Subcutaneous Administratio

Home , Catumaxomab, Trifunctional antibody

Published OnlineFirst June 10, 2015; DOI: 10.1158/1535-7163.MCT-15-0156

Large Molecule Therapeutics Molecular Cancer Therapeutics Efﬁcacy and Tolerability of a GD2-Directed Trifunctional Bispeciﬁc Antibody in a Preclinical Model: Subcutaneous Administration Is Superior to Intravenous Delivery Nina Deppisch1, Peter Ruf2, Nina Eissler1, Frauke Neff3, Raymund Buhmann4, Horst Lindhofer2, and Ralph Mocikat1

Abstract

Trifunctional bispecific antibodies (trAb) are novel antican- the standard application route for therapeutic antibodies. cer drugs that recruit and activate different types of immune Despite lower plasma levels after subcutaneous administration, effector cells at the targeted tumor. Thus, tumor cells are the same tumor-protective potential was observed in vivo com- effectively eliminated and a long-lasting tumor-specificT-cell pared with intravenous administration of Surek. However, memory is induced. The trAb Ektomab is directed against subcutaneously delivered Surek showed better tolerability. This human CD3 on T cells and the tumor-associated ganglioside could be explained by a continuous release of the antibody GD2, which is an attractive target for immunotherapy of leading to constant plasma levels and a delayed induction of melanoma in humans. To optimize clinical applicability, we proinflammatory cytokines. Importantly, the induction of studied different application routes with respect to therapeutic counter-regulatory mechanisms was reduced after subcutane- efficacy and tolerability by using the surrogate trAb Surek (anti- ous application. These findings are relevant for the clinical GD2 anti-murine CD3) and a murine melanoma engineered application of trifunctional bispecific antibodies and, possibly, to express GD2. We show that subcutaneous injection of the also other immunoglobulin constructs. Mol Cancer Ther; 14(8); trAb is superior to the intravenous delivery pathway, which is 1877–83. 2015 AACR.

Introduction simultaneously stimulates accessory cells via activating Fcg recep- tors (2). Thereby, trAbs recruit and activate different types of Although treatment of cancer has substantially improved dur- immune effector cells at the targeted tumor where they mediate ing the past years, the prognosis of many malignancies is still effective tumor cell destruction (3). The interplay between T cells, poor. Recent developments in the field of immunotherapy, how- tumor cells, and accessory cells mainly leads to the induction of a ever, may pave the way for new therapeutic approaches. Trifunc- Th1 response, which is a prerequisite for potent tumor rejection tional bispecific antibodies (trAb) are promising reagents that (4, 5). Furthermore, it has been shown that antigen-presenting harness the immune system to reject cancer (1). These novel cells recruited to the tumor site subsequently present the phago- therapeutic antibodies consist of two different binding arms, cytosed and processed tumor-associated antigens to T cells (2). which are directed against a tumor-associated antigen and CD3 This leads to the induction of tumor-specific T cells and a long- on T cells, respectively. In addition, they are endowed with an lasting immunologic memory, which was not achieved by using intact Fc region (comprising a rat and a mouse moiety) that 0 the corresponding bispecific F(ab )2 fragment (6, 7). Therefore, trAbs not only eliminate tumor cells directly, but also exert a long- 1Helmholtz-Zentrum Munchen,€ Institut fur€ Molekulare Immunologie, term vaccination effect. This unique characteristic enables a long- Munich, Germany. 2Trion Research GmbH, Martinsried, Germany. 3Helm- lasting antitumor response and offers new perspectives for anti- holtz-Zentrum Munchen,€ Institut fur€ Pathologie, Munich, Germany. cancer immunotherapy. 4 € € Ludwig-Maximilians-Universitat Munchen, Klinikum Großhadern, Med- fi izinische Klinik III und Abteilung fur€ Transfusionsmedizin, Munich, Catumaxomab, which has dual speci city for epithelial cell Germany. adhesion molecule (EpCAM) and CD3, is the first trAb routinely Note: Supplementary data for this article are available at Molecular Cancer used in the clinic. It was approved in 2009 for the treatment of Therapeutics Online (http://mct.aacrjournals.org/). malignant ascites resulting from EpCAM-expressing tumors (8, 9). fi Corresponding Authors: Ralph Mocikat, Institut fur€ Molekulare Immunologie, Other trAbs targeting different speci cities are being developed and fi Helmholtz-Zentrum Munchen,€ Deutsches Forschungszentrum fur€ Gesundheit have already shown therapeutic bene ts for cancer patients in und Umwelt, Marchioninistr. 25, Munich D-81377, Germany. Phone: 4989-3187- clinical studies. A promising trAb, named Ektomab, is directed 1302; Fax: 4989-3187-1300; E-mail: [email protected]; and against human CD3 on T cells and the tumor-associated gangli- Horst Lindhofer, Trion Research GmbH, Am Klopferspitz 19, D-82152 Martinsried, oside GD2 (10). This cell-surface molecule is an attractive target for Germany. Phone: 4989-700766-24; Fax: 4989-7007-6611; E-mail: cancer immunotherapy in humans because it is strongly expressed [email protected] on small-cell lung cancer and on malignancies of neuroectodermal doi: 10.1158/1535-7163.MCT-15-0156 origin such as neuroblastoma, glioma, and melanoma, while 2015 American Association for Cancer Research. it shows a highly restricted pattern of expression in normal tissue

www.aacrjournals.org 1877

Deppisch et al.

(11, 12). Using the B16F0-derived murine melanoma cell line B78- anti-Ki-67 (SolA15; eBioscience) and CTLA-4 (UC10-4B9; Bio- D14, which is engineered to express GD2 (13), and trAb Surek Legend) was carried out without stimulation. Cells were fixed and (anti-GD2 anti-murine CD3) as a surrogate construct for Ekto- permeabilized (eBioscience) according to the manufacturer's mab, we previously demonstrated that this approach is highly instructions and subsequently analyzed by FACS using an LSRII effective in terms of eliminating melanoma cells (6, 14, 15). flow cytometer (BD Biosciences). A critical factor that may determine the antitumor efficacy as well as possible adverse side effects exerted by trAbs is the appli- Weight monitoring cation route. Several clinical trials indicated that both intraperi- After premonitoring, mice were treated as mentioned under the toneal (8, 9, 16, 17) and intravenous (18–20) administration is heading "animals." Then, body weight was monitored on a daily feasible. However, we argued that for broader clinical applicability basis over a period of 7 days by using the laboratory balance of trAbs, the subcutaneous delivery may offer several benefits, for Quintix 5101-1S (Sartorius). example, a shortened application time (minutes vs. hours) and a further improved safety profile. Therefore, we directly compared Histopathology and immunohistochemistry intravenous administration, which has been mostly used in the After shaving mice, 50 mg Surek was injected subcutaneously. clinic, and subcutaneous application. As this regimen will be The skin was continuously monitored macroscopically. In addi- particularly relevant for future trAb-mediated treatment of mela- tion, formalin-fixed skin biopsies (2 and 7 days after trAb injec- noma, we selected the GD2-targeting trAb Surek for preclinical tion) were processed and embedded in paraffin. For histologic evaluation. Here, we show that trAb-mediated therapy can indeed analysis, 4 mm (for hematoxylin and eosin stain) and 1 mm (for be optimized by choosing the appropriate application route. immunohistochemistry) sections were made from paraffin blocks. Slides were then stained with hematoxylin/eosin or immunohistochemical evaluation was done by using antibodies Materials and Methods against CD3 (Dianova), Mac3 (DCS), and B220 (BD Biosciences). Cell lines and tumor model The GD2-positive B78-D14 mouse melanoma cell line (kindly Measurement of plasma concentrations of antibodies provided by J.C. Becker, Julius-Maximilians-Universit€at, Wurzburg,€ After injecting Surek (50 mg) intravenously and subcutaneous- Germany) was derived from the B16F0 melanoma by transfection of ly, respectively, blood samples were taken at different time points genes coding for the GD3 and GD2 synthases as described previ- and sera were stored at 20C. Surek plasma concentrations were ously (13). B78-D14 cells were cultured in RPMI1640 medium measured by ELISA. Briefly, Surek was captured by an anti-rat supplemented with 8, 9% FCS, 2 mmol/L L-glutamine, 0.4 mg/mL IgG2b antibody (RG7/11.1; BD Biosciences) and detected via a G418, 0.5 mg/mL hygromycin B, sodium pyruvate, and nonessen- biotin-labeled anti-mouse IgG2a antibody (R19-15; BD Bio- tial amino acids. Before in vivo application, cells were extensively sciences). Subsequently, streptavidin-b-galactosidase (Roche washed in PBS. On a regular basis, the identity of the cell line was Diagnostics) and the substrate chlorophenol red-b-D-galactopyr- confirmed by morphology, in vivo growth behavior, and antigen anoside (Roche Diagnostics) were added. The colorimetric reac- expression. tion was measured at 570 nm. Surek concentration was calculated by interpolation on a standard curve. TrAb construct To measure anti-antibody induction, blood samples were col- The trifunctional bispecific antibody Surek is derived from the lected at day 15, 30, and 45 after injection. The titers of mouse anti- parental antibodies 17A2 (anti-mouse CD3, rat IgG2b) and mouse antibodies (MAMA) as well as mouse anti-rat antibodies Me361 (anti-GD2, mouse IgG2a; ref. 10). Surek was generated (MARA) were determined by ELISA. Briefly, anti-antibodies in by quadroma technology and purified by affinity and ion serially diluted sera were captured by parental antibody 17A2 (21) exchange chromatography (14). for analyzing MARAs or Me361 (22) for analyzing MAMAs. Bound anti-antibodies were detected by a biotinylated goat-anti mouse Animals IgG (HþL) antibody (MARAs) or a biotinylated goat-anti mouse Female C57BL/6 mice, purchased from Taconic and kept under IgG1 antibody (MAMAs), both obtained from SouthernBiotech. specific pathogen-free conditions in our animal facility, were used Then, streptavidin-b-galactosidase and substrate were added and at the age of 9–12 weeks in groups of 5 animals. Mice were injected optical density was measured at 570 nm. Titers were calculated as intraperitoneally with 1 105 B78-D14 cells and 50 mg Surek those reciprocal serum dilutions that yielded an extinction exceed- intravenously or subcutaneously at the indicated time points. ing the mean background signal by 3 SD. Control groups received only 1 105 B78-D14 cells or no treatment. All experiments were in accordance with relevant reg- Cytokine quantification ulations and have been approved by Regierung von Oberbayern. For analyzing cytokine production, mice were treated as mentioned before. One, 3, 24, and 48 hours after treatment, blood T-cell phenotyping by FACS samples were taken. Cytokine levels in sera of mice were analyzed For analyses of T cells, mice were treated as mentioned before. by using the Bio-Plex cytokine assay system (Bio-Rad). Two and 3 days after treatment, splenocytes were analyzed by staining with directly labeled monoclonal antibodies against CD4 Statistical analysis (RM4-5; BD Biosciences), CD8 (53-6.7; eBioscience), and CD69 GraphPad Prism software version 5.01 (GraphPad Software) (H1.2F3; BD Biosciences). Intracellular staining with anti-IFNg was used for generation of survival curves, data plots, and statis- (XMG1.2; eBioscience) and anti-IL10 (JES5-16E3; eBioscience) tical calculations. Statistical analyses were done by using the was performed after 4 hours of stimulation with PMA/ionomycin unpaired, two-tailed Student t test or the log-rank test. P values and Brefeldin A (eBioscience), whereas intracellular staining with < 0.05 were considered as statistically significant.

1878 Mol Cancer Ther; 14(8) August 2015 Molecular Cancer Therapeutics

Optimization of trAb-Mediated Tumor Therapy

Results Therapeutic efficacy of subcutaneously delivered Surek As already shown, the in vivo efficacy of trAbs is reflected by T- cell activation in vivo and a predominant release of Th1 cytokines, particularly IFNg, following trAb injection (6). To evaluate the efficacy of different application routes, we examined T-cell activation and proliferation after delivery of Surek intravenously and þ þ subcutaneously. Both CD8 and CD4 T cells showed a strong upregulation of CD69 and Ki-67 in mice treated with B78-D14 and Surek as compared with control mice (Fig. 1A and B), indicating trAb-induced T-cell activation and proliferation. Figure 2. Importantly, no differences were seen between intravenous and Therapeutic effectiveness of trAb Surek. After injection of 105 B78-D14 tumor subcutaneous application. Likewise, after intravenous and sub- cells intraperitoneally (i.p.) at day 0, mice either received three injections of þ cutaneous treatment, CD8 cells were equally capable of produc- 50 mg Surek (days 2, 7, and 11) or two injections of 50 mg Surek (days 2 and 7) or no therapy. Surek was administered intravenously (i.v.) and ing IFNg 2 days after treatment (Fig. 1C). subcutaneously (s.c.), respectively. Each group comprised of 5 to 10 mice. The results encouraged us to compare the tumor-protective Survival of mice treated with Surek was significantly prolonged in comparison potential of Surek subcutaneously and Surek intravenously, with the tumor control (log-rank test, , P < 0.01; , P < 0.001, respectively). respectively, in vivo (Fig. 2). Subcutaneous and intravenous The group injected three times intravenously had to be euthanized treatment showed the same antitumor efficacy with a median immediately due to trAb-related toxicity after the third trAb injection. As survival of >100 days when Surek was injected twice (day 2 and demonstrated previously, the nonspecific trifunctional control antibody TRBs01 revealed no therapeutic effect (14). 7 after B78-D14 challenge). As already published, a trAb with an irrelevant tumor specificity (anti-HER-2 anti-CD3), which did not bind to B78-D14 cells, was completely ineffective (14). It is was further increased up to 85% when Surek was administered established that the potential of trAbs depends on the number three times (day 2, 7, and 11) after tumor challenge, while the of treatment cycles but increasing the number of intravenous intravenous group had to be euthanized immediately after the injections is often limited by possible side effects. Therefore, we third antibody injection due to severe adverse effects, which asked whether the subcutaneous route allows to extend the were most likely caused by aggregation of Surek and anti-drug treatment cycles and to ameliorate the therapeutic effect. antibodies (see below). This suggests a better tolerability of Indeed, the overall survival after subcutaneous administration Surek subcutaneously versus intravenously.

Local and systemic tolerability of Surek subcutaneously For using subcutaneous injection of trAbs as an appropriate alternative to conventional intravenous application in a clinical setting, it is pivotal to provide data not only on efficacy, but also on safety. As subcutaneous trAb administration could be associated with inflammatory reactions at the injection site, we performed a close skin monitoring before and after Surek administration. One, 24, 48 hours, as well as 7 days after trAb administration, no notable skin reaction was observed, indicating that Surek subcutaneously was not associated with significant adverse events at the injection site. These observa- tions were verified by histo- and immunopathologic analyses 2 and 7 days after subcutaneous administration of Surek. All skin biopsies showed a normal distribution of immune cells (data not shown). Taken together, no inflammatory reactions or other alterations were detected after subcutaneous administration of trAb Surek. Systemic tolerability of subcutaneous trAb treatment is of special interest for clinical use. In mice, weight loss is used as a sensitive marker for health monitoring (23). Therefore, we Figure 1. Activation and proliferation of T cells after treatment with trAb Surek. Mice performed a close weight monitoring following intraperitoneal received 50 mg Surek intravenously (i.v.) and subcutaneously (s.c.), injection of GD2-positive B78-D14 mouse melanoma cells and respectively, together with 105 B78-D14 cells intraperitoneally or tumor cells trAb (Fig. 3). After injection of Surek intravenously and sub- alone or were left untreated. After 48 to 72 hours, T cells were analyzed in cutaneously, respectively, a massive weight loss in both treat- spleens. A, staining of the activation marker CD69 on T cells. B, intracellular ment groups was observed by day 2. While weight loss con- staining of the proliferation marker Ki-67 in T cells. C, intracellular staining of þ tinued in intravenously treated mice until day 3, subcutane- IFNg in CD8 T cells. All graphs show means and SEM from at least 4 individual mice. The differences between Surek-treated and nontreated animals are ously treated mice started earlier to recover. At day 3, the statistically significant with at least P < 0.05 (Student t test), whereas there is relative weight loss of intravenously treated animals was about no significant change in activation and proliferation between intravenous and 40% higher than that of the subcutaneously treated group subcutaneous administration of Surek. , P < 0.05; , P < 0.01; , P < 0.001. (Student t test, P ¼ 0.0032).

www.aacrjournals.org Mol Cancer Ther; 14(8) August 2015 1879

Deppisch et al.

ment with Surek intravenously and subcutaneously, respectively. As shown in Fig. 5A, we observed an increase of all Th1 cytokines tested after Surek administration. However, intravenous and subcutaneous treatment differed in the kinetics of cytokine response. Intravenously injected Surek led to an immediate increase of IL6, IFNg, and TNF within 1 hour after injection, whereas trAb administered subcutaneously induced a delayed rise with the maximum concentration only 3 hours after injection and a more rapid decline at later time points compared with intravenous treatment. The moderate induction of cytokine release after Figure 3. subcutaneous administration in comparison with the rapid Systemic tolerability of trAb Surek evidenced by weight loss of mice. Animals were injected with 105 B78-D14 tumor cells intraperitoneally (i.p.) and 50 g release after intravenous injection may have a favorable effect on m fi Surek intravenously (i.v.) and subcutaneously (s.c.), respectively, at day 0. the safety pro le. Nonetheless, as shown above, T-cell activation Weight at day 0 was defined as 100%. Weight of control groups remained (Fig. 1A) and especially IFNg secretion of T cells (Fig. 1C) after stable, while Surek-treated mice experienced a severe weight loss. At day 3, subcutaneous treatment were obviously sufficient to secure a weight of intravenously treated animals was statistically different from beneficial therapeutic outcome. P t n ¼ subcutaneously treated mice: , < 0.01 (Student test, 8/group). Given the lower bioavailability of Surek (50%) following subcutaneous application as compared with intravenous injec- Bioavailability of trAb Surek and anti-drug antibody induction tion (Fig. 4A), the effective tumor killing by subcutaneously The data show that subcutaneous administration of Surek delivered Surek (Fig. 2) as well as the high expression of T-cell results in equal efficacy and even better tolerability compared activation parameters (Fig. 1) were surprising. We hypothesized with intravenous injection of trAb. As this outcome might be related to the bioavailability of the trAb, we measured plasma concentrations over 10 days following injection and determined bioavailability by calculating the areas under the curves (Fig. 4A). Surek injected once intravenously was available immediately, reached its maximum concentration 10 minutes after injection, and was then eliminated constantly. In contrast, subcutaneously administered Surek showed a delayed absorption. A plateau concentration was observed between 8 and 24 hours, followed by trAb elimination. At day 5 after injection, Surek could not be detected any more in either setting. Taken together, bioavailability was only 50% after subcutaneous administration as compared with intravenous injection. Previous studies showed that the antitumor efficacy of Surek was dose-dependent (14). To mimic the clinical situation more closely, we therefore measured the plasma levels of Surek in a treatment schedule involving three injections (day 0, 2, and 4; Fig. 4B). While each intravenous injection of Surek led to a concentration peak that subsequently declined rapidly, multiple subcutaneous administrations entailed a steadily continuing rise in plasma concentration. For clinical use, application of Surek subcutaneously looks promising because plasma peaks, which may cause systemic adverse effects, were absent. The therapeutic potential of Surek may be limited by the development of neutralizing anti-drug antibodies. As high concentrations of trAb in the skin and the presence of numerous antigen-presenting cells may favor presentation of trAb epitopes to the immune system in the subcutaneous setting, we measured MAMAs and MARAs in mice sera. However, comparing Surek subcutaneously and intravenously did not reveal any difference in anti-antibody titers 15, 30, and 45 days after trAb injection. In 5 6 Figure 4. both settings, anti-antibody titers were in the range of 10 to 10 Bioavailability of trAb Surek. A, plasma concentration of Surek after injection (Supplementary Fig. S1). of 50 mg Surek intravenously (i.v.) and subcutaneously (s.c.), respectively. Samples were taken over a period of 10 days and trAb concentrations were Differential cytokine milieu induced by Surek injected measured by ELISA. Typical result from three experiments. Bioavailability intravenously versus subcutaneously (BV) was calculated on the basis of the area under the curve (AUC) and related to the bioavailability obtained after intravenous injection, which was Adverse systemic effects associated with trAb treatment are defined as 100%. B, plasma levels of Surek in a treatment schedule with three induced by cytokines and closely correlate with the systemic 20 mg injections (days 0, 2, and 4) intravenously and subcutaneously, cytokine profile after trAb administration. Therefore, we studied respectively. Each data point represents the mean concentration SEM. Four the cytokine pattern in sera of tumor-bearing animals after treat- mice were included in each group.

1880 Mol Cancer Ther; 14(8) August 2015 Molecular Cancer Therapeutics

Optimization of trAb-Mediated Tumor Therapy

most potent immunologic antitumor reagents hitherto used in cancer patients. These Ig constructs, which are effective at nano- gram/mL concentrations (18, 27), may raise particular safety concerns (18, 19). Although trAb-induced cytokine-related adverse events are manageable in the clinics by using low starting doses and subsequent dose escalation (18, 28), there must be constant efforts to further improve patients' safety thus increasing the treatment options for trAbs in the near future. In the current study, we directly compared conventional intravenous treatment and subcutaneous application of trAbs by using trAb Surek as a preclinical model, which targets GD2-expressing melanoma cells and CD3 on murine T cells (14). Many antibody-induced side effects as observed in humans (like chills, nausea, or headache; refs. 18) cannot be documented in mice. A reliable indicator of adverse health effects in mice, however, is the body weight (23). Both weight loss in animals and the clinical symptoms observed in humans are equally affected by cytokines released upon antibody delivery. Therefore, we assessed weight loss as a marker for systemic tolerability although specific effects that will be seen in humans after intravenous or subcutaneous injection of trAbs cannot be predicted. Our results suggest a better tolerability of the subcutaneous application. It should be noted that T-cell activation requires binding of Surek to CD3 but not simultaneous engagement of the tumor- Figure 5. specific binding arm. The latter is only necessary for redirecting the Differential cytokine milieu induced by Surek subcutaneously (s.c.) vs. intravenously (i.v.). Cytokine concentrations in sera were measured in a cytotoxic activity of activated T cells towards the target cells (15). Bioplex assay 1, 3, 24, and 48 hours after a single injection of 105 B78-D14 Therefore, Surek also induced T-cell activation, weight loss, and tumor cells and 50 mg Surek intravenously or subcutaneously. A, concentration of IL6, IFNg, and TNF in sera. B, concentration of IL10 in sera. C, þ expression of IL10 in CD8 cells from spleens as measured by intracellular FACS staining 48 hours after treatment of mice as indicated. Plots show means and SEM from at least 4 individual mice. Statistical significance in comparison with intravenously treated group according to the Student t test: , P < 0.05; , P < 0.01, respectively; ns, not significant. that the reduced release of counter-regulatory cytokines could be responsible for the higher trAb efficiency in the subcutaneous situation. Indeed, IL10, which may exert a suppressive effect, showed a high concentration in sera of intravenously treated mice during the entire observation period of 48 hours (Fig. 5B). In sera of subcutaneously treated mice, by contrast, significantly less IL10 could be detected. In line with this, we found an þ enhanced IL10 production of CD8 T cells in intravenously þ þ treated mice, whereas less IL10 CD8 T cells could be detected in subcutaneously treated animals at day 2 after injecting B78- D14 and Surek (Fig. 5C). Furthermore, in all T-cell subsets, CTLA- 4 expression was significantly higher in the intravenously treated group than in subcutaneously injected mice (Fig. 6). Consequent- ly, we assume that after subcutaneous administration of Surek, there are less counter-regulatory effects that interfere with the activity of effector T cells and the release of proinflammatory cytokines, both of which are crucial for specific tumor cell killing.

Discussion Figure 6. Reduced CTLA-4 expression after Surek subcutaneous (s.c.) administration. Antibody-mediated approaches of cancer immunotherapy Mice received 50 mg Surek intravenously (i.v.) and subcutaneously, 5 have widely been used in the clinic (24, 25). However, such respectively, together with 10 B78-D14 cells intraperitoneally or tumor cells regimens are often associated with undesired adverse effects such alone or were left untreated. After 72 hours, T cells from spleens were analyzed. Intracellular staining of CTLA-4 in CD4þ (A) and CD8þ (B) cells. as ﬂulike symptoms, chills, and fever that are due to off-site T-cell Left, representative experiments. Right, cumulative results with means and activation and systemic cytokine release (18, 19, 26). TrAbs, which SEM (n ¼ 5 per group). Statistical signiﬁcance in comparison with recruit different types of immune effector cells to the tumor cells intravenously treated group according to the Student t test: , P < 0.05; and thereby also induce a T-cell memory (6, 7), are one of the , P < 0.01, respectively.

www.aacrjournals.org Mol Cancer Ther; 14(8) August 2015 1881

Deppisch et al.

þ þ cytokine release in the absence of tumor cells (Supplementary Figs. CD4 and CD8 T cells of subcutaneously treated animals express S2–S4), which, of course, does not reflect the clinical situation. significantly less CTLA-4 (Fig. 6), which is upregulated as a While intravenous delivery of trAb Surek gave rise to sharp plasma counter-regulatory mechanism during an immune response peaks followed by a rapid decline after each injection, subcutaneous (31, 32). These findings indicate that the balance between T-cell administration led to a continuous building up of plasma levels activation and suppression is shifted after subcutaneous admin- (Fig. 4B). Peaks observed after intravenous injection correlated with istration, thus enabling potent tumor rejection despite lower an immediate rise of all Th1 cytokines tested, whereas sera of bioavailability. subcutaneously treated animals exhibited the maximum concen- Other monoclonal antibodies, which are already routinely tration of thesecytokines only 3 hours after administration (Fig. 5A). used in the clinic, like trastuzumab or alemtuzumab, were also As shown earlier (6), T-cell activation and the release of Th1 effective despite slower absorption rates when delivered subcu- cytokines, especially IFNg, are crucial for the induction of an taneously (33, 34). To facilitate absorption of trastuzumab, a effective and long-lasting antitumor response, but proinflammatory subcutaneous formulation containing recombinant hyaluroni- cytokines also convey the undesired side effects. Although the dase was developed (35). For rituximab, the subcutaneous appli- bioavailability of Surek in the subcutaneous setting was consider- cation has also been approved (36). In our study, we provide for ably decreased (about 50% compared with intravenous the first time, a systematic comparison of the application routes by injection; Fig. 4A), the IFNg induction was apparently still sufficient using a trAb. Given the safety concerns emanating from the high to provide the same beneficial therapeutic effects as after intravenous tumoricidal efficacy of these constructs, our results are of partic- delivery (Fig. 1C). Nonetheless, systemic side effects were greatly ular relevance for refining therapeutic approaches using other reduced (Fig. 3), which may be explained by the absence of trAb monoclonal antibodies. plasma peaks and a hereby conveyed delay of cytokine release, In summary, the efficacy of subcutaneously delivered Surek is which is also reduced at most time points (Figs. 4B and 5A). comparable with intravenous administration (Fig. 2), and the The development of anti-drug antibodies could pose a possible safety profile may be even better than observed after intravenous limitation to subcutaneous therapy with trAbs because the skin injection (Fig. 3). The availability of a subcutaneous formulation contains numerous antigen-presenting cells, which could engulf dramatically simplifies trAb administration and thus opens subcutaneously injected trAb and present trAb-derived epitopes completely new perspectives of redirecting immunity via trAbs to the immune system. However, no increased anti-antibody titers in the future. As subcutaneous administration presents a well- were observed following subcutaneous therapy as compared with tolerated treatment, trAb may additionally be applicable in a intravenous application (data not shown). After several therapy home setting after dose escalation (37, 38). This would increase cycles, anti-drug antibodies may hamper the clinical use not only patients' quality of life and might also reduce costs in the health by neutralizing the trAb-dependent therapeutic effect, but also by care sector. forming aggregates that could lead to severe adverse events in vivo. However, as trAbs are typically used in extremely low amounts, Disclosure of Potential Conflicts of Interest such effects have not been observed in clinical trials so far H. Lindhofer is a CEO/CSO at Lindis Biotech and has ownership interest (8, 18, 19). Our preclinical studies demonstrate that application (including patents) in a patent application. No potential conflicts of interest of much higher doses in humans would be feasible, provided that were disclosed by the other authors. the appropriate application route is chosen. In contrast to intravenously treated animals, subcutaneously treated mice did never Authors' Contributions show any reaction related to immunogenicity of Surek despite Conception and design: N. Deppisch, N. Eissler, R. Buhmann, H. Lindhofer, identical amounts of anti-antibodies. This can be explained by the R. Mocikat different maximum concentrations reached after trAb injection: Development of methodology: N. Deppisch, P. Ruf, N. Eissler Acquisition of data (provided animals, acquired and managed patients, After intravenous injection, high trAb doses are immediately provided facilities, etc.): N. Deppisch, P. Ruf, F. Neff available and can form aggregates with anti-antibodies that were Analysis and interpretation of data (e.g., statistical analysis, biostatistics, induced before and are circulating in the bloodstream. After computational analysis): N. Deppisch, F. Neff, R. Buhmann, R. Mocikat subcutaneous delivery of Surek, by contrast, the absence of Writing, review, and/or revision of the manuscript: N. Deppisch, P. Ruf, plasma peaks and lower trAb concentrations lead to formation R. Buhmann, H. Lindhofer, R. Mocikat of less aggregates despite identical amounts of anti-antibodies. Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): H. Lindhofer, R. Mocikat These data further support the assumption that subcutaneous Study supervision: R. Buhmann, R. Mocikat application of Surek is better tolerable than intravenous delivery and open the possibility to apply high trAb concentrations in Acknowledgments multiple treatment cycles. Indeed, a third subcutaneous injection The authors thank A. Geishauser, M. Hagemann, N. Homberg,€ J. Muller,€ of Surek was feasible and increased the overall survival (Fig. 2), Y. Suckstorff, L. Thurmann, and S. Wosch for expert technical assistance and while a third intravenous injection of trAb led to severe adverse Josef Mysliwietz for critically reading the manuscript and helpful discussions. events in mice. The regulatory cytokine IL10 is known to mediate counter- Grant Support regulatory signals and to suppress Th1-driven antitumor The work was supported by a grant of Bayerische Forschungsstiftung to responses (29, 30). Interestingly, the plasma concentration of N. Deppisch and Trion Research. The costs of publication of this article were defrayed in part by the payment of IL10 was higher in intravenously treated animals compared with page charges. This article must therefore be hereby marked advertisement in subcutaneously injected mice (Fig. 5B). Intracellular FACS stain- accordance with 18 U.S.C. Section 1734 solely to indicate this fact. þ þ ing revealed that this difference was accounted for by CD8 IL10 T cells, which were increased in intravenously treated animals Received February 17, 2015; revised May 22, 2015; accepted June 3, 2015; compared with subcutaneously treated mice (Fig. 5C). Besides, published OnlineFirst June 10, 2015.

1882 Mol Cancer Ther; 14(8) August 2015 Molecular Cancer Therapeutics

Optimization of trAb-Mediated Tumor Therapy

References 1. Hess J, Ruf P, Lindhofer H. Cancer therapy with trifunctional antibodies: (FBTA05), a trifunctional anti-CD3 x anti-CD20 antibody and donor linking innate and adaptive immunity. Future Oncol 2012;8:73–85. lymphocyte infusion. Bone Marrow Transplant 2009;43:383–97. 2. Zeidler R, Mysliwietz J, Csanady M, Walz A, Ziegler I, Schmitt B, et al. The Fc- 20. Sebastian M, Passlick B, Friccius-Quecke H, J€ager M, Lindhofer H, region of a new class of intact bispecific antibody mediates activation of Kanniess F, et al. Treatment of non-small cell lung cancer patients with accessory cells and NK cells and induces direct phagocytosis of tumour the trifunctional monoclonal antibody catumaxomab (anti-EpCAM x cells. Br J Cancer 2000;83:261–6. anti-CD3): a phase I study. Cancer Immunol Immunother 2007;56: 3. Zeidler R, Reisbach G, Wollenberg B, Lang S, Chaubal S, Schmitt B, et al. 1637–44. Simultaneous activation of T cells and accessory cells by a new class of intact 21. Miescher GC, Schreyer M, MacDonald HR. Production and characteriza- bispecific antibody results in efficient tumor cell killing. J Immunol 1999; tion of a rat monoclonal antibody against the murine CD3 molecular 163:1246–52. complex. Immunol Lett 1989;23:113–8. 4. Egeter O, Mocikat R, Ghoreschi K, Dieckmann A, Rocken€ M. Eradication of 22. Thurin J, Thurin M, Herlyn M, Elder DE, Steplewski Z, Clark WH Jr, et al. disseminated lymphomas with CpG-DNA activated T helper type 1 cells GD2 ganglioside biosynthesis is a distinct biochemical event in human from nontransgenic mice. Cancer Res 2000;60:1515–20. melanoma tumor progression. FEBS Lett 1986;208:17–22. 5. Ziegler A, Heidenreich R, Braumuller H, Wolburg H, Weidemann S, Mocikat 23. Ullman-Cullere MH, Foltz CJ. Body condition scoring: a rapid and accurate R, et al. EpCAM, a human tumor-associated antigen promotes Th2 devel- method for assessing health status in mice. Lab Anim Sci 1999;49:319–23. opment and tumor immune evasion. Blood 2009;113:3494–502. 24. Carter P. Improving the efficacy of antibody-based cancer therapies. Nat 6. Eissler N, Ruf P, Mysliwietz J, Lindhofer H, Mocikat R. Trifunctional Rev Cancer 2001;1:118–29. bispecific antibodies induce tumor-specific T cells and elicit a vaccination 25. Waldmann TA. Immunotherapy: past, present and future. Nat Med 2003; effect. Cancer Res 2012;72:3958–66. 9:269–77. 7. Ruf P, Lindhofer H. Induction of a long-lasting antitumor immunity by a 26. Teachey DT, Rheingold SR, Maude SL, Zugmaier G, Barrett DM, Seif AE, trifunctional bispecific antibody. Blood 2001;98:2526–34. et al. Cytokine release syndrome after blinatumomab treatment related to 8. Heiss MM, Murawa P, Koralewski P, Kutarska E, Kolesnik OO, Ivanchenko abnormal macrophage activation and ameliorated with cytokine-directed VV, et al. The trifunctional antibody catumaxomab for the treatment of therapy. Blood 2013;121:5154–7. malignant ascites due to epithelial cancer: results of a prospective random- 27. Ruf P, Kluge M, Jager M, Burges A, Volovat C, Heiss MM, et al. Pharma- ized phase II/III trial. Int J Cancer 2010;127:2209–21. cokinetics, immunogenicity and bioactivity of the therapeutic antibody 9. Strohlein€ MA, Siegel R, J€ager M, Lindhofer H, Jauch KW, Heiss MM. catumaxomab intraperitoneally administered to cancer patients. Br J Clin Induction of anti-tumor immunity by trifunctional antibodies in patients Pharmacol 2010;69:617–25. with peritoneal carcinomatosis. J Exp Clin Cancer Res 2009;28:18. 28. Heiss MM, Strohlein€ MA, J€ager M, Kimmig R, Burges A, Schoberth A, et al. 10. Ruf P, J€ager M, Ellwart J, Wosch S, Kusterer E, Lindhofer H. Two new Immunotherapy of malignant ascites with trifunctional antibodies. Int J trifunctional antibodies for the therapy of human malignant melanoma. Cancer 2005;117:435–43. Int J Cancer 2004;108:725–32. 29. Naujoks M, Weiß J, Riedel T, Homberg€ N, Przewoznik M, Noessner E, et al. 11. Ragupathi G, Livingston PO, Hood C, Gathuru J, Krown SE, Chapman PB, Alterations of costimulatory molecules and instructive cytokines expressed et al. Consistent antibody response against ganglioside GD2 induced in by dendritic cells in the microenvironment of an endogenous mouse patients with melanoma by a GD2 lactone-keyhole limpet hemocyanin lymphoma. Cancer Immunol Immunother 2014;63:491–9. conjugate vaccine plus immunological adjuvant QS-21. Clin Cancer Res 30. O'Garra A, Vieira P. T(H)1 cells control themselves by producing inter- 2003;9:5214–20. leukin-10. Nat Rev Immunol 2007;7:425–8. 12. Navid F, Santana VM, Barfield RC. Anti-GD2 antibody therapy for GD2- 31. Krummel MF, Allison JP. CD28 and CTLA-4 have opposing effects on the expressing tumors. Curr Cancer Drug Targets 2010;10:200–9. response of T cells to stimulation. J Exp Med 1995;182:459–65. 13. Haraguchi M, Yamashiro S, Yamamoto A, Furukawa K, Takamiya K, Lloyd 32. Walunas TL, Lenschow DJ, Bakker CY, Linsley PS, Freeman GJ, Green JM, KO, et al. Isolation of GD3 synthase gene by expression cloning of GM3 et al. CTLA-4 can function as a negative regulator of T cell activation. alpha-2,8-sialyltransferase cDNA using anti-GD2 monoclonal antibody. Immunity 1994;1:405–13. Proc Natl Acad Sci U S A 1994;91:10455–9. 33. Leveque D. Subcutaneous administration of anticancer agents. Anticancer 14. Ruf P, Sch€afer B, Eissler N, Mocikat R, Hess J, Ploscher M, et al. Ganglioside Res 2014;34:1579–86. GD2-specific trifunctional surrogate antibody Surek demonstrates thera- 34. Hale G, Rebello P, Brettman LR, Fegan C, Kennedy B, Kimby E, et al. Blood peutic activity in a mouse melanoma model. J Transl Med 2012;10:219. concentrations of alemtuzumab and antiglobulin responses in patients 15. Eissler N, Mysliwietz J, Deppisch N, Ruf P, Lindhofer H, Mocikat R. with chronic lymphocytic leukemia following intravenous or subcutane- Potential of the trifunctional bispecific antibody surek depends on den- ous routes of administration. Blood 2004;104:948–55. dritic cells: rationale for a new approach of tumor immunotherapy. Mol 35. Hamizi S, Freyer G, Bakrin N, Henin E, Mohtaram A, Le Saux O, et al. Med 2013;19:54–61. Subcutaneous trastuzumab: development of a new formulation for treat- 16. Sebastian M, Kiewe P, Schuette W, Brust D, Peschel C, Schneller F, et al. ment of HER2-positive early breast cancer. Onco Targets Ther 2013;6: Treatment of malignant pleural effusion with the trifunctional antibody 89–94. catumaxomab (Removab) (anti-EpCAM x anti-CD3): results of a phase 1/2 36. Davies A, Merli F, Mihaljevic B, Siritanaratkul N, Solal-Celigny P, Barrett M, study. J Immunother 2009;32:195–202. et al. Pharmacokinetics and safety of subcutaneous rituximab in follicular 17. Strohlein€ MA, Lordick F, Ruttinger€ D, Grutzner€ KU, Schemanski OC, J€ager lymphoma (SABRINA): stage 1 analysis of a randomised phase 3 study. M, et al. Immunotherapy of peritoneal carcinomatosis with the antibody Lancet Oncol 2014;15:343–52. catumaxomab in colon, gastric, or pancreatic cancer: an open-label, mul- 37. Lundin J, Kimby E, Bjorkholm M, Broliden PA, Celsing F, Hjalmar V, et al. ticenter, phase I/II trial. Onkologie 2011;34:101–8. Phase II trial of subcutaneous anti-CD52 monoclonal antibody alemtu- 18. Kiewe P, Hasmuller€ S, Kahlert S, Heinrigs M, Rack B, Marme A, et al. Phase I zumab (Campath-1H) as first-line treatment for patients with B-cell trial of the trifunctional anti-HER2 x anti-CD3 antibody ertumaxomab in chronic lymphocytic leukemia (B-CLL). Blood 2002;100:768–73. metastatic breast cancer. Clin Cancer Res 2006;12:3085–91. 38. Launay-Vacher V. An appraisal of subcutaneous trastuzumab: a new 19. Buhmann R, Simoes B, Stanglmaier M, Yang T, Faltin M, Bund D, et al. formulation meeting clinical needs. Cancer Chemother Pharmacol 2013; Immunotherapy of recurrent B-cell malignancies after allo-SCT with Bi20 72:1361–7.

www.aacrjournals.org Mol Cancer Ther; 14(8) August 2015 1883

Efficacy and Tolerability of a GD2-Directed Trifunctional Bispecific Antibody in a Preclinical Model: Subcutaneous Administration Is Superior to Intravenous Delivery

Nina Deppisch, Peter Ruf, Nina Eissler, et al.

Mol Cancer Ther 2015;14:1877-1883. Published OnlineFirst June 10, 2015.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-15-0156

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2015/06/11/1535-7163.MCT-15-0156.DC1

Cited articles This article cites 38 articles, 13 of which you can access for free at: http://mct.aacrjournals.org/content/14/8/1877.full#ref-list-1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/14/8/1877. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.