Precision Medicine Modeling Helps Advance Colorectal Cancer Care

Oncology Live®, September 2015, Volume 16, Issue 9

Preclinical mouse models and cell lines link colorectal cancer subtypes subtypes with potential therapies.

Dustin A. Deming, MD

Assistant Professor

University of Wisconsin Carbone Cancer Center

Madison, WI

My Own Cancer Diagnosis: A Doctor’s Story

In treating my patients with colorectal cancer, I see daily how research can provide hope and impact patient outcomes in the clinic. But my understanding goes beyond what I’ve learned as a gastrointestinal oncologist and translational researcher.

Two weeks after starting as a medical oncologist specializing in the treatment of colorectal cancer, I was diagnosed with colorectal cancer at age 31. Having completed radiation, surgery, and chemotherapy, I now understand how a colorectal cancer diagnosis changes every aspect of your life.

I am honored to care for people diagnosed with cancer. In addition, I now better understand the importance of research and the urgency with which further advances in cancer treatment are needed.

With this urgency in mind, my laboratory focuses on fundamentally changing the way in which colon cancer is treated to a more individualized approach. Our lab has developed multiple models to investigate novel therapies and we are now examining rational combinations of therapeutic agents that are already in clinical development for the treatment of subtypes of colon cancer.

Although significant progress has been made during the last 20 years, colorectal cancer (CRC) is a leading cause of cancer-related death in the United States, accounting for nearly 50,000 deaths every year.1 Improved treatment options and biologic markers to predict treatment response are clearly needed.

Colorectal Cancer Mutations

Several key mutations are important in tumor initiation, progression, metastasis, and the response to some therapeutic agents. These mutations include APC (found in 80% of CRC tumors), TP53 (50%), KRAS (35%-45%), PIK3CA (20%-30%), and BRAF (10%), among others.2 Each CRC has been shown to possess two to six driver mutations per tumor.3 To date, KRAS, NRAS, and BRAF mutations are utilized clinically because their presence predicts resistance to the anti-epidermal growth factor receptor antibodies cetuximab and panitumumab.4

With advances in our ability to perform clinical genomic profiling, each histological type of cancer is now better understood as a collection of multiple subtypes characterized by unique mutation profiles. To make precision medicine a reality for patients with CRC, a better understanding of how the molecular profile can help select the best therapies is needed. Giant strides toward achieving this goal have been made with improvements in the CRC preclinical models used for developing treatment strategies.

Figure. Preclinical Models for Investigating Subtypes of Colorectal Cancer.

Colon cancers in mice can be identified with an endoscope and monitored for response to therapeutics (A and B). Fluorescent murine colon cancer spheroids can be generated with specific mutation profiles for translational studies (C and D). We have now generated human CRC spheroids to grow cells that are not able to otherwise be cultured (E and F).

Clinical advancements in targeting subtypes of CRC are starting to be realized as recent advances have been described for tumors that are BRAF mutant, HER2 expressing, or demonstrate microsatellite instability (MSI).

The combination of vemurafenib, cetuximab, and irinotecan demonstrated a 35% response rate in a phase I study in patients with BRAF mutant CRC (5%-10% of all patients).5 The combination of trastuzumab and lapatinib resulted in a 34.7% response rate in patients with HER2-overexpressing CRC.6 In addition, pembrolizumab demonstrated a 40% response rate for patients with high-MSI CRC.7 These exciting studies are demonstrating the promise of targeting subtypes of CRC, and these early results will be investigated further in upcoming clinical trials.

Improved Preclinical Modeling

Genetically Engineered Murine Model

Mouse models have been vital in studying the biology and response to pharmacologic agents in CRCs. However, ApcMin mouse and models using tissue-specific promoters, which are commonly utilized for preclinical studies, are limited in their abilities to address the genetic complexity seen in human cancers.

Our laboratory has developed an innovative adaptable transgenic mouse model of CRC that allows for multiple combinations of mutations to be expressed, thus enabling multiple mutations to be initiated within a single colon cancer.8 Using this model system, we are able to control when, where, and with which mutations the cancers form. These cancers can be followed by murine colonoscopy, and biopsies can be performed pre- and posttreatment for biomarker identification (Figure A and B).

This model has already proved effective in developing targeted strategies that exploit the cancer's mutation profile. Our laboratory has demonstrated benefit for targeting the PI3K pathway in colon cancers with PIK3CA mutations and these therapies are now being evaluated in new clinical trials.9-11

Spheroid Cultures

The capability to investigate subtypes of CRC has also been limited by classic cell culture techniques. The commonly utilized cell lines were selected for their ability to be grown in adherent cultures and not based upon their mutation profile.

To further our investigations into subtypes of CRC, we are developing a library of CRC spheroids in 3-dimensional culture media. We are culturing colon cancers from our novel murine models (Figure C and D), as well as human cancers (Figure E and F). The cancer cells in these cultures form hollow spheres, develop infolding, and can even develop crypt-like structures.

We have now developed multiple lines of spheroids derived from colon cancers with defined mutation profiles. These cells cannot be cultured using adherent cell culture techniques, but can be grown in spheroid culture. This is important because the commonly utilized panels of colon cancer cells do not fully represent the mutation profiles of human cancers.

References

  1. American Cancer Society. Cancer Facts & Figures 2015. Atlanta: American Cancer Society; 2015. http://www.cancer.org/acs/groups/content/@editorial/documents/document/acspc-044552.pdf.
  2. Markowitz SD, Bertagnolli MM. Molecular origins of cancer: molecular basis of colorectal cancer. N Engl J Med. 2009;361(25):2449-2460.
  3. Vogelstein B, Papadopoulos N, Velculescu VE, et al. Cancer genome landscapes. Science. 2013;339(6127):1546-1558.
  4. Tran NH, Cavalcante LL, Lubner SL, et al. Precision medicine in colorectal cancer: the molecular profile alters treatment strategies [published online July 5, 2015]. Ther Adv Med Oncol. 2015;7(5):252-262.
  5. Hong DS, Morris VK, El Osta BE, et al. Phase Ib study of vemurafenib in combination with irinotecan and cetuximab in patients with BRAF-mutated metastatic colorectal cancer and advanced cancers. J Clin Oncol. 2015;33(suppl; abstr 3511).
  6. Siena S, Sartore-Bianchi A, Lonardi S, et al. Trastuzumab and lapatinib in HER2-amplified metastatic colorectal cancer patients (mCRC): the HERACLES trial. J Clin Oncol. 2015;33(suppl; abstr 3508).
  7. Le DT, Uram JN, Wang H, et al. PD-1 blockade in tumors with mismatch-repair deficiency [published online May 30, 2105]. N Engl J Med. 2015;372(26):2509-2520.
  8. Hadac JN, Leystra AA, Paul Olson TJ, et al. Colon tumors with the simultaneous induction of APC, KRAS and PIK3CA mutations still progress through the adenoma-to-carcinoma sequence [published online August 14, 2015]. Cancer Prev Res (Phila). pii:canprevres.0003.2015.
  9. Leystra AA, Deming DA, Zahm C, et al. Mice expressing activated PI3K rapidly develop advanced colon cancer [published online April 23, 2012]. Cancer Res. 2012;72(12):2931-2936.
  10. Deming DA, Leystra AA, Nettekoven L, et al. PIK3CA and APC mutations are synergistic in the development of intestinal cancers [published online May 27, 2013]. Oncogene. 2014;33(17):2245-2254.
  11. Deming DA, Leystra AA, Farhoud M, et al. mTOR inhibition elicits a dramatic response in PI3K-dependent colon cancers. PLoS One. 2013;8(4):e60709. doi:10.1371/journal.pone.0060709.

At the University of Wisconsin Carbone Cancer Center, we are developing improved preclinical models that will help precision medicine become a reality for patients with CRC, as we aim to identify the patient populations most likely to benefit from new therapies.