A high performance liquid chromatography-tandem mass spectrometry assay for therapeutic drug monitoring of 10 drug compounds commonly used for antimicrobial therapy in plasma and serum of critically ill patients: Method optimization, validation, cross-validation and clinical application.

BACKGROUND AND AIMS: Intensive care antibiotic treatment faces challenges due to substantial pharmacokinetic differences in critically ill patients. Individualized antibiotic dosing guided by therapeutic drug monitoring (TDM) is considered to minimize the risk of treatment failure and toxicity. This study aimed to develop a valid method for simultaneous LC-MS/MS quantification of 10 drugs frequently used in intensive care antibiotic therapy for which TDM-guided dosing is recommended: piperacillin, meropenem, flucloxacillin, cefuroxime, vancomycin, colistin A and B, linezolid, ciprofloxacin and tazobactam. METHODS AND RESULTS: Thorough optimization of sample preparation and chromatography resulted in a fast and simple method based on protein precipitation of 50 µL plasma or serum and gradient elution using an Acquity UPLC HSS-T3 column. Electrospray ionization-triple quadrupole mass spectrometry in dynamic multiple reaction monitoring was used for quantification, covering the therapeutic range of each drug compound. Validation following EMA and FDA recommendations, including inter-platform validation and inter-laboratory comparison, demonstrated high accuracy, precision and robustness of the new method. The assay was successfully used to monitor plasma antibiotic levels of critically ill patients (n = 35). CONCLUSION: The established multiplex method covers major drug classes with documented dosing challenges, provides a reliable basis for the implementation of high-throughput TDM, and its application confirmed the clinical impact of TDM in a real-world setting.

Targeted and untargeted screening for impurities in losartan tablets marketed in Germany by means of liquid chromatography/high resolution mass spectrometry.

Technical advances in the field of quality analysis allow an increasingly deeper look into the impurity profile of drugs. The ability to detect unexpected impurities in addition to known impurities ensures the supply of high-quality drugs and can prevent recalls due to the detection of harmful unexpected impurities, as has happened recently with the N-nitrosamine and azido impurities in losartan (LOS) drug products. In the present study, the LC-MS/HRMS approach described by Backer et al. was applied to an even more complex system, being the investigation of 35 LOS drug products and combination preparations purchased in 2018 and 2022 in German pharmacies. The film-coated tablets were analysed by means of four LC-MS/HRMS method variants. For the separation a Zorbax RR StableBond C18 column (3.0 ×100 mm, particle size of 3.5 µm, pore size of 80 Å), a gradient elution and for mass spectrometric detection a qTOF mass spectrometer with electrospray ionization in positive and negative mode was used. An information-dependent acquisition method was applied for the acquisition of high-resolution mass spectrometry data. The combination of an untargeted and a targeted screening approach revealed the finding of eight impurities in total. Beside the five LOS related compounds, LOS impurity F, J, K, L, M, and related compound D from amlodipine besilate, LOS azide and an unknown derivative thereof were detected. Identification and structure elucidation, respectively, were successfully performed using in silico fragmentation. Differences in the impurity profiles of drug products from 2018 and 2022 could be observed. This study shows that broad screening approaches like this are applicable to the analysis of drug products and can be an important enhancement of the quality assurance of medicinal products.

Understanding adefovir pharmacokinetics as a component of a transporter phenotyping cocktail.

PURPOSE: Adefovir (as dipivoxil) was selected as a probe drug in a previous transporter cocktail phenotyping study to assess renal organic anion transporter 1 (OAT1), with renal clearance (CLR) as the primary parameter describing renal elimination. An approximately 20% higher systemic exposure of adefovir was observed when combined with other cocktail components (metformin, sitagliptin, pitavastatin, and digoxin) compared to sole administration. The present evaluation applied a population pharmacokinetic (popPK) modeling approach to describe adefovir pharmacokinetics as a cocktail component in more detail. METHODS: Data from 24 healthy subjects were reanalyzed. After establishing a base model, covariate effects, including the impact of co-administered drugs, were assessed using forward inclusion then backward elimination. RESULTS: A one-compartment model with first-order absorption (including lag time) and a combination of nonlinear renal and linear nonrenal elimination best described the data. A significantly higher apparent bioavailability (73.6% vs. 59.0%) and a lower apparent absorption rate constant (2.29 h-1 vs. 5.18 h-1) were identified in the combined period compared to the sole administration period, while no difference was seen in renal elimination. The population estimate for the Michaelis-Menten constant (Km) of the nonlinear renal elimination was 170 nmol/L, exceeding the observed range of adefovir plasma maximum concentration, while the maximum rate (Vmax) of nonlinear renal elimination was 2.40 µmol/h at the median absolute estimated glomerular filtration rate of 105 mL/min. CONCLUSION: The popPK modeling approach indicated that the co-administration primarily affected the apparent absorption and/or prodrug conversion of adefovir dipivoxil, resulting in the minor drug-drug interaction observed for adefovir as a victim. However, renal elimination remained unaffected. The high Km value suggests that assessing renal OAT1 activity by CLR has no relevant misspecification error with the cocktail doses used.

Application of advanced high resolution mass spectrometric techniques for the analysis of losartan potassium regarding known and unknown impurities.

Recalls of medicinal products can cause supply bottlenecks. This is often due to the findings of unexpected impurities that pose a health risk to patients. A recent example is losartan potassium which was contaminated with azido-impurities. The choice of the analytical method determines which substances can be detected and thus controlled. In this study a combination of an untargeted screening approach for impurities and a targeted evaluation of high-resolution mass spectrometry data was applied to search for impurities not described so far, leaving out a precise quantification. Six losartan potassium samples were studied regarding known and unknown impurities and hence highlight the applicability and capability of the approach. For separation a Zorbax RR StableBond C18 column (3.0 ×100 mm, particle size of 3.5 µm, pore size of 80 Å), a gradient elution and an electrospray ionization in positive and negative mode for mass spectrometric detection was used. An information-dependent acquisition method was applied for the measurement of losartan potassium samples. The untargeted data evaluation using general unknown comparative screening revealed the presence of N-methyl-2-pyrrolidone (NMP) and another impurity from synthesis. The identity of NMP was corroborated by a spiking experiment and the amount was estimated by means of standard addition. A targeted data evaluation by generating extracted ion chromatograms resulted in finding of four additional impurities. Combined approaches like this are needed to detect and respond to changes in the quality of drugs precociously.

Clinical and demographic factors affecting trough levels of isavuconazole in critically ill patients with or without COVID-19.

BACKGROUND: The broad-spectrum antifungal isavuconazole is administered to treat invasive aspergillosis and mucormycosis. OBJECTIVES: Isavuconazole plasma concentrations in critically ill ICU patients with or without COVID-19 and invasive fungal infection were determined, and factors for sub-therapeutic drug levels (<1 µg/mL) were evaluated. PATIENTS AND METHODS: Isavuconazole plasma levels were measured as part of therapeutic drug monitoring (TDM) in ICUs of a tertiary hospital. Concentrations determined 20-28 h after previous dosing were defined as trough (Cmin ) levels. A total of 160 Cmin levels from 62 patients with invasive fungal infections were analysed, 30 of which suffering from COVID-19. Patient characteristics included into univariable and multivariable analyses were gender, age, COVID-19 status, body mass index (BMI), sepsis-related organ failure (SOFA) score, renal replacement therapy (RRT) and extracorporeal membrane oxygenation (ECMO) requirement. RESULTS: The mean Cmin of isavuconazole in all patients was 1.64 µg/mL (interquartile range 0.83-2.24 µg/mL, total range 0.24-5.67 µg/mL). In total, 34.4% of the Cmin values (corresponding to 46.8% of patients) were below a threshold concentration of 1 µg/mL. Drug concentrations between patients with or without COVID-19 did not differ (p = .43). In contrast, levels were significantly lower in patients with female sex (p = .0007), age ≤ 65 years (p = .002), BMI > 25 (p = .006), SOFA score > 12 (p = .026), RRT (p = .017) and ECMO requirement (p = .001). CONCLUSIONS: Isavuconazole plasma levels can be negatively affected by patients' risk factors, supportive renal replacement and ECMO therapy. Future prospective studies analysing the relevance of isavuconazole drug levels in ICU patient outcome are urgently needed.

Identification of low-level impurities in drug prototypes of carbocisteine by means of liquid chromatography-high-resolution mass spectrometry and general unknown comparative screening.

High-resolution tandem quadrupole time-of-flight mass analysers enable new automated workflows for untargeted data evaluation of complex samples like drug products. An example of such procedure is the so-called general unknown comparative screening (GUCS), which is used for software-assisted, automated identification of components that are only present in a sample and not in a reference. The GUCS approach has been employed for the first time to detect both degradation products and reaction products in drug products. Two different carbocisteine containing syrup prototypes - one with sucrose and the other with artificial sweeteners - were selected as examples after nine months of storage at 40 °C and 75% relative humidity. The samples were analysed chromatographically using a Coresep SB mixed-mode column and high-resolution MS and MS/MS data were recorded in information dependant acquisition mode on a Sciex X500R quadrupole time-of-flight mass spectrometer. Data analysis was considerably facilitated using the corresponding placebo formulation as reference samples. With the GUCS approach two hitherto unknown degradation products of carbocisteine, i.e. the carbocisteine lactam of the sulfoxides and the disulfide between l-cysteine and thioglycolic acid, were detected at low concentrations in both of the syrup formulations. The presumed structures were confirmed by in silico analysis of the fragment spectra and high-resolution LC-MS experiments with reference substances. Two additional impurities were found in the sucrose-containing sample and identified as the N-glycosides of carbocisteine and its lactam, respectively, using binary mixtures with a 13C-labelled monosaccharide.

Pharmacokinetics of Nirmatrelvir and Ritonavir in COVID-19 Patients with End-Stage Renal Disease on Intermittent Hemodialysis.

Nirmatrelvir/ritonavir is an effective antiviral therapy against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Use is not recommended in patients with end-stage renal disease (ESDR) due to a lack of data. We investigated the pharmacokinetics of nirmatrelvir/ritonavir (150 mg/100 mg twice a day) in four patients with ESRD undergoing hemodialysis. Nirmatrelvir peak concentrations ranged from 4,563 to 7,898 ng/mL and declined after hemodialysis. Concentrations were up to 4-fold higher but still within the range known from patients without ESRD, without accumulation of nirmatrelvir after the end of treatment.

SARS-CoV-2 viral load dynamics in immunocompromised critically ill patients on remdesivir treatment.

The relationship between SARS-CoV-2 quantitative viral load and risk of disease progression, morbidity such as long- COVID or mortality in immunosuppressed, remains largely undefined in COVID-19 patients. Critically ill immunosuppressed patients potentially benefit from remdesivir treatment because of the prolonged course of their infection. Four critically ill immunocompromised patients and the impact of remdesivir on viral dynamics in lower respiratory samples were studied. Bronchoalveolar lavage (BAL) samples were assessed to measure SARS-CoV-2 quantitative viral load using real-time PCR. Corresponding plasma levels of remdesivir and its metabolite GS-441524 were determined. Mean virus load of 39.74 x 107 geq/ml (±33.25 x 107 geq/ml) on day 1 dropped significantly (p<0.008) to 3.54 x 106 geq/ml (±6.93 x 106 geq/ml) on day 3 and to 1.4 x 105 geq/ml (±2.35 x 105 geq/ml) on day 5 of remdesivir treatment. Mean virus load dropped below <1% between day 1 and 5 of remdesivir treatment. Parent prodrug remdesivir and also GS441524 metabolite levels of antiviral activity in our patients were far in excess of EC 50. Our data present that remdesivir treatment potentially reduces the SARS-CoV-2 viral load in immunosuppressed critically ill patients. However, the implication of viral load reduction on morbidity and mortality needs further investigation.

Caffeine, Paraxanthine, Theophylline, and Theobromine Content in Human Milk.

This study aimed to assess the content of caffeine and its metabolites-paraxanthine, theophylline, and theobromine-in breast milk according to selected factors. Samples of human milk were collected from 100 women living in the east-northeast region of Poland. Information on the consumption of beverages and foods containing caffeine was collected using a 3 day food record. The determination of caffeine and its metabolite content was performed using liquid chromatography-mass spectrometry (LC-MS/MS). This study research showed that more caffeine was found in the milk of women living in cities, with secondary education, aged 34-43, and also in milk from the 3rd and 4th lactation periods (p ≤ 0.05). Factors such as place of residence, level of education, age, and stage of lactation influenced the nutritional choices of breastfeeding women, which had an impact on the level of caffeine and its metabolites in breast milk. A positive correlation was found between the consumption of caffeine with food and drinks and its level in human milk.

Isavuconazole therapeutic drug monitoring in critically ill ICU patients: A monocentric retrospective analysis.

BACKGROUND: The broad-spectrum triazole isavuconazole is used for the treatment of invasive aspergillosis and mucormycosis. Data regarding human plasma concentrations in clinical routine of the drug are rare. OBJECTIVES: Plasma concentrations of isavuconazole were determined in critically ill ICU patients while considering different patients' characteristics. METHODS: Retrospective analysis of isavuconazole plasma concentrations were obtained as part of routine therapeutic drug monitoring (TDM) of ICU patients with invasive aspergillosis or other fungal infections treated with isavuconazole. Plasma levels 0-4 h after last dosing were defined as peak levels (Cmax ), those 20-28 h after last dosing as trough levels (Cmin ). RESULTS: Overall, 223 isavuconazole levels of 41 patients were analysed, divided into 141 peak levels and 82 trough levels. The overall median Cmax was 2.36 µg/ml (mean 2.43 µg/ml, range 0.41-7.79 µg/ml) and the overall median Cmin was 1.74 µg/ml (mean 1.77 µg/ml, range 0.24-4.96 µg/ml). In total, 31.7% of the Cmin values of the total cohort were below the plasma target concentrations of 1 µg/ml, defined as EUCAST antifungal clinical breakpoint for Aspergillus fumigatus. Both peak and trough plasma levels of isavuconazole were significantly lower among patients with a body mass index (BMI) ≥25. In addition, a significant correlation was observed between isavuconazole trough levels and sepsis-related organ failure assessment (SOFA) score. CONCLUSIONS: This study shows that isavuconazole plasma concentrations vary in critical ill ICU patients. Significantly lower isavuconazole levels were associated with elevated BMI and higher SOFA score indicating a need of isavuconazole TDM in this specific patient population.

Methylprednisolone Plasma Concentrations During Cardiac Surgery With Cardiopulmonary Bypass in Pediatric Patients.

Introduction: To our knowledge, methylprednisolone pharmacokinetics and plasma concentrations have not been comprehensively investigated in children with congenital heart disease undergoing cardiac surgery with cardiopulmonary bypass. It is unknown whether there is a significant influence of cardiopulmonary bypass on the plasma concentrations of methylprednisolone and whether this may be an explanation for the limited reported efficacy of steroid administration in cardiac surgery with cardiopulmonary bypass. Methods: The study was registered in the Dutch Trial Register (NTR3579; https://www.trialregister.nl/trial/3428). Methylprednisolone 30 mg/kg was administered as an intravenous bolus after induction of anesthesia. Methylprednisolone concentration was measured with liquid chromatography tandem mass spectrometry and analyzed using linear mixed-effects modeling. Results: Thirty-nine patients were included in the study, of which three were excluded. There was an acute decrease in observed methylprednisolone plasma concentration on initiation of cardiopulmonary bypass (median = 26.8%, range = 13.9-48.14%, p < 0.001). We found a lower intercept (p = 0.02), as well as a less steep slope of the model predicted methylprednisolone concentration vs. time curve for neonates (p = 0.048). A lower intercept (p = 0.01) and a less steep slope (p = 0.0024) if the volume of cell saver blood processed was larger than 91 ml/kg were also found. Discussion: We report similar methylprednisolone plasma concentrations as earlier studies performed in children undergoing cardiopulmonary bypass, and we confirmed the large interindividual variability in achieved methylprednisolone plasma concentrations with weight-based methylprednisolone administration. A larger volume of distribution and a lower clearance of methylprednisolone for neonates were suggested. The half-life of methylprednisolone in our study was calculated to be longer than 6 h for neonates, 4.7 h for infants, 3.6 h for preschool children and 4.7 h for school children. The possible influence of treatment of pulmonary hypertension with sildenafil and temperature needs to be investigated further.

Penetration of Isavuconazole in Ascites Fluid of Critically Ill Patients.

Fungal peritonitis is a life-threatening condition which is not only difficult to diagnose, but also to treat. Following recent guidelines, echinocandins and azoles are the recommended antimycotics for the management of intra-abdominal Candida spp. infections, with a favor for echinocandins in critically ill patients. However, the new extended spectrum triazole isavuconazole also has a broad spectrum against Candida spp. Data on its target-site penetration are sparse. Therefore, we assessed isavuconazole concentrations and penetration ratios in ascites fluid of critically ill patients. Obtaining of Isavuconazole plasma and ascites fluid levels as well penetration ratios using paracentesis in critically ill patients. Isavuconazole concentrations were quantified in human plasma and ascites by a liquid chromatography/tandem mass spectrometry (LC-MS/MS) method. Isavuconazole concentrations in plasma and ascites fluid were measured in sixteen critically ill patients. Isavuconazol levels in ascites fluid (1.06 µg/mL) were lower than plasma levels (3.08 µg/mL). Penetration ratio was 36%. In two out of sixteen patients, Candida spp., in detail C. glabrata and C. tropicalis, could be isolated. Cmax/MIC Ratio in plasma of 560 for C. glabrata and 2166 for C. tropicalis could be observed. Following our results, isavuconazole penetrates into ascites. Successful treatment in Candida spp. peritonitis depends on pathogen susceptibility.

The nitrosamine contamination of drugs, part 3: Quantification of 4-Methyl-1-nitrosopiperazine in rifampicin capsules by LC-MS/HRMS.

Upon emergence of nitrosamines in various drugs, e.g in valsartan, metformin and ranitidine, 4-methyl-1-nitrosopiperazine (MeNP) was found in rifampicin in August 2020. Rifampicin is used, amongst others, for post-exposure prophylaxis of leprosy. The occurrence of MeNP can be explained by the synthesis, because 1-amino-4-methylpiperazine is concomitantly used with the organic oxidizing reagent isoamyl nitrite. According to a method reported by the FDA, the quantification of MeNP in rifampicin capsules was performed by LC-MS/HRMS. A significant contamination with MeNP was found in all samples, ranging from 0.7 to 5.1 ppm and exceeding the acceptable intake limit proposed by the FDA up to 32-fold. However, the severity of a possible leprosy infection outweighs the risks, which are concomitant with the intake of a single dose of rifampicin for post-exposure prophylaxis. Nevertheless, the extent of contamination is alarming, and countermeasures are needed to minimize public health risks. The presence of nitrosamines in rifampicin illustrates the need for better strategies in impurity profiling and compendial testing once again.

Central nervous system infections and antimicrobial resistance: an evolving challenge.

PURPOSE OF REVIEW: Antimicrobial resistance is an increasing threat to patients also in nosocomial central nervous system (CNS) infections. The present review focusses on optimizing intravenous treatment in order to achieve sufficient concentrations of antibiotics in the different compartments of the CNS when the causative pathogens have reduced sensitivity to antibiotics or/and the impairment of the blood-cerebrospinal fluid (CSF) and blood-brain barrier is mild. RECENT FINDINGS: Experience has been gathered with treatment protocols for several established antibiotics using increased doses or continuous instead of intermittent intravenous therapy. Continuous infusion in general does not increase the average CSF concentrations (or the area under the concentration-time curve in CSF) compared to equal daily doses administered by short-term infusion. In some cases, it is postulated that it can reduce toxicity caused by high peak plasma concentrations. In case reports, new ß-lactam/ß-lactamase inhibitor combinations were shown to be effective treatments of CNS infections. SUMMARY: Several antibiotics with a low to moderate toxicity (in particular, ß-lactam antibiotics, fosfomycin, trimethoprim-sulfamethoxazole, rifampicin, vancomycin) can be administered at increased doses compared to traditional dosing with low or tolerable adverse effects. Intrathecal administration of antibiotics is only indicated, when multiresistant pathogens cannot be eliminated by systemic therapy. Intravenous should always accompany intrathecal treatment.

Physiologically Based Population Pharmacokinetic Modeling Approach for Ciprofloxacin in Bone of Patients Undergoing Orthopedic Surgery.

Ciprofloxacin is highly active against bacteria that commonly cause bone infections. However, the time-course of ciprofloxacin in bone has not been characterized using population pharmacokinetic modeling. Thirty-nine patients received a 1-h infusion of 400 mg of ciprofloxacin before orthopedic surgery. Blood and bone samples were collected at 0.5 to 20 h following the start of the infusion. Bone samples were separated into cortical and cancellous bone and pulverized under liquid nitrogen using a cryogenic mill. Ciprofloxacin in plasma, and cortical and cancellous bone was quantified by liquid chromatography-tandem mass spectrometry. A physiologically based pharmacokinetic modeling approach was utilized to describe the concentration-time profiles in plasma and bone. Ciprofloxacin concentrations ranged from 0.176 to 5.98 mg/L (median, 1.67; density, 1.99 g/cm3) in cortical, and 0.224 to 14.6 mg/L (median, 1.22; 1.92 g/cm3) in cancellous bone. The average observed cortical bone/plasma concentration ratio was 0.67 at 0.5 to 2 h (n = 7) and 5.1 at 13 to 20 h (n = 9). For cancellous bone the respective average ratios were 0.77 and 4.4. The population PK model included a central (blood) compartment, two peripheral tissue compartments, and compartments for the organic and inorganic (hydroxyapatite) matrix in cortical and cancellous bone. The population mean ciprofloxacin clearance was 20.7 L/h. The estimated partition coefficients of the organic bone matrix were 3.39 for cortical and 5.11 for cancellous bone. Ciprofloxacin achieved higher concentrations in bone than plasma. Slow redistribution from bone to plasma may have been due to binding to the inorganic bone matrix. The developed model presents a step toward optimized antibiotic dosing in osteomyelitis.

A Novel Study Design Using Continuous Intravenous and Intraduodenal Infusions of Midazolam and Voriconazole for Mechanistic Quantitative Assessment of Hepatic and Intestinal CYP3A Inhibition.

The extent of a drug-drug interaction (DDI) mediated by cytochrome P450 (CYP) 3A inhibitors is highly variable during a dosing interval, as it depends on the temporal course of victim and perpetrator drug concentrations at intestinal and hepatic CYP3A expression sites. Capturing the time course of inhibition is therefore difficult using standard DDI studies assessing changes in area under the curve; thus, a novel design was developed. In a 4-period changeover pilot study, 6 healthy men received intraduodenal or intravenous infusions of the CYP3A substrate midazolam (MDZ) at a rate of 0.26 mg/h for 24 hours. This was combined with intraduodenal or intravenous infusion of the CYP3A inhibitor voriconazole (VRZ), administered at rates of 7.5 mg/h from 8 to 16 hours and of 15 mg/h from 16 to 24 hours, after starting midazolam administration. Plasma and urine concentrations of VRZ, MDZ, and its major metabolites were quantified by liquid chromatography-tandem mass spectrometry and analyzed by semiphysiological population pharmacokinetic nonlinear mixed-effects modeling. A model including mechanism-based inactivation of the metabolizing enzymes (maximum inactivation rate constant kinact , 2.83 h-1 ; dissociation rate constant KI , 9.33 µM) described the pharmacokinetics of VRZ well. By introducing competitive inhibition by VRZ on primary and secondary MDZ metabolism, concentration-time profiles, MDZ and its metabolites were captured appropriately. The model provides estimates of local concentrations of substrate and inhibitor at the major CYP3A expression sites and thus of the respective dynamic extent of inhibition. A combination of intravenous and intraduodenal infusions of inhibitors and substrates has the potential to provide a more accurate assessment of DDIs occurring in both gut wall and liver.

Sorafenib and everolimus in patients with advanced solid tumors and KRAS-mutated NSCLC: A phase I trial with early pharmacodynamic FDG-PET assessment.

BACKGROUND: Treatment of patients with solid tumors and KRAS mutations remains disappointing. One option is the combined inhibition of pathways involved in RAF-MEK-ERK and PI3K-AKT-mTOR. METHODS: Patients with relapsed solid tumors were treated with escalating doses of everolimus (E) 2.5-10.0 mg/d in a 14-day run-in phase followed by combination therapy with sorafenib (S) 800 mg/d from day 15. KRAS mutational status was assessed retrospectively in the escalation phase. Extension phase included KRAS-mutated non-small-cell lung cancer (NSCLC) only. Pharmacokinetic analyses were accompanied by pharmacodynamics assessment of E by FDG-PET. Efficacy was assessed by CT scans every 6 weeks of combination. RESULTS: Of 31 evaluable patients, 15 had KRAS mutation, 4 patients were negative for KRAS mutation, and the KRAS status remained unknown in 12 patients. Dose-limiting toxicity (DLT) was not reached. The maximum tolerated dose (MTD) was defined as 7.5 mg/d E + 800 mg/d S due to toxicities at previous dose level (10 mg/d E + 800 mg/d S) including leucopenia/thrombopenia III° and pneumonia III° occurring after the DLT interval. The metabolic response rate in FDG-PET was 17% on day 5 and 20% on day 14. No patient reached partial response in CT scan. Median progression free survival (PFS) and overall survival (OS) were 3.25 and 5.85 months, respectively. CONCLUSIONS: Treatment of patients with relapsed solid tumors with 7.5 mg/d E and 800 mg/d S is safe and feasible. Early metabolic response in FDG-PET was not confirmed in CT scan several weeks later. The combination of S and E is obviously not sufficient to induce durable responses in patients with KRAS-mutant solid tumors.