In 28 subjects given either sevoflurane + nitrous oxide or enflurane + nitrous oxide anesthesia, sevoflurane caused fewer cardiodepressant effects than enflurane (3). Nevertheless, in 10 healthy subjects atrial contraction and left ventricular diastolic function, including active relaxation, passive compliance, and elastic recoil were impaired by sevoflurane (1 MAC) (4).
Sevoflurane has a similar effect on regional blood flow to other halogenated anesthetics, although it is perhaps slightly less of a coronary artery vasodilator than isoflur-ane. It reduces myocardial contractility and does not potentiate adrenaline-induced cardiac dysrhythmias (5).
It also reduces baroreflex function, and in that respect is similar to other halogenated anesthetics. Coronary artery disease is not a risk factor for the use of these agents (6).
In contrast to isoflurane and desflurane, sevoflurane tends not to increase the heart rate, and is usually well tolerated for induction of anesthesia in young children. However, profound bradycardia was reported in four unpremedicated children aged 6 months to 2 years during anesthesia induction with sevoflurane 8% and nitrous oxide 66% (7). The episodes were not associated with loss of airway or ventilation. In three of the children there was spontaneous recovery of heart rate when the sevoflurane concentration was reduced; the other child received atropine because of evidence of significantly reduced cardiac output. In a previous study of sevoflurane induction of anesthesia in children with atropine pre-medication there was also a low incidence of this complication (8), which is probably due to excessive sevoflurane concentrations.
The effects of sevoflurane on cardiac conduction have been studied in 60 healthy unpremedicated infants (9). They received sevoflurane either as a continuous concentration of 8% from a primed circuit or in incrementally increasing doses. Nodal rhythm occurred in 12 cases. The mean duration of the nodal rhythm was 62 seconds in the incremental group and 90 seconds in the 8% group. All of the dysrhythmias were self-limiting and there were no ventricular or supraventricular dys-rhythmias. No adverse events occurred as a result of the dysrhythmias. This study highlights the importance of using electrocardiographic monitoring when inducing anesthesia with volatile agents.
• Complete atrioventricular block occurred in a 10-year-old child with a history of hypertension, severe renal dysfunction, incomplete right bundle branch block, and a ventricular septal defect that had been repaired at birth (10). After slow induction with sevoflurane and nitrous oxide 66%, complete atrioventricular block occurred when the inspired sevoflurane concentration was 3% and reverted to sinus rhythm after withdrawal of the sevoflurane. The dysrhythmia recurred at the end of the procedure, possibly caused by lidocaine, which had infiltrated into the abdominal wound, and again at 24 hours in association with congestive cardiac failure following absorption of peritoneal dialysis fluid.
Congenital or acquired forms of the long QT syndrome can result in polymorphous ventricular tachycardia (torsade de pointes). Many drugs, including inhalational anesthetics, alter the QT interval, and sevoflurane prolongs the rate-corrected QT interval (QTc). In a randomized study of whether sevoflurane-associated QTc prolongation was rapidly reversed when propofol was used instead, 32 patients were randomly allocated to one of two groups (11). All received sevoflurane induction and maintenance for the first 15 minutes. In one group, sevoflurane was then withdrawn, and anesthesia was maintained with propofol for another 15 minutes; the other group continued to receive sevoflurane for 30 minutes. Sevoflurane-associated QTc prolongation was fully reversed within 15 minutes when propofol was substituted.
A case of torsade de pointes has been attributed to sevoflurane anesthesia (12).
• A 65-year-old woman, who had had normal preoperative serum electrolytes and a normal QT interval with sinus rhythm, received hydroxyzine and atropine premedication followed by thiopental and vecuronium for anesthetic induction. Endotracheal intubation was difficult and precipitated atrial fibrillation, which was refractory to disopyramide 100 mg. Anesthesia was then maintained with sevoflurane 2% and nitrous oxide 50%. Ten minutes later ventricular tachycardia ensued, refractory to intravenous lidocaine, disopyra-mide, and magnesium. DC cardioversion resulted in a change to a supraventricular tachycardia, which then deteriorated to torsade de pointes. External cardiac massage and further DC cardioversion were initially unsuccessful, but the cardiac rhythm reverted to atrial fibrillation 10 minutes after the sevoflurane was switched off. Two weeks later she had her operation under combined epidural and general anesthesia, with no changes in cardiac rhythm.
In this case the role of excessive sympathetic drive as a result of the difficult intubation and the lack of opioid use during induction must be considered, even if sevoflurane played a role in precipitating the dysrhythmia.
In a randomized study of the respiratory effects of high concentrations of halothane and sevoflurane in 21 healthy boys undergoing inguinal or penile surgery, there was similar respiratory depression with each agent (13). Minute ventilation fell by about 50% as a result of a reduction in tidal volume, despite an increase in respiratory rate.
Despite a fall in mean arterial pressure, with a consequent reduction in cerebral perfusion pressure, sevoflurane should be a suitable agent for neuroanesthesia (14). Even in patients with ischemic cerebrovascular diseases, both the CO2 response and cerebral autoregulation were well maintained during sevoflurane anesthesia (0.88 MAC) (15).
A case of acute dystonia has been reported during induction of anesthesia with sevoflurane (16).
• A 19-year-old man with schizophrenia, who was taking cyamemazine (a phenothiazine) 75 mg/day, and dihydroergotamine 180 mg/day to avoid neuroleptic drug-induced hypotension, had no history of involuntary movements, and neurological examination was normal preoperatively. Anesthesia was induced with midazolam 5 mg oral premedication and an inhalational induction using 4-5 maximum breaths of sevoflurane 8% and nitrous oxide 50% in oxygen. One minute after loss of consciousness, he developed a torticollic posture and stiffness, rapidly extending to the left tra-pezius and scalene muscles. There was severe rotation of the head accompanied by trismus and opisthotonos. An intravenous injection of the muscle relaxant atra-curium 30 mg resolved the muscle spasms. Subsequent anesthesia was uneventful.
Dystonia after inhalational anesthesia is rare and is presumably due to an alteration in the dynamic relation between dopaminergic and other receptors in the brain.
Sevoflurane can cause epileptiform activity on the electroencephalogram, especially during emergence from anesthesia. It has also been associated with epilepti-form discharges in volunteer studies, but clinical convulsions appear to very be rare. Two cases of epileptiform activity during sevoflurane anesthesia have been reported in healthy volunteers (17). They were taking part in a study of the effects of sevoflurane on regional cerebral blood flow and received twice the minimum alveolar concentration (MAC) of sevoflurane (4.4%). The only other drug administered was rocuronium, a muscle relaxant. Sevoflurane was used at up to twice its MAC, to induce burst suppression of the electroencephalogram. One of the subjects had partial motor seizure activity in the form of slight clonic movements in the right and then later in the left leg. There was an associated increase in heart rate (from 65 to 79 beats/minute) and systolic blood pressure (from 85 to 106mmHg) and rhythmic epileptiform discharges on the electroencephalogram. The second subject had epileptiform activity on his electroencephalogram, consisting of partial and secondarily generalized discharges lasting for 2 and 3 minutes respectively. There were no clinical signs of an epileptic seizure. Burst suppression appeared on the electroencephalogram in both subjects before the seizure activity, and the Bispectral Index increased dramatically during the epileptiform discharge to maximum values of 44 and 73 respectively. As expected, regional cerebral blood flow and regional metabolism of the epileptic focus fell inter-ictally and increased ictally. Although the concentrations of sevoflurane used in this study were high compared with usual anesthetic practice, further human studies are warranted, because prolonged epileptiform discharge is known to be harmful.
In another case, epileptiform activity was reported during sevoflurane anesthesia, but not with propofol in the same individual (18).
• A 62-year-old woman with no personal or family history of seizures had general relaxant anesthesia for plastic surgery using a total intravenous anesthetic technique with propofol, remifentanil, and cisatracurium, after benzodiazepine premedication. Routine electroence-phalographic monitoring showed continuous slowing followed by burst suppression (consistent with very deep anesthesia), but no epileptiform activity. At a second procedure, and following identical benzodiaze-pine premedication and induction with propofol, anesthesia was maintained with sevoflurane (plus remi-fentanil for analgesia and cisatracurium for neuromus-cular blockade). During the procedure, sevoflurane was increased from 2% to 8%. After 5 minutes, at an end-tidal concentration of 5.9%, there was epileptiform activity on the electroencephalogram. There were no hemodynamic changes.
Epileptiform activity on the electroencephalogram in association with sevoflurane induction has also been reported in a prospective study of 20 non-premedicated healthy children in whom electroencephalographic monitoring was started before sevoflurane induction (19). At 2
MAC there was epileptiform activity in two boys, with spontaneously resolving myoclonic movements.
Epileptiform activity on the electroencephalogram in association with sevoflurane has also been reported in two children aged 3 and 5 years in a center in which electroencephalographs monitoring is routine (20). In both cases the activity occurred after several minutes of anesthesia, when the sevoflurane concentrations were increased to 7-8%. The epileptiform activity resolved after a reduction in sevoflurane concentrations. No seizure activity was noted.
In a prospective, observational study in 30 children undergoing adenoidectomy anesthesia was induced with midazolam and thiopental (both potent anticonvulsants) and maintained with sevoflurane; no electroencephalo-graphic epileptiform activity was observed (21).
Generalized tonic-clonic seizures have been reported in association with the use of sevoflurane.
• Generalized tonic-clonic seizure-like movements lasting 40 seconds occurred in a healthy 32-year-old man after emergence from sevoflurane-based anesthesia (22).
• A 19-year-old man with a history of metamfetamine abuse 3 weeks earlier, but no personal or family history of seizure activity had anesthesia induced with midazo-lam 1mg, nitrous oxide 50%, and sevoflurane 8% (23). The sevoflurane was subsequently reduced to 2%. After radical orchidectomy the sevoflurane and nitrous oxide were withdrawn and oxygen 100% was given and 2 minutes later rhythmic jerking movements began in the legs and quickly spread to the rest of the body. The movements were accompanied by an arched back and a stiff neck. Arterial oxygen saturation dropped to 50% and ventilation was controlled, again using sevoflurane 8%. The duration of the seizure was about 4 minutes. The sevoflurane was again withdrawn 3 minutes later, and a similar seizure occurred. This time it was controlled with midazolam 1mg and propofol 30 mg. Recovery was marked only by mild disorientation. Postoperative computerized tomography showed a ganglioneuroma in the posterior cortex. The electroencephalogram was normal.
These reports show that clinicians need to be aware of the possibility of generalized seizures, especially in patients who are predisposed to seizures.
Peripheral neuropathy has been reported in two healthy men anesthetized with 1.25 MAC sevoflurane at 2 l/minute fresh gas flow for 8 hours. Their average concentrations of compound A were 45 and 28 ppm. Both had had previous minor injuries in the regions in which the neuropathies were reported. The authors suggested that compound A, or other factors associated with sevoflurane anesthesia, may predispose patients to peripheral neuropathy. Both men were volunteers for earlier published studies comparing the nephrotoxic properties of sevoflurane and desflurane, sponsored by Baxter PPD, New Jersey, the manufacturer of desflurane, a rival inhalational anesthetic agent; these reports need to be regarded with caution.
Prolongation of rapacuronium-induced neuromuscular blockade by sevoflurane has been studied in a randomized, placebo-controlled comparison with suxa-methonium in 40 children (24). Patients received sevo-flurane and nitrous oxide anesthesia followed by rapacuronium 2 mg/kg. The study was stopped after only seven patients had been recruited, because the mean time to return of twitch height to 25% of baseline was 26 minutes. This time represents a stage at which neuromus-cular blockade can be reversed and in this case it was twice as long as predicted from experience in adult patients. The authors suggested that the prolonged neu-romuscular relaxation was due to the interaction of sevo-flurane with rapacuronium, because such prolongation has not been observed using other inhalation agents.
Delirium during emergence from sevoflurane anesthesia has often been documented. Four patients, an adult and three children aged 3-8 years, who were able to recount the experience, have been reported (25). They had full recall of postoperative events, were terrified, agitated, and distressed, and hence presented with acute organic mental state dysfunction which was short-lived. Two were disoriented and had paranoid ideation. They were not in any pain or were not distressed by pain if it was present. The authors hypothesized that misperception of environmental stimuli associated with sevoflurane's particular mode of action may have been the underlying cause of this phenomenon. Anxiolytic premedication and effective analgesia did not necessarily prevent the problem.
The effect of intravenous clonidine 2 micrograms/kg on the incidence and severity of postoperative agitation has been assessed in a double-blind, randomized, placebo-controlled trial in 40 boys who had anesthetic induction with sevoflurane after oral midazolam premedication
(26). There was agitation in 16 of those who received placebo and two of those who received clonidine; the agitation was severe in six of those given placebo and none of those given clonidine.
The effects of intravenous and caudal epidural cloni-dine on the incidence and severity of postoperative agitation have been assessed in a randomized, double-blind study in 80 children, all of whom received sevoflurane as the sole general anesthetic for induction and maintenance
(27). A caudal epidural block was performed before surgery for analgesia with 0.175% bupivacaine 1 ml/kg. The children were assigned randomly to four groups: (I) clonidine 1 microgram/kg added to the caudal bupiva-caine; (II) clonidine 3 micrograms/kg added to the caudal bupivacaine; (III) clonidine 3 micrograms/kg intravenously; and (IV) no clonidine. The incidences of agitation were 22, 0, 5, and 39% in the four groups respectively. Thus, clonidine 3 micrograms/kg effectively prevented agitation after sevoflurane anesthesia independent of the route of administration.
The effect of a single preoperative dose of the opioid oxycodone on emergence behavior has been studied in a randomized trial in 130 children (28). Oxycodone prophylaxis had no effect on post-sevoflurane delirium.
The effect of a single bolus dose of midazolam before the end of sevoflurane anesthesia has been investigated in a double-blind, randomized, placebo-controlled trial in 40 children aged 2-7 years (29). Midazolam significantly reduced the incidence of delirium after anesthesia. However, when it was used for severe agitation midazolam only reduced the severity without abolishing agitation. The authors concluded that midazolam attenuates, but does not abolish, agitation after sevoflurane anesthesia.
Sevoflurane can be used to induce hypotension during neurosurgery. Hypotensive anesthesia has little effect on postoperative liver function (30).
• A 3-day-old boy underwent inguinal herniorrhaphy under sevoflurane anesthesia, and 2 days later developed vomiting, anorexia, and fever (31). His aspartate transaminase, alanine transaminase, and lactate dehy-drogenase activities were increased and peaked 12-16 days after the operation. Viral markers were negative, as was a lymphocyte stimulation test with sevoflurane. Toxic (not allergic) liver damage due to exposure to sevoflurane was considered to be the most probable diagnosis.
In a randomized study of the renal and hepatic effects of prolonged low-flow anesthesia with sevoflurane or iso-flurane in patients undergoing prolonged operations (over 8 hours), using a technique that maximized compound A production, there were no differences in markers of hepatocellular injury at 24 or 72 hours (32).
The effect of minimal-flow (as opposed to low-flow) anesthesia with sevoflurane and isoflurane has been examined in a randomized trial in 76 patients (33). There were no significant differences between the groups in blood chemistry markers of hepatic function, despite high exposure to Compound A in the patients who received sevoflurane.
Plasma activity of alpha-glutathione S-transferase activity (aGT) is a more sensitive and specific marker of hepatocellular injury than transaminase activity and it correlates better with hepatic histology. Anesthesia with halothane leads to transiently raised aGT activity, but propofol and isoflurane do not. In a randomized study of plasma aGT activity during and after low-flow anesthesia with sevoflurane or isoflurane, there were no significant differences in aGT activities between the two groups during or after anesthesia (34).
Thus, the evidence suggests that sevoflurane is as safe as isoflurane in low-flow anesthesia with respect to liver dysfunction.
The effects of sevoflurane, isoflurane, and desflurane on macroscopic renal structure have been studied in 24 patients undergoing nephrectomy (35). All anesthetics were administered using a fresh gas flow of 1 l/minute and a sodium hydroxide absorber and had an average duration of 3 hours. No injury to nephrons was observed by pathologists blinded to which anesthetic agent had been used. Postoperative creatinine concentrations and urine volumes did not differ significantly between the groups.
• Transient renal tubular dysfunction has been reported in a patient with asthma requiring mechanical ventilation who received sevoflurane for 9 days (36). Soda lime was not used, and the cumulative dose was 298 MAC-hours. Serum and urinary inorganic fluoride concentrations reached maximum concentrations of 71 and 2047 mmol/l respectively. Markers of renal tubular injury were also greatly raised (urinary N-acetyl-beta-d-glucosaminidase and beta2-microglobulin). However, urine volume, creatinine clearance, and serum creati-nine and urea concentrations were unaffected.
There has been a meta-analysis of 22 controlled trials in 3436 patients (82% ASA I or II, 16% ASA III, and 2% ASA IV) (37). The trials had compared sevoflurane for anesthesia maintenance with isoflurane, propofol, or enflurane. Serum creatinine and blood urea nitrogen were used to assess preoperative and postoperative renal function. The duration of anesthesia was 0.5-11 hours. Most patients (97%) were exposed to less than 4 MAChours of volatile agent. Falls in the serum creatinine and blood urea nitrogen were significantly smaller with iso-flurane than with sevoflurane. In patients who received concurrent aminoglycosides, sevoflurane was associated with a small increase in serum creatinine. The following factors had no effect on renal function: the type of anesthetic circuit, the choice of carbon dioxide absorber, the inorganic fluoride ion concentration, the duration of anesthesia, the use of nitrous oxide, or how sick patients were. When all patients were considered, the incidences of clinically significant increases in serum creatinine were the same between agents. In patients with baseline creatinine values greater than 132 mmol/l (1.5mg/dl), the incidence of clinically important increases in serum crea-tinine was significantly higher in both treatment groups compared with baseline. This meta-analysis has provided strong evidence that sevoflurane does not contribute to clinically significant renal insufficiency.
Renal impairment often follows cardiac surgery, but in a randomized trial in elective coronary artery surgery in 354 patients, sevoflurane did not produce greater increases in serum creatinine concentrations than isoflur-ane or propofol (38).
Sevoflurane is metabolized to compound A by carbon dioxide absorbers. It is nephrotoxic in rats, but nephro-toxicity in humans has not been proven. The accumulation of compound A is greatest with low fresh gas flows and barium hydroxide absorbers, both of which cause higher temperatures in the absorber. Current anesthetic practice is to use sodium hydroxide for carbon dioxide absorption, because it produces less compound A than barium hydroxide.
There has been controversy over whether compound A causes significant renal damage in humans. The potential for renal damage using sevoflurane was investigated in 42 patients without renal disease scheduled for surgery lasting more than 4 hours (39). The patients were given low-flow sevoflurane or isoflurane (fresh gas flow 1 l/minute/m2) or high-flow sevoflurane (6 l/minute/m2). None of these increased blood urea nitrogen concentrations, creatinine concentrations, or creatinine clearance.
There were no significant differences in beta2-microglo-bulin, a marker of tubular function, or urinary glucose concentrations. However, there was an increase in the 24hour urinary excretion of N-acetyl-beta-glucosaminidase, a marker of proximal tubular necrosis, with both doses of sevoflurane but not with isoflurane. There were no significant differences in the serum and urinary fluoride concentrations between the two sevoflurane groups, despite the higher concentration of compound A (29 versus 3.9 ppm) in the expired gases of those who received low-flow sevoflurane. The maximum 24-hour protein excretion was higher with low-flow sevoflurane compared with the other two groups.
The effect of the nephrotoxic aminoglycoside antibiotic amikacin on renal function during low-flow sevoflurane anesthesia has been studied in a randomized study in 37 men undergoing orthopedic surgery (40). Markers of renal tubular injury (urinary N-acetyl-beta-d-glucosami-nidase and beta2-microglobulin) were not abnormally raised, and urine volume, creatinine clearance, and serum creatinine and urea concentrations were unaffected. The duration of anesthesia and compound A concentrations were similar in the two groups.
In a randomized study of the renal and hepatic effects of prolonged low-flow anesthesia with sevoflurane or isoflurane in patients undergoing prolonged operations (over 8 hours), using a technique that maximized compound A production, there were no significant differences between the groups in serum creatinine or urea concentrations, creatinine clearance, or urinary protein or glucose excretion at 24 or 72 hours (32). Proteinuria and glycosuria were common in both groups. There was no correlation between exposure to compound A and any measure of renal function. There were no differences in markers of hepatocellular injury. There was no evidence of nephrotoxicity of sevoflurane even at high degrees of exposure to compound A for as long as 17 hours.
The effect on renal function of minimal-flow (as opposed to low-flow) anesthesia with sevoflurane and isoflurane has been examined in a randomized trial in 76 patients (33). There were no significant differences between the groups in blood chemistry markers of renal or hepatic function or in urinary markers of tubular injury, despite high exposure to compound A in the patients who received sevoflurane.
These studies have confirmed earlier findings that although there is biochemical evidence of renal damage after sevoflurane anesthesia, there are no clinically significant effects.
Serum and urinary inorganic fluoride concentrations can rise after inhalation of sevoflurane, because of hepatic metabolism (41). The authors concluded that lengthy sevoflurane anesthesia could alter renal function, although there was no other evidence of nephrotoxicity. Although patients with normal renal function are probably not at risk during normal anesthesia with sevoflur-ane, those with pre-existing renal impairment may be at risk.
A randomized, open study in 26 patients with renal dysfunction who received either isoflurane or sevoflurane for operations lasting up to 6 hours showed no significant differences in postoperative creatinine clearances. However, there was a significant increase in the plasma fluoride ion concentration with sevoflurane (42). In 10 adults who were given repeat high-flow sevoflurane anesthesia there was no evidence of renal or hepatic injury and no increases in serum or urine fluoride concentrations that would indicate an increase in sevoflurane metabolism with repeated use (43).
Renal function has been assessed after low fresh gas flow anesthesia (ll/minute or less) with either sevoflurane or isoflurane in a multicenter study of 254 patients (44). The mean duration of anesthesia was 3.0 MAC-hours in both groups. Peak serum fluoride concentrations were significantly higher (40 mmol/l) after sevoflurane compared with isoflurane (3 mmol/l), and 26 patients had peak fluoride concentrations over 50 mmol/l, a concentration that is associated with renal dysfunction after methoxyflurane anesthesia. There were no significant differences in the renal function of the two groups, as measured by serum creatinine, urea, glycosuria, proteinuria, urine pH, or specific gravity. Absence of renal dysfunction, despite high serum fluoride concentrations after sevoflurane anesthesia, was consistent with previous reports. It appears that low fresh gas flow anesthesia with sevoflurane is not associated with clinically significant renal damage.
Aquaporin-2 is a protein involved in regulation of water permeability in the kidneys. The effects of sevoflurane- and propofol-based anesthesia on urine concentrating ability and aquaporin-2 concentrations have been compared in 30 patients undergoing major surgical procedures given sevoflurane + nitrous oxide or propofol + nitrous oxide (45). Sevoflurane caused a transient 25% fall in aqua-porin-2 concentrations 90 minutes after surgery, rather than the usual 40% increase, which occurred in the propo-fol group. By 3 hours after surgery the aquaporin concentrations in the sevoflurane group had increased and were similar to those in the propofol group. There was a 40% fall in urine osmolarity in the sevoflurane group, but recovery occurred by 3 hours postoperatively. This effect is the likely cause of the occasional cases of polyuria reported in association with sevoflurane anesthesia, rather than nephro-toxicity caused by fluoride ion or compound A.
There have been reports of rhabdomyolysis after anesthesia with halothane, enflurane, and isoflurane in patients with muscular dystrophy, in whom suxamethonium was not used. Rhabdomyolysis has also been reported after sevoflurane anesthesia (46).
• An 11-year-old boy with Duchenne's muscular dystrophy underwent strabismus repair. He also had asthma, for which he was taking prednisone 25 mg/day and theophylline. He underwent inhalational induction with sevoflurane 4% and nitrous oxide 64%; tracheal intubation was then performed without the use of a muscle relaxant. Anesthesia was maintained using sevo-flurane 1.5-3.0% and nitrous oxide 64%. He also received hydrocortisone 100 mg and diclofenac 25 mg. The operation lasted 51 minutes and anesthesia was uneventful. He suffered heel pain during the first few hours postoperatively, and 3 hours postoperatively passed 300 ml of dark red urine, containing large amounts of myoglobin. His serum enzymes increased from preoperative values, serum aspartate transaminase from 76 to 458 IU/l, alanine transaminase from 136 to 254 IU/l, and creatine kinase from 4430 to 55 700 IU/l. He was treated with dantrolene 1 mg/kg and recovered over the next day.
The history and finding in this case are strongly diagnostic of rhabdomyolysis. The most likely cause of rhabdomyo-lysis in this patient was thought to be inhalation of sevoflurane.
Rhabdomyolysis triggered by sevoflurane in a child with Duchenne muscular dystrophy has been reported in one case (47).
Was this article helpful?