Prospective evaluation of the efficacy, safety, and optimal biomarker enrichment strategy for nangibotide, a TREM-1 inhibitor, in patients with septic shock (ASTONISH): a double-blind, randomised, controlled, phase 2b trial

BACKGROUND

Activation of the triggering receptor expressed on myeloid cells-1 (TREM-1) pathway is associated with septic shock outcomes. Data suggest that modulation of this pathway in patients with activated TREM-1 might improve survival. Soluble TREM-1 (sTREM-1), a potential mechanism-based biomarker, might facilitate enrichment of patient selection in clinical trials of nangibotide, a TREM-1 modulator. In this phase 2b trial, we aimed to confirm the hypothesis that TREM1 inhibition might improve outcomes in patients with septic shock.

METHODS

This double-blind, randomised, placebo-controlled, phase 2b trial assessed the efficacy and safety of two different doses of nangibotide compared with placebo, and aimed to identify the optimum treatment population, in patients across 42 hospitals with medical, surgical, or mixed intensive care units (ICUs) in seven countries. Non-COVID-19 patients (18-85 years) meeting the standard definition of septic shock, with documented or suspected infection (lung, abdominal, or urinary [in patients ≥65 years]), were eligible within 24 h of vasopressor initiation for the treatment of septic shock. Patients were randomly assigned in a 1:1:1 ratio to intravenous nangibotide 0·3 mg/kg per h (low-dose group), nangibotide 1·0 mg/kg per h (high-dose group), or matched placebo, using a computer-generated block randomisation scheme (block size 3). Patients and investigators were masked to treatment allocation. Patients were grouped according to sTREM-1 concentrations at baseline (established from sepsis observational studies and from phase 2a change to data) into high sTREM-1 (≥ 400 pg/mL). The primary outcome was the mean difference in total Sequential Organ Failure Assessment (SOFA) score from baseline to day 5 in the low-dose and high-dose groups compared with placebo, measured in the predefined high sTREM-1 (≥ 400 pg/mL) population and in the overall modified intention-to-treat population. Secondary endpoints included all-cause 28-day mortality, safety, pharmacokinetics, and evaluation of the relationship between TREM-1 activation and treatment response. This study is registered with EudraCT, 2018-004827-36, and Clinicaltrials.gov, NCT04055909.

FINDINGS

Between Nov 14, 2019, and April 11, 2022, of 402 patients screened, 355 were included in the main analysis (116 in the placebo group, 118 in the low-dose group, and 121 in the high-dose group). In the preliminary high sTREM-1 population (total 253 [71%] of 355; placebo 75 [65%] of 116; low-dose 90 [76%] of 118; high-dose 88 [73%] of 121), the mean difference in SOFA score from baseline to day 5 was 0·21 (95% CI -1·45 to 1·87, p=0·80) in the low-dose group and 1·39 (-0·28 to 3·06, p=0·104) in the high-dose group versus placebo. In the overall population, the difference in SOFA score from baseline to day 5 between the placebo group and low-dose group was 0·20 (-1·09 to 1·50; p=0·76),and between the placebo group and the high-dose group was 1·06 (-0·23 to 2·35, p=0·108). In the predefined high sTREM-1 cutoff population, 23 (31%) patients in the placebo group, 35 (39%) in the low-dose group, and 25 (28%) in the high-dose group had died by day 28. In the overall population, 29 (25%) patients in the placebo, 38 (32%) in the low-dose, and 30 (25%) in the high-dose group had died by day 28. The number of treatment-emergent adverse events (111 [96%] patients in the placebo group, 113 [96%] in the low-dose group, and 115 [95%] in the high-dose group) and serious treatment-emergent adverse events (28 [24%], 26 [22%], and 31 [26%]) was similar between all three groups. High-dose nangibotide led to a clinically relevant improvement in SOFA score (of two points or more) from baseline to day 5 over placebo in those with higher cutoff concentrations (≥532 pg/mL) of sTREM-1 at baseline. Low dose nangibotide displayed a similar pattern with lower magnitude of effect across all cutoff values.

INTERPRETATION

This trial did not achieve the primary outcome of improvement in SOFA score at the predefined sTREM-1 value. Future studies are needed to confirm the benefit of nangibotide at higher concentrations of TREM-1 activation.

FUNDING

Inotrem.

Nangibotide in patients with septic shock: a Phase 2a randomized controlled clinical trial

PURPOSE

Nangibotide is a specific TREM-1 inhibitor that tempered deleterious host-pathogens interactions, restored vascular function, and improved survival, in animal septic shock models. This study evaluated the safety and pharmacokinetics of nangibotide and its effects on clinical and pharmacodynamic parameters in septic shock patients.

METHODS

This was a multicenter randomized, double-blind, two-stage study. Patients received either continuous infusion of nangibotide (0.3, 1.0, or 3.0 mg/kg/h) or placebo. Treatment began < 24 h after shock onset and continued for up to 5 days. Safety primary outcomes were adverse events (AEs), whether serious or not, and death. Exploratory endpoints evaluated nangibotide effects on pharmacodynamics, organ function, and mortality, and were analyzed according to baseline sTREM-1 concentrations.

RESULTS

Forty-nine patients were randomized. All treatment emergent AEs (TEAEs) were collected until Day 28. No significant differences were observed in TEAEs between treatment groups. No drug withdrawal linked to TEAE nor appearance of anti-drug antibodies were reported. Nangibotide pharmacokinetics appeared to be dose-proportional and clearance was dose-independent. Nangibotide did not significantly affect pharmacodynamic markers. Decrease in SOFA score LS mean change (± SE) from baseline to Day 5 in pooled nangibotide groups versus placebo was - 0.7 (± 0.85) in the randomized population and - 1.5 (± 1.12) in patients with high baseline plasma sTREM-1 concentrations (non-significant). This pattern was similar to organ support end points.

CONCLUSION

No significant increases in TEAEs were detected in nangibotide-treated patients versus placebo. These results encourage further evaluation of nangibotide and further exploration of plasma sTREM-1 concentrations as a predictive efficacy biomarker.

A first-in-man safety and pharmacokinetics study of nangibotide, a new modulator of innate immune response through TREM-1 receptor inhibition

AIMS

The peptide nangibotide is the first clinical-stage agent targeting the immunoreceptor TREM-1 (triggering receptor expressed on myeloid cells-1) and is being investigated as a novel therapy for acute inflammatory disorders such as septic shock. This first-in-man, randomized, double-blind, ascending dose, placebo-controlled Phase I study evaluated the safety, tolerability and pharmacokinetics of nangibotide.

METHODS

Twenty-seven healthy subjects (aged 18-45 years) were randomized into eight groups. Nangibotide was administered as a single continuous intravenous infusion. The first two groups received a single i.v. dose of 1 and 10 mg, respectively, over 15 min. Subsequent groups were randomized in a product : placebo ratio of 3:1 at doses ranging from 0.03 to 6 mg kg^-1  h^-1 over 7 h 45 min, preceded by a 15-minute loading dose of up to 5 mg kg^-1 .

RESULTS

Nangibotide was safe and well tolerated up to the highest dose tested. There were only few adverse events and they were mild in severity and considered unrelated to treatment. Nangibotide displayed dose-proportional PK properties, with a clearance of 6.6 l kg^-1  h^-1 for a subject of 70 kg and a 3 min effective half-life, which are compatible with extensive enzymatic metabolism in blood. Central and peripheral volumes of distribution were 16.7 l and 15.9 l respectively, indicating limited distribution of the drug mainly in blood and interstitial fluid. No circulating anti-drug antibodies were detectable up to 28 days after administration.

CONCLUSIONS

The novel immunomodulator nangibotide displayed favourable safety and PK profiles at all doses, including expected pharmacologically active doses, and warrants further clinical development.

Cardiovascular safety of lumiracoxib: a meta-analysis of all randomized controlled trials > or =1 week and up to 1 year in duration of patients with osteoarthritis and rheumatoid arthritis

BACKGROUND

The cardiovascular (CV) safety of non-steroidal anti-inflammatory drugs (NSAIDs) and selective cyclooxygenase-2 inhibitors has been the subject of considerable debate.

OBJECTIVE

The objective of this study was to determine the risk of CV events with lumiracoxib by meta-analysis of all completed, randomized controlled trials (RCTs) of > or =1 week and up to 1 year in duration of patients with osteoarthritis and rheumatoid arthritis.

METHODS

The Novartis Lumiracoxib Clinical Trial Database, which includes all clinical studies conducted to date with lumiracoxib, was reviewed. Data were extracted from RCTs of > or =1 week and up to 1 year in duration, the maximum study duration; 34,668 patients were included in standard and cumulative meta-analyses. Twenty-two RCTs of lumiracoxib 100 to 1200 mg daily were identified; 22,781 patients were included in 1-year trials. Mean age of the patients was 61.5 years and 74% were female. More than 50% of the patients in these studies had hypertension at baseline and 6% had diabetes. Parameters analyzed were the Antiplatelet Trialists' Collaboration (APTC) composite CV end point of myocardial infarction (MI), stroke (ischemic and hemorrhagic), and CV death; MI alone; and stroke alone. Twenty-one of the 22 RCTs have been published.

RESULTS

For all 3 parameters, relative risk (RR) was calculated versus non-naproxen NSAIDs, naproxen, and placebo. The results were as follows: for the APTC end point versus non-naproxen NSAIDs: RR 0.83, 95% CI, 0.46-1.51; versus naproxen: RR 1.49, 95% CI, 0.94-2.36; versus placebo: RR 1.08, 95% CI, 0.41-2.86; for MI alone versus non-naproxen NSAIDs: RR 0.80, 95% CI, 0.28-2.25; versus naproxen: RR 1.69, 95% CI, 0.82-3.48; versus placebo: RR 1.27, 95% CI, 0.25-6.56; and for stroke alone versus non-naproxen NSAIDs: RR 0.91, 95% CI, 0.35-2.35; versus naproxen: RR 1.42, 95% CI, 0.70-2.91; versus placebo: RR 0.59, 95% CI, 0.13-2.74. Cumulative meta-analyses of lumiracoxib versus all comparators (placebo, diclofenac, ibuprofen, celecoxib, rofecoxib, and naproxen) did not find any significant differences in APTC, MI alone, or stroke alone.

CONCLUSION

This meta-analysis of 34,668 patients receiving > or =1 week and up to 1 year of treatment found no evidence that lumiracoxib was associated with a significant increase in CV risk compared with naproxen, placebo, or all comparators (placebo, diclofenac, ibuprofen, celecoxib, rofecoxib, and naproxen).

Adherence to combination therapy enhances sustained response in genotype-1-infected patients with chronic hepatitis C

BACKGROUND & AIMS

Patient adherence to prescribed antiviral therapy in human immunodeficiency virus infection enhances response. We evaluated the impact of adherence to combination therapy with interferon or peginterferon plus ribavirin in chronic hepatitis C patients.

METHODS

We assessed the effect of dose reduction on sustained virologic response (SVR) from prior trials with interferon alpha-2b plus ribavirin (n = 1010) or peginterferon alpha-2b 1.5 microg/kg/week plus ribavirin (n = 511). The actual treatment administered was verified from drug dispensing/return records and patient diaries. Two groups were defined: (1) patients who received >or=80% of both their total interferon and ribavirin doses for >or=80% of the expected duration of therapy and (2) patients who received reduced doses (<80% of one or both drugs for >or=80% of the expected duration of therapy). A statistical model provided comparative estimates of the response rates in compliant patients.

RESULTS

Most patients were at least 80% compliant with interferon alpha-2b/ribavirin or peginterferon alpha-2b/ribavirin therapy and had SVR rates of 52% and 63%, respectively, for the 2 regimens. This was most apparent for HCV-1-infected patients. The impacts of adherence on efficacy from subgroup analysis and the statistical modeling approach were similar.

CONCLUSIONS

HCV-1-infected patients who can be maintained on >80% of their interferon or peginterferon alpha-2b and ribavirin dosage for the duration of treatment in the setting of a clinical trial exhibit enhanced sustained response rates. Our results suggest that adherence will enhance the likelihood of achieving an initial virologic response. Adherence beyond 12-24 weeks will be advantageous only for those patients who have achieved such an early virologic response.

Predicting response to initial therapy with interferon plus ribavirin in chronic hepatitis C using serum HCV RNA results during therapy

In patients with chronic hepatitis C, 48 weeks of therapy with interferon (IFN) plus ribavirin results in a sustained virologic response of 40%. Preliminary analysis suggests that measuring HCV RNA at week 24, rather than week 12, might provide the best prediction of treatment response. To assess the clinical utility of serum HCV RNA determinations at different times during therapy as a predictor of a sustained virologic response we evaluated 912 treatment-naïve patients. Patients were randomized to receive IFN-alpha2b, 3 million units (MU) three times weekly (tiw), for 24 or 48 weeks with either ribavirin or placebo, and then followed for 24 weeks. Serum HCV RNA was measured at weeks 4 and 12 in patients treated for 24 weeks; at 4, 12, and 24 weeks during therapy in those treated for 48 weeks, and week 24 post-therapy in all patients. Sustained response was defined as loss of serum HCV RNA at week 24 follow-up. Other patients were considered virologic nonresponders. For patients receiving 48 weeks of combination therapy, detectable serum HCV RNA at week 24 predicted nonresponse (positive predictive value) in 99% of patients compared to 89% at week 12. In patients treated for 24 weeks, testing at week 12 was more predictive of nonresponse than testing at week 4 in the combination-therapy group but not in the monotherapy group. Hence, for combination therapy, testing for serum HCV RNA as a predictor of nonresponse is most accurate at week 24 of therapy; a positive test correctly identified 99% of nonresponders.

Mechanisms of alcoholic liver disease: cytokines

This article represents the proceedings of a workshop at the 2000 ISBRA Meeting in Yokohama, Japan. The chair was Manuela G. Neuman. The presentations were (1) New aspects of hepatic fibrosis, by D. A. Brenner; (2) Cellular immune response in hepatitis C models, by B. Rehermann; (3) The role of interleukin-10 in acute alcoholic hepatitis, by J. Taieb, S. Chollet-Martin, M. Cohard, J. J. Garaud, and T. Poynard; (4) Cytokine-mediated apoptosis in vitro, by M. G. Neuman; (5) Signaling for apoptosis and repair in vitro, by G. G. Katz, R. G. Cameron, N. H. Shear, and M. G. Neuman; (6) Interferons activate the P42/44 mitogen-activated protein kinase and Janus Kinase signal transducers and activation of transcription (JAK-STAT) signaling pathways in hepatocytes: Differential regulation by acute ethanol via a protein kinase C-dependent mechanism, by B. Gao; (7) Genetic polymorphisms of interleukin-1 in association with the development of Japanese alcoholic liver disease, by M. Takamatsu, M. Yamauchi, M. Ohata, S. Saito, S. Maeyama, T. Uchikoshi, and G. Toda; and (8) Increased levels of macrophage migration inhibitory factor in sera from patients with alcoholic liver diseases, by T. Kumagi, S. M. F. Akbar, M. Abe, K. Michitaka, N. Horiike, and M. Onji.

Disposition of recombinant human interleukin-10 in subjects with various degrees of renal function

The pharmacokinetics of intravenously administered recombinant human interleukin-10 (rHuIL-10) were evaluated in 18 subjects with creatinine clearances (Clcr) between 2.7 and 116.7 mL/min/1.73 m2. Serum samples for rHuIL-10 were obtained over a 48-hour period after a single 25 micrograms/kg i.v. bolus infusion. AUC, total body clearance (Clp), and steady-state volume of distribution (Vdss) were derived by compartmental methods. Analysis of serum concentrations showed statistically significant group differences for log-transformed AUC and original scale Clp (p < 0.01). The AUC and effective half-life increased, while the mean Clp of rHuIL-10 decreased as renal function declined. A linear relationship between AUC and Clcr as well as Clp and Clcr demonstrates that the disposition of rHuIL-10 is altered in subjects with renal insufficiency. No serious adverse events were noted.

Moderate inhibitory effect of interleukin-10 on human neutrophil and monocyte chemotaxis in vitro

Neutrophils and monocytes are the major classes of phagocytes that migrate from the blood stream and accumulate in inflamed tissues in response to various chemoattractants. Because IL-10 is a potent anti-inflammatory cytokine, we analyzed its in vitro effect on chemotaxis using an under agarose method. We found that, as compared to prednisolone, IL-10 alone was a modest inhibitor of C5a, fMLP and IL-8-induced neutrophil chemotaxis, and C5a-induced monocyte chemotaxis. However, GM-CSF pretreatment of the cells potentiated this inhibitory effect. Similarly, the IL-10 induced modulation of the beta2 integrin CD11b/CD18 adhesion molecule expression was only observed on GM-CSF-preactivated neutrophils and monocytes. Taken together, these results suggest that the migration and accumulation of phagocytes at infection sites would not be significantly affected by IL-10 given as an immunomodulatory therapy.

Teicoplanin: a well-tolerated and easily administered alternative to vancomycin for gram-positive infections in intensive care patients

A prospective, randomized multicentre study was conducted in order to evaluate the potentially superior tolerability profile of teicoplanin plus netilmicin compared with vancomycin plus netilmicin in patients in ICUs. We considered that these glycopeptides have been shown to have comparable efficacy and that comparative tolerability is of paramount importance, particularly in severely ill patients. A total of 56 patients were enrolled into the study (36 males and 20 females). Twenty-four patients were included in the teicoplanin plus netilmicin group (15 males, 9 females: mean age 56.8 years). The mean simplified acute physiological score (SAPS) was 9.4 (range 4-20). Thirty-two patients were randomized to receive vancomycin plus netilmicin (21 males, 11 females; mean age 56.4 years). The mean SAPS was 9.3 (range 2-16). Septicaemia was the most common infection (14 cases in each group). Most infections were caused by Staphylococcus aureus or coagulase-negative staphylococci. The mean daily doses were: for teicoplanin, 457 mg (6.7 mg/kg); for vancomycin, 1678 mg (24.4 mg/kg); and for netilmicin 263.3 mg (3.9 mg/kg) in the teicoplanin group and 248 mg (3.8 mg/kg) in the vancomycin group. The trough levels of teicoplanin in the serum remained mostly between 7 and 10 mg/l, while more fluctuation was seen in patients receiving vancomycin. The mean trough levels of netilmicin in the serum were 1.2 (SD 0.9) mg/l in the teicoplanin group, compared with 1.7 (SD 1.4) mg/l in the vancomycin group (NS: p > 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

A phase I, double-blind, placebo-controlled study of the tolerance and pharmacokinetic behaviour of RP 59500

The tolerance and pharmacokinetic behaviour of a new injectable streptogramin antibiotic, RP 59500 were evaluated on 26 healthy male volunteers in a double-blind, placebo-controlled, phase I study. The doses used were 1.4, 2.8, 4.6, 7.0, 9.8, 12.6, 16.8, 22.4 and 29.4 mg/kg; each dose was administered as a 1 h infusion to eight subjects, two of them receiving placebo. Blood levels of RP 59500 were measured by both microbiological and HPLC assays. At the end of the infusion, Cmax for RP 59500 ranged from 0.95 +/- 0.22 to 24.20 +/- 8.82 mg/L. The apparent elimination half-life of the compound ranged from 1.27 to 1.53 h. RP 59500 did not significantly affect any of the laboratory or clinical assessments (blood pressure, pulse rate, ECG, peak expiratory flow rate). Reported adverse effects were of mild intensity; the most frequent event was mild pain or burning at the infusion site with doses of 7 mg/kg or higher. These results support further studies of RP 59500 in phase II clinical trials.

Evaluation of teicoplanin for treatment of endocarditis caused by gram-positive cocci in 20 patients

Teicoplanin, a new glycopeptide antibiotic similar to vancomycin, was evaluated for the treatment of bacterial endocarditis in an open multicenter study from May 1985 to August 1987. A total of 20 patients with positive blood culture endocarditis received teicoplanin once daily as a mean intravenous injection of 7.3 mg/kg of body weight (range, 4.8 to 10.6 mg/kg); in 17 patients, teicoplanin was combined with another antibiotic, usually an aminoglycoside. The mean duration of therapy was 28 days (range, 7 to 66 days). The diagnosis of endocarditis was confirmed by echocardiography or anatomical findings in 15 patients and established on the basis of clinical manifestations plus positive blood cultures in 5 patients. The tricuspid valve was involved in 11 of the 20 patients. Isolates from blood were 12 Staphylococcus aureus, 1 Staphylococcus hominis, 1 Micrococcus sedentarius, 1 Enterococcus faecalis, 3 Streptococcus bovis, and 2 nongroupable Streptococcus sp. At the end of therapy, bacterial eradication was achieved in 17 of 20 patients (85%), and a favorable clinical outcome had occurred in 14 of 17 evaluable patients (82%). Of these 14 patients, one relapsed 4 months after the end of treatment. Thus, teicoplanin was effective in 13 of 17 patients (76%). Mean peak levels of teicoplanin in serum were lower, 23.1 +/- 2.9 micrograms/ml, in patients who failed than in those who were cured (45.8 +/- 8.4 micrograms/ml). Side effects occurred in 7 of 20 patients (35%), and required premature discontinuation of teicoplanin in 3 patients. These side effects were fever in three patients, rash in three patients, hearing loss in two patients, and increased serum transaminase levels in two patients. This study demonstrates the efficacy of teicoplanin in the treatment of endocarditis and the need for achieving peak levels in serum close to 40 micrograms/ml. Teicoplanin should now be further evaluated in endocarditis caused by gram-positive cocci means of controlled comparative study with standard therapy.

A multicentre open clinical trial of teicoplanin in infections caused by gram-positive bacteria

A multicentre open trial of teicoplanin in 81 centres in nine European countries included 1431 cases: 531 female, 900 male; mean age 49.4 years, range 1-93 years. These were hospitalized patients most of whom had infections caused by Staphylococcus aureus (816 isolates). Of a total of 1427 Gram-positive pathogens 280 (19.6%) were methicillin resistant. There were 536 skin and soft tissue infections, 263 septicaemias, 135 lower respiratory tract infections, 179 joint and bone infections and 83 endocarditides. More than a third of the infections were severe. Complicating medical factors were present in 69% of cases, including malignant disease in 14% and diabetes mellitus in 11%. Mean teicoplanin dose was 289 mg/day; mean duration of treatment was 14 days. A total of 471 patients received a high dose regimen, 400 mg teicoplanin daily for at least five days. Monotherapy with teicoplanin was used in 1037 cases and combination with other antibiotics in 394. Overall 91.7% of the 1333 evaluable cases were clinical cured or improved. The MIC of teicoplanin was less than or equal to 1 mg/l for 90% of Gram-positive isolates. Adverse events were reported in 189 cases (13.2%). The most common drug-related event was an allergic type skin reaction which occurred in 35 cases (2.4%). Transient hepatic dysfunction was reported in 28 patients (2.0%).

Teicoplanin in the treatment of infections caused by coagulase-negative staphylococci

A total of 88 patients were treated with teicoplanin for infections caused by coagulase-negative staphylococci in open clinical studies in France and the UK. Teicoplanin was administered once daily, with a mean dose of 323 mg, for a mean duration of 16 days. Thirty-nine patients received teicoplanin alone while 49 received combination treatment. Clinical cure or improvement occurred in 79 of 82 evaluable cases (96.3%) and bacteriological elimination in 82 cases (95.1%). Treatment with teicoplanin alone was clinically successful in all cases, including 20 septicaemias, and the elimination rate was 93%. There was no apparent correlation between clinical and bacteriological outcomes and results of in-vitro testing. The variation in MICs between France and the UK was attributed to differences in the methods used.

Pharmacokinetics of teicoplanin in critically ill patients with various degrees of renal impairment

The pharmacokinetics of teicoplanin were studied in 15 adult patients in the acute phase of severe infections caused by gram-positive cocci. All the subjects were given a daily intravenous bolus dose of 6 mg of teicoplanin kg-1 (body weight). The pharmacokinetic study was performed over a 48-h period after injection 4. The subjects were categorized according to their mean creatinine clearances (ml.min-1.kg-1) during the study period: group 1 (n = 3), greater than 1.6; group 2 (n = 6), 0.8 to 1.6; and group 3 (n = 6), 0.15 to 0.8. Mean concentrations of teicoplanin in serum at 1, 24, and 48 h were 33 +/- 8, 9 +/- 3, and 6 +/- 2.5 micrograms.ml-1, respectively. The mean half-lives of the concentration-time curve from 12 to 48 h were 28 +/- 4, 44 +/- 24, and 48 +/- 14 h in groups 1, 2, and 3, respectively (group 3 versus group 1: P less than 0.05). The mean area under the serum concentration-time curve from time zero to 24 h was 344 +/- 92 mg.h.liter-1, and the mean hybrid volume of distribution was 1.09 +/- 0.46 liter.kg-1. These values were similar for the three groups, with a trend for larger areas under the curve in group 3. Creatinine clearance correlated directly with the total body clearance of teicoplanin (r = 0.70) and with the renal clearance of teicoplanin (r = 0.82). However, in critically ill patients, the wide interindividual variations in pharmacokinetic parameters are more relevant than those related to the variations in renal function when creatinine clearance is above 0.30 ml.min-1.kg-1. We concluded that, in such conditions, monitoring of concentrations of teicoplanin in serum is mandatory.

[Staphylococcus aureus endocarditis]

Among 409 cases of bacterial endocarditis (BE) observed from 1972 to 1985, 142 were caused by Staphylococcus aureus. Of these 142 cases, 59 affected native valves of the left heart (left BE), 47 affected native valves of the right heart (right BE) (including 36 drug-addicts), and 36 involved prosthetic valves (BEP) and were associated with mediastinitis in 11 cases. Symptoms were acute in 122 cases and subacute in 2 cases. 91 of the BE on native valves were primary (86%). Cutaneous manifestations were present in 38 cases (27%). Of the 67 patients who died (47%), 28/59 had left BE (47%), 7/47 had right BE (15%) (including 2 drug addicts) and 32/36 had BEP (86%); all differences were statistically significant. Complications consisted of: heart failure in 78 cases (55%), including 40 cases of left BE, 8 cases of right BE and 30 cases of BEP; systemic peripheral embolism in 29 cases (left BE 17, BEP 12) and neurological accidents in 58 cases (left BE 34, right BE 24). Thirty of these accidents occurred before the 4th day (left BE 13, BEP 17). Documented neurological accidents included cerebral haemorrhage (13 cases), cerebral infarction (14 cases) and cerebral abscess (4 cases); 4 of the 12 patients who underwent arteriography were found to have one or several aneurysms. Thirty-nine of these 58 patients died, death being directly due to a neurological cause in 20 cases (left BE 10, right BE 10).(ABSTRACT TRUNCATED AT 250 WORDS)

Alteration of gentamicin and cefazolin kinetics with control of the hypothyroid state in humans

We studied the effect of the control of the hypothyroid state in humans on the pharmacokinetic parameters of cefazolin and gentamicin after a single intravenous injection. These two antibiotics were chosen because of their different patterns of binding to serum albumin (0% for gentamicin and 82% for cefazolin). In hypothyroidism, the behavior of cefazolin only was altered, with a decrease in the total body clearance, possibly due to reduced urinary excretion, and a decrease in the volume of distribution without a significant alteration of cefazolin binding to serum protein.

Kinetics and bactericidal effect of gentamicin and latamoxef (moxalactam) in experimental Escherichia coli endocarditis

The pharmacokinetics of gentamicin and latamoxef (moxalactam) were examined in serum, normal heart valves, sterile cardiac vegetations and vegetations infected with Escherichia coli. Penetration of antibiotics into heart valves and vegetations was rapid; the maximum concentration was achieved in both sites at the end of a 20 min iv infusion. However, both antibiotics penetrated better into vegetations than into normal heart valves. In rabbits with left-sided endocarditis, similar antibiotic levels were found in infected vegetations after one or 22 im injections. After 11 im injections (one every 8 h) of latamoxef (15 mg/kg) the bacterial titre (cfu/g) was significantly reduced, but not nil, despite concentrations about 40 times the MBC for this antibiotic in the vegetations. Afer 22 im injections, vegetations in rabbits receiving latamoxef were sterile and were significantly reduced in those receiving gentamicin (1.5 mg/kg/im), concentrations of which in the vegetations were inferior to the MBC. Our results suggest that the in-vivo antibacterial effect depends on local antibiotic levels, kinetics of killing and duration of contact between antibiotic and bacteria.

[Treatment of severe staphylococcal infections. Failure of rifampicin in combination therapy. 4 cases]

Seven patients with severe methicillin-resistant Staphylococcus aureus infections (5 with bacterial endocarditis, 1 with mediastinitis and 1 with meningitis and septicaemia) were treated with rifampicin combined with vancomycin in 6 cases and with gentamicin in 1 case. The 7 strains initially isolated from haemocultures were resistant to methicillin and sensitive to rifampicin. In 4 of these patients (3 of whom received vancomycin and 1 gentamicin) clinical and bacteriological failure was observed, with selection of rifampicin-resistant mutants. Bactericidal activity was always mediocre in both serum and cerebrospinal fluid, with insufficient vancomycin and gentamicin concentrations in 3 patients. An in vitro study of the combined antibiotics by the chequer-board method suggested antagonism in 3 of these 4 cases. Thus, in spite of its excellent activity and unquestionable effectiveness, rifampicin should be used with caution in severe staphylococcal infections.

Renal disposition of gentamicin, dibekacin, tobramycin, netilmicin, and amikacin in humans

The tubular disposition of five aminoglycosides was studied in humans to establish a possible relationship between tubular reabsorption and the nephrotoxicity that has been described in the literature. Thirty-three healthy male volunteers received a continuous intravenous infusion of isotonic saline with inulin, followed 1 h later by inulin plus gentamicin, dibekacin, tobramycin, netilmicin, or amikacin (1 mg/kg per h) or amikacin (4 mg/kg per h) over a period of 2 h. Brain-stem-evoked response audiometry was performed both before and at the end of each infusion. The latency of wave V remained constant whichever antibiotic was considered. The glomerular filtration rate did not vary significantly during the infusion of each drug. The percent fractional excretion was 79 +/- 6, 81 +/- 22, 85 +/- 5, and 99 +/- 9 for gentamicin, dibekacin, tobramycin, and netilmicin, respectively, and 83 +/- 4 and 124 +/- 13 for amikacin at concentrations of 1 and 4 mg/kg per h, respectively. Net balance and renal clearance were similar for the five aminoglycosides when administered at a rate of 1 mg/kg per h. With gentamicin only, fractional excretion was correlated with the urinary flow rate. We can conclude that (i) gentamicin, generally considered the most nephrotoxic agent, had the highest degree of net reabsorption; (ii) netilmicin exhibited a net zero tubular balance; (iii) amikacin had different patterns of tubular disposition according to the dose, i.e., reabsorption at 1 mg/kg per h and secretion at 4 mg/kg per h, raising the hypothesis of a saturable process of reabsorption; and (iv) these differences in tubular reabsorption could account at least in part for the known different nephrotoxic potentials of these five aminoglycosides in humans.

Vancomycin pharmacokinetics in critically ill patients

Peak and trough vancomycin serum levels were measured in 37 severely ill patients following dosing using the Moellering nomogram. The peak and trough serum concentrations were 61.2 +/- 23 and 22.6 +/- 16.5 mg/l, respectively, and higher than expected. Vancomycin pharmacokinetics obtained from 10 other patients were also studied. In five patients with a creatinine clearance greater than 1 ml min-1 kg-1, a mean plasma elimination half-life of 7.8 +/- 2.8 h was calculated. In the five other patients with markedly reduced renal function (creatinine clearance less than or equal to 1 ml min-1 kg-1), the mean elimination half-life was 18.3 +/- 10.2 h. A correlation was observed between the vancomycin and creatinine clearance. Important inter-patient variations of vancomycin clearance for the same creatinine clearance was also noted. Tubular damage in critically ill patients with severe sepsis may explain our results of the decreased vancomycin elimination.

Pefloxacin in rabbits: protein binding, extravascular diffusion, urinary excretion and bactericidal effect in experimental endocarditis

The serum protein binding, extravascular diffusion and urinary excretion of pefloxacin were studied in rabbits. The effect of furosemide on the urinary excretion of pefloxacin was investigated. In an experimental model of Escherichia coli endocarditis, diffusion into heart valves and infected vegetations and bactericidal effect of pefloxacin were also studied. We observed a serum protein binding of 25%. Extravascular concentrations found were within the range of the minimal inhibitory concentrations for most susceptible strains. Pefloxacin appeared to be reabsorbed by renal tubules (fractional excretion: 61 +/- 21%). Furosemide significantly increased the renal excretion of pefloxacin through a tubular process. We observed a good penetration of pefloxacin into normal heart valves and infected vegetations. Pefloxacin reduced the colony counts in infected vegetations after seven im injections of the drug (given as 15 mg/kg/12 h).