The research program of the Div. of Molecular Oncology of Gastrointestinal Cancer aims at the development of immunotherapeutic strategies that can be deployed in conjunction with surgery and cytostatic anti-cancer treatment.

Scientific approach

The median survival time for patients with non-resectable pancreatic cancer (80% of cases diagnosed) is approx. 6 months. In our opinion, this leaves insufficient time to boost endogenous T-cell immunity through, for instance, vaccines or the CTLA-4 blocking antibody Ipilumumab. Consequently, our strategy with respect to immunotherapy primarily involves two approaches:

  • Stimulation of the endogenous immune potential, in particular in the tumor stroma, by means of agonist immunostimulatory monoclonal antibodies (ISmAbs)
  • Replenishing the patient’s immune system with ex vivo engineered autologous T-cells.

Since the median survival time of patients diagnosed with resectable disease is longer, approximately 1.5 years, modulation of endogenous T-cell immunity will be considered in this patient group.

Although phase I safety studies generally involve end stage patients, it is important that further studies aimed at harnessing the endogenous immune response are implemented in the context of first-line treatment, before the patient’s immune system has deteriorated due to progressive disease and cytotoxic anti-cancer regimens. In practical terms, this currently means combination with standard of care chemotherapy (gemcitabine) for non-resectable disease, and (neo)adjuvant treatment for resectable disease.

Agonist ISmAbs


In the context of a collaborative effort with several European experts, we are aiming at the rapid clinical development of agonist antibodies targeting receptor molecules on myeloid cells and T-lymphocytes.


Recently, two immunotherapeutic drugs became available for the treatment of cancer patients as a result of their efficacy and safety in clinical studies. In April 2010, the FDA approved a cell-based vaccine (Provenge) for use in metastatic prostate cancer. In March 2011, a monoclonal antibody (mAb) blocking the function of the immune inhibitory molecule CTLA-4 (Ipilimumab) was approved for use in metastatic melanoma. Both drugs aim at enhancing the patient’s endogenous T-cell response against cancer. This success boosted the clinical development of other vaccines and antagonistic immunostimulatory mAbs (ISmAbs) as cancer therapeutics.

Nevertheless, the potential of agonist ISmAbs remains largely unexplored, in spite of ample evidence from pre-clinical tumor models that this drug concept can be at least as efficacious. One of the main setbacks to agonist ISmAb development was a serious incident in 2006, when administration of a mAb against the T-cell surface molecule CD28 (TGN1412) to human volunteers resulted in a life threatening cytokine-release syndrome. Importantly, TGN1412 is a highly potent super-agonist mAb, while many of the ISmAbs successfully used in pre-clinical studies do not exhibit this super-agonist potency. Furthermore, currently available preclinical analyses allow much better assessment of the risk for a cytokine release syndrome. Finally, it is now evident that the human study with TGN1412 was not conducted in a proper, careful manner. As such, many experts feel that the disqualification of agonist ISmAbs as a drug concept due to the outcome of the TGN1412 study was premature.


In view of these considerations, we will conduct a systematic, ‘safety-first’ trajectory for the development of agonist ISmAbs for treatment of solid cancers, in particular pancreatic cancer. The primary aim is to obtain proof of concept for the safety and efficacy of agonist ISmAbs in the clinic. Initial clinical studies will involve single agent dose escalation in end stage cancer patients. Emphasis will be on measurement of toxicity parameters and blood/PBMC biomarkers. Subsequent clinical trials, if warranted by the prior safety studies, will focus on applying the agonist mAbs in first line treatment of non-resectable disease in conjunction with surgery and cytotoxic drugs. The rationale for using these combinations is that striking synergistic effects have been observed in pre-clinical tumor models. If these combination regimens are well tolerated, this will open the way to their application as (neo)adjuvant treatment for patients with resectable disease.

Adoptive immunotherapy using ex vivo engineered, autologous T-lymphocytes

The failure of many prior clinical studies concerning immunotherapy of cancer has taught us that progressive cancers feature a profound state of immune subversion. In view of this, we must come to the realization that any attempt to mobilize the endogenous immune response in end-stage cancer patients, the typical test population of phase I and II clinical trials, is likely to fail. Furthermore, pancreatic cancer is very aggressive disease, so time to get immunity going is limited. In view of this, replenishing the immune system through infusion of tumor-targeting T-lymphocytes may offer a better likelihood for therapeutic efficacy within the available time window.

Ex vivo revitalized tumor-infiltrating lymphocytes (TILs) offer a source of such effector cells. A more sophisticated alternative involves infusion of genetically engineered autologous T-lymphocytes expressing a tumor-targeting T-cell receptor (TCRs). A hurdle with respect to application of this powerful approach to pancreatic cancers, as opposed to melanoma, is the lack – at present – of truly suitable target antigens. The choice of proper target antigens is pivotal for achieving efficacy in the absence of life-threatening auto-immune pathology in normal tissues. We therefore aim at obtaining proof of concept for safety and clinical efficacy using TCRs targeting pancreas lineage-specific antigens, in the context of total pancreatectomy, a procedure applied to approximately 15% of cases of resectable pancreatic cancer. If successful, subsequent studies will aim at the use of tumor-specific TCRs in the context of patient-tailored adoptive T-cell therapy.

Combining immunotherapy with cytotoxic drugs.

Pancreatic cancer offers a prime setting for combining immunotherapeutic strategies with cytotoxic drugs, because the standard of care chemotherapeutic drug used for this indication is Gemcitabine, a mild cytotoxic agent that does not disable the immune system. Although combining gemcitabine with immunostimulatory drugs was shown to result in synergy, the ultimate therapeutic efficacy has so far been disappointing. It is therefore important to test combinations of more potent cytotoxic drugs, in particular the latest small molecule inhibitors of the Ras pathway, with either agonistic ISmAbs or adoptively transferred T-lymphocytes.

Complementary and interlocking research in three areas

In order to achieve our goal, we conduct research in three complementary areas:

  • Human T-cell immunology
  • Genetically engineered mouse models (GEMMs)
  • Biomarkers

Patient-based research; human T-cell immunology

Although human pancreatic tumors may be infiltrated with CD3 cells, most of these reside in the stroma and fail to infiltrate the tumor cell areas.
© E. Ryschich, University of Heidelberg

Surgical treatment of pancreatic cancer commonly involves removal of at least half of the pancreas, including multiple lymph nodes. Consequently, different tissues from the same patient are available for systematic patient-based research. For instance, primary tumor tissue can be compared with normal and transitional pre-malignant areas, as well as lymph node metastases to study tumor evolution. Furthermore, immune cells can be retrieved from tumor and tumor-draining lymph nodes. Notably, the latter tissue is the best source of tumor-reactive T-lymphocytes, as it represents the interface between tumor and adaptive immune system.

Genetically engineered mouse models

Various immunotherapeutic approaches were shown to have striking efficacy in mouse tumor models, but very few of these have translated into clinical success. A main reason for this is that the commonly used mouse models feature small, transplantable tumors that differ with human cancers in three essential aspects:
  • Tumor burden
  • Lack of long-term tumor immune interaction resulting in profound immune subversion
  • Artificial priming of the immune system due to the tumor cell death and pro-inflammatory signals that are intrinsic to the implantation of tumor cells in the host.

Due to these issues, the capacity of the endogenous T-cell compartment to fight ‘real’ cancer may be overestimated. At present, genetically engineered mouse models offer the best way to mimic the stealthy nature of cancer. In our research, we make use of such models for melanoma and pancreatic cancer. Both models are conditional rather than constitutive, allowing the induction of primary tumors in adult mice with a fully matured immune system.

Biomarker research

Treatment of pancreatic cancer cannot be improved without increasing our knowledge of disease-associated biomarkers, in particular the following classes:

  • Diagnostic markers, allowing earlier diagnosis and, thereby, treatment
  • Prognostic markers, allowing patient stratification and more personalized treatment options
  • Tumor antigens, allowing monitoring of adaptive immunity in context of disease and treatment, as well as selection of tumor targeting TCRs for adoptive T-cell therapy
  • Pharmodynamic markers for tracking the mechanism of action (MOA)-related impact of cytotoxic drugs, immunotherapeutic treatments and combinations thereof.

Therefore, biomarker research is an integrate aspect of our pre-clinical and clinical research program.

to top