PORT: A Randomized, Cross-Over, Phase 2 Study of Melflufen Peripheral Versus Central Intravenous Administration in Patients With Relapsed/Refractory Multiple Myeloma.

BACKGROUND: Melflufen, a first-in-class alkylating peptide-drug conjugate, rapidly enters tumor cells and metabolizes to melphalan. In previous studies, melflufen was administered via central venous catheter (CVC). However, administration by peripheral venous catheter (PVC) may be preferable. PATIENTS AND METHODS: PORT was a two-period, phase 2 crossover study of CVC versus PVC melflufen administration in patients with relapsed/refractory multiple myeloma. Adults with ≥ 2 prior therapies refractory to/intolerant of an immunomodulatory drug and a proteasome inhibitor were randomized 1:1 to weekly oral dexamethasone plus melflufen (40 mg) via CVC or PVC infusion on day 1 of 28-day cycle 1. In cycle 2, patients continued dexamethasone and crossed over to the other melflufen administration route. In cycle 3, all patients received melflufen until progression; PVC or CVC routes were allowed based upon investigator decision. Pharmacokinetic sampling was performed during and after melflufen infusion. Primary endpoints were melphalan pharmacokinetic parameters (Cmax, AUC(0-t), and AUC(0-∞)) and frequency and severity of PVC-related local reactions. RESULTS: The 90% CIs for adjusted geometric mean ratios for pharmacokinetic parameters following CVC versus PVC administration were within the 0.8-1.25 bioequivalence range (Cmax 0.946 [90% CI: 0.849, 1.053]; AUC(0-t) 0.952 [90% CI: 0.861, 1.053]; AUC(0-∞) 0.955 [90% CI: 0.863, 1.058]). In both arms, adverse events were primarily hematological and similar; no phlebitis or local infusion-related reactions occurred. CONCLUSION: Melflufen PVC and CVC administrations are bioequivalent based on melphalan pharmacokinetic parameters. Melflufen via PVC was well tolerated, with no infusion-related reactions or new safety signals and may represent an alternative route of administration.

Pharmacokinetics and Metabolism of Melflufen, an Alkylating Peptide-Drug Conjugate, in Patients with Relapsed Refractory Multiple Myeloma.

Melphalan flufenamide (melflufen) is a novel lipophilic peptide-drug conjugate recently approved in the European Union and the United Kingdom for the treatment of relapsed refractory multiple myeloma. Melflufen rapidly crosses the cell membrane, and inside tumor cells, melflufen utilizes peptidases and esterases to release entrapped hydrophilic metabolites with alkylating activity. In vitro, in whole blood, melflufen was rapidly distributed into blood cells and quickly converted to its main metabolite melphalan, with maximum cellular concentrations of noncovalently bound melflufen and melphalan after 1 and 6 minutes, respectively. Melphalan outflow from blood cells was slow, with peak concentrations in plasma after 25 minutes. The pharmacokinetics of melflufen was best described by a 2-compartment model. Following a 30-minutes intravenous infusion of 40 mg in 27 patients with relapsed refactory multiple myeloma, mean half-life in the α phase of the curve was 1.24 minutes, half-life in the ß phase of the curve 26.7 minutes, and clearance 13.4 L/min. Desethyl-melflufen exposure was below 20% compared to melflufen. Based on population analysis (298 patients with relapsed refactory multiple myeloma), the melphalan pharmacokinetics were well characterized by a 3-compartment model with melflufen dosing into a peripheral compartment, assuming instantaneous distribution of melflufen into cells and subsequent rapid metabolism to melphalan. Mean clearance and central and deep peripheral volumes of distribution were 22.4 L/h, 2.70 L, and 51.3 L, respectively. Clearance increased and maximum concentration decreased with increasing body weight and estimated glomerular filtration rate. In conclusion, melflufen administration differs from melphalan administration by a more rapid distribution into cells, which, in conjunction with a rapid intracellular metabolism, allows for higher maximum concentrations of alkylating agents, and by a more extensive distribution of melphalan to peripheral tissues.

Chemically modified recombinant human sulfamidase (SOBI003) in mucopolysaccharidosis IIIA patients: Results from an open, non-controlled, multicenter study.

PURPOSE: Mucopolysaccharidosis IIIA (MPS IIIA) is an inherited lysosomal storage disorder caused by mutations in the N-sulfoglucosamine sulfohydrolase gene that result in deficient enzymatic degradation of heparan sulfate (HS), resulting in progressive neurodegeneration in early childhood and premature death. A chemically modified variant of recombinant human sulfamidase, SOBI003, has shown to cross the blood-brain barrier (BBB) in mice and achieve pharmacologically relevant levels in cerebrospinal fluid (CSF). We report on a phase 1/2, open-label, first-in-human (FIH) study (NCT03423186) and its extension study (NCT03811028) to evaluate the long-term safety, tolerability, pharmacokinetics/pharmacodynamics (PK/PD) and clinical efficacy of SOBI003 in patients with MPS IIIA for up to 104 weeks. METHODS: Six patients aged 1-6 years with confirmed MPS IIIA with developmental age ≥ 12 months received weekly intravenous injections of SOBI003 at 3 mg/kg (Cohort 1, n = 3) or 10 mg/kg (Cohort 2, n = 3). During the extension study, the individual dose of SOBI003 could be adjusted up to 20 mg/kg at the discretion of the investigator. RESULTS: SOBI003 was generally well tolerated. Serum concentrations of SOBI003 increased in proportion to dose, and presence in CSF confirmed that SOBI003 crosses the BBB. Anti-drug antibodies (ADA) were detected in serum and CSF in all patients, with subsequent reductions in serum SOBI003 exposure at high ADA titers. SOBI003 exerted a clear PD effect: a mean reduction in HS levels in CSF of 79% was recorded at the last assessment, together with reductions in HS levels in serum and urine. Neurocognitive development age-equivalent scores showed a stabilization of cognition for all patients, whereas no clear overall clinical effect was observed on adaptive behavior, sleep pattern or quality of life. CONCLUSION: SOBI003 was well tolerated when administered as weekly intravenous infusions at doses of up to 20 mg/kg for up to 104 weeks. ADA development was common and likely affected both PK and PD parameters. SOBI003 crossed the BBB and showed pharmacological activity on HS in CSF.

Non randomized study on the potential of nitisinone to inhibit cytochrome P450 2C9, 2D6, 2E1 and the organic anion transporters OAT1 and OAT3 in healthy volunteers.

PURPOSE: Nitisinone inhibits the cytochrome P450 (CYP) subfamilies CYP2C9, CYP2D6, and CYP2E1 and the organic anion transporter (OAT) isoforms OAT1 and OAT3 in vitro. Since the effect of nitisinone on these enzymes and transporters in humans is still unknown, the purpose of this study was to evaluate the effect of nitisinone on these CYP subfamilies and OAT isoforms. METHODS: This was an open-label, nonrandomized, two-arm, phase 1 study (EudraCT: 2016-004297-17) in healthy volunteers. The substrates (tolbutamide, metoprolol, and chlorzoxazone for the respective CYPs and furosemide for the OATs) were administered as single doses, before and after 15 days of once daily dosing of 80 mg nitisinone, to determine the AUC∞ ratios ([substrate+nitisinone]/[substrate]). Nitisinone pharmacokinetics, safety, and tolerability were also assessed, and blood and urine were collected to determine substrate and nitisinone concentrations by LC-MS/MS. RESULTS: Thirty-six subjects were enrolled with 18 subjects included in each arm. The least square mean ratio (90% confidence interval) for AUC∞ was 2.31 (2.11-2.53) for tolbutamide, 0.95 (0.88-1.03) for metoprolol, 0.73 (0.67-0.80) for chlorzoxazone, and 1.72 (1.63-1.81) for furosemide. Clinically relevant nitisinone steady-state concentrations were reached after 12 days: mean Cav,ss of 94.08 µM. All treatments were well tolerated, and no safety concerns were identified. CONCLUSIONS: Nitisinone did not affect CYP2D6 activity, was a weak inducer of CYP2E1, and was a weak inhibitor of OAT1 and OAT3. Nitisinone was a moderate inhibitor of CYP2C9, and treatment may therefore result in increased plasma concentrations of comedications metabolized primarily via this enzyme. CLINICAL TRIAL REGISTRY IDENTIFICATION: EudraCT 2016-004297-17.

The BACE-1 inhibitor CNP520 for prevention trials in Alzheimer's disease.

The beta-site amyloid precursor protein cleaving enzyme-1 (BACE-1) initiates the generation of amyloid-ß (Aß), and the amyloid cascade leading to amyloid plaque deposition, neurodegeneration, and dementia in Alzheimer's disease (AD). Clinical failures of anti-Aß therapies in dementia stages suggest that treatment has to start in the early, asymptomatic disease states. The BACE-1 inhibitor CNP520 has a selectivity, pharmacodynamics, and distribution profile suitable for AD prevention studies. CNP520 reduced brain and cerebrospinal fluid (CSF) Aß in rats and dogs, and Aß plaque deposition in APP-transgenic mice. Animal toxicology studies of CNP520 demonstrated sufficient safety margins, with no signs of hair depigmentation, retina degeneration, liver toxicity, or cardiovascular effects. In healthy adults ≥ 60 years old, treatment with CNP520 was safe and well tolerated and resulted in robust and dose-dependent Aß reduction in the cerebrospinal fluid. Thus, long-term, pivotal studies with CNP520 have been initiated in the Generation Program.

Population pharmacokinetics and safety of AZD1446, a neuronal nicotinic receptor agonist, administered in healthy volunteers.

AZD1446 is a highly selective agonist of central α4ß2 and α2ß2 neuronal nicotinic receptors. The compound has been shown to improve cognition in preclinical studies and thus has potential for treatment of cognitive disorders, including Alzheimer's disease (AD). This report presents the pharmacokinetics of AZD1446 based on a pooled population pharmacokinetic analysis of five studies in Caucasian and Japanese healthy volunteers. The model described the inter-individual and inter-occasion variability as well as identified the impact of covariates such as age and ethnicity on the parameters. Single doses of AZD1446 ranged between 0.5 and 350 mg and the multiple-dose regimens ranged between 10 and 100 mg four times daily. The maximum duration was 4 weeks. AZD1446 exhibited modest variability in CL/F. Compared with Caucasian subjects, Japanese subjects had approximately 25% higher rate of absorption and higher renal clearance as well as volume of distribution, resulting in a similar half-life. Compared with elderly subjects, young subjects had approximately 25% lower rate of absorption. Due to lower creatinine clearance, renal clearance was lower in elderly subjects. AZD1446 was safe and well tolerated, with nausea, headache and dizziness as the most frequently reported adverse events.

Population pharmacokinetic and pharmacodynamic analysis of plasma Aß40 and Aß42 following single oral doses of the BACE1 inhibitor AZD3839 to healthy volunteers.

Modulating deposition of Aß-containing plaques in the brain may be beneficial in treating Alzheimer's disease. ß-site amyloid precursor protein cleaving enzyme 1 (BACE1) inhibitors have been shown to reduce Aß in plasma and CSF in healthy volunteers. In this study safety, pharmacokinetics and pharmacodynamics that is reduction of the plasma biomarkers Aß40 and Aß42 , of the BACE1 inhibitor AZD3839 were evaluated. Single oral ascending doses (1-300 mg) of AZD3839 were administered to 54 young healthy volunteers in a randomized, double-blind, placebo-controlled study. The data was analyzed using non-linear mixed effects modeling. AZD3839 reduced Aß40 and Aß42 in plasma with estimated potencies (EC50 ) of 46 and 59 nM, respectively, and a maximum effect of approximately 55%. This was in excellent agreement with the concentration-response relationships obtained in mouse and guinea pig. AZD3839 exposure displayed non-linear kinetics, described by a three-compartment model with a saturated binding compartment and an increase in bioavailability with dose. AZD3839 was safe, although, a dose-dependent QTcF prolongation was observed (mean 20 milliseconds at 300 mg). In conclusion, AZD3839 reduced plasma Aß40 and Aß42 , demonstrating clinical peripheral proof of mechanism. Pre-clinical models were predictive for the effect of AZD3839 on the human plasma biomarker in a strictly quantitative manner.

The effect of a long term epidural infusion of ropivacaine on CYP2D6 activity.

BACKGROUND: Ropivacaine and one of its metabolites, pipecoloxylidide, inhibit CYP2D6 in. human liver microsomes in vitro with K(i) values of 5 microM (1.4 mg/L) and 13 microM (3.6 mg/L), respectively. We investigated the effect of a 50 h continuous epidural infusion of ropivacaine 2 mg/mL at a rate of 14 mL/h on CYP2D6 activity. METHODS: Nineteen patients (41-85 yr) undergoing hip or knee replacement, all extensive metabolizers with respect to CYP2D6 activity, were included. Medications known to inhibit or be metabolized by CYP2D6, or known to be strong inhibitors/inducers of CYP1A2 or CYP3A4 were not allowed. Patients received 10 mg debrisoquine (a marker for CYP2D6 activity) before surgery and after 40 h epidural infusion. The metabolic ratio (MR) for debrisoquine hydroxylation was calculated as the amount of debrisoquine/amount of 4-OH-debrisoquine excreted in 0-10 h urine. RESULTS: The median (range) of MR before and after ropivacaine were 0.54 (0.1-3.4) and 1.79 (0.3-6.7), respectively. The Hodges Lehman estimate of the ratio MR after/MR before ropivacaine was 2.2 with a 95% confidence interval 1.9-2.7 (P < 0.001). CONCLUSION: A continuous epidural infusion of ropivacaine inhibits CYP2D6 activity in patients who are extensive metabolizers resulting in a twofold increase in the MR for debrisoquine hydroxylation. However, since none of the patients was converted into a functional poor metabolizer (MR >12.6), the effect on the metabolism of other drugs metabolized by CYP2D6 is unlikely to be of major clinical importance.

Effect of NXY-059, a novel neuroprotectant, on the pharmacokinetics of a single dose of digoxin in healthy subjects.

OBJECTIVE: NXY-059 is a novel free-radical trapping neuroprotectant. Digoxin treatment is common in acute ischaemic stroke, the intended patient population for NXY-059. Since both digoxin and NXY-059 are eliminated primarily renally, with a substantial contribution by active renal secretion, and because digoxin has a narrow therapeutic window, this open, randomised, crossover, two-period study investigated whether NXY-059 affects the pharmacokinetics (PK) of digoxin. RESEARCH DESIGN AND METHODS: Twenty-two healthy subjects received 0.5 mg oral digoxin 2 h after the start of 60-h intravenous infusions of NXY-059 and placebo separated by a 14-day washout. Blood and urine were collected for 60 h. Digoxin concentrations were measured by a novel liquid chromatography-mass spectrometry assay. MAIN OUTCOME MEASURES: The ratio of the geometric mean (90% confidence interval) between NXY-059 and placebo for the digoxin area under the concentration-versus-time curve was 0.91 (0.83-0.99) and was within the predefined range for no interaction (0.80-1.25). No safety concerns were raised in the study. No serious adverse events were recorded. The most common adverse event was headache with similar frequencies in the two treatments. CONCLUSIONS: NXY-059 had no clinically significant effect on the PK of digoxin.

Pharmacokinetics and efficacy of ropivacaine for continuous epidural infusion in neonates and infants.

INTRODUCTION: The primary objective of this noncomparative study was to evaluate the pharmacokinetics of ropivacaine during a 48-72-h continuous epidural infusion of ropivacaine in children under 1 year. The secondary objectives were to assess efficacy and safety. METHODS: Neonates and infants (ASA I-III, gestational age > or =37 weeks, > or =2.5 kg, scheduled for major abdominal or thoracic surgery) were included and separated into age groups: 0-30 (neonate), 31-90, 91-180, and 181-365 days. Ethics committee approval and informed parental consent were obtained before inclusion. An epidural catheter was introduced under general anesthesia at the appropriate dermatomal level. An initial bolus dose (0.9-2.0 mg.kg(-1) of ropivacaine 0.2%) was followed by an epidural infusion (0.2 mg.kg(-1).h(-1) for infants <180 days or 0.4 mg.kg(-1).h(-1) for infants >180 days). Plasma samples were collected every 12 h from 24 h, and on termination of the epidural infusion. Postoperative pain was evaluated using both the Objective Pain Scale and a four-graded descriptive scale. RESULTS: Forty-five infants, median age 116 (0-362) days, were included. Forty-three and 19 patients received an infusion for at least 48 and 72 h, respectively. Satisfactory analgesia was provided in the majority, only 20 patients were given supplementary medication during the infusion. In all age groups, plasma concentrations of unbound ropivacaine leveled at 24 h, without any further increase at 48 and 72 h. Because of lower clearance of unbound ropivacaine in neonates (mean 33 ml.min(-1).kg(-1)) than in infants above the age of 30 days (80, 124, and 163 ml.min(-1).kg(-1), respectively, in the age groups 31-90, 91-180, and 180-365 days), unbound ropivacaine concentrations at the end of infusion were higher in neonates [median 0.10 mg.l(-1) (0.04-0.21 mg.l(-1))] than in infants >30 days [median 0.03 mg.l(-1) (0.003-0.10 mg.l(-1))]. CONCLUSION: Epidural infusions (0.2-0.4 mg.kg(-1).h(-1) ropivacaine) provided satisfactory pain relief in neonates and infants under 1 year. As plasma concentrations of unbound ropivacaine were not influenced by the duration of the infusion, ropivacaine can be safely used for postoperative epidural infusion for 48-72 h. Levels of unbound ropivacaine were higher in the neonates than in the infants, but were below threshold concentrations for CNS toxicity in adults (> or =0.35 mg.l(-1)). This should not preclude the use of ropivacaine infusions in neonates but suggests a need for caution during the first weeks of life.

Renal effects of the cyclooxygenase-inhibiting nitric oxide donator AZD3582 compared with rofecoxib and naproxen during normal and low sodium intake.

BACKGROUND: Nonsteroidal anti-inflammatory drugs (NSAIDs) inhibit cyclooxygenase (COX) and can thereby reduce renal function, especially with respect to solute excretion and renal perfusion. AZD3582 [4-(nitrooxy)butyl-(2S)-2-(6-methoxy-2-naphthyl)propanoate] is a COX-inhibiting nitric oxide donator. Donation of nitric oxide by AZD3582 could preserve blood flow and thereby counteract the deleterious effects of COX inhibition in the gastrointestinal tract and possibly in other organ systems, including the kidney. The aim of this single-dose study was to assess the hypothesis that AZD3582 would not adversely affect renal function compared with NSAIDs. METHODS: In a parallel, randomized, double-blind fashion, a total of 60 healthy subjects (age range, 20-44 years) received 2 single doses of 750 mg AZD3582, 1500 mg AZD3582, 50 mg rofecoxib, 500 mg naproxen, or placebo (n = 12 per group). The first dose was given after a 5-day normal-sodium diet (150 mmol/d), and the second was given after a consecutive 3-day low-sodium diet (10 mmol/d). Urinary sodium excretion during normal sodium intake and glomerular filtration rate (GFR) (assessed by iohexol clearance) during sodium depletion were the primary variables measured. RESULTS: Urinary sodium excretion was reduced in all active treatment groups (maximal reduction of approximately 11 mmol/h during normal sodium intake, P < .05 versus placebo for all groups). GFR was also reduced in all active treatment groups. In sodium-depleted subjects, the mean (SD) maximal reduction in GFR during 0 to 6 hours for 750 mg AZD3582, 1500 mg AZD3582, 50 mg rofecoxib, and 500 mg naproxen was 28.1 mL/min (13.5 mL/min), 33.7 mL/min (23.3 mL/min), 25.2 mL/min (29.2 mL/min), and 41.7 mL/min (30.7 mL/min), respectively, with a statistically significant difference between 500 mg naproxen and placebo. Relative changes in sodium excretion and GFR were similar during normal sodium intake and sodium depletion during active treatment. CONCLUSION: The renal effects of 750 mg and 1500 mg AZD3582 were similar to those of 500 mg naproxen and 50 mg rofecoxib. Thus the potential for nitric oxide donation to create a renal-sparing agent was not found for AZD3582.

Ropivacaine in neonates and infants: a population pharmacokinetic evaluation following single caudal block.

BACKGROUND: The aims of this study were to evaluate pharmacokinetics, efficacy and safety of ropivacaine in infants aged 0-12 months following a single caudal injection. METHODS: Term ASA I-III patients, scheduled for surgery, with a body weight of > or = 2500 g received a caudal block with ropivacaine 2 mg x ml(-1), 1.0 ml x kg(-1). Plasma samples were collected at different time intervals up to 30 h, for analysis of total and unbound ropivacaine and alpha-1-acid glycoprotein (AAG). Pharmacokinetic data were characterized by population analysis. Unbound and total concentrations from 35 patients, median (min-max) postnatal age of 66 (4-351) days, were included in the nonlinear mixed effects modeling to provide estimates of pharmacokinetic parameters and the exploration of covariate relationships. Simulations were made to test the predictive performance of the final model and to describe the effect of significant covariates on systemic exposure. RESULTS: The mean (min-max) peak plasma concentration of total ropivacaine was 0.83 (0.05-1.57) mg x l(-1) at 0.5-5.7 h (median: 1.0 h) and the plasma concentration of unbound ropivacaine was 0.042 (0.012-0.081) mg x l(-1) within 0.5-1 h. The observed unbound fraction in plasma was 6% (1%-14%). A one-compartment open model with first-order absorption and elimination, incorporating a linear-binding model of ropivacaine to AAG best described the data. The only significant covariate relationship was that of age on Clu/F according to the following relationship Clu/F = 3.01 x e0.00474 x Age. This predicts a Clu/F of 3.5 l x h(-1) x kg(-1) at 30 days and 10.8 l x h(-1) x kg(-1) at 270 days with corresponding terminal half-lives of 6.7 and 2.2 h. The interindividual variability (coefficient of variation, CV) in Clu/F was 39%. The population estimate (CV) of ka was 1.65 h(-1) (30%), Vu/F was 33.6 (l x kg(-1)) (45%) and Ka was 1.78 l x mg(-1) (14%). Thirty-five infants received supplementary analgesics (mostly paracetamol). The median time to first supplementary analgesic (based on all 37 patients) was 3.9 h. No safety concerns or signs of systemic toxicity were observed. CONCLUSIONS: Following a caudal block with ropivacaine 2 mg x kg(-1) plasma concentrations of unbound ropivacaine were well below threshold levels for toxicity in adults. Apparent volume of distribution is unchanged, apparent unbound clearance increases and the terminal half-life decreases with age in 0-12-month-old neonates and infants. The postoperative pain management provided adequate analgesia and was well tolerated.

Plasma levels of lidocaine, o-toluidine, and prilocaine after application of 8.5 g Oraqix in patients with generalized periodontitis: effect on blood methemoglobin and tolerability.

Oraqix, a novel non-injectable anesthetic gel containing lidocaine and prilocaine and a thermosetting agent has been developed to provide localized anesthesia in periodontal pockets during scaling/root planing (SRP). The aim of this open study was to determine the plasma levels of lidocaine and prilocaine following application of 8.5 g Oraqix (5 cartridges) to 11 patients with generalized periodontitis (> or = 49% of tooth pockets > or = 5 mm and > or = 23% of pockets > or = 6 mm). Oraqix was applied to the pockets during periodontal probing and SRP over a 2.6 3.4 h period. Blood samples were collected up to 10 h after the start of application of Oraqix. Peak plasma levels of lidocaine (0.16-0.55 mg/L) and prilocaine (0.05-0.18 mg/L) occurred 2.0-3.7 h and 2.0-3.3 h, respectively, after the start of application of Oraqix. These levels are well below threshold levels for initial signs of central nervous system (CNS) toxicity. In conclusion, application of 8.5 g Oraqix (212.5 mg of lidocaine base and 212.5 mg of prilocaine base) in periodontal pockets was well tolerated and displayed a wide safety margin with respect to plasma levels normally associated with systemic toxicity.

Continuous interscalene analgesia with ropivacaine 2 mg/ml after major shoulder surgery.

BACKGROUND: In this open, randomized study, the pharmacokinetics, clinical efficacy, and safety of a 48-h continuous interscalene infusion of 2 mg/ml ropivacaine for postoperative pain relief were investigated in patients undergoing open major shoulder surgery. METHODS: An initial interscalene block with 30 ml ropivacaine, 7.5 mg/ml (225 mg), was performed. After completion of interscalene block, all patients (n = 24) received general anesthesia, and 6 h after interscalene block, a 48-h continuous interscalene infusion of 12 or 18 mg/h using 2 mg/ml ropivacaine was started. Total and unbound plasma concentrations of ropivacaine and 2.6-pipecoloxylidide (PPX; a major active metabolite) were determined during and up to 6 h after the interscalene infusion. Postoperative pain at rest was assessed by a visual analog scale. Supplementary analgesics and adverse events were recorded. RESULTS: Plasma concentrations of total and unbound ropivacaine were proportional to the total dose. At the end of the interscalene infusion of 9 ml/h, the mean +/- SD plasma concentrations of total and unbound ropivacaine were 1.40 +/- 0.54 and 0.03 +/- 0.01 mg/l, respectively, and of total and unbound PPX were 0.70 +/- 0.38 and 0.30 +/- 0.20 mg/l, respectively. Plasma concentrations of unbound ropivacaine and unbound PPX, added together, remained well below threshold levels for systemic central nervous system toxicity. There were no significant differences between the groups for postoperative pain (median maximum of about 20 mm on the visual analog scale in both groups), analgesic consumption, or quality of pain relief assessed by the patient. No signs or symptoms of systemic local anesthetic toxicity were observed. CONCLUSION: A 48-h continuous interscalene infusion of 6 or 9 ml/h ropivacaine, 2 mg/ml, started 6 h after an initial interscalene block of 30 ml ropivacaine, 7.5 mg/ml, provided satisfactory postoperative pain relief after major shoulder surgery and was well tolerated. Unbound plasma concentrations of ropivacaine and PPX remained well below threshold levels for systemic central nervous toxicity.