The glucagon-like peptide-1-based therapeutics exenatide and saxagliptin did not cause detrimental effects on the pancreas in mice, rats, dogs and monkeys.

AIMS: Recent reports in the literature have suggested that glucagon-like peptide-1 (GLP-1)-based therapies may lead to increased risk of pancreatic pathology leading to chronic pancreatic injury and pancreatic neoplasia. Extensive non-clinical and clinical safety testing was conducted to support the global development of exenatide twice daily, exenatide once weekly and saxagliptin. Our aim was to integrate these non-clinical data obtained with both mechanisms of GLP-1-based drugs to provide complementary data regarding the potential for drug-induced pancreatic safety signals. METHODS: More than 70 regulated non-clinical toxicology studies in rodents and non-rodents were conducted in accordance with International Conference on Harmonisation and US Food and Drug Administration guidance documents, current industry standards, animal welfare regulations and in compliance with Good Laboratory Practice regulations. Treatment duration was up to 2 years in rodents and up to 12 months in non-rodents using high doses representing large multiples of human exposures (up to 130× for exenatide and 2200× for saxagliptin). Comprehensive pancreas assessments involved more than 2400 pancreata from animals exposed to exenatide and over 1700 pancreata from animals exposed to saxagliptin. RESULTS: Neither exenatide nor saxagliptin treatment resulted in drug-related microscopic changes indicative of acute or chronic adverse effects (including neoplasia) in the endocrine or exocrine pancreas, at doses far exceeding the maximum human systemic exposures. CONCLUSIONS: These data substantially add to the weight of evidence supporting the lack of non-clinical drug-induced pancreatic safety signals in animals exposed to GLP-1-based therapies.

Aspects of pathogenesis of serious group A streptococcal infections in Sweden, 1988-1989.

Serotypes of serious, sometimes fatal, streptococcal infections in Sweden during 1988-1989 were analyzed. The T1M1 type totally dominated, representing almost 70% of all group A streptococci from serious and uncomplicated infections at the peak of the outbreak. Immunoblots of isolates from various patient groups showed that all isolates produced high amounts of erythrogenic toxin (ET) B and high amounts of ET-C, whereas ET-A was released only in small amounts and from few isolates. ELISAs showed high antibody levels to these toxins and to the M1 antigen in patients with uncomplicated infections. Low antibody levels against M1 were seen in patients with bacteremia and in fatal cases; the latter also had low antibody levels against ET-B. It seems likely that a combination of production of large amounts of toxin and low antibody titers to it and to the M antigen of the infecting isolate are determining factors for the outcome of the infection. No signs of primary immune deficiency were noted.

Synergy and cumulated killing effect of the penems FCE 22101 and FCE 25199 in combination with gentamicin against bacteria isolated from septicaemia.

Blood isolates of Enterococcus faecalis, Streptococcus sanguis, Staphylococcus aureus, E. coli and Klebsiella oxytoca were tested for their synergistic and cumulated killing effect (CKE) with the new penems FCE 22101 or FCE 25199 in combination with gentamicin. The tissue cage model in rabbits was used to study the CKE in vivo after antibiotic treatment of the bacteria in vitro. Synergy was observed within two to seven h with all isolates in early logarithmic phase, except with S. aureus, which was rapidly killed by the penems alone. After one h treatment with the antibiotic combinations in vitro, a CKE was demonstrated for up to six h both in vitro and in vivo. The magnitude of the CKE differed between strains and in vitro vs. in vivo.

Postantibiotic effect of the penem FCE 22101 against selected gram-positive and gram-negative bacteria in vitro and in vivo by the use of a tissue cage model in rabbits.

Isolates of Staphylococcus aureus, Streptococcus pneumoniae, Haemophilus influenzae and Klebsiella pneumoniae were tested for their bactericidal activity and postantibiotic effect (PAE) with the new penem FCE 22101. The tissue cage model in rabbits was used to study PAE in vivo. The bactericidal activity against all four species was shown to be in the range of 0.05-4.0 mg/l. A 99.9% killing effect at MBC concentrations was reached within 2 hours with S. pneumoniae and K. pneumoniae and within 6-8 hours with S. aureus and H. influenzae. After in vitro exposure by FCE 22101 a PAE in vitro and in vivo was obtained against S. aureus, S. pneumoniae and H. influenzae strains but no PAE could be demonstrated against K. pneumoniae. FCE 22101 showed a good bactericidal activity and PAE against the strains investigated, except for K. pneumoniae.

A streptococcal plasminogen activator in the focus of infection and in the kidneys during the initial phase of experimental streptococcal glomerulonephritis.

Strains of group A streptococci known to secrete the nephritis strain-associated protein (NSAP), a plasminogen activator, were studied for their ability to produce APSGN in rabbits. A tissue cage model was used to monitor the secretion of NSAP at the focus of infection and histopathological examination of kidney tissue was used to determine glomerular pathology. Animals infected with NSAP positive strains exhibited NSAP deposits in the glomerular tissue by day 7 in the absence of antibody to this molecule with progressive pathology indicative of APSGN three weeks later. Animals infected with the NSAP negative streptococcal strain exhibited no abnormal pathology. The ability of NSAP to bind to kidney tissue suggested that it has unique nephrotropic properties; and its ability to activate plasminogen to plasmin, possibly in situ, suggests that much of the pathological events associated with APSGN may be initiated by plasmin activity.

Factors influencing the biological inactivation of high protein bound beta-lactam antibiotics in vitro.

Unpredictable inactivation of antimicrobial agents may cause erratic results in pharmacokinetic studies. In this study we followed the inactivation of the high protein bound beta-lactams flucloxacillin, dicloxacillin and ceftriaxone in vitro. The antibiotics were added to pools of human and rabbit sera, ultrafiltrates of these pools, rabbit interstitial fluid, phosphate buffered saline (PBS), rabbit albumin in PBS and sodium dodecyl sulphate (SDS) treated preparations of human sera. Ceftriaxone was relatively stable but different serum pools varied significantly in their flucloxacillin and dicloxacillin inactivating capacity. The dominating inactivation took place within five minutes after the addition of antibiotics to serum. The inactivating factor was heat stable at 56 degrees C, 0.5 h, of relatively high molecular weight, and not related to albumin. The inactivating capacity could be diminished by SDS-treatment of serum suggesting a lipoprotein nature.

Properties of high and low density subpopulations of group B streptococci: enhanced virulence of the low density variant.

From the group B streptococcus (GBS) reference strain 090 la Colindale two subpopulations, which differed markedly regarding their capacities for biosynthesis of type-specific polysaccharide, were obtained by separation on a hypotonic Percoll density gradient. In the original strain and the high and low density variants, there was a negative correlation between buoyant density and bio-synthesis of type-specific polysaccharide as determined by ultrastructure and quantitative assays. The invasiveness of these variants was investigated by infecting rabbits via subcutaneously implanted tissue cages. In the animals infected with highly encapsulated bacteria, heavy bacteremia was detected 8 h post-infection, whereas in the animals which received high density bacteria with small amount of capsule, heavy bacteremia was not detected until after five days. All isolates recovered from the blood or organs of these rabbits were of the capsule rich phenotype, indicating a phenotypic shift in the subpopulation of high density bacteria. An apparently similar phenotypic shift was noted in an isolate from a baby with early onset septicemia. There was a dominance of low density bacteria in the isolate obtained from the baby as compared with the colonizing population of bacteria isolated from the cervix of the mother. From these type III isolates, subpopulations with different density maxima were obtained. A reversed shifting towards dominance of less encapsulated, high density bacteria was observed during in vitro passage of these subpopulations.

Demonstration of streptococcal antigens produced in vivo and identification of a nephritis associated protein.

An experimental infectious model in rabbits was used to study the release of bacterial substances in vivo in relation to the development of renal lesions, using three different group A Streptococci (type M56, M12 and T3). Three weeks post infection, renal lesions were confirmed in animals infected with M56 and M12, but not T3 streptococci, by histological and immunohistological analyses of the kidneys. Tissue-cage fluid (TCF) from the potentially nephritogenic strains (M56 and M12) showed a time-related increase of proteins and a production of streptococcal antibodies in serum. Isoelectrofocusing of TCF from nephritogenic and non-nephritogenic strains showed the presence of "nephritogenic-strain-specific" proteins in the most cationic region as well as in pH region 5.8-6.4. Immunoblotting after SDS-PAGE of TCF revealed a nephritogenic-strain-restricted protein that, by means of a mouse-monoclonal antibody, identified with the NSAP (Nephritic-Strain-Associated Protein). The biological activity demonstrated in the focus of infection using the M56 and M12 strains thus seems to be related in time to the induction of the nephritic process.

A sialic-acid-specific lectin from Cepaea hortensis that promotes phagocytosis of a group-B, type-Ia, streptococcal strain.

Group-B streptococci that possess a type-specific surface polysaccharide undergo phagocytosis only in the presence of antibodies to this, and complement. The snail Cepaea hortensis forms a lectin that is specific for sialic acid; treatment with this promoted the phagocytosis of a group-B streptococcus of serotype Ia (strain O90) in the absence of opsonic antibodies. The effect of the lectin was dose-dependent and required the presence of complement. The specificity of the lectin reaction for sialic acid was proved by the inhibition of phagocytosis by bovine submaxillary mucin. The participation of complement in the reaction was confirmed by demonstrating that C3 was bound to the surface of lectin-treated cells.

The influence of penicillin on growth and morphology of Streptococcus pyogenes in vivo.

The influence of penicillin (pc) on the growth, phagocytosis and killing of Streptococcus pyogenes was studied for an M protein positive (M+) and an M protein negative (M-) strain in vivo as well as in vitro. In vivo studies were based on a tissue cage model and the analyses were performed by CFU determinations and electron microscopic investigations. The M- strain was easily phagocytized with and without pc, but killing only occurred after pc treatment and thus the number of viable bacteria rapidly decreased under the influence of pc. M+ streptococci were not reduced in numbers by pc-treatment in vivo, but morphological changes and at high pc concentrations, phagocytosis could be seen. When this strain (M+) was cultivated in the absence of pc, the phagocytic cells were totally destroyed - a reaction that was prevented by penicillin. Variations in surface morphology of the two strains seem to influence the differences in sensitivity to penicillin, phagocytosis and killing.

Experimental infection with group B streptococci in tissue cages implanted in rabbits.

Using tissue cages subcutaneously implanted in rabbits an experimental group B streptococcal (GBS) infection was induced. The type Ia strain differed from the type III strain in producing a septicemic infection leading to the death of the animals on day 4 of infection. Active immunization with whole cells of the homologous strain resulted in phagocytosis and killing of the bacteria, whereas passive immunization only resulted in inhibition of the spread of the type Ia infection from the cages. Electron microscopy showed that in unprotected animals infection with either strains was followed by a lysis of erythrocytes and polymorphonuclear cells in the cage fluid. Opsonization by type-specific antibodies already present in the cage fluid preceded the phagocytosis of streptococci in the immunized animals. The tissue cage model offers a suitable way for studies of the pathogenesis of GBS infection. The different factors influencing the process of infection can also be easily monitored.

Static and dynamic properties of tissue cage fluid in rabbits.

The properties of tissue cage fluid in a steel net tissue cage model in rabbits were compared to those of serum by determination of the protein profile, the cell contents, and the pharmacokinetics of 125I albumin, 3H sucrose and 3H fusidic acid. The dominating serum proteins demonstrated by crossed immunoelectrophoresis were also detected in tissue cage fluid but at lower levels and at various ratios. The cell pattern gradually changed from an initial dominance of polymorphonuclear cells to lymphocytes during the five weeks following the subcutaneous implantation of the cages. The distribution of the highly protein-bound fusidic acid was markedly slower and the maximal tissue cage fluid level significantly lower than that of sucrose. Equilibrium of 125I albumin between serum and tissue cage fluid was slowly achieved during the following two weeks. The advantages and disadvantages of tissue cage models for studies of drug pharmacokinetics are discussed. The properties of tissue cage fluid and the possibility of repeated sampling make the model suitable for studies of experimental local infections. The influence of therapeutic agents and the host's response to the infectious process may also be elucidated.

Experimental poststreptococcal glomerulonephritis in rabbits.

Viable group A, M56 streptococci were inoculated into subcutaneously implanted steel net cages in rabbits, thereby establishing a local infection. The infectious process was monitored by quantitating the number of bacteria in the cage fluids. Functional signs of renal lesion were followed by urine analysis and measurement of serum creatinine. All animals were sacrificed 2-8 weeks after the inoculation and showed histological and immunological signs of kidney lesions. The morphological and functional abnormalities observed closely mimic those of post-streptococcal glomerulonephritis in man.

Effect of early penicillin treatment on the development of experimental poststreptococcal glomerulonephritis.

A newly developed experimental model in which poststreptococcal glomerulonephritis (PSGN) was established in rabbits utilizing viable group A streptococci, was used to study the possibility of hindering the development of renal disease by early penicillin treatment. The development of renal injury was followed by proteinuria, creatinine clearance, histological and immunological examinations of the kidneys. Therapy within the first 3 days of infection with i.m. pc-G prevented the nephritic process.

Distribution of lymecycline in interstitial fluid and paranasal sinus mucosa. An experimental and clinical study.

A comparison between different methods of determining the levels of lymecycline in tissue is presented. The drug levels have been compared with those registered, during the steady state condition, in interstitial fluid obtained from subcutaneously implated tissue cages in rabbits. The 'tissue piece diffusion method' developed is reproducible; it allows antibiotic levels to be determined in minute pieces of tissues, and it seems to measure the total diffusable drug in the interstitial fluid phase. This method was employed in the analyses of lymecycline concentrations in the maxillary sinus mucosa in 12 patients, and it demonstrated a good penetration into the tissue. The level reached was well above the Minimal Inhibitory Concentration of most bacteria causing maxillary sinusitis.

Studies on the penetration of lymecycline into the paranasal sinus in man.

Concentration of lymecycline, a lipophilic tetracycline, in the human sinusal mucosa tissue, as measured by the "punch method" technique are fairly high. The median value of all examined pieces of mucosa amounts to 1.2 mcg/g.

Penetration of lymecycline (tetracycline) into the tissues of the Fallopian tube.

On the basis of a comparison of different methods used to determine the concentration of active lymecycline in Fallopian tube tissue, a method was adopted which allowed the determination of lymecycline in serosa and mucosa tissue of the Fallopian tube in 16 women undergoing hysterectomy. A difference in the concentration of active drug was found in serosa (greater than 1.7 mg/kg tissue) and mucosa (greater than 0.8 mg/kg tissue) in 16 patients on lymecycline therapy for 24 hours. These figures were fairly constant for the 7.5 hours following the administration. The figures are discussed in relation to the MIC values of the commonly found microorganisms in salpingitis.