[Special aspects of implant-associated infection in orthopedic surgery. From the pathophysiology to custom-tailored prevention strategies]. / Besonderheiten der implantatassoziierten Infektion in der orthopädischen Chirurgie. Von der Pathophysiologie zu massgeschneiderten Präventionsstrategien.

One of the most important risk factors in orthopedic surgery is implant-associated infection. Adhesion and colonization mediated implant infections are extremely resistant to antibiotics and host defences and frequently persist until the biomaterial or foreign body is removed, which is standard therapy. Tissue damage caused by surgery and foreign body implantation increases the susceptibility to infections, activates host defences and stimulates the generation of inflammatory mediators including radicals that are further aggravated by bacterial activity and toxins. Nearly one third of implant-related infections can be prevented by strictly following established infection control guidelines. However, a significant number of implant-associated infections remains. The escape of bacteria from host defence and antibiotic therapy makes the development of infection-resistant materials as anti-microbial drug delivery systems feasible. This concept consists of the sustained delivery of antimicrobial drugs into the local microenvironment of implants avoiding systemic side effects exceeding usual systemic concentrations by magnitudes of order.

Pharmacokinetics of the antimicrobial agents rifampicin and miconazole released from a loaded central venous catheter.

The time course of rifampicin and miconazole concentrations after insertion of a polyurethane catheter loaded with these antibiotics were studied. Data from controlled release experiments in vitro were used, and the concentration time courses of the antimicrobials in serum were calculated by pharmacokinetic simulations. Systemic therapy using typical dosages (rifampicin 600 mg/day iv, miconazole 3 x 200 mg/day iv) results in rifampicin concentrations between 54 and 8424 microg/L, and miconazole concentrations between 3567 and 4676 microg/L. After insertion of a polyurethane catheter loaded with these antibiotics, the maximal concentrations after catheter placement were determined as 6 microg/L at 10.7h for rifampicin, and 13 microg/L at 28.6 h for miconazole. Assuming that the total amount of antibiotics incorporated in the catheter matrix were bioavailable ('worst case'), the resulting maximal concentrations calculated by simulation are 10 microg/L for rifampicin and 65 microg/L for miconazole. Maximal concentrations of rifampicin or miconazole resulting from the insertion of a polyurethane catheter loaded with these antibiotics are, therefore, far below the concentrations resulting from a systemic therapy with the same antimicrobial agents. Even in the worst case, the danger of selecting resistant bacterial strains seems remote because the systemic drug levels are magnitudes of order below subinhibitory concentrations.

Pharmacokinetics of intra-arterial mitomycin C in the chemoembolization treatment of liver metastases with polyvinylalcohol or degradable starch microspheres.

OBJECTIVES: In cancer treatment, arterial blood flow reduction by embolization combined with intra-arterial chemotherapy may be advantageous by achieving high and prolonged drug concentrations in the tumor, and lower systemic drug exposure. The pharmacokinetics of intra-arterial mitomycin C (MMC) was investigated in patients with liver metastases undergoing chemoembolization treatment. METHODS: The chemoembolization treatment consisted of the use of polyvinylalcohol microspheres (ITC-Contour, diameter 150-250 micro m, irreversible vessel occlusion, 20 mg MMC, 15 patients) followed by sealing of the supplying artery with Ethibloc. Alternatively, starch microspheres (Spherex, diameter 45 micro m, biologically degradable, 10 ml of suspension containing 60 mg starch/ml, 20 mg MMC in 7 patients, 15 mg/m2 body surface in 3 patients) were used. MMC was infused over 6 min into the artery supplying the tumor. Serum MMC concentrations were determined from peripheral venous blood samples [protein precipitation with acetonitrile, reverse-phase HPLC with ultraviolet detection (C18 column, elution with 0.01 M, pH 6.5 phosphate buffer/methanol, v/v 70:30, 365 nm)]. The pharmacokinetic parameters were computed assuming an open two-compartment model and linear kinetics. RESULTS: The disposition parameters for MMC in patients treated with polyvinylalcohol microspheres were comparable to data from the literature (C(max)=913+/-98 ng/ml, T(max)=7.7+/-0.3 min, V(c)=0.27+/-0.03 l/kg, V(ss)=0.59+/-0.07 l/kg, Cl=757+/-67 ml/min, T(1/2 alpha)=5.8+/-0.8 min, T(1/2 beta)=50.4+/-4.1 min). There was no significant difference in the disposition parameters for MMC between patients treated with polyvinylalcohol microspheres and those treated with starch microspheres ( P>0.05). In particular, there was no significant difference in the standardized AUC between the groups; this implies that the systemic toxicity of MMC is comparable when polyvinylalcohol microspheres and ethibloc (AUC 1472+/-123 microg min/l per 1 mg MMC) or starch microspheres (AUC: 1448+/-172 microg min/l per 1 mg MMC) are used. CONCLUSION: The AUC values found in this study do not indicate a reduction in the systemic toxicity of MMC if applied intra-arterially in combination with embolizing microspheres, when compared to the AUC values in the literature for intra-arterial application without embolization, or intravenous MMC application. The amount of starch microspheres may have been too small to cause marked effects. On the other hand, there is a very wide range of AUC values reported in the literature for different application modes, and the use of such historical controls is not adequate for detecting more subtle advantages of the chemoembolization procedure, which may, however, exist.

Antiinfective and encrustation-inhibiting materials--myth and facts.

Catheters, urethral and ureteral stents and other urological implants are frequently affected by encrustration and infection due to their permanent contact with urine. Indwelling urinary catheters provide a haven for microorganisms and thus require extensive monitoring. Several surface modification techniques have been proposed to improve the performance of devices including the immobilization of biomolecules, the incorporation of hydrophilic grafts to reduce protein adsorption, the creation of hydrophobic surfaces, the creation of microdomains to regulate cellular and protein adhesion, new polymers and antimicrobial coatings. Physico-chemical explanation to elucidate the mechanism of such encrustation or infection inhibiting materials is still not available. Our series of experiments showed a marked decrease of silver-activity in biological fluids which corresponds with the controversial clinical results obtained with silver coated urinary catheters. Rifampicin/minocycline coated catheters had very low activity against Gram-negative rods, enterococci and Candida spp., the main causing organisms of urinary catheter infection. Surface engineered materials and antimicrobial drug delivery systems will be the next generation of sophisticated urinary catheters and stents, if both efficacy as well as efficiency has been proved clinically.

Studies on the role of superoxide anion radicals for the cardiotoxicity of adrenochrome.

Adrenochrome is an oxidative product of adrenaline and possesses cardiotoxic properties. As oxygen free radicals play a role in the cytotoxic effects of catecholamines, the role of superoxide anion radicals, as mediators of adrenochrome toxicity, was investigated using electrically-driven Langendorff rabbit hearts with depleted catecholamine stores. Repetitive regional myocardial ischemia (MI) was induced by coronary artery branch ligature, and MI was quantitated from epicardial NADH-fluorescence photography. Adrenochrome (10(-6) mol/l) was added to the perfusion solution after a reperfusion period of 20 min, 30 min before the 2nd coronary occlusion, with or without the additional application of SOD (30 U/ml). Left ventricular pressure was significantly enhanced by adrenochrome (p < 0.05), but it fell rapidly down below its initial value (p < 0.05). Coronary flow was significantly decreased by adrenochrome (p < 0.05). Whereas epicardial NADH-fluorescence was similar after repetitive coronary occlusions in untreated controls, it was significantly enhanced by adrenochrome (p < 0.05). The deleterious effects of adrenochrome on MI were not inhibited by SOD. Thus, there is no evidence for superoxide anion radicals as mediators of the deleterious effects of adrenochrome on MI in isolated rabbit hearts. The deleterious effects of adrenochrome on MI in isolated rabbit hearts might be caused by functional effects, impairing the oxygen consumption/oxygen supply balance.

Studies of the toxic effects of norepinephrine on isolated cardiomyocytes of guinea-pigs.

Catecholamines have been demonstrated to possess direct cardiotoxic effects mediated by oxygen free radicals in isolated organ preparations. In order to assess direct cytotoxic properties, the influence of exogenous noradrenaline (norepinephrine, CAS 51-41-2) (10(-6) mol/l) on isolated guinea-pigs cardiomyocytes was examined, in the presence of propranolol (10(-6) mol/l) and phentolamine (10(-6) mol/l) to inhibit adrenoceptor-mediated effects. Cell viability was assessed by morphologic examination (% of striated, rod-shaped cells), before and after a treatment period of 15 and 60 min by the measurement of intracellular enzyme activities in the supernatant of the suspension (lactate dehydrogenase, creatine kinase, aspartate aminotransferase, alanine aminotransferase, glutamate dehydrogenase). The proportion of viable, rod-shaped cardiomyocytes (21.6% +/- 7.6% after preparation, before starting the treatments) significantly decreased over the experimental time (p < 0.05) and, concomitantly, the activity of intracellular enzymes in the supernatant increased. There was no difference between controls and treated suspensions. Thus, there is no evidence for direct toxic effects of norepinephrine in micromolar concentration on isolated cardiomyocytes of guinea-pigs. However, cytoprotective effects by propranolol and/or phentolamine cannot be excluded in this model.

[Hyperbaric oxygen therapy (HBO): current standing]. / Hyperbare Oxygenationstherapie (HBO): eine Standortbestimmung.

For more than thirty years hyperbaric oxygen therapy (HBO) has been an important and ultimate therapeutic tool in special indications. Hyperbaric oxygen improves tissue oxygenation, stimulates important mechanisms in wound healing and exerts beneficial effects on other biochemical and cellular processes. The properties of hyperbaric oxygen have built the rationale for its use as therapy of choice in patients with severe carbon monoxide poisoning, decompression sickness and arterial gas embolism, and as adjunctive therapy for the treatment of osteoradionecrosis, necrotizing fasciitis and compromised skin grafts and flaps. The efficacy of adjunctive hyperbaric oxygen in the treatment of lower extremity problem wounds in diabetic patients seems to be proven. There is little scientific support for other uses of hyperbaric oxygen and its therapeutical benefit should be further investigated. When used according to standard protocols hyperbaric oxygen treatment is a safe therapy with little adverse effects.

Lidocaine pharmacokinetics during hyperbaric hyperoxia in humans.

METHODS: The disposition of drugs may be influenced by hyperbaric conditions, in particular by changes of liver perfusion. The effect of hyperbaric hyperoxia on the pharmacokinetics of lidocaine, a drug eliminated in the liver with a perfusion-limited clearance, was investigated in human volunteers in a crossover trial. METHODS: A single dose lidocaine i.v. bolus (0.69 or 0.75 mg x kg(-1)) was administered to two volunteers under normobaric conditions (NB: 1 bar or 0.1 MPa, air) and under hyperbaric/hyperoxic conditions (HBO: 2.5 bar or 0.25 MPa, alternating 100% O2-breathing for 20 min and air breathing for 5 min). Blood samples were serially collected for 5 h (NB) or 75 min (HBO), and lidocaine concentration in serum was measured by immunoassay. Data were analyzed assuming linear kinetics and an open two-compartment model. RESULTS: At 1 bar or 0.1 MPa, lidocaine injection caused only slight dizziness and buzzing in the ear. Heart rate and blood pressure were not influenced. Under HBO, lidocaine injection caused marked dizziness and buzzing in the ears, sweating, tremor and coordination-disturbances, even though maximal lidocaine concentrations (0.63 mg x L(-1) and 0.70 mg x L(-1)) were far below therapeutic serum concentrations (1.5-5.0 mg x L(-1)). Pharmacokinetic parameters of lidocaine were similar to those published earlier (T1/2beta: 110+/-16 min; CI: 12.6+/-2.9 ml x min(-1) x kg(-1); Vss: 1.73+/-0.18 L x kg(-1)). There was no indication for effects of HBO on the disposition of lidocaine (p > 0.05). CONCLUSION: The pharmacokinetics of lidocaine do not seem to be influenced in a clinically relevant way in humans by a single HBO-exposure under usual therapeutic conditions. Side effects of lidocaine at 2.5 bar or 0.25 MPa may be caused by pharmacodynamic interactions between lidocaine and hyperbaric/hyperoxic conditions.

Pseudocholinesterase-activity reduction during cardiopulmonary bypass: the role of dilutional processes and pharmacological agents.

UNLABELLED: 1. Pseudocholinesterase (ChE) activity is a determinant of the elimination kinetics of several drugs used in anesthesia. The time course of ChE activity was investigated in 16 patients undergoing cardiosurgery for a cardiopulmonary bypass (CPB) in normothermia or hypothermia. 2. The onset of the CPB was accompanied by a decrease in ChE activity (-37%) (P<0.05) and protein concentration (-24%) (P<0.05). The quotient ChE activity/protein concentration was numerically reduced to a smaller extent (-15%) (P>0.05). After the CPB was finished, ChE activity and the protein concentration remained low for the remaining operation time. 3. There was no difference in ChE activity, measured in vitro at 37 degrees C, between the normothermic and hypothermic group (P>0.05). 4. There was no correlation between heparin concentration in serum and reduction of ChE activity in vitro (P>0.05). In vitro, the ChE activity was not affected by either heparin in doses as high as 10,000 U/ml or aprotinin in doses as high as 10,000 U/ml (P>0.05). 5. CONCLUSIONS: (1) ChE activity is reduced by CPB mainly by hemodilution and (2) the pharmacological agents used in the present anesthetic technique (heparin, aprotinin, midazolam, fentanyl, propofol and mivacurium) do not inhibit ChE activity at therapeutic serum concentrations.

Effects of hyperbaric and hyperoxic conditions on the disposition of drugs: theoretical considerations and a review of the literature.

1. Hyperbaric or hyperoxic or both conditions may affect the disposition of drugs by (1) changes in the catalytic activity of drug metabolizing enzymes, (2) hemodynamic changes and (3) changes in membrane permeability, affecting drug distribution. 2. In isolated microsome preparations from rat liver, the metabolism rate of aniline, but not of amidopirin, is reduced by hyperoxia. In vivo, the clearance of salicylic acid is enhanced in the dog at 2.8 ATA and 100% O2, but not at 6 ATA and air, for reasons that are still unknown. The disposition of theophylline, pentobarbital or pethidine is not affected in dogs by hyperbaric or hyperoxic conditions. 3. In human volunteers, hyperbaric or hyperoxic or both conditions do not affect the disposition of gentamycin (2.4 bar, 100% O2), caffeine or lidocaine (2.5 bar, 100% O2). 4. In conclusion, a single exposure to hyperbaric or hyperoxic conditions does not seem to affect single-dose pharmacokinetics of drugs eliminated by the kidney (gentamycin) or by the liver with a capacity-limited clearance (pentobarbital, theophylline, caffeine) or with a perfusion-limited clearance (pethidine, lidocaine). The enhancement of salicylic acid clearance in dogs under hyperoxic conditions remains unclear.

Anti-infective catheters: novel strategies to prevent nosocomial infections in oncology.

Intravenous access contributes significantly to the therapeutical success and to the comfort of oncologic patients. The highest risk for bloodstream infections, however, is vascular catheter-mediated. In oncology high mortality is associated with Pseudomonas aeruginosa, Candida albicans and Staphylococcus aureus sepsis. Besides established hygienic measures, the coupling or incorporation of antimicrobial substances to or into catheter materials may be a suitable way to prevent the development of catheter-associated infections. Here we present a risk- benefit evaluation of different models of antimicrobial catheter coated with silver, antiseptics or antibiotics. The controversial reports on clinical efficacy and the potential of adverse reactions due to silver and antiseptic coated catheters are discussed. The microbiological, pharmaceutical and physicochemical backgrounds of different types of coating are discussed in detail. Incorporation of antimicrobial agents into long-term silicon catheters providing a slow release of those substances through the external and internal surfaces of catheters may be the most effective technological innovation for reducing biomaterial-mediated nosocomial infections.

Studies of the cardioprotective effects of ascorbic acid in isolated rabbit hearts.

Ascorbic acid (CAS 50-81-7) might mediate cardioprotective effects by scavenging free oxygen radicals. The effects of exogenous ascorbic acid on acute myocardial ischemia (MI) was investigated in isolated electrically-driven rabbit hearts (Langendorff, constant pressure: 70 cm H2O, Tyrode solution, Ca2+ 1.8 mmol/l, 37 degrees C). Repetitive MI, separated by a reperfusion period of 50 min, was induced by coronary artery branch ligature and quantitated from epicardial NADH-fluorescence photography. Starting after a reperfusion period of 20 min, isolated hearts were treated with ascorbic acid (10(-5) or 10(-4) mol/l). Ascorbic acid had no significant influence on the left ventricular left ventricular pressure or the coronary flow (p > 0.05). Ascorbic acid had no significant effect on epicardial NADH-fluorescence area or intensity (p > 0.05). Free radical scavenging properties reported for ascorbic acid do not mediate cardioprotective effects at the concentrations used in isolated rabbit hearts.

Ocular concentrations of mitomycin C using different delivery devices.

Hypotony and its sequelae are a frequent complication of trabeculectomies performed with mitomycin C (MMC), possibly related to intraocular toxicity of the substance. In an animal model with rabbits, we used different devices for the application of MMC and measured extra- and intraocular concentrations by HPLC. In addition, the concentrations of MMC remaining in the devices were determined. The devices were (1) a regular surgical sponge, (2) a scleral shield, (3) a presoaked soft contact lens, (4) a soft contact lens with MMC application, and (5) subconjunctival injection. Ocular concentrations of MMC were similar within the first 4 groups and were highest in the last. The measurements suggest that MMC penetrates intraocularly regardless of the device used. The variability of remaining MMC concentrations in the devices was lowest in the soft contact lenses suggesting an improved delivery system compared to the usually used surgical sponges.

Controlled release of antibiotics from biomedical polyurethanes: morphological and structural features.

Polymer-associated infections are of increasing importance. Antistaphylococcal antimicrobial substances (ciprofloxacin, gentamycin, fosfomycin, flucloxacillin) were incorporated into polyurethanes by the solvent casting technique. Drug release rates, bacterial colonization and morphological features were evaluated to predict and understand the antimicrobial activity of these delivery systems. Drug release characteristics were investigated by standard bioassay and high-performance liquid chromatography (HPLC), and the physico-chemical mechanisms of the delivery were discussed. Ciprofloxacin hydrochloride showed a fast initial release rate, whereas gentamicin-base was characterized by a more continuous release type of behaviour. Bacterial colonization to the antibiotic-loaded polyurethanes was inhibited effectively by preparations showing a slower but more sustained antimicrobial delivery. Polyurethane-antibiotic combinations were most homogeneous for gentamicin-base and flucloxacillin as shown by scanning electron microscopy (SEM). In polymers loaded with fosfomycin and ciprofloxacin a granular structure of the crystallized drug embedded in the polyurethane matrix could be demonstrated. Physico-chemical similarity of the polymeric material and the antibiotics is important for the homogeneity of polymer-antibiotic combinations. High homogeneity is required for a sustained and prolonged release over time and effective inhibition of bacterial colonization.

[Clinical and preclinical efficacy of antimicrobial catheters]. / Klinische und präklinische Effizienz antimikrobieller Katheter.

Frequent use of polymeric material for diagnostic and therapeutic use has provoked an increase of so-called polymer-associated foreign body infections. The difficulties in correctly diagnosing catheter-related infections and the necessity to remove the device in case of suspected infection require preventive procedures in the first place. Therefore, several studies have dealt with surface treatment by coating with antibiotics, antiseptics and silver. Some of them remain in clinical trials or are already sold in the USA. Clinical failure of silver-sulfadiazine coated CVCs has been reported recently. Recently a rifampicinminocyclin-catheter has been shown to reduce efficiently material-associated infection in preclinical and clinical studies. Most of the substances retained on catheter surfaces are released only in small quantities resulting in ineffective suppression of bacterial growth. If substantial amounts of antimicrobials are incorporated into the biomaterial, release pattern and killing efficacy of each substance group has to be determined in order to reach bactericidal concentrations in the environment of adherent bacteria. Killing kinetics of antibiotics, antiseptics and silver to bacteria are strongly different. On the one hand killing kinetics of microorganism due to substantial quantities of antiseptics are fast, on the other hand antiseptics destroy structures of both, bacteria and blood cells. Bactericidal concentrations of Ag-ions show fibroblast inhibition. Antibiotics, however, show lowest toxicity due to the specific action to microorganisms. However, sustained release over time is needed to reach bactericidal concentrations, considering the slower bactericidal action of typical antibiotics. Development of allergic reaction is possible for all substances, whereas development of bacterial resistance can be prevented by use of antimicrobial combinations.

Caffeine pharmacokinetics during hyperbaric hyperoxia in humans.

The effect of hyperbaric hyperoxia on the pharmacokinetics of caffeine was investigated in human volunteers. Some 600 ml of coffee were administered to 2 volunteers and blood samples were serially collected for 24 h. The volunteers entered a hyperbaric chamber 2.5 h following coffee ingestion for a total period of 110 min (0.25 MPa, alternating 100% O2-breathing for 20 min and air breathing for 5 min). The concentrations of caffeine in serum was determined by high pressure liquid chromatography. The caffeine amount ingested was determined by analyzing an aliquot of the coffee beverage. Data were analyzed assuming linear kinetics and an open one-compartment model. Effects of hyperbaric hyperoxia on caffeine disposition were investigated using a runs test. Moreover, a one-population t-test was applied to residuals, separately for data from the initial normobaric period, the hyperbaric period and the terminal normobaric period. Pharmacokinetic parameters were similar to established literature data on caffeine [Volunteer 1: maximal concentration (Cmax: 6.13 mg.L-1 at Tmax: 55 min, half-time of elimination (T1/2: 180 min, total clearance (Cl): 3.41 ml.min-1.kg-1; volume of distribution (Vd: 0.88 I.kg-1; Volunteer 2: Cmax: 6.23 mg.L-1, Tmax: 94 min, T1/2: 283 min, Cl: 1.90 ml.min-1.kg-1, Vd: 0.77 L.kg-1). The runs test as well as the analysis of residuals gave no evidence for alterations of caffeine disposition by hyperoxia (p > 0.05). The pharmacokinetics of caffeine do not seem to be influenced in a clinically relevant way in humans during a stay for 100 min at 0.25 MPa, alternating 100% O2 and air breathing.

Pharmacokinetics of intraarterial mitomycin C in the chemoembolisation treatment of liver metastases.

1. The pharmacokinetics of mitomycin C (MMC) were investigated in 12 colorectal cancer patients with liver metastases undergoing chemoembolisation. Hepatic artery branches were embolized using polyvinylalcohol microspheres (150-250 microns) before applying 20 mg MMC in 4-8 min. 2. Serum MMC concentrations were determined from peripheral venous blood samples by reverse-phase HPLC using ultraviolet detection. Pharmacokinetic parameters were computed assuming an open two-compartment model. 3. Pharmacokinetic parameters were similar to values given in the literature for intravenous (IV) or intraarterial (IA) bolus MMC injections (Tmax = 7.0 min following the beginning of MMC infusion, Vss = 0.57 1/kg, C1 = 8.9 ml/min.kg, T1/2 alpha = 8.3 min, T1/2 beta = 58.6 min). 4. The area under the serum concentration-time-curve (AUC), standardized by the MMC amount injected, was similar to values reported in the literature for IV or IA bolus injections. There is no evidence for reduced systemic MMC exposure following the embolization procedure used.

Influence of angiotensin-converting enzyme inhibition by fosinopril on myocardial perfusion in streptozotocin-diabetic rats.

We examined the influence of the new angiotensin-converting enzyme inhibitor (ACEI) fosinopril on function and perfusion of the diabetic rat heart. Streptozotocin-diabetic rats (60 mg/kg body weight) were treated with fosinopril (10 mg/kg body weight/day) for 4 months. Cardiac performance was analyzed in the isolated heart perfused at constant volume. Epicardial perfusion was determined by measuring epicardial fluorescence changes after injection of FITC-dextrane (3 kDa) as described previously. As compared with controls, fosinopril prevented or diminished the increase in end-diastolic pressure (EDP), coronary perfusion pressure (CPP), and vascular resistance in diabetes. The intravascular volume strongly reduced in diabetes was increased, and the epicardial perfusion rate was accelerated in hearts of diabetic rats treated with fosinopril. The vascular exchange diminished in hearts of untreated diabetic rats was enhanced, and the transcapillary permeability was slightly accelerated at low flow rates. These data indicate that treatment of streptozotocin-diabetic rats with fosinopril prevents severe disturbances of coronary autoregulation and at least partly the impairment of cardiac perfusion generally observed in diabetic rats. Together with previously published morphological data demonstrating an inhibition of interstitial and perivascular fibrosis in hearts of diabetic rats with fosinopril, our observations suggest that ACE inhibition by fosinopril is cardioprotective in diabetes.

Mitomycin-C concentration in human ocular aqueous humor after topical administration during trabeculectomy.

UNLABELLED: The concentration of the antitumor antibiotic mitomycin (CAS 50-07-7, mitomycin C, MMC), used in ophthalmic surgery for its antiproliferative effects, was measured in the aqueous humor of 7 glaucoma patients undergoing trabeculectomy. Sponges soaked with MMC-solution (100 microliters of MMC-solution 0.2 mg/ml: 20 micrograms) were applied intraoperatively on the scleral flap for 5 min. 100 to 200 microliters of aqueous humor were drawn with a needle 10 min following the end of topical MMC-treatment. Samples were assayed for MMC using a reverse-phase HPLC-system with ultraviolet detection (C18-column, elution: phosphate buffer (0.01 mol/l, pH: 6.5): methanol, v:v = 70:30, 365 nm). Swabs were extracted in phosphate-buffer (0.1 mol/l, pH: 7.0) before HPLC-analysis. External calibration was used for MMC quantitation. Quantitation limit was 10 ng/ml. In all aqueous humor samples MMC-concentration was below 10 ng/ml. MMC in the swabs amounted to 37% of the MMC amount applied. CONCLUSION: After intraoperative topical application, the MMC concentration in the aqueous humor of patients is very low. The substantial loss of MMC from the swabs used for the topical MMC-treatment suggests 1. rapid systemic absorption of MMC and/or 2. a loss through irrigation of the operative field following topical MMC-application.