e19015 Background: CY-503 is a recombinant mistletoe lectin ( INN: aviscumine) leading to catalytical inactivation of ribosomes, thereby inhibiting protein synthesis. Dependent on dosage immunomodulatory and cytotoxic effects are induced. In patients (pts.) with stage IV metastatic melanoma CY-503 could potentially lead to an antitumoral response. Methods: Pts. with stage IV metastatic melanoma (ECOG performance score 0-1), who had failed from one or more previous systemic chemotherapies, were eligible in this trial. Pts. with severe organ dysfunction or brain metastasis were excluded. Pts. received CY-503 350ng SC twice weekly for a minimum of 8 weeks (2 cycles). Pts. without disease progression were continued. The primary study endpoints were progression free survival (PFS) and overall survival (OS). Tumor assessment included CT and MRI scans every 8 weeks, which were centrally reviewed. Results: Between April 2008 and June 2009, 32 pts. were enrolled, one being ineligible. 31 pts. (16m; 15f) with a median age of 67 years (range: 20-85y) were evaluable. Pts. were belonging to the following stage IV categories: M1a/b/c = 4/5/22, and median number of prior treatments was 2 (range: 1- 6), 7 pts. (23%) were pretreated with ≥ 3 lines. (84%) received at least 8 weeks of treatment; 4 pts. withdrew prior to 8 weeks due to clinically overt disease progression, and 1 pt. due to a severe adverse event (SAE). At 8 weeks 11 pts. with stable disease (SD) were observed (disease control rate: 36% ITT) with a median PFS of 3.75 months (95% CI: 3,6-3,9 months). The median PFS in the ITT population was 2.0 months (95% CI: 1.9–2.1 months). Median OS had not been reached at the time of evaluation. Local reactions at the injection site were the main toxicities. Two severe adverse events with fatal outcome (1 pt. with cardiac arrhythmia, 1 pt. with stroke) could not clearly be related to the study medication. Conclusions: CY-503 showed a modest clinical activity in pts. with advanced melanoma. Side effects were mostly limited to local reactions at injection site. Given this heavily pretreated population, further investigations on the use of CY-503 in the treatment of malignant melanoma might be warranted. Author Disclosure Employment or Leadership Position Consultant or Advisory Role Stock Ownership Honoraria Research Funding Expert Testimony Other Remuneration Cytavis Biopharm GmbH Cytavis Biopharm GmbH Cytavis Biopharm GmbH Cytavis Biopharm GmbH

Background: Lectins, carbohydrate proteins, bind to glycoconjugates of all mammalian cells, including cancer cells. Aberrant glycosylation, detected by lectin histochemistry, can predict outcome in some tumour entities. One such lectin is aviscumine (recombinant mistletoe lectin). Aviscumine has cytotoxic effects and can therefore be used as anti-tumour therapy. Materials and Methods: Lectin histochemistry with aviscumine was performed on primary tumour sections from resected adenocarcinoma of the lung. Staining results were then correlated with the clinical course of the patients. Results: Most of the adenocarcinomas (92.5%) bound aviscumine. Kaplan-Meier analysis revealed no correlation between aviscumine binding and progression-free survival or overall survival. Conclusion: These results suggest that for the selected group of patients with adenocarcinoma of the lung aviscumine binding activity can not serve as a prognostic factor. More strikingly, however, aviscumine binds to malignant cells in 92.5% of the patients. This is an indicator for the use of aviscumine as a possible target for tumour therapy. Lectins are carbohydrate-binding proteins of nonimmunological origin, which often multivalently bind to carbohydrate residues of glycoconjugates. This study focuses on the mistletoe lectins (MLs) derived from the European mistletoe (Viscum album). Three different MLs (ML-I, MLII, ML-III) have been isolated and all three belong to the type-2 ribosome-inactivating proteins (RIPs), and are thus highly cytotoxic. ML-I is one of the best characterised lectins and has attracted considerable interest as it was shown to be the main biologically active component of aqueous mistletoe extracts (2-4). In addition to its cytotoxicity, ML-I displays a broad spectrum of immunomodulatory effects, which are additionally thought to be of benefit to cancer patients treated with aqueous mistletoe extracts (4). Because of these broad anti-tumour actions, ML-I has now been made available as a recombinant mistletoe lectin (rML) called rViscumin or aviscumine (INN). Aviscumine differs from the plantderived ML-I, mainly in its lack of glycosylation, while the other features, such as carbohydrate-binding specificity and cytotoxicity, are identical for plant-derived ML-I and aviscumine (5, 6). Aviscumine has successfully passed phase I clinical trial and is presently being considered for phase II studies (7). Type-2 RIPs consist of two protein chains, the B-chain which contains the carbohydrate-recognising domain (i.e. the lectin activity) and the A-chain, which exhibits rRNAN-glycosidase activity (1). For the cytotoxic effect of avisumine, the B-chain initially has to bind to the carbohydrate coat of the cell, the glycocalyx. By a so-far unknown mechanism, the aviscumine molecule crosses the lipid bilayer of the cell membrane in the endosome/ lysosome compartment and enters the cytoplasm, where the A-chain irreversibly inactivates the ribosomes. As surface 3303 Abbreviations: MLs, mistletoe lectins; NSCLC, non-small cell lung cancers; RIP, ribosome-inactivating proteins; rML, recombinant mistletoe lectin; SC, Spearman coefficient. Correspondence to: Katharina Blonski, Department of Anatomy II: Experimental Morphology, University Hospital HamburgEppendorf, Martinistrasse 52, D-20246 Hamburg, Germany. Tel: 0049-40-42803-3587, Fax: 0049-40-42803-5427, e-mail: k.blonski@ uke.uni-hamburg.de

An immuno-polymerase chain reaction (IPCR) assay is used to evaluate the kinetic behaviour of the novel anti-cancer drug Aviscumine in plasma samples taken from 41 patients during a 3-year clinical trial. The ultrasensitive IPCR assay employed the amplification of a detection-antibody linked marker-DNA and an internal competitor DNA for standardization, thus enabling the detection of the antigen in concentrations far below the detection limit of conventional enzyme-linked immuno-sorbent assay (ELISA). The quantification of Aviscumine was carried out using external calibration curves obtained from individual patient plasma samples, collected previous to the administration of Aviscumine, which were spiked with known amounts of the reference substance Aviscumine. Additional controls were measured containing standardized human serum spiked with Aviscumine to assure the continuous general reproducibility of the assay as well as to estimate differences between individual patients. Average recovery was found to be 95+/-19% and the average deviation in precision of the assay was determined to be 9+/-5%. Data for the quantification of Aviscumine were obtained from all patient samples investigated with the exception of a single patient. The collected data provided the basis for the valid routine quantification of patient samples for the calculation of the pharmacokinetic behaviour of Aviscumine in patient plasma.

Background: Lectins, carbohydrate proteins, bind to glycoconjugates of all mammalian cells, including cancer cells. Aberrant glycosylation, detected by lectin histochemistry, can predict outcome in some tumour entities. One such lectin is aviscumine (recombinant mistletoe lectin). Aviscumine has cytotoxic effects and can therefore be used as anti-tumour therapy. Materials and Methods: Lectin histochemistry with aviscumine was performed on primary tumour sections from resected adenocarcinoma of the lung. Staining results were then correlated with the clinical course of the patients. Results: Most of the adenocarcinomas (92.5%) bound aviscumine. Kaplan-Meier analysis revealed no correlation between aviscumine binding and progression-free survival or overall survival. Conclusion: These results suggest that for the selected group of patients with adenocarcinoma of the lung aviscumine binding activity can not serve as a prognostic factor. More strikingly, however, aviscumine binds to malignant cells in 92.5% of the patients. This is an indicator for the use of aviscumine as a possible target for tumour therapy.

Mistletoe lectin I (ML-I) is a type II ribosome-inactivating protein, which inhibits the protein biosynthesis at the ribosomal level. ML-I is composed of a catalytically active A-chain with rRNA N-glycosidase activity and a B-chain with carbohydrate binding specificities. Using comparative solid-phase binding assays along with electrospray ionization tandem mass spectrometry, ML-I was shown to preferentially bind to terminally alpha2-6-sialylated neolacto series gangliosides from human granulocytes. IV(6)Neu5Ac-nLc4Cer, VI(6)Neu5Ac-nLc6Cer, and VIII(6)Neu5Ac-nLc8Cer were identified as ML-I receptors, whereas the isomeric alpha2-3-sialylated neolacto series gangliosides were not recognized. Only marginal binding of ML-I to terminal galactose residues of neutral glycosphingolipids with a Galbeta1-4Glc or Galbeta1-4GlcNAc sequence was determined, whereas a distal Galalpha1-4Gal, GalNAcbeta1-3Gal, or GalNAcbeta1-4Gal disaccharide did not bind at all. Among the glycoproteins investigated in Western blot and microwell adsorption assays, only those carrying Neu5Acalpha2-6Galbeta1-4GlcNAc residues, exclusively, predominantly, or even as less abundant constituents in an assembly with Neu5Acalpha2-3Galbeta1-4GlcNAc-terminated glycans, displayed high ML-I binding capacity. From our data we conclude that (i) ML-I has to be considered as a sialic acid- and not a galactose-specific lectin and (ii) neolacto series gangliosides and sialoglycoproteins with type II glycans, which share the Neu5Acalpha2-6Galbeta1-4GlcNAc terminus, are true ML-I receptors. This strict preference might help to explain the immunostimulatory potential of ML-I toward certain leukocyte subpopulations and its therapeutic success as a cytotoxic anticancer drug.

The anticancer drug rViscumin, currently under clinical development, has been shown in previous studies to be a sialic acid specific ribosome inactivating protein (RIP). Comparative binding assays with the CD75s-specific monoclonal antibodies HB6 and J3-89 revealed rViscumin to be a CD75s-specific RIP due to identical binding characteristics toward CD75s gangliosides. The receptor gangliosides are IV6nLc4Cer, VI6nLc6Cer, and the newly characterized ganglioside VIII6nLc8Cer, all three carrying the Neu5Acalpha2-6Galbeta1-4GlcNAc motif. To elucidate the clinical potential of the rViscumin targets, CD75s gangliosides were determined in several randomly collected gastrointestinal tumors. The majority of the tumors showed an enhanced expression of CD75s gangliosides compared with the unaffected tissues. The rViscumin binding specificity was further investigated with reference glycoproteins carrying sialylated and desialylated type II N-glycans. Comparative Western blots of rViscumin and ricin, an rViscumin homologous but galactoside-specific RIP, revealed specific recognition of type II N-glycans with CD75s determinants by rViscumin, whereas ricin failed to react with terminally sialylated oligosaccharides such as CD75s motifs and others. This strict binding specificity of rViscumin and the increased expression of CD75s gangliosides in various tumors suggest this anticancer drug as a promising candidate for an individualised adjuvant therapy of human tumors.

Pulmonary delivery is an attractive alternative route to deliver protein drugs that are currently delivered by injection. Inhalation therapy via nebulizers is a well accepted way for pulmonary application of proteins considering the formulation difficulties of MDIs or DPIs. This research presents the effect of variable excipients on the stability and aerosol performance of freeze-dried aviscumine after reconstitution and nebulization. Aviscumine formulations containing 100 mmol/L Tris buffer, 0.1% (w/v) Polysorbate 80, 0.01% (w/v) Na(2)-EDTA and 8% (w/v) Hydroxyethyl starch have been lyophilized and reconstituted with a buffered isotonic solution pH 8. The aviscumine activity was determined by a binding assay directly after reconstitution and after nebulization with a PariBoy air-jet nebulizer, a Multisonic and a Systam ultrasonic nebulizer. The stabilization of aviscumine by the addition of variable buffer salts to the reconstitution medium, such as 50, 100, and 200 mmol/L Tris buffer, 20 and 100 mmol/L phosphate buffer, and 20 and 100 mmol/L Tricine buffer, was studied. About 50% of aviscumine activity was lost after 20 min nebulization time without any additives. Nevertheless, higher buffer concentrations confer greater stability. About 70% of the aviscumine activity could be retained by the addition 0.03% octanoyl-N-methylglucamide and 100 mmol/L Tricine to the reconstitution medium.

The pulmonary drug delivery of proteins present an alternative to parenteral and oral administration. Nebulization of aqueous protein solutions is an ideal method for pulmonary application of therapeutic proteins considering the difficulties of their formulation as MDIs or DPIs. This research presents the effect of variable excipients on the stability of freeze-dried aviscumine after reconstitution and nebulization. Formulations containing different lyoprotectants have been lyophilized and reconstituted with isotonic salt solution. The loss of aviscumine activity in the nebulizer reservoir and after nebulization with a PariBoy air-jet nebulizer, a Multisonic ultrasonic nebulizer and a Systam ultrasonic nebulizer was determined by a binding assay. The effect of variable lyoprotectants such as 8% (w/v) Dextran T1, HES130, HES450, HP-beta-CD and 6% (w/v) HES450 plus 2% (w/v) mannitol on the stability of aviscumine to air-jet and ultrasonic nebulization has been evaluated. Only 50% of aviscumine activity was retained after 20 min nebulization, where 8% (w/v) HES450 was shown to be the best stabilizer. Stabilization of aviscumine by the addition of variable surfactants as 0.01 and 0.1% (w/v) Poloxamer 188, 0.03 and 0.1% (w/v) PEG 8000, 0.03 and 0.1% (w/v) Solutol HS15 and 0.03 and 0.1% (w/v) octanoyl-N-methyl-glucamide to the reconstitution solution has also been studied. By the addition of 0.03% (w/v) octanoyl-N-methyl-glucamide, 70% of the activity was retained after 20 min nebulization.

To allow for pharmacokinetic studies in adjunction with the current clinical developments of the potent cytostatic anti-cancer drug rViscumin, a sandwich immuno-PCR (IPCR) assay was developed for the detection of rViscumin in blood plasma. The IPCR was carried out with a commercially available reagent kit, consisting of pre-assembled rViscumin-specific antibody–DNA conjugates as well as a specific competitor DNA fragment to be amplified by PCR. Various combinations of capture- and detection-antibodies were compared for performance in IPCR. Using the optimized assay, as few as 50 zeptomol (approx. 100fg/ml) rViscumin (MW 57kDa) was detectable in standardized human serum samples. The IPCR assay was very selective for rViscumin and in spiking experiments in proband plasma samples, signal recovery rates between 70% and 120% were obtained. The linear sensitivity range of the assay covered more than five orders of magnitude. Repeated measurements of rViscumin resulted in a mean standard deviation value of 14.2%.

Nebulization of aqueous drug solutions is a suitable delivery system for pulmonary application of proteins because it can easily produce droplets small enough to reach the alveolar region. However, proteins are sensitive to nebulization. Therefore, stabilizers need to be added which on the other hand influence the aerosol performance, such as average droplet size or mass output. This research presents the effect of various cryoprotectants such as Na-polyphosphate, CaCl(2) x 6H(2)O and MgSO(4) x 7H(2)O on the stability and aerosol performance of freeze-dried aviscumine after reconstitution and nebulization using three different nebulizers. Formulations containing Tris-buffer, polysorbate 80, Na(2)-EDTA and HES450 were lyophilized and reconstituted with a buffered isotonic solution containing 100 mmol/l Tricine-buffer pH 8, 0.03% (w/v) octanoyl-N-methylglucamide, 150 mmol/l NaCl and a cryoprotectant. The aviscumine activity was determined by a binding assay. The addition of 0.2% Na-polyphosphate to the reconstitution medium led to retention of approx. 73% of the aviscumine activity after 20 min nebulization with the Systam ultrasonic nebulizer. It has been observed that 84 and 72% of the activity were retained by the addition of 10 mmol/l CaCl(2) x 6H(2)O using PariBoy air-jet and Multisonic ultrasonic nebulizer, respectively. In addition, a decrease in the mean droplet size with increasing the cryoprotectant concentration has been observed. A relationship between the average droplet size, surface tension and viscosity depending on the used cryoprotectant type and concentration could be established.

The aim of this research was to ascertain whether dextrans with low molecular weight will stabilize aviscumine. During freeze-drying increasing concentrations of dextran T1 (MW 1000) stabilized aviscumine. Eight percent of dextran resulted in a nearly 100% recovery of the activity and in addition a complete amorphous structure of the solid phase was obtained. By decreasing the molecular weight of the dextran from 75 to 1 kDa, the protein activity was increased by 20% in the lyophilisate. Combinations of dextran with either trehalose or mannitol showed no additional effects on stability. The improved stabilization of aviscumine using low molecular weight dextrans is explained by an increased interaction between the protein and the dextran molecules (like hydrogen bonds), whereas they are sterically hindered if larger dextran molecules are used. When the protein concentration was increased from 10 to 100 microg/ml (in formulas with 8% dextran T1), no influence on the protein activity could be found. With regard to the carbohydrate-binding activity of the protein, it was shown that the optimal content of residual water in the lyophilisate should be about 2%. Above and below this percentage a destabilization of the protein was observed. The often discussed failure of dextran as a stabilizing excipient in the freeze-drying of proteins seems to be a question of the selection of the correct molecular weight.

A method for the analysis of the rViscumin heterodimer (recombinant mistletoe lectin) based on two-dimensional (2-D) poiyacrylamide gel electrophoresis and mass spectrometry was developed and used for quality control concerning purity and homogeneity of the recombinant protein processed under Good Manufacturing Practice (GMP) conditions. A series of spots with different p/-values in the pH-gradient of both rViscumin A- and B-chain were observed independently from the experimental conditions like urea concentration, heat treatment or the use of cysteine alkylating agents. Comparative studies of the major spots using matrix assisted laser desorption/ionization-mass spectrometry (MALDI-MS), liquid chromatography-electrospray ionization (LC-ESI)-MS and LC-ESI-tandem MS (MS/MS) after tryptic in-gel digestion resulted in a sequence coverage of 92% for the A-chain and 95% for the B-chain. No molecular differences like common chemical or post-translational modifications or nonenzymatic deamidation were found to cause the different charge values of the separated spots. Therefore, these protein spots were extracted from the 2-D gel and separated again by 2-D gel electrophoresis (termed Re-2-DE). Each of the single spots tested in the Re-2-DE experiment split up in the same heterogeneous pattern concerning the pl-values. We suggest that the observed charge variants of rViscumin are the result of conformational protein variants, existing in an equilibrium during sample preparation and/or isoelectric focusing and are not caused from microheterogeneity in the primary structure of rViscumin.

Mistletoe lectin I (pML) and its isoforms ML II and III constitute the active principle in extract preparations from mistletoe, commonly used as immunomodulator in adjuvant tumour therapy. The heterodimeric disulfide-linked cytotoxic protein is classified as type II ribosome inactivating protein (RIP). Recently, the sequence coding for the mistletoe lectin prepro-protein was identified and the existence of a single intron-free gene was shown [Eck, J., Langer, M., Möckel, B., Baur, A., Rothe, M., Zinke, H. & Lentzen, H. (1999) Eur. J. Biochem. 264, 775-784]. The aim of this study was to prepare pure and homogeneous rMLB-chain as well as rML heterodimer for studying the carbohydrate binding specificity of recombinant versus natural protein and its contribution to the observed cytotoxic effect. Expression in E. coli resulted in the production of insoluble protein (inclusion bodies). A procedure for generating correctly folded, biochemically and biologically active rMLB was established starting from the insoluble single chain. Carbohydrate binding and specificity of pMLB and rMLB were analysed by a competitive enzyme linked lectin assay (ELLA). Asialofetuin was able to compete with binding of both chains (50% at 0.8 microM). The specificity of the B-chains to lactose was more distinct with halfmaximal competition at 4.9 mM (pMLB) and > 90 mM (rMLB), respectively. Furthermore, in a coassociation process rMLA- and rMLB inclusion bodies were associated in one step by defined dilution yielding active rML-heterodimer. The activities of recombinant (rML) and plant derived mistletoe lectin (pML) were compared. Cytotoxicity was determined using MOLT-4 cells and enzymatic rRNA N-glycosidase activity was measured in a coupled transcription/translation assay. The IC50 values of the two heterodimers were similar in both assays; rMLB-chain did not show any cytotoxic effect. In the ELLA with lactose as a competitor 50% competition of binding to asialofetuin was achieved at 1.6 mM (rML) and 1.8 mM (pML). Hence, using three different assays we found no significant differences between the recombinant protein and the glycosylated form of ML. Comparing the biological activities of the single chains with those of the heterodimer we conclude, that both, lectin activity and the rRNA N-glycosidase activity, are prerequisites for the cytotoxic effects on target cells.

It could be shown from several experiments that carbohydrate-binding mistletoe lectins represent the pharmacologically active constituents of mistletoe extracts. On the basis of these findings, it was possible to develop an extract preparation standardized with respect to the mistletoe lectin concentration. This drug is the first mistletoe preparation that fulfills the criteria of the guidelines for the development of drugs (1) regarding its quality and stability of the active ingredients under certain storage conditions. The quality of this preparation has also been shown in several animal models to demonstrate antitumoral potencies.