Tumor oxygenation and radiation resistance (clinical)

Research group: PD Dr. Oliver Riesterer & Prof. Dr. Gabriela Studer

For the preclinical part of this project please see project Pruschy

Clinical investigations will first focus on feasibility studies as well as validation of oxygenation/perfusion imaging in comparison to the best available alternative (e.g. immunohistochemistry of initial tumor biopsy). We envisage to adapt radiochemotherapy according to microenvironmental/tumor vasculature-related endpoints, including tumor oxygenation. Dose escalation or de-escalation trials are envisaged in head and neck cancer since tumor hypoxia is a robust predictive biomarker for response to radiotherapy in this cancer type (Quelle?). Modern radiation technology, i.e. intensity modulated radiotherapy, allows to give different dose levels to different structures at the same time thereby allowing to specifically target hypoxic tumor areas.

For example, in clinical routine we already give different dose levels in parallel to the tumor region, i.e. 70 Gy to the primary tumor and lymph node metastases, 60 Gy close to the primary tumor (where there is high risk for micrometastasis), and 54 Gy to elective lymph node regions. All dose levels are given in 35 fractions with 2 Gy, 1.7 Gy, 1.5 Gy fractions at the same time, 5 fractions per week. This technique, which is called dose painting, could also be used to distribute the radiation dose according to the level of hypoxia within the tumor. Hypoxic areas as small as 2-5mm could be boosted. Current (chemo-) radiation standard is given up to the tolerance of the surrounding tissue in these patients and is therefore very toxic. Further steps in clinical development are to exploit the novel technology for adaptive radiotherapy, e.g., change of the radiotherapy concept during treatment according to the course of tumor oxygenation, or to combine radiotherapy with molecular therapeutics that interact with tumor oxygenation. Further, preclinical research from us and other groups indicates that tumor persistence after irradiation or locally recurrent tumors growing in preirradiated tumor bed conditions are overly hypoxic (Milas, Hunter, Peters, 1989; Riesterer, Mason, Raju, Yang, Wang, Hittelman, Ang, Milas, 2009; Zips, Eicheler, Bruchner, Jackisch, Geyer, Petersen, van der Kogel, Baumann). Characterization of such biological conditions by using hypoxia imaging tools on the clinical level is a highly promising new approach to select patients for suitable adjuvant treatments.

The clinical radiation oncology part will be divided into two work packages. In work package 1, feasibility studies of serial non-invasive imaging of tumor oxygenation are performed. These studies will also translate the preclinical findings into the clinical situation. In parallel, biological validation of the imaging tools will be performed in preclinical models. Additionally, we plan to investigate innovative candidate tissue and serum biomarkers in material stored from previously treated patients. In work package 2, we envisage to translate the findings from the phase 1 studies and the preclinical groups into a phase II study of hypoxia-adapted radiochemotherapy.

Work Package 1: Master Protocol for phase 1 studies of hypoxia imaging Several hypoxia imaging tools will be developed in parallel within the KFSP (PET, MRI, NIRI based). Therefore, a phase 1 master protocol will be designed with the aim to integrate and investigate all imaging tools for biology guided radiochemotherapy. This master protocol will include several phase 1 studies, one for each imaging tool (PET, MRI, NIRI). The aims of these phase 1 studies are the following:

- Investigate the dynamics of tumor hypoxia before and during chemoradiation

- Establish hypoxia imaging in patients

- Learn how to integrate hypoxia imaging into radiation treatment planning protocols

- Design hypoxia adapted radiation schedules

In each phase 1 study ten patients with intermediate and advanced stage squameous cell carcinoma of the head and neck (oral cavity, oropharynx, hypoxpharynx, larynx) will be included. In case of large primary tumors (e.g. > 6cm) or lymph nodes (e.g. > 3cm) necrotic areas can often be seen on the staging CT. In contrast, intermediate size tumors (e.g. primary < 6cm, lymph nodes < 3cm) often do not show necrotic areas but still might be hypoxic. Patients with intermediate size tumors that are not hypoxic, or that are hypoxic but do quickly reoxygenate during radiotherapy, are potential candidates for deescalation of the total dose. Therefore, in the phase I study 5 patients with large primary tumors (> 6cm) and/or large lymph nodes (> 3cm) and 5 patients with intermediate size primary tumors (< 6cm) and/or lymph nodes (< 3cm) will be included. In order to assess feasibility of hypoxia dependent treatment planning the hypoxia images will be imported into the radiation treatment planning software and hypoxia images will be matched with the treatment planning CT. Feasibility planning studies will be performed including modeling of tumor response dependent on the dose prescribed and the tumor hypoxic fraction. In order to investigate the dynamics of tumor hypoxia, repetitive hypoxia imaging in week 4 of the 7 weeks long treatment regimen, and, if possible, at one week after end of radiotherapy is planned. Our clinical experience shows that in week four the tumors often start to shrink. Therefore, microenvironmental changes do likely occur at this time point. Fractionated radiotherapy is expected to induce reoxygenation within the tumor as itinerative process by killing well oxygenated tumor cells with every fraction thereby reducing oxygen demand in the remaining tumor cells. Reoxygenation at week 4 might be predictive for a responsive tumor. In contrast, persisting hypoxia could be indicative for treatment resistance. Repetitive hypoxia imaging at week 4 allows for hypoxia-dependent adaptation of the treatment plan or the treatment schedule. In case the first imaging does not show any hypoxia no further O2-PETs will be performed. Since repetitive O2 PET might logistically be difficult, patient accrual will be continued beyond 10 patients until 10 patients are included with one PET before start of RT and one PET at week 4, respectively. As soon as hypoxia imaging with NIRI becomes clinically available we can then monitor tumor hypoxia more densely (e.g. weekly). Preclinical research results from us and other groups demonstrate that the irradiated tumor environment immediately after radiochemotherapy is overly hypoxic and novel targeted pharmaceuticals are already in development as adjuvant therapy after chemoradiation to target these residual tumor cells. We therefore envisage to also perform hypoxia imaging of the irradiated tumor / tumor bed during follow-up after chemoradiation by using NIRI and / or other imaging techniques.

The introduction of highly conformal radiation techniques such as intensity modulated radiotherapy allows to increase the dose to the tumor and at the same time spare the surrounding normal tissue. As a consequence clinicians are now increasingly able to shorten radiation schedules by increasing the dose per fraction. By shortening the total treatment time in days we increase tumor control based on the results of clinical studies and radiobiological tumor response models. By integrating IMRT with novel image guidance techniques and high precision linear accelerators it now is even possible to treat selected cancers with only one to four fractions of high doses. This technique is referred to as stereotactic radiotherapy and is mostly used for brain, lung and liver tumors but can be used for many other tumor locations as well. The concept of giving radiotherapy in only one fraction intuitively contradicts the concept of fractionation, which exploits the idea of itinerative reoxygenation after every fraction. In the Division of Radiation Oncology of the University Hospital Zurich stereotactic radiotherapy of small lung tumors has been introduced two years ago with rapidly increasing numbers of patients. Tumor control rates after ablative sterotactic radiotherapy of lung tumors (including single fractions) reach 80 – 100% in multiple phase II clinical trials. In these trials mostly 50-60 Gy in 3-4 fractions were given. So far it is widely unknown if such high doses, which are biologically equivalent to up to 150 Gy in 2 Gy fractions, are really necessary for small tumors. In contrast, large tumors, which are hypoxic, might be resistant even to large single fractions. Such tumors should be better treated with fractionated radiotherapy or radiotherapy should be combined with a hypoxic cell sensitizer. Therefore in an additional phase 1 study we aim to investigate hypoxia imaging before and after large fractions. In this study 5 patients with primary tumors or metastases in the lung and who are treated with stereotactic radiotherapy of 3 x 16 Gy will undergo hypoxia imaging before treatment and one day after the first fraction of 16 Gy. Movement of lung tumors can be compensated by acquisition of 4 dimensional CT, PET or MRI images. Insights into the microenvironmental response to high single fractions might lead to the design of biological response adapted radiotherapy trials also in lung cancer.
In parallel to the prospective phase 1 studies retrospective tissue biomarker analysis will be performed in tissue from previously treated patients with head and neck cancer. Our institution treated appr. 1200 patients with head and neck cancer with modern intensity modulated radiotherapy between 2002 and 2012. The aims of these retrospective studies are:

- Translate innovative tissue biomarkers from preclinical research into the clinic

- Identify candidate biomarkers for validation in prospective phase II study (see work package 2)

- Learn more about prevalence and distribution of tumor hypoxia in different patient cohorts.

Work Package 2: Phase 2 study of hypoxia adapted radiochemotherapy in patients with H&N cancer
Derived from the results of the phase I studies and the results of preclinical groups a hypoxia-adapted radiotherapy regimen will be designed and clinically investigated in a subsequent phase II study. Adaptation of radiotherapy will include:

Hypoxia dependent distribution of radiation dose within the tumor

Increased total dose if tumor hypoxia persists in week 4 after start of radiochemotherapy

If available: combination of radiotherapy with hypoxic cell sensitizer

De-Escalation of total radiation dose to non-hypoxic tumors

In this phase II study 60 patients with T2-4 Nx HNSCC will be randomized to hypoxia-adapted versus standard chemoradiation. Initial hypoxia imaging will be performed as part of staging procedures. Starting with chemoradiation a simultaneous integrated boost will be given to hypoxic tumor and/or metastatic lymph node regions. In case tumor and lymph node metastases are not hypoxic, a lower total dose will be given. All patients with hypoxic tumors will receive repetitive hypoxia imaging. In case of reoxygenation at week four of radiotherapy the treatment plan will be adapted (e.g. the boost to hypoxic tumor regions will be cancelled) and a reduction of the number of fractions will be performed. A reduction of only one or two fractions could already substantially reduce late toxicity and the rate of complications (e.g. risk of osteonecrosis), because patients are usually treated up to the tolerance of the normal tissue.
The major endpoint of the phase II study will be locoregional tumor control. Initial oxygenation status and serial changes of tumor and tumor bed oxygenation will be analyzed and correlated with local and regional tumor control and with tumor persistence. Innovative tissue and serum biomarkers from preclinical research part will be integrated into this study with the aim to identify predictive biomarkers for future clinical use.

Qualitative and quantitative characterization of these (preclinical and) clinical situations with the respective high-resolution approaches will lead to improved understanding on the relevance of tumor hypoxia and reoxygenation for radiation resistance on the individual level, and eventually to personalized radiotherapy protocols with the tumor microenvironment as a critical parameter.