Abstract
We describe the successful anesthetic management of cesarean section in a patient with Eisenmenger’s syndrome secondary to an atrial septal defect. Although concep-tion is discouraged in women with Eisenmenger’s syndrome, in inevitable circum-stances, careful and meticulous planning of anesthesia can help the parturient survive the ordeal of a cesarean section. The cardiac output must be maintained and sys-temic vascular resistance must not be allowed to fall. This should ensure that there is minimal change in the right to left shunt. In our patient, the scenario of Eisenmenger’s syndrome was complicated by biventricular hypertrophy. We achieved the goals in our patient by using general anesthesia with the Proseal® laryngeal mask airway, a com-bination of ketamine and propofol, and epidural analgesia.
Keywords
analgesia, epidural; anesthesia; Eisenmenger complex; laryngeal masks, Proseal®;
1. Introduction
Eisenmenger’s syndrome was originally described in 1897 and the definition was revised by Wood in 1958.1 It includes pulmonary artery hypertension, reversed or bidirectional shunt associated with an atrial sep-tal defect (ASD), ventricular septal defect and pat-ent ductus arteriosus.2−4 Death can occur at any time during pregnancy as well in the puerperium.5 Here, we describe the anesthetic management of a preg-nant woman at term with Eisenmenger’s syndrome secondary to ASD for lower segment cesarean section (LSCS).
2. Case Report
A 24-year-old, 47-kg pregnant woman (gravida 2, para 1) was admitted to our hospital at 35 weeks of gestation with the major complaint of breathless-ness. On examination, the patient was found to be cyanotic with clubbed fingers and raised jugular ve-nous pressure. Her respiratory rate was 18 breaths/min, blood pressure (BP) was 140/85 mmHg and pulse rate was 75 beats/min. On auscultation, a loud S2 and grade-III/IV systolic murmur was heard over the precordium with right ventricular lift, but the lungs were clear.
Two years previously, during her first pregnancy, ostium secundum type of ASD with severe pulmo-nary hypertension was diagnosed. The patient was advised to have regular follow-ups with the cardi-ologist, but her visits were irregular.
With regard for her health, the patient had ear-lier been advised to undergo medical termination of the second pregnancy by her physicians. Misoprostol was administered orally but the termination was not successful, following which dilatation and curettage was attempted but the procedure was abandoned because of the presence of cervical stenosis. Therefore, the pregnancy continued. The patient had dyspnea of New York Heart Association (NYHA) grade 1 in the first trimester, but it increased to NYHA-III in the third trimester. Two days after ad-mission, she developed premature labor pains and in view of a previous LSCS and cervical stenosis, an LSCS was thus planned. All risks were explained and written informed consent was gained.
Preoperative investigations including a hemo-gram and biochemical tests revealed no abnormal-ities. Electrocardiography (ECG) showed normal sinus rhythm. Chest X-ray revealed cardiomegaly. Echocardiography revealed biventricular enlarge-ment, a 3.2-cm ostium secundum type ASD with right to left shunt, a dilated pulmonary artery with severe pulmonary hypertension (70 mmHg), mod-erate tricuspid regurgitation and right ventricular systolic pressure of 135 mmHg with ejection fraction of 55%. Preoperative atmospheric arterial blood gas analysis revealed pH 7.381, PCO2 27.2 mmHg, PO2 45.5 mmHg, base excess 6.6 mmol/L, HCO3 15.7 mmol/L and SaO2 79.4%.
Preoperatively, prophylaxis for infective endocar-ditis and aspiration were administered. The patient was conveyed to the operating room in a 30º head-up tilt left lateral position with oxygen sup-plementation by facemask. Standard monitoring in-cluding pulse oximetry (SpO2), noninvasive BP monitoring, ECG, and capnography were established. An epidural catheter was inserted in the left lat-eral position via the L3,4 interspace. Under ultra-sonographic guidance and local anesthesia the right internal jugular vein was cannulated with a triple lumen central venous catheter and the central ve-nous pressure (CVP) then measured 12 mmHg. A 20G radial artery cannula and two 16G venous cannulas were inserted and secured.
Anesthesia was induced with ketamine 50 mg, propofol 70 mg and fentanyl 50 μg, and neuromus-cular blockade was achieved with vecuronium 5 mg. Ventilation was assisted and anesthesia was maintained with isoflurane (1%) in oxygen (100%) conveyed by a size 3 Proseal® laryngeal mask air-way (PLMA). Anesthesia was maintained with iso-flurane (1−1.5%) in O2 and air (80:20 mixture), and vecuronium. A 2-kg baby was delivered with an Apgar score of 9 at 1 minute and 10 at 5 minutes. Uterine massage and a slow oxytocin infusion were started.
After the delivery, fentanyl (50 μg) was injected, and 3 mg morphine diluted in 8 mL of 0.1% bupi-vacaine was administered epidurally. As the patient remained hemodynamically stable throughout sur-gery, neuromuscular block was reversed with ne-ostigmine (2.25 mg) and glycopyrrolate (0.4 mg) and the trachea was extubated. The patient was sent to the intensive care unit (ICU) with oxygen (10 L/min) supplementation by facemask. For mainte-nance of fluid, intravenous crystalloids were given at 2 mL/kg/hr. Her vital signs were acceptably maintained with heart rate 78−84 beats/min, BP 140−150/80−90 mmHg, CVP 10−12 mmHg, and SpO2 90−95%. Analgesia was provided by epidural ad-ministration of morphine 3 mg diluted in 8 mL of 0.1% of bupivacaine every 12 hours. For thromboproph-ylaxis, low molecular weight heparin 20 mg was given subcutaneously at 12-hour intervals.
After an uneventful 2-day stay in the ICU, she was transferred to the high dependency unit. Epidural and central venous catheters were removed on the 5th postoperative day. The patient was returned to the ward on the 6th postoperative day where oxygen supplementation (5−6 L/min) was continued. She was weaned from oxygen gradually in the next 2 weeks and discharged home on postoperative day 21.
3. Discussion
We report the successful anesthetic management of a pregnant woman with Eisenmenger’s syndrome, who underwent cesarean section under general anesthesia with isoflurane conveyed by PLMA, in conjunction with epidural analgesia.
The primary anesthetic goal in such a patient is to minimize pulmonary vascular resistance (PVR) sec-ondary to hypercarbia, hypoxia, acidosis, stress, pain and to avoid any hemodynamic changes that might increase the right to left shunt, thereby increasing the severity of hypoxemia.5
Most patients with this syndrome survive until the second or third decade of life.2 Pregnancy should be avoided,2,4,5 or medical termination is preferred for women who become symptomatic in early pregnancy.2,5 The maternal mortality rate associ-ated with cesarean section in the presence of Eisenmenger’s syndrome is 30−50%.5−7 Heart fail-ure, thromboembolism, and bleeding at delivery have been considered as portents of an unfavora-ble prognosis.1,2 Sudden death is probably the re-sult of an abrupt rise in PVR or fall in systemic vascular resistance (SVR).2,6
The thromboembolic phenomenon results from a hypercoagulable state.2 In the intrapartum and early postpartum period, widespread fibrin and platelet thrombi occlude the already narrowed pul-monary arteries resulting in an increased right to left shunt.2 This has been attributed to reduced levels of antithrombin III, protein C and protein S, as well as to increased levels of fibrinogen, pro-thrombin, factor VII, VIII, X, antithrombin II and III.8 Heparin is the anticoagulant of choice.2,3,6 Cyanosis is another factor carrying a thromboembolic risk because of polycythemia and increases the risk of fetal loss (if oxygen saturation is ≤ 85%) or intra-uterine growth restriction. Worsening of maternal hypoxemia is explained by the decrease in SVR, pronounced pulmonary arterial hypertension and consequent increase in right to left flow through intercavitary communication.2
We planned a cesarean section for delivery to avoid rigorous hemodynamic changes during labor and to spare the fetus the risk of hypoxemia second-ary to maternal physical exertion (bearing down efforts during normal delivery) and hypoxia.
A sympathetic response to laryngoscopy and intu-bation is undesirable in these patients as it increases PVR thus worsening the right to left shunt and hence we preferred placement of PLMA over conventional endotracheal intubation. Epidural anesthesia can produce patchy or incomplete sensory block ne-cessitating general anesthesia as reinforcement.5
In our patient, we planned anesthetic induction with propofol mixed with ketamine (ketofol), as propofol reduces both PVR and SVR, and the reduc-tion in SVR is sometimes so abrupt and severe that it may worsen the right to left shunt, while ketamine may increase SVR more than PVR and thus reduce the right to left shunt. Ketamine was not used as a sole induction agent as it causes tachycardia, an in-crease in myocardial oxygen consumption and de-crease in diastolic filling causing low coronary perfusion and decreased cardiac output. Also, it hampers uteroplacental blood flow. Thus, we admin-istered ketofol to counteract propofol-induced side effects such as bradycardia and hypotension and, in addition, the analgesic effect of ketamine reduces the opioid requirement.
We used isoflurane in place of sevoflurane be-cause of its less marked myocardial depressive and vasodilatory effects.9 Isoflurane causes less de-pression of the baroreceptor reflex as compared with other inhalational agents such as halothane and enflurane.10 At high concentration, sevoflu-rane decreases coronary flow as it induces hypo-tension without coronary vasodilation. Sevoflurane induces more pronounced tachycardia than isoflu-rane at low concentrations which would have been undesirable in this case.11
We used CVP and not pulmonary artery pressure as the reference parameter in our patient as she had right ventricular failure (right ventricular pres-sure 133 mmHg) because of severe pulmonary hy-pertension (70 mmHg). The right ventricle was continuously working against high PVR thus increas-ing the risk of failure. The left ventricle was work-ing against a normal SVR and so was not at risk of failure. In such a patient, a right to left shunt may increase when CVP is less than 8 mmHg, which may be further exaggerated by pain. Thus, by providing optimal preload (CVP 8−12 mmHg) and adequate analgesia, we were able to minimize the magni-tude of the right to left shunt.
In a patient with pulmonary hypertension and an ASD, the change from breathing air to breathing 100% oxygen is nearly always associated with an increase in the flow of blood through the pulmonary circuit and decrease of pressure in the pulmonary artery.2 As our patient was in right heart failure and had tricuspid regurgitation, we kept her in the left lateral position with 30º head-up tilt to keep the right heart pressure lower by decreas-ing the preload, and to maintain uteroplacental circulation.
The PLMA was inserted in the semirecumbent left lateral position. This avoided a sudden rise of pressure in the failing right ventricle from an in-creased preload, which would occur when the pa-tient was positioned supinely or head down.
Epidural analgesia was chosen for perioperative pain6 as it reduces PVR and SVR by sympathetic block and reduces catecholamine levels, thus caus-ing less tachycardia, less myocardial oxygen con-sumption, and reduction of the right to left shunt.12 It also induces vasodilatation to accommodate autotransfusion.
Aside from oxygen administration (as in our case), patients may be treated with various spe-cific medications to reduce PVR and improve oxy-genation.13 With long-term oxygen administration, the use of specific medications in patient manage-ment is generally palliative. Sildenafil and L-arginine are effective in decreasing PVR. L-arginine is con-verted into nitric oxide in the body which decreases PVR. Prostacyclin and its analogs such as epoproste-nol have also been used. Bosentan, an endothelin receptor antagonist, has been used and it decreases PVR by 25%.14 Phenylephrine, norepinephrine and metaraminol have been recommended to prevent a decrease in SVR during anesthesia,15 but they also increase PVR and hence they were not used in our patient. Prophylactic use of vasopressors has not been recommended because of associated side ef-fects on these patients with an already compromised cardiovascular system and an unpredictable and probably increased response to potent vasoactive rugs. However, their immediate availability must be ensured in the perioperative period.16
We chose not to give oxytocin in the form of a bolus injection because it causes direct va-sodilatation and reduces SVR with a compensatory increase in heart rate and right to left shunt, and so uterine massage followed by slow oxytocin infu-sion was likely to be effective and safer in this case.5
In conclusion, although pregnancy should be dis-couraged in women with Eisenmenger’s syndrome, careful and meticulous perioperative planning is essential for a successful outcome. The cardiac out-put must be maintained and SVR must not be al-lowed to fall. This should ensure that there is minimal change in the magnitude of the right to left shunt. In our patient, the scenario of Eisenmenger’s syndrome was complicated by biventricular hyper-trophy. We achieved the goals in our patient by application of inhalation general anesthesia con-veyed by PLMA in addition to intravenous anesthe-sia (combination of propofol and ketamine) and epidural analgesia.