
After a meticulous planning that involved a multidisciplinary team, including neurosurgeons, plastic surgeons, interventional radiologists, anesthesiologists, paediatric intensivists, nurses and technicians, the separation of the twins was undertaken. The algorithm of management is given in [Figure 4].
The anesthetist was faced with the arduous task of providing anesthesia to the twins with a combined weight of 5.2 kg. The administration of medications presented a unique problem with the effect of drugs affecting the children simultaneously but with differential effects. For example, the administration of ionotropes caused the heart rate to rise to 40 beats/min in the child into whose intravenous access the medication was introduced but caused an increase of 15 beats/min in the other child. The reverse effect was noted when the medication was administered to the other side.
For induction, a combination of halothane and oxygen were used prior to intubation of the twins in a sequential manner. Sedation was administered to avoid stress to one twin while the other was being intubated.
Intubation was done using atracurium with an immediate injection of 0.5 mg/kg and a maintenance dose of 5 μg/kg/min. This was followed by the administration of midazolam (0.1 mg/kg), thiopentone (0.5 mg/kg) and fentanyl (2 μg/kg).
Stage 1 of surgery: Skin expansion
The children were taken up for 2 stages of skin expansion, one from the front and the other from behind. During this process, the planning and assessing of different surgical positions to determine the optimum surgical position in which the surgical separation would be performed was also done. This was undertaken taking into consideration factors that included the ease of administration of anesthesia, and the achievement of access to maximum circumference of the conjoined portion of the head. On evaluation, it was found that the position with the children one above the other was the worst one, as it produced haemodynamic instability in the form of increased pressure on the child at the lower level. Hence, this position was ruled out. The position where the children lay side-by-side was also ruled out due to the scarcity of space and the difficulties encountered by the anaesthetic team. In the end, the children lay with their torso in opposite directions. This served the purpose of providing space for the two anaesthetic teams who were managing the individual children, as the teams stood in opposite directions to one another. In addition, access was achieved to the maximum circumference of the conjoined portion of the head [Figure 5].
The skin expanders were initially placed anteriorly and expanded with 10 ml saline at a 2-weekly interval. When the anterior scalp was expanded to about half of the desired capacity of expansion, a posterior expander was also placed. On injection of saline into the expanders, a major problem encountered was severe pain. When the skin expanders were rapidly injected with saline, there was thinning of the skin seen in certain areas, which would pose a problem of implant extrusion due to skin ulceration. This phenomenon eventually occurred from the posterior side. The total time taken for adequate expansion of the scalp was 1.5 years. In the present scenario, a microvascular skin flap would have been considered. During this interim period, the children started walking and were able to speak a few words [Figure 6].
Stage 2 of surgery: Final separation planned in 4 steps
Step 1: MRI
An MRI was performed after positioning the twins in the position planned for surgery. This helped the surgeons to get oriented to the structures as they would appear on the table during surgery. At the present time, this procedure could possibly have been obviated by using the neuronavigation system and registering the preoperative images with the actual orientation of the twins' anatomical structures based upon their actual surgical positioning.
Step 2: Interventional balloon occlusion of the communication between the venous sinuses of the twins
A mobile endovascular catheterization table was shifted into the operation room and a digital subtraction angiogram (DSA) was performed under general anesthesia with the twins positioned in the proposed surgical position. Using access via the venous route of the child who had the fully formed sinus (Sou), balloon occlusion of the communication present between the venous sinuses was performed. No hemodymanic instability was noted following occlusion of the communication between the venous sinuses for one hour [Figure 7].
Step 3: Exposure
The plastic surgery team planned and elevated the skin flaps to ensure that an adequate cover would be available following separation of the twins. The dura was also incised in a manner so as to facilitate maximum coverage of the brain. To prevent contraction of the skin flap during the prolonged surgery, the flaps were kept stretched with moist surgical mops [Figure 8]. Bony reconstruction was deferred to a later date to accommodate for the development of brain edema.
Step 4: Surgical separation
According to the pre-operative planning, it had been decided to initiate the separation of the two heads from the posterior region of the conjoined portion of the head of the craniopagus twins. This was because a good arachnoidal plane was present and the superior sagittal sinus was separate in the posterior portions of the conjoined area; this helped in ensuring that the maximum portions of the brain got separated prior to reaching the region where the superior sagittal sinus was less well defined on its anterior aspect. A craniectomy, rather than a craniotomy, was undertaken to prevent an untoward risk of damage to the brain or the dural venous sinuses during the raising of the bone flap
During the procedure, it was observed that the pulsations of the brain tissue of the twin brains were moving the brain tissue of twins in an opposite direction relative to each other thus widening the cleavage between the brains. A few blood vessels were crossing over from Sou's brain to Af's brain, which were coagulated and divided. A posterior separation of 3/4th of the circumference of the conjoined part and more than 5 cm depth was achieved till the point where a portion of the brain was merging with each other [Figure 9]. In order to separate the anterior third of the conjoined portion of the brain, the children were then turned face up.
During the process of the craniectomy of the anterior-third portion of the twin's heads, there was torrential bleeding from the superior sagittal sinus, which was initially controlled with gelfoam and by suturing the adjacent dural flap over the site of bleed. Hemostasis was achieved. However, due to the prolonged bleeding, the brains became edematous and the subsequent dissection was rendered difficult. A second bout of sinus bleeding occurred during the subsequent dissection. In the course of these events, there was a drop in the blood pressure of both the children. During the resuscitative measures, the use of inotropes led to a significant increase in Af's blood pressure. This, in turn, led to further intraoperative hemorrhage, due to the increased venous return from Af's to Sou's veins (a phenomenon referred to as 'shunting'). The further loss of blood caused profound hypotension in the case of Sou and hypertension in the case of Af. Due to this, it was decided to withhold the ionotropes, which resulted in a drop of Af's blood pressure but it stabilized in the case of Sou. During this process, Af expired. Immediately after this phenomonon, Sou's blood pressure normalized and the surgical separation of the brain could be completed. During the separation of final portions of the brain, a portion of Af's brain was left behind in order to ensure that there was no impairment in Sou's cognitive function. The fascia and skin harvested from the scalp incision of the two twins was utilized for reconstruction and closure of Sou's operative wound.
In the immediate post-operative period, Sou was electively ventilated for 24 hours and gradually weaned off from the ventilator. Her neurological evaluation did not reveal any neurological deficits. She was hospitalized for 3 months and a repeat CT scan was done [Figure 10]. During this period, she developed a cerebrospinal fluid leak, which was managed conservatively.
Sou was followed up at 3 monthly intervals and was assessed for her cognitive status by a clinical psychologist, which was found to be normal for her age. She had no focal neurological deficit. She was subjected to a CT scan and an MRI. Her craniectomy skull defect persisted. As a result of this, cranioplasty was planned but was deferred in view of her thin skin cover. At the 9- month follow-up visit, she was noted to have developed a pseudomeningocele through her craniectomy defect [Figure 11] and [Figure 12]. Hence, she was re-admitted and a surgical site re-exploration with excision of the gliotic tissue and repair of the scalp incision was performed [Figure 13].
After a year, further clinical evaluation was performed on an yearly basis on her birthday. A detailed neurological and psychological evaluation was performed to assess for the status of her higher mental and cognitive functions at each visit. The patient had no neurological deficits and is now attending school. She has a normal intelligence and a good scholastic performance Her CT scan shows a large cerebrospinal fluid filled space with a mild pseudomeningocoele persisting but with an intact skin cover [Figure 14].
Craniopagus twins occur in 10 to 20 cases per billion births. These twins are of the same gender and a female preponderence has been noted (3:2)., The first partially successful separation of craniopagus twins was carried out in 1952 by a team of neurological surgeons led by Dr. Oscar Sugar. Even though both the twins survived the operation that lasted for twelve hours, only one regained consciousness postoperatively while other remained in a coma for 34 days and eventually expired., Stone and Goodrich classified craniopagus twins broadly into two types, partial and total. They chose to use the word 'total'when both the twins depended upon a shared dural venous sinus. [Table 1] shows the classification for craniopagus twins proposed by Stone and Goodrich.
Separation of craniopagus twins is usually an elective, multidisciplinary and staged surgery that involves the well co-ordinated efforts from a multi-disciplinary team. However, this elective surgery can become an emergency in case of death of one of the twins as it exposes the live twin to many hazards. The credit of separation of the youngest craniopagus twins till date goes to Nejat et al., who separated craniopagus twins aged 32 weeks of gestation and weighing 1250 grams each, after the death of one of the twins. Over the last 22 years (1995 to 2017), 20 cases of craniopagus separation have been reported with mixed results. Of the 20 reported cases, only 15 were considered separable and the other 5 were deemed inseparable. The success of their separation hinges on many factors including the age and weight of the children at surgery, additional comorbidities or anomalies present, the requirement of a single surgery or of multiple staged surgeries, the angular or vertical variety of the fusion present, and whether or not a partial or total craniopagus is present. Normally, a partial craniopagus is associated with a better chance of separation. However, Harvey et al., suggested that only the presence of a vertical craniopagus was a statistically significant favourable factor in the successful separation of craniopagus twins. [Table 2] gives details of the craniopagus twins in chronological order in whom surgical separation has been attempted.,,,,,,,,,,,,,, We report the first case of partially successful craniopagus twin separation from India, that was performed in the year 2002.
A few salient features have emerged from our experience. A meticulous attention to the patients' positioning during surgery should be given, taking into account that none of the twins is getting compromised as a result of the weight of the other twin or the torsion induced at the site of the craniopagus by an unnatural position. It is also a prerequisite to ensure that the position exposes maximum surface of the conjoined area for surgical dissection. Although the circulation between the craniopagus twins is shared, many of the medications, especially the ionotropes have a differential effect on the twins. In case a common and major intracranial dural venous sinus is shared between the craniopagus twins, the possibility of a venous bypass graft must be considered at an earlier stage, prior to the time when the actual procedure to separate the two children is being undertaken. Enough time must be given to permit the venous bypass graft to start functioning. It is imperative to plan for an adequate scalp cover for both the children even if it entails using skin expanders, planning an myocutanous skin flap or a transposed scalp flap using microvascular techniques. While undertaking the digital subtraction angiogram to assess for the circulation of the twins, the dynamic circulation and the percentage and direction of blood flow through the circulation of the twins must be evaluated in detail. While undertaking this surgery, the surgeon should not go with the predetermination of wanting to save one child over the other. Although the circulation of both the children is often common, the twins may react differentially to the process of separation and it is impossible to determine prior to the actually developing situation during surgery, which twin has a better chance of survival. Therefore, a meticulous planning should be done in order to save both the children. During the surgical procedure, even if one child is lost, every attempt must be made to save the other child, who often has a good chance of a long-term good outcome with an excellent preservation of cognitive and motor functions, as was seen in our patient. The presence of newer tools such as neuronavigation, advances in endovascular technique and hardware, increased knowledge of the anatomical variations and the ability of the modern radiological tools to clearly demonstrate them, and a deeper understanding of the hemodynamic and the anesthetic difficulties anticipated, will no doubt result in a better outcome while managing this complex entity.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8], [Figure 9], [Figure 10], [Figure 11], [Figure 12], [Figure 13], [Figure 14]
[Table 1], [Table 2]