Two previous articles on this blog have included discussions of the “co-clinical mouse/human trial” strategy for improving mouse models of human cancer, and simultaneously improving human clinical trials of drugs for these cancers. Now comes an article on the use of a co-clinical trial strategy in personalized treatment of non-small cell lung cancer (NSCLC) in the 29 March 2012 issue of Nature. In the same issue of Nature is a News and Views article by Genentech’s Leisa Johnson Ph.D. that provides a minireview of the research article.
As we discussed in our April 15, 2010 article on this blog, the co-clinical trial strategy has been developed by Pier Paolo Pandolfi, MD, PhD (Director, Cancer and Genetics Program, Beth Israel-Deaconess Medical Center Cancer Center and the Dana-Farber/Harvard Cancer Center) and his colleagues.
As discussed in that article, these researchers constructed genetically engineered transgenic mouse strains that have genetic changes that mimic those found in human cancers. These mouse models spontaneous develop cancers that resemble the corresponding human cancers. In Dr. Pandolfi’s ongoing co-clinical mouse/human trial project, researchers simultaneously treat a genetically engineered mouse model and patients with tumors that exhibit the same set of genetic changes with the same experimental targeted drugs. The goal of this two-year project is to determine to what extent the mouse models are predictive of patient response to therapeutic agents, and of tumor progression and survival. The studies may thus result in validated mouse models that are more predictive of drug efficacy than the currently standard xenograft models.
The human clinical trials being “shadowed” by simultaneous studies in mice included Phase 3 trials of several targeted therapies for lung and prostate cancer. Xenograft models in which tumor tissue from the patients had been transplanted into immunosuppressed mice were also being tested in parallel with the genetically engineered mouse models. This project represents the most rigorous test to date of how well genetically engineered mouse models of cancer can predict clinical outcomes.
Our October 28, 2011 blog article, which is mainly a review of a 29 September 2011 Nature article by Nature writer Heidi Ledford, Ph.D., focuses on ways to fix the clinical trial system. Our article includes a discussion of a co-clinical trial published in January 2011. This trial utilized two genetically-engineered PDGF (platelet-derived growth factor)-driven mouse models of the brain tumor glioblastoma multiforme (GBM), one of which had an intact PTEN gene and the other of which was PTEN deficient. In this trial, researchers tested the Akt inhibitor perifosine (Keryx Biopharmaceuticals, an alkylphospholipid) and the mTOR inhibitor CCI-779 (temsirolimus; Pfizer’s Torisel), both alone and in combination, in vitro and in vivo. The drugs and drug combinations were tested in cultured primary glioma cell cultures derived from the PTEN-null and PTEN-intact mouse PDGF-driven GBM models, and in the animal models themselves.
The studies showed that both in vitro and in vivo, the most effective inhibition of Akt and mTOR activity in both PTEN-intact and PTEN-null cells in animals was achieved by using both inhibitors in combination. In vivo, the decreased Akt and mTOR signaling seen in mice treated with the combination therapy correlated with decreased tumor cell proliferation and increased cell death; these changes were independent of PTEN status. The co-clinical animal study also suggested new ways of screening GBM patients for inclusion in clinical trials of treatment with perifosine and/or CCI-779.
The new co-clinical trial reported in the March 2012 issue of Nature
The March 2012 Nature report describes research carried out by a large, multi-institution academic consortium, which included Dr. Pandolfi. It focuses on strategies for treatment of patients with non-small-cell lung cancer (NSCLC) with activating mutations in KRAS (Kirsten rat sarcoma viral oncogene homolog). These mutations occur in 20–30% of NSCLC cases, and patients whose tumors carry KRAS driver mutations have a poor prognosis. Moreover, KRAS is a “hard” or “undruggable” target, and no researchers have thus been able to discover inhibitors of oncogenic KRAS.
Because of the intractability of oncogenic KRAS as a target, researchers have been attempting to develop combination therapies for mutant-KRAS tumors (including, for example, colorectal cancers as well as NSCLCs) that address downstream pathways controlled by KRAS. We discussed examples of these strategies in our book-length report Multitargeted Therapies: Promiscuous Drugs and Combination Therapies, published by Cambridge Healthtech Institute/Insight Pharma Reports in 2011. Strategies discussed in that report are based on the finding that KRAS controls signal transduction via two key pathways: the B-Raf-MEK-ERK pathway and the PI3K-Akt pathway. This is illustrated in the figure at the top of this article. As discussed in our 2011 report, researchers are attempting to develop treatments of mutant-KRAS tumors that involve combination therapies with an inhibitor of the mitogen-activated protein kinase (MEK) together with an inhibitor of phosphatidylinositol 3-kinase (PI3K). Researchers are also attempting to develop combination therapies of MEK inhibitors with standard cytotoxic chemotherapies, which if successful will avoid having to use combinations of two expensive targeted therapies.
In the co-clinical trial that is the focus of the 29 March 2012 Nature research report and News and Views commentary, researchers developed a genetically-engineered mouse model to study treatment of mutant-KRAS NSCLCs with either the antimitotic chemotherapy drug docetaxel alone, or docetaxel in combination with the MEK kinase inhibitor selumetinib (AZD6244, AstraZeneca). In the parallel human clinical trial, researchers are also studying treatment of patients with mutant-KRAS NSCLC with docetaxel alone or docetaxel plus selumetinib. (There is no treatment arm in the human clinical trial in which patients are treated with selumetinib alone, since selumetinib monotherapy of NSCLC patients had shown no efficacy in a previous Phase 2 study; this was confirmed in mouse model studies.)
In humans with mutant-KRAS NSCLC, many tumors with mutations in KRAS have concomitant genetic alterations in other genes that may affect response to therapy. Therefore, the co-clinical trial researchers wished to design mouse models with lung tumors with either Kras mutations alone or with mutations in both Kras and another gene that is often concomitantly mutated in mutant-KRAS NSCLCs in humans. The researchers therefore constructed mouse models with cancers bearing the activating Kras(G12D) mutation, either alone or together with an inactivating mutation in either p53 or Lkb1. The researchers achieved this via a conditional mutation system using nasal instillation of specifically genetically-engineered adenoviruses. As result, a small percentage of lung epithelial cells harbored these mutations. It is from these cells that the NSCLC-like tumors arose, analogous to the clonal origin of sporadic lung tumors in humans.
Of the two tumor suppressor genes that are frequently mutated in human mutant-KRAS NSCLCs and that were modeled by the co-clinical trial researchers, p53, often called the “guardian of the genome”, is familiar to most of you. The other gene, Lkb1 [liver kinase B1, also known as serine/threonine kinase 11 (STK11)], was discussed in an earlier article on the Biopharmconsortium Blog, entitled “The great metformin mystery–genomics, diabetes, and cancer.”
LKB1 (whether in regulation of gluconeogenesis in the liver or in its role as a tumor suppressor) acts by activation of AMPK (AMP-activated kinase, a sensor of intracellular energy status.) In lung cancer (as shown by the same group that performed the 2012 co-clinical trial), LKB1 acts to modulate lung cancer differentiation and metastasis. Germline mutations in LKB1 are associated with the familial disease Peutz-Jegher syndrome, in which patients develop benign polyps in the gastrointestinal tract. Studying a mouse model of mutant-LKB1 Peutz-Jeger syndrome, Reuben J. Shaw (Howard Hughes Medical Institute, The Salk Institute for Biological Studies, La Jolla, CA, who was prominently mentioned in our “great metformin mystery” article) and his colleagues showed that the LKB1-AMPK pathway downregulates the mTOR pathway–specifically the rapamycin-sensitive mTOR complex 1 (mTORC1) and its downstream effector hypoxia-inducible factor-1α (HIF-1α). HIF-1α expression in turn upregulates the expression of its downstream effectors hexokinase II and glucose transporter 1 (GLUT1), which are involved in cellular utilization of glucose. LKB1-deficient polyps in this mouse model thus show increased expression of hexokinase II and GLUT1, resulting in dramatically increased glucose utilization.
In the new co-clinical trial, genetically-engineered mice that showed established lung tumors [as determined via magnetic resonance imaging (MRI)] were randomized to receive either docetaxel, selumetinib, or a combination of the two drugs. For tumors with only a Kras mutation, treatment with docetaxel alone resulted in a modest rate of response, with 30% of mice showing a partial response. Mice that bore mutant-Kras tumors that also had mutations in either p53 or Lkb1 had much lower rates of response to docetaxel monotherapy (5% and 0%, respectively), and more of these mice showed progressive disease on MRI or died of their disease. Of mice treated with the docetaxel/selumetinib combination, those with single-mutant Kras tumors showed a 92% overall response rate, and those with mutant Kras/p53 tumors showed a 61% overall response rate. However, mice with mutant Kras/Lkb1 cancers showed only a modest response to the docetaxel/selumetinib combination; 33% of mice achieved a partial response. The difference in response rate of mice with Kras/Lkb1 tumors to docetaxel/selumetinib compared to the other two genotypes was found to be statistically significant.
Using the genetically-engineered NSCLC mouse model in biomarker development
In human patients in clinical trials or in treatment for their cancers, performing repeated biopsies to monitor treatment is difficult. The co-clinical trial researchers therefore wished to develop less invasive means of monitoring both co-clinical and clinical trials of docetaxel/selumetinib treatment of NSCLC. They therefore tested the use of positron emission tomography (PET) with 18F-fluoro-2-deoxyglucose (FDG-PET) as an indicator of early response to therapy that could be used in the clinic. Prior to its radioactive decay (109.8 minute half -life), 18F-FDG is a nonmetabolizable glucose analogue that moves into cells that is preferentially taken up by high-glucose utilizing cells. The researchers found that both Kras/p53 and Kras/Lkb1 tumors showed higher FDG uptake in vivo in the mouse model than did single-mutant Kras tumors. As expected from the earlier study, GLUT1 expression was elevated in Kras/Lkb1 mutant tumors. In human patients, pre-treatment, mutant KRAS/LKB1 tumors also showed a higher FDG uptake that did KRAS tumors negative for LKB1.
Treatment of the mice with docetaxel alone gave no significant changes in FDG uptake in Kras, Kras/p53, or Kras/Lkb1 tumors in vivo. However, within 24 hours of the first dosing of docetaxel/selumetinib, FDG uptake was markedly inhibited in Kras and Kras/p53 tumors. In contrast, treatment of mice with Kras/Lkb1 mutant tumors gave no appreciable decrease in FDG uptake in these tumors. These results show that early changes in tumor metabolism, as assessed by FDG-PET, predict antitumor efficacy of docetaxel/selumetinib treatment.
The FDG-PET study in this mouse model supports the use of this imaging method as a biomarker to monitor the course of treatment in humans.
Cellular signaling in mutant Kras, Kras/p53, and Kras/Lkb1 tumors
The researchers assessed activation of relevant intracellular pathways in tumors in treated and untreated mice with mutant Kras, Kras/p53, and Kras/Lkb1 lung cancers. They performed these studies using two different methods–immunostaining of cancer nodules for phosphorylated ERK, and immunoblotting of tumor lysates. In untreated mice, Kras/Lkb1 tumors show less activation of ERK than do Kras and Kras/p53 tumors, with Kras/p53 tumors showing the greatest amount of activation of the MEK-ERK pathway. Docetaxel had no discernible effect on signaling via the MEK-ERK pathway. Selumetinib alone resulted in decreased ERK activation in Kras and Kras/p53 tumors, but there was still residual activity. The docetaxel/selumetinib combination, however, was more effective in eliminating ERK activation. Pharmacokinetic studies indicated that selumetinib levels were higher in both serum and tumors of mice treated with docetaxel/selumetinib that in those treated with selumetinib alone; this might account for the more potent suppression of MEK-ERK signaling by the combination therapy as compared to selumetinib monotherapy. The researchers studied MEK-ERK activation (as determined by phospho-ERK staining) in a set of 57 human NSCLC tumors with known RAS, p53 and LKB1 mutation status. As with the tumors in the mouse model, of seven patients whose tumors harbored the KRAS activating mutation, the three patients with concurrent p53 mutations showed higher levels of ERK activation.
The decreased activation of ERK in Kras/Lkb1 tumors suggested that these tumors utilize other pathways to drive their proliferation. On the basis of their prior studies of signal transduction in mutant-Lkb1 lung tumors, the researchers focused on AKT and SRC. Immunoblotting studies showed that Kras/Lkb1 mutant tumors had elevated activation of both AKT and SRC. As one can see from the figure at the top of this article, AKT is a downstream effector of PI3K; since the PI3K/AKT pathway regulates expression of GLUT1 and hexokinase, increased activation of the PI3K/AKT pathway is consistent with the increased uptake of FDG of mutant Kras/Lkb1 tumors. In the figure, SRC (which is not shown) represents one of the major “other effectors” controlled by RAS. These results indicate that concomitant mutation of Lkb1 in mutant-Kras NSCLCs may shift the signaling pathways that drive tumor proliferation from MEK-ERK to PI3K/AKT and/or SRC. This shift would result in primarily resistance of Kras/Lkb1 tumors to treatment with docetaxel/selumetinib.
Long-term benefits of treatment of mice with mutant-Kras and Kras/p53 tumors with docetaxel/selumetinib as opposed to docetaxel monotherapy
The researchers studied long-term treatment of mice with mutant-Kras and Kras/p53 tumors with docetaxel monotherapy versus docetaxel/selumetinib. In mice with mutant-Kras tumors, treatment with docetaxel monotherapy gave stable disease for several weeks, while docetaxel/selumetinib treatment resulted in tumor regression and slower regrowth of tumors. The addition of selumetinib to docetaxel significantly prolonged progression-free survival in these mice. In mice with Kras/p53 tumors, treatment with docetaxel alone resulted in progressive disease, but docetaxel/selumetinib treatment resulted in initial disease regression followed by progression, resulting in prolonged progression-free survival. These results indicate that treatment with combination therapy as opposed to docetaxel alone results in improved progression-free survival, but not cure, in mice with Kras– and Kras/p53-mutant tumors.
The researchers also investigated mechanisms of acquired tumor resistance in mice with mutant-Kras and Kras/p53 tumors, which had been treated long-term with docetaxel/selumetinib. In moribund animals that had received this treatment, all tumor nodules examined showed a recurrence of ERK phosphorylation. This suggested that acquired resistance could be at least in part due to reactivation of MEK–ERK signaling despite ongoing treatment with selumetinib. Evaluation of resistant tumor nodules suggested that more than one mechanism for pathway reactivation was occurring; study of these mechanisms is ongoing.
Conclusions of the new co-clinical study
The results of this co-clinical study predict that docetaxel/selumetinib combination therapy will be more effective than docetaxel monotherapy in several sub-classes of mutant-KRAS NSCLC. This prediction is consistent with the early results of a Phase 2 clinical trial of these two drug combinations in second-line treatment of patients with KRAS-mutant NSCLC.
However, the co-clinical trial also predicts that concurrent mutation of LKB1 in mutant-KRAS tumors will result in primary resistance to docetaxel/selumetinib combination therapy, perhaps via activation of parallel signaling pathways such as AKT and SRC. Since LKB1 status is not being prospectively assessed in the ongoing human clinical trial, the presence of patients with cancers having concurrent LKB1 mutations may diminish the differences between treatment arms based solely on KRAS status. The results of the co-clinical trial suggests that researchers perform retrospective analysis of p53 and LKB1 status in samples from the concurrent human clinical trial. Future clinical trials should then be designed that involve prospective analysis to ensure sufficient enrollment of patients with all three genotypes to enable sufficiently powered sub-group analyses.
Although the results of the co-clinical trial indicate that patients with mutant KRAS/LBK1 tumors be excluded from trials of docetaxel/selumetinib treatment, the research group that has been conducting the co-clinical trial has also been conducting studies that may lead to treatment strategies for KRAS/LBK1 tumors.
The co-clinical trial also allowed researchers to design and validate biomarker strategies, specifically the potential use of the less-invasive FDG-PET to predict efficacy and to monitor treatment. The co-clinical animal-model study also enabled the discovery of mechanisms of both primary and acquired resistance that might benefit future clinical trials and discovery/development of drugs. (The studies on acquired resistance are in a early stage and are ongoing). Any mechanisms of acquired resistance discovered in co-clinical studies should be confirmed in human clinical trials by examining biopsy samples from patients who relapse on therapy. The ability to assess mechanisms of resistance in preclinical or co-clinical animal studies may enable researchers to design rational drug combination strategies that can be implemented in future clinical studies.
The results of the new co-clinical trial strengthens the contention that co-clinical trials in genetically-engineered mice can provide data that can predict the outcome of parallel human clinical trials. Co-clinical trials can also be used to generate new hypotheses for use in analyzing concurrent human trials, and for design of future clinical studies. Moreover, co-clinical trials can result in the validation of improved animal models for human cancers, which can be used in research and preclinical testing of oncology agents, and in validation of biomarkers for clinical studies in oncology. Given the inadequacy of “standard” xenograft models, which is a major factor in the high attrition rate of pipeline oncology drugs, the availability of validated genetically-engineered animal models may be expected to enable improved oncology drug development.
As the producers of this blog, and as consultants to the biotechnology and pharmaceutical industry, Haberman Associates would like to hear from you. If you are in a biotech or pharmaceutical company, and would like a 15-20-minute, no-obligation telephone discussion of issues raised by this or other blog articles, or of other issues that are important to your company, please click here. We also welcome your comments on this or any other article on this blog.