|MAIN MENU||INTRODUCTION TO LAPAROSCOPIC COLECTOMY||HISTORY OF LAPAROSCOPIC COLECTOMY||
CARDIOVASCULAR CHANGES DURING
LAPAROSCOPIC COLECTOMY FOR
LAPAROSCOPIC COLECTOMY AT A
|REVIEW OF 752 LAPAROSCOPIC COLECTOMIES||
Stephen N. Harris, MD*, Garth H. Ballantyne, MDt, Martha A. Luther, MS*, and Albert C. Perrino Jr., MD*
Departments of *Anesthesiology and tsurgery, Veterans Affairs Healthcare System, West Haven Campus, Yale University School of Medicine, New Haven, Connecticut.
|HOME||CENTER (CALS)||GERD (REFLUX)||Rx OF GERD||NISSEN FUNDO|
|Dr BALLANTYNE||LAPAROSCOPY||HERNIA REPAIR||GALLBLADDER||ALT MEDICINE|
|MAIN MENU||INTRODUCTION TO LAPAROSCOPIC COLECTOMY||HISTORY OF LAPAROSCOPIC COLECTOMY||CARDIOVASCULAR
COLECTOMY AT A|
|REVIEW OF 752 LAPAROSCOPIC COLECTOMIES||LAPAROSCOPIC
We investigated cardiovascular performance in 12 patients (mean age 66 -- 12 yr) with significant coexisting cardiopulmonary disease (hypertension, coronary artery disease, chronic obstructive pulmonary disease) during laparoscopic colectomy under general anesthesia. Hemodynamic monitors included arterial and pulmonary artery catheters in combination with transesophageal echocardiography. Hemodynamic and echocardiographic data were obtained at five epochs: baseline (after induction of anesthesia), insufflation (after pneumoperitoneum, supine position), Trendelenburg 5 (5 min after placement into Trendelenburg's position), Trendelenburg 20 (at 20 min in Trendeleriburg's position), and end (after release of the pneumoperitoneum, supine position). Hemodynamic responses to peritoneal insufflation resulted in significant increases in systemic vascular resistance (SVR) as well as endsystolic area (ESA) and significant decreases in cardiac index (CI) and ejection fraction area (EFa) compared with baseline. Trendelenburg's positioning augmented ventricular preload and performance, resulting in significant increases in pulmonary capillary wedge pressure, Cl, end-diastolic area, and EFa COMpared with insufflation. At the final epoch, end, a hyperdynamic state occurred as evidenced by a significantly decreased ESA and SVR while heart rate, Cl, and EFa increased significantly compared to baseline and insufflation. In an elderly population with significant coexisting cardiopulmonary disease, intraoperative maneuvers required for laparoscopic colectomy resulted in previously undescribed alterations of cardiovascular performance, which persisted after release of the pneumoperitoneum. (Anesth Analg 1996;83:482-7)
Advances in instrumentation and surgical techniques are expanding the application of laparoscopic approaches to more extensive surgeries, including colorectal surgery (1). This approach may permit reduced surgical trauma and postoperative pain, allowing earlier ambulation and hospital discharge (2). However, the growing experience of laparoscopic approaches to gynecological (3) and binary tract pathology (4), have brought about an increased awareness of the anesthetic implications of laparoscopic surgery.
Investigators have demonstrated significant alterations of cardiac performance after peritoneal insufflation during laparoscopic procedures (5). These changes were reported in otherwise healthy populations presenting for elective laparoscopic gynecological or binary tract surgery (6). Cardiovascular collapse during laparoscopic procedures in young patients undergoing gynecological laparoscopy has similarly been documented (7).
The extension of a laparoscopic approach to a more complicated intraabdominal procedure, colon resection, presents additional concerns related to intraoperative anesthetic management. In contrast to the younger patient population historically served by laparoscopic surgery, these patients are often elderly, with significant coexisting cardiopulmonary disease (hypertension, coronary artery disease, chronic obstructive pulmonary disease). In addition to establishment of a pneumoperitoneum, laparoscopic colectomy uses steep Trendelenburg's positioning to facilitate operative exposure. The capacity of an elderly population to compensate for these potential cardiopulmonary stressors has not been reported previously. This study was undertaken to determine how an elderly population with significant cardiopulmonary disease would respond to routine maneuvers encountered during laparoscopy. In addition to arterial and pulmonary artery catheters, this study used two-dimensional transesophageal echocardiography (TEE) to determine the alterations in cardiovascular performance occurring during elective laparoscopic colectomy.
After an investigational review board approval, and after obtaining informed consent, we studied 16 consecutive patients presenting for elective laparoscopic colectomy. The same surgeon (GHB) performed all the surgeries. All patients had routine clinical monitors placed in addition to a 20-gauge radial artery catheter and a flow-directed pulmonary artery catheter. After the placement of invasive monitoring, a 10-mL/kg isotonic crystalloid fluid bolus was infused, followed by a continuous infusion at 4-8 mL - kg-3 - h-1 through a fluid warmer for the duration of the surgery. A standardized anesthetic induction was then administered consisting of etomidate 0.3-0.5 mg/kg, fentanyl 3-5 tkg/kg, and vecuronium 0.1 mg/kg. The anesthetic was maintained with a heated and humidified air/oxygen mixture (fraction of inspired oxygen 0.40). Isoflurane was titrated to end-tidal concentrations -1.2 vol %. Additional fentanyl was infused at a rate of 1-3 jig - kg-' - h-1. Patients received a minute ventilation of 10-12 mL/kg with a respiratory rate of 8-12 breaths/min by an OhmedaO model 7000 ventilator (BOC Healthcare Co., Madison, WI). Ventilator variables were changed only if the desired minute ventilation was not delivered.
After laryngoscopy and tracheal intubation, a 5.0MHz phased array (TEE) transducer probe (HewlettPackard, Andover, MA) was then inserted. A nasogastric tube and an indwelling urinary drainage catheter were also placed. Prior to surgical preparation, all patients had Flowtron DVTO (HNE Healthcare, Manalapan, NJ) prophylactic deep venous thrombosis units placed on both calves and a Bair Huggerg patient heating unit (Augustine Medical Inc., Eden Prarie, MN) was placed on the chest and arms, set on the high (43.3 ± 2.8'C) setting.
Hemodynamic and echocardiographic data were obtained at five epochs: baseline-after anesthetic induction, prior to skin incision, supine position; insufflation-5 min after the creation and maintenance of the pneumoperitoneum at 14-16 mm Hg, supine position; Trendelenburg 5-5 min after placement in steep Trendelenburg's position (20'), pneumoperitoneum, during the intraoperative dissection; Trendelenburg 20-20 min after placement in steep Trendelenburg's position (20'), pneumoperitoneum, continuing the intraoperative dissection; end-after the release of the pneumoperitoneum, during the construction of the extracorporeal anastomosis, supine position.
At each study epoch, a hemodynamic profile was obtained at end-expiration during a brief period of apnea. Thermodilution cardiac outputs were obtained in triplicate. Cardiac outputs within 10% were averaged and the mean used for final calculations. Simultaneously, all vital signs including heart rate (HR), arterial and pulmonary arterial pressures, pulmonary artery capillary wedge pressure (PCWP), and derived hemodynamic indices were recorded. An arterial blood sample was sent for a temperature-corrected analysis. ETco2, duration, and total C02 used for insufflation were similarly noted. Core temperature was obtained from the pulmonary artery catheter. The fraction of inspired oxygen and ETC02 tension were measured by an Ohme@ RGM monitor (model 5250, BOC Healthcare Co., Madison WI). Due to the extremes in patient positioning, arterial and pulmonary arterial pressure transducers were maintained at the phlebostatic axis (midaxillary line, fourth intercostal space) throughout the entire procedure.
After inserting the TEE probe, a standardized examination of the heart and great vessels was performed. The probe was then positioned to monitor a transgastric transverse plane short axis view of the midpapillary level of the left ventricle. At each study epoch, the view of the left ventricle was optimized (circular ventricular cavity/symmetric wall thickness) and taping began simultaneously with capture of the invasive hemodynamic data. Analysis was performed off line via manual planimetry by a skilled observer (ACP) blinded to intraoperative events. End-diastolic (EDA) and end-systolic areas (ESA) were planimetered on two consecutive cardiac cycles. The average of two separate cardiac cycles were used for analysis. Left ventricular EDA and ESA were manually planimetered from the leading edge of the left ventricular endocardial border. EDA was determined by the greatest cross-sectional area obtained at the peak of the electrocardiogram ECG R wave. ESA was determined by the smallest left ventricular area obtained. Anterolateral and posterolateral papillary muscles were not included in the area calculations. Percent ejection fraction area (%EFa) was determined from the formula %EFa - (EDA - ESA/EDA) - 100.
After surgical skin preparation, the abdomen was insufflated with C02 to create and maintain a pneumoperitoneum between 14 and 16 mm Hg. The patient was then placed in the supine position for the initiation of the second intraoperative epoch. Postoperatively, all patients were admitted to the surgical intensive care unit and had creatine phosphokinase isoenzymes and a 12-lead ECG obtained every 8 h for 24 h.
Statistical analysis was determined using analysis of variance and Duncan's multiple range test with P < 0.05 considered statistically significant. All values expressed are mean ± SD unless otherwise noted.
Four patients (4/16) were excluded from data analysis due to poor quality echocardiographic examinations; therefore, there were 12 patients in our final study group. All patients studied were male. Our patient population was elderly (66 -- 12 yr) with significant coexisting cardiopulmonary disease (Table 1). Patients weighed 75 ± 6.5 kg with a body surface area of 1.90 0.15 M2 and received a total of 2710 ± 886 mL (range 1500-4200 mL) of crystalloid. Duration of insufflation was 115 -- 79.7 min (range 36-317 min). No inotropes, vasodilators, blood products, or additional fluid boluses were administered. Most of the operations (8/12) were performed to resect adenocarcinoma of the colon, including two laparoscopic-assisted abdominal perineal resections and two laparoscopic-assisted low anterior resections. Two patients (2/12) had their laparoscopy converted to an open procedure due to difficulty of dissection and/or presence of metastatic disease. These patients were not excluded from data analysis, since all of their intraoperative epochs were completed as per study protocol. All patients were discharged within 14 days. No patient experienced a perioperative myocardial infarction.
Compared with baseline values, hemodynamic responses at insufflation resulted in significant (P < 0.05) increases in mean arterial pressure (MAP), central venous pressure (CVP), mean pulmonary artery pressure (mPA), pulmonary capillary wedge pressure (PCWP), and systemic vascular resistance (SVR) (Table 2). Cardiac index (CI) decreased significantly, while heart rate (HR) and left ventricular stroke work index (LVSWI) remained unchanged. The simultaneous acquisition of echocardiographic data obtained during this epoch revealed significant decreases (P < 0.05) in %EFa and increases in ESA.
After patients were placed in Trendelenburg's position, MAP, CVP, mPA, and PCWP remained significantly higher (P < 0.05) compared with baseline values. Cl was significantly greater (P < 0.05) than the values obtained at insufflation. SVR decreased to values similar to baseline, while HR remained unchanged. Echocardiographic analysis demonstrated no significant changes from the previous epoch.
During the epoch Trendelenburg 20, MAP, CVP, mPA, and PCWP remained significantly higher than baseline values. Cl remained unchanged. Echocardiographic analysis revealed EDA increased to a significantly greater (P < 0.05) value compared with baseline. ESA and %EF. remained unchanged from previous epochs.
At end, HR and Cl increased significantly (P < 0.05) compared with baseline and insufflation values. LVSWI was increased compared with baseline (P < 0.05). MAP decreased significantly (P < 0.05) from the three prior epochs, but remained significantly greater than baseline values. SVR decreased to a value significantly less (P < 0.05) than the previous three epochs. CVP, mPA, and PCWP all returned to values similar to those obtained at baseline. Echocardiographic analysis showed EDA decreasing to a value similar to baseline and ESA decreasing significantly (P < 0.05) below baseline values. %EFa increased to a level that was significantly (P < 0.05) higher than all epochs.
Establishment of the pneumoperitoneum resulted in significant increases of PacO2 and ETcO2 (P < 0-05); however, these values remained within normal limits. Throughout the entire study period there were no significant changes in PaO2- Patient temperature, observed from Trendelenburg 5 to end, decreased significantly (P < 0.05) compared with baseline values (Table 3).
Through the combined use of TEE and invasive hemodynamic monitoring, this investigation provided a detailed examination of the hemodynamic responses which occurred during laparoscopic colectomy in an elderly population with significant cardiopulmonary disease. Our data divides laparoscopic colectomy into three phases, each with distinct hemodynamic findings.
The first phase, peritoneal insufflation, resulted in significant increase in SVR and central filling pressures with a significant decrease in CI. Decreased cardiac performance was confirmed by TEE findings of a significantly decreased %EFa (16% less than baseline) and increased ESA (25% greater than baseline). The unchanged LVSWI despite an increased PCWP suggests that ventricular systolic dysfunction occurred during an acute increase in afterload (8,9).
These findings during peritoneal insulation show clinically important differences from previous reports. Cunningham et al. (10), studying a younger patient population presenting for laparoscopic biliary tract surgery, reported significant increases in MAP and end-systolic wall stress with no decreases in %EFa. Dorsay et al. (11) evaluating a similar population, demonstrated preserved %EFa and CI during pneumoperitoneum. Gannendahl et al. (12) in a study of patients without cardiovascular disease, reported significant increases in MAP, PCWP, and SVR, without significant decreases in cardiac output or fractional area shortening.. In contrast, our patient population evidenced ventricular dysfunction induced by an acute increase in SVR during peritoneal insufflation.
Cardiovascular changes seen at insufflation may have been in part due to inadequate anesthetic depth for the corresponding stimulus, pneumoperitoneum, as well as lack of available preload and intrinsic myocardial disease. Although insufflation occurred within 20 minutes of anesthetic induction, we think that the anesthetic depth should have been sufficient for a surgical incision. Despite attempts to replace fluid deficits arising from preoperative hospitalization and bowel preparation, our data suggests there was insufficient preload available to compensate for the increased stroke work necessary to maintain cardiac output (13). This was evidenced by the baseline EDA's being considerably smaller than previously reported values (10,14). The normal left ventricle responds to this stress with little change in stroke volume, an increase in stroke work, and a small increase in ventricular end-diastolic pressure and volume. When left ventricular contractility is impaired, filling pressures increase markedly but stroke volume decreases, so that LVSWI either remains constant or declines (15). Our data are consistent with this finding. In addition, a majority of our population was receiving 0-adrenergic and/or calcium channel blocker therapy, which also may have contributed to an impaired inotropic response observed during this epoch (16).
The second distinct phase of circulatory responses to laparoscopic colectomy occurred during Trendelenburg's positioning. Increased preload (CVP and EDA) facilitated ventricular function to meet the demands of an increased SVR, previously encountered during the epoch, insufflation. While ESA remained stable, %EFa and Cl improved, placing the heart on a higher portion of the Frank Starling curve. However, LVSWI did not significantly improve despite significant increases in PCWP and EDA. This is consistent with preexisting myocardial disease. Despite increased filling pressures, oxygenation and ventilation remained unchanged and no patients exhibited ECG signs of myocardial ischemia.
The third distinct phase of circulatory responses during laparoscopic colectomy was a hyperdynamic state seen during the final epoch, end. During this epoch, the pneumoperitoneum was released while the surgeon completed the extracorporeal anastomosis through a small infraumbilical incision. This phase of the procedure resulted in significant increases in HR, Cl, and LVSWI with a significant decrease in SVR. Echocardiographic evaluation confirmed the hemodynamic data as evidenced by an EDA returning to baseline values and significantly lower ESA resulting in a significantly higher %EFa.
The resultant hyperdynamic state may be attributed to several events which occurred during this phase. The previously increased cardiopulmonary blood volume observed during Trendelenburg's positioning, was abruptly diminished by supine positioning. Simultaneously, the pneumoperitoneum was released, thus reducing the venous resistance and/or the hormonal stimulation created by the pneumoperitoneum. Cumulative effects of a prolonged C02 load on the circulatory system, hypothermia, and a relative excess of anesthesia may also have been contributing factors. Effects of excess C02 on the circulation may result in arteriolar dilation, myocardial depression, and enhanced sympathoadrenal activity (17,18). Additionally, anesthetic depth could have been greater than needed for the remaining stimulus, a small incision the surgeon was working through to complete the anastomosis. Clearly, there are several factors influencing the hemodynamic responses observed during this epoch. Further investigation is needed to clearly elucidate the etiology of these hemodynamic alterations.
In addition to hemodynamic changes occurring during laparoscopy, management issues arose concerning adequate ventilation and maintenance of patient temperature. After peritoneal insufflation Of C02, the PaCO2 increased significantly from baseline values, but remained within normal limits for the entire procedure. The increase in ETC02 paralleled increases in PaCO2, indicating no significant difficulties with C02 accumulation and ventilation in our population.
Temperature loss is routinely encountered during an open colectomy. However, our data suggests that despite a "closed abdomen" during laparoscopic colectomy, significant decreases in temperature can still occur. the patient population exhibited a progressive decrease in-vore temperature despite the use of fluid warmers and forced hot air heating units. This decrease in core temperature in part may be due to the continued insufflation of a dry, room-temperature gas during general anesthesia. Trocar site number and patency, as well as surgical expertise, contribute to the amount and duration Of C02 exposure. The potential for significant heat loss during laparoscopic surgery has not been widely reported in the literature, perhaps due to the fact that most studies have been in younger, healthier patients undergoing considerably shorter procedures. Two patients, aged 69 and 71 years, each experienced over 135 minutes of insufflation resulting in core temperatures of 32 and 33'C at the termination of the laparoscopy. Both patients required more than two hours of care in the postanesthesia care unit for ventilatory support and additional rewarming prior to extubation. Avoidance of hypothermia is of particular importance in our population, because an increased carbon dioxide load and decreased cardiac output can alter circulation to vital organs and impair wound healing (19,20).
This study relied on two distinct monitoring techniques, TEE and pulmonary artery pressure monitoring, for evaluation of the cardiovascular effects during laparoscopic colectomy. Laparoscopic surgery has unique implications regarding the use and interpretation of data derived from these monitoring techniques. Positive pressure ventilation, patient positioning, and peritoneal insufflation result in increases in intraabdominal pressures. Transmission of intraabdominal pressure to the thoracic cavity results in increased pulmonary arterial pressures, which in turn could be misinterpreted as increased central blood volume or decreased ventricular compliance. Joris et al. (6) reported an increase of intrathoracic pressures by 9 mm Hg during a pneumoperitoneum of 14 mm Hg. A measure of cardiac transmural pressures, relative to pleural pressures, by an esophageal manometer or intrapleural catheter would be needed to more accurately measure central filling pressures (21). Although increases in abdominal pressures were transmitted to the thoracic cavity, off-line analysis of the TEE data, i.e., increases in EDA and ESA, paralleled the hemodynamic findings. Additionally, intraoperative maneuvers required steep Trendelenburg's positioning, which made the zero reference point change for our pressure transducers. This required proper alterations of transducer location to maintain a consistent reference point.
TEE provides additional information not readily obtainable from pulmonary artery catheter monitoring. EDA offers a reliable measure of ventricular preload and is not influenced by increases in intraabdominal pressures (14). TEE-derived data, in contrast to pulmonary artery diastolic pressures, have been shown to correlate with radionuclide estimates of preload and ejection fraction during changes in ventricular compliance (22). However, peritoneal insufflation introduces C02 gas between the stomach and the diaphragm, making transgastic imaging of the heart technically difficult. This required the exclusion of four patients whose TEE examinations were unsatisfactory. Image quality was optimized by adjustments in gain settings and saline irrigation into the upper abdomen while in Trendelenburg's position.
In conclusion, our results demonstrate marked alterations in loading conditions and ventricular performance occurring during laparoscopic colectomy in an elderly population with significant cardiopulmonary disease. The use of invasive monitoring techniques in conjunction with TEE in this study provided additional insight as to the sequence, magnitude, and duration of the cardiovascular responses occurring during laparoscopic surgery. Patients without significant cardiovascular disease may require only noninvasive monitoring and an increase in preload prior to peritoneal insufflation for maintenance of Cl. However, in patients with impaired ventricular function undergoing laparoscopic procedures, pulmonary artery catheterization and/or TEE may prove useful in detecting adverse hemodynamic responses. The use of increasing concentrations of volatile anesthetics in addition to intravenous P-adrenergic blockade or CA 21 channel blocker therapy for treatment of the hypertension associated with insufflation, may further exacerbate diminished ventricular function. Intravenous vasodilators, i.e., sodium nitroprusside and nicardipine, may offer an alternative means of optimizing the loading conditions and hemodynamic performance. In addition to the cardiovascular challenges posed intraoperatively, maintenance of patient temperature is of similar importance. In summary, these findings suggest that anesthetic management may play a major role in ensuring the success of laparoscopic procedures in a patient population with significant coexisting cardiopulmonary disease.
1. Fowler DL, White SA. Laparoscopy-assisted sigmoid resection. Surg Lajparosc Endosc 1991;1:183-8.
2. Senagol'e Aj, Luchtefeld MA, MacKeigan JM, Mazier WP. Open colectomy versus laparoscopic colectomy: are there differences? Am Surg 1993;59:549-54.
3. Lenz Rj, Thomas TA, Wilkins DG. Cardiovascular changes during laparoscopy. Anaesthesia 1976;31:4-12.
4. Westerband A, Van De Water JM, Amzallag M, et al. Cardiovascular changes during laparoscopic cholecystectomy. Surg Gynecol Obstet 1992;175:535-8.
5. Motew M, Ivankovich AD, Bienarz J, et al. Cardiovascular effects and acid-base and blood gas changes during laparoscopy. Am j Obstet Gynecol 1973;115,7:1002-12.
6. Joris JL, Noirot DP, Legrand Mj, et al. Hemodynamic changes during laparoscopic cholecystectomy. Anesth Analg 1993;76: 1067-71.
7. Ivankovich AD, Albrecht RF, Zahed B, Bonnet RF. Cardiovascular collapse during gynecological laparoscopy. III Med J 1974; 145:58-61.
8. Ross J, Braunwald E. The study of left ventricular function in man by increasing resistance to ventricular ejection with angiotensin. Circulation 1964;24:739-49.
9. Ross J. Afterload mismatch and preload reserve: a conceptual framework for the analysis of ventricular function. Prog Cardiovasc Dis 1976;28:4255-63.
10. Cunningham Aj, Turner J, Rosenbaum S, Rafferty T. Transesophageal echocardiographic assessment of haemodynamic function during laparoscopic cholecystectomy. Br j Anaesth 1993;70:621-5.
11. Dorsay DA, Greene FL, Baysinger CL. Hemodynamic changes during laparoscopic cholecystectomy monitored with transesophageal echocardiography. Surg Endosc 1995;9:128-34.
12. Gannendahl P, Odeberg S, Brodin L-A, Sollevi A. Effects of posture and pneumoperitoneum during anaesthesia in the indicies of left ventricular filling. Acta Anaesthesiol Scand 1996; 40:160-6.
13. Lee JD, Tsukasa T, Patritti J, Ross J. Preload reserve and mechanisms of afterload mismatch in normal conscious dog. Am j Physiol 1986;250:H464-73.
14. Cheung AT, Savino JS, Weiss Sj, et al. Echocardiographic and hemodynamic indexes of left ventricular preload in patients with normal and abnormal ventricular function. Anesthesiology 1994;81:376 - 87.
15. Braunwald E. Assessment of cardiac function. In: Braunwald E, ed. A textbook of cardiovascular medicine. Philadelphia: WB Saunders, 1992:419-43.
16. Marshall RL, jebson PJR, Davie IT, Scott DB. Circulatory effects of carbon dioxide insufflation of the peritoneal cavity for laparoscopy. Br j Anaesth 1972;44:680-4.
17. Rasmussen JP, Dauchot Pj, De Palma RG, et al. Cardiac function and hypercarbia. Arch Surg 1978;113:1196-1200.
18. Van Den Bos GC, Drake Aj, Noble MIM. The effect of carbon dioxide upon myocardial contractile performance, blood flow and oxygen consumption. J Physiol (Lend) 1979;287:149-61.
19. Frank SM, Beattie C, Christopherson R, et al. Unintentional hypothermia is associated with postoperative myocardial ischemia. Anesthesiology 1993;78:468-76.
20. Kurz A, Sessler DI, Lenhardt R, et al. Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. N Engl J Med 1996;334:19,1209-15.
21. Diamant M, Benumof JL, Saidman Lj. Hemodynan-dcs of increased intra-abdominal pressure: interaction with hypovolemia and halothane anesthesia. Anesthesiology 1978;18:23-7.
22. Harpole DH, Clements FM, Quill T, et al. Right and left ventricular performance during and after abdominal aortic aneurysm repair. Ann Surg 1989;3:209,356-62.