Conjoined twins occur once in every 200,000 live births (1) and result from the incomplete fission of the embryonic disk before the third week of pregnancy. There are numerous reports in the literature documenting anesthetic management strategies for the separation of conjoined twins (2–8). There are also reports in the literature detailing anesthetic approaches for surgical procedures not involving separation (9), surgical separation where one or both twins had congenital heart disease (10,11), and surgical procedures on the heart before separation (12). This is the first report of the anesthetic management of a set of inseparable conjoined twins presenting for palliative open-heart surgery on one twin with single ventricle physiology.
A 4.12 kg (combined weight) set of dicephalous parapagus twins were born at 36 wk gestational age; the diagnosis having been made in utero at 25 weeks. Parapagus, sometimes called “diprosopus,” signifies the lateral union of the lower half of the body, extending a variable distance cephalad. This constitutes about 5% of all conjoined twins, and has been associated with congenital heart disease in some cases.
Using the heads as reference points, the right-sided baby was designated Twin A and the baby on the left was called Twin B. Preoperative investigation included echocardiography, cardiac catheterization, and computed tomography scans. The twins had separate hearts and pericardial sacs, separate esophagi and stomachs, and a shared liver, pancreas, small intestine below the duodenum, colon, bladder, and single kidney (Fig. 1). They each had one normal lung and one trachea; however, the medially located lungs of each were hypoplastic. This was particularly true for Twin B, whose right pulmonary artery had a very abnormal branching pattern. Although there were two vertebral columns within one trunk, the twins had only one pelvis, one anus, a single set of male genitalia and one pair of legs. Each had one arm on the lateral side of the body, but shared a fused upper limb medially that protruded superiorly from between the two heads. This fused limb contained two humeri, two ulnae, two radii, and two hands.
Twin A was diagnosed with complex congenital heart disease including atrial inversion, secundum atrial septal defect, bilateral superior vena cavae (SVC) with the large right SVC draining into the coronary sinus, a canal-type ventricular septal defect with a cleft in the mitral valve, a hypoplastic right-sided morphological left ventricle, L-loop transposition of the great vessels (I,L,L), and long segment pulmonary atresia with ductal dependent pulmonary blood flow. The vascular anatomy of Twin A was complex, consisting of a right aortic arch and an interrupted inferior vena cava (IVC) with azygos continuation to the left SVC. The aortic arch vessels from proximal to distal were: left common carotid, a right brachiocephalic artery giving rise to a right common carotid and a right subclavian and then an aberrant left subclavian artery that arose distal to the ductus arteriosus and coursed leftward towards the fused central arm. There were no other vessels originating from this aorta, which, distal to the aberrant left subclavian, became very hypoplastic; ultimately fusing with the descending aorta from Twin B just proximal to the single renal artery (Fig. 2). All other major visceral arteries originated from the descending aorta of Twin B (e.g., celiac, superior mesenteric).
Twin B had a very small patent ductus arteriosus (PDA), a patent foramen ovale, and an interrupted IVC with azygos continuation into the SVC (Fig. 2).
After a complete evaluation, the twins were deemed inseparable. Despite extensive counseling relating to potentially poor long-term cardiopulmonary prognosis, the parent’s expressed a strong desire to palliate the twin with congenital heart disease. The twins presented for palliative open-heart surgery via a median sternotomy in Twin A. Two anesthesia teams, using two sets of anesthesia machines and hemodynamic monitors, provided synchronized care for the infants. Standard monitors were used, including separate electrocardiographs, noninvasive arterial blood pressure monitors, capnography monitors, and pulse oximeters. They had separate esophageal stethoscopes and temperature probes, but shared a pulse oximeter (placed on the foot) and one Foley catheter and bladder temperature probe. Regional cerebral hemoglobin oxygen saturation (rSo2) was monitored on each twin’s head with two INVOS cerebral oximeters (Somanetics Corporation, Troy MI). Anesthesia was induced with fentanyl and vecuronium via a peripheral IV in the left saphenous vein. Anesthesia was promptly induced and both twins were tracheally intubated consecutively with 4.0 endotracheal tubes. Dual 22-gauge radial arterial lines were placed in the common arm because of technical difficulties in obtaining the arterial access in the twin’s separate right and left upper extremities. A double-lumen catheter was inserted in Twin A’s right internal jugular vein, in addition to a double-lumen catheter in the common left femoral vein. Sevoflurane was administered to Twin B, who showed normal inspiratory and expiratory values on the agent analyzer; however, no exhaled anesthetic was noted on the agent analyzer of Twin A after 15 min of administration at a concentration of 1%.
Anesthesia was maintained with a combination of fentanyl (a total of 100 mcg was given to each twin), vecuronium, and sevoflurane. Arterial oxygen saturation was consistently more than 97% in Twin A, despite the administration of 21% oxygen and maintenance of a high normal Paco2. In the pre-bypass period Twin B required endotracheal tube suctioning twice to clear mucous in the airways. Twin B was hyperventilated with 100% oxygen both before and after these two episodes. Both these maneuvers adversely impacted Twin A’s, but not Twin B’s, hemodynamics. Twin A suffered transient decreases in arterial blood pressure from 70/40 to 45/20 mm Hg during these episodes. During the pre-bypass period, the heart rates and the arterial blood pressures varied from one twin to the other by as much as 20%. Twin B’s cerebral oximetry reading was consistently higher than Twin A’s, occasionally by as much as 30%.
After a median sternotomy, Twin A was cannulated and placed on cardiopulmonary bypass (CPB). The surgical procedure involved an extensive central pulmonary arterioplasty using autologous pericardium, ligation of the PDA, and creation of a modified Blalock-Taussig shunt from the innominate artery to the main pulmonary artery (Fig. 3). Twin B was continuously ventilated during this period; Twin A was not. Twin A remained in normal sinus rhythm at normothermia throughout the 155-min bypass period. There was a marked difference in cerebral oximetry readings during the bypass run with consistently higher values in Twin B. These values differed by as much as 35 percentage points. The hematocrit was maintained in the low 20s for both.
Neither twin suffered from myocardial dysfunction or hypotension on termination of bypass. Small-dose dopamine (3 μg · kg−1 · min−1) was started on Twin B. Twin A was ventilated on room air to a high normal Paco2; Twin B was maintained on 100% oxygen owing to continued ventilatory difficulties secondary to ventilation/perfusion (V/Q) mismatching from copious secretions. The highest Po2 measured was 73 mm Hg. The infants were transfused cell saver blood to a hematocrit of 43%. The marked discrepancy in cerebral oximeter readings diminished after separation from CPB and the values remained within 10% of each other thereafter.
After surgery the twins were transferred to the cardiac intensive care unit in critical but stable condition. After numerous unsuccessful attempts at tracheal extubation, the infants were ultimately tracheally extubated 4 wk after surgery. At 4 months of age Twin B contracted pneumonia. The twins developed multisystem organ failure and died.
This unique case illustrates the challenge of anesthetizing inseparable conjoined twins when one has single ventricle congenital heart disease and the other has active pulmonary disease. The management strategies for these two disease states were often in complete conflict. Twin B’s pulmonary pathology and normal cardiac anatomy required a management strategy (high inspired oxygen concentration, increased respiratory rate) that adversely impacted on the physiology of Twin A, who required a single ventricle management strategy. A fundamental principle in the management of a patient with single ventricle anatomy and ductal dependant pulmonary blood flow is to preserve the balance between the systemic and pulmonary circulations (Qs and Qp respectively) (13). In the anesthetic setting, pulmonary vascular resistance (PVR) is most easily manipulated with the goal of obtaining a Qp:Qs ratio of 1. This equilibrium is evidenced by the maintenance of an adequate cardiac output with normal arterial blood pressures and pulse oximetry values of 75%–85%. Should PVR decrease, the resultant increase in pulmonary blood flow will “steal” from the systemic blood flow and cause a decrease in arterial blood pressure, an increase in arterial oxygen saturation, and the development of metabolic acidosis. Decreasing the concentration of inspired oxygen to room air (21%), or even to subambient concentrations (17%–18%) by the addition of nitrogen, and allowing the Paco2 to increase through relative hypoventilation or the addition of carbon dioxide to the fresh gas mixture will increase PVR and restore the balance between the circulations (14).
Evidence of this physiologic management conflict was apparent during the open-heart surgery. Anesthesia was successfully induced in both twins using a lower limb vein—evidence of the shared circulation in the lower half of the body. However, in the upper bodies, despite the presence of duplicate organs above the level of the renal artery, there were physiologic conflicts. When Twin B required hyperventilation on 100% oxygen to treat a period of desaturation secondary to excessive pulmonary secretions, the resultant increased mixed venous oxygen saturation returning to Twin A effectively decreased PVR, causing hypotension. Although there may have been other factors contributing to the hypotension in Twin A during the airway suctioning of Twin B (e.g., autonomic and hormonal stress responses to noxious stimulation, upsetting the delicate balance of Qp:Qs in Twin A), the temporal relationship of the hypotension occurring at the time of hyperventilation on 100% oxygen before suctioning leads us to believe that the principal mechanism involved was that described above.
The absence of sevoflurane in the exhaled gas of Twin A, when it was administered to Twin B at a concentration of 1% for 15 minutes, is an enigma. The twins shared all the abdominal organs and vasculature below the level of the stomach. Although there was limited cross-circulation of the pulmonary vasculature, any inhaled anesthetic absorbed into the bloodstream of Twin B would be distributed to the brain and the rest of the body. This anesthetic should have entered the systemic venous blood of both twins in the lower half of the body, destined to return to the lungs of both babies for exhalation. It is possible that the amount of anesthetic crossing over was simply less than the threshold of detection by the agent analyzer.
Another interesting conjoined twin physiologic response was the marked differences in arterial blood pressure, arterial oxygen saturation, and particularly cerebral oximetry readings between the twins despite the partially shared circulations. The arterial saturation of Twin A was expected to be less than Twin B’s because of the intracardiac shunting secondary to the anatomic defects. The arterial blood pressure of Twin A was also lower as a result of the pulmonary overcirculation that was occurring, despite the administration of 21% oxygen and a high-normal Paco2. The difference in cerebral oximetry readings was most evident during CPB, when rSo2 was consistently less in Twin A. This was likely a reflection of the lower mean arterial blood pressure in Twin A, resulting in reduced cerebral blood flow. It could also have been, in part, attributable to some of the CPB perfusion being shared with Twin B. Careful calibration of arterial flow rates and venous drainage was required while on CPB to optimize the continuous mixed venous oxygen saturation reading in Twin A and the cerebral oximetry readings in both twins. Changing arterial flow rates and venous drainage while on CPB definitely impacted Twin B.
In summary, we report the anesthetic management of a set of inseparable conjoined twins with two hearts, presenting for palliative open-heart surgery on one twin with complex single ventricle congenital heart disease. Because of their shared circulation, it was impossible to isolate the physiological and pharmacological management of the two babies, and a synchronized dual anesthetic management strategy was essential to achieve and maintain a safe anesthetic state.
1. Spitz L, Kiely EM. Conjoined twins. JAMA 2003;289:1307–10.
2. Moscicka A, Uchman J, Moscicki A, et al. Thoracoventropagus: case presentation [in Polish]. Ginekol Pol 2002;73:460–5.
3. Furman EB, Roman DG, Hairabet J, et al. Management of anesthesia for surgical separation of newborn conjoined twins. Anesthesiology 1971;34:95–101.
4. Bloch EC, Karis JH. Cariopagus in neonatal thoracopagus twins: anesthetic management. Anesth Analg 1980;59:304–7.
5. Patel R, Fox K, Dawson J, et al. Cardiovascular anomalies in thoracopagus twins and the importance of preoperative cardiac evaluation. Br Heart J 1977;39:1254–8.
6. Simpson JS. Separation of conjoined thoracopagus twins, with the report of an additional case. Can J Surg 1969;12:89–96.
7. Towey RM, Kisia AK, Jacobacci S, Muoki M. Anaesthesia for the separation of conjoined twins. Anaesthesia 1979;34:187–92.
8. Mulcare RJ, Bhokakul P, Potitung P, Wheeler B. The surgical separation of the thoracopagus conjoined twins of Korat, Thailand. Ann Surg 1970;172:91–7.
9. Seefelder C, Hill DR, Shamberger RC, Holzman RS. Awake caudal anesthesia for inguinal surgery in one conjoined twin. Anesth Analg 2003;96:412–3.
10. Chen TL, Lin CJ, Lai HS, et al. Anaesthetic managements for conjoined twins with complex cardiac anomalies. Can J Anaesth 1996;43:1161–7.
11. Izukawa T, Kidd BS, Moes CA, et al. Assessment of the cardiovascular system in conjoined thoracopagus twins. Am J Dis Child 1978;132:19–24.
12. Cordoba RA, Juaneda E, Alday LE. Surgical ligation of a persistent arterial duct in one of conjoined thoracopagus twins prior to surgical separation. Cardiol Young 1999;9:203–6.
13. Nichols K, Cameron D, Greeley W. Critical heart disease in infants and children. St. Louis: Mosby, 1995.
14. Keidan I, Mishaly D, Berkenstadt H, et al. Combining low inspired oxygen and carbon dioxide during mechanical ventilation for the Norwood procedure. Paediatr Anaesth 2003;1:58–62.