Advertisements
Feeds:
Posts
Comments

Posts Tagged ‘CHK or BCL2 inhibitors as agents’


DISCUSSION – Genomics-driven personalized medicine for Pancreatic Cancer

Reporter: Aviva Lev-Ari, PhD, RN

[bold face added, ALA]

Integrated Patient-Derived Models Delineate Individualized Therapeutic Vulnerabilities of Pancreatic Cancer –>>> Personalized Tumor Models Could Help Identify Combination Therapies for Hard-to-Treat Cancers

 

Original article

Pancreatic Cancer – Genomics-driven personalized medicine

 

PDAC has a particularly poor prognosis, and even with new targeted therapies and chemotherapy, the survival is poor. Here, we show that patient-derived models can be developed and used to investigate therapeutic sensitivities determined by genetic features of the disease and to identify empirical therapeutic vulnerabilities. These data reveal several key points that are of prime relevance to pancreatic cancer and tumor biology in general.

The Challenges of Using Genetic Analysis to Inform Treatment in PDAC

Precision oncology is dependent on the existence of known vulnerabilities encoded by high-potency genetic events and drugs capable of exploiting these vulnerabilities. At present, the repertoire of actionable genetic events in PDAC is limited.

  • Rare BRAF V600E mutations are identified in PDAC and could represent the basis for targeted inhibition, as our group and others have previously published (Collisson et al., 2012; Witkiewicz et al., 2015).

Similarly,

  • germline BRCA deficiency is the basis for ongoing poly(ADP-ribose) polymerase (PARP) inhibitor clinical trials (Lowery et al., 2011).

As shown here, out of 28 cases, only one genetic event was identified that yielded sensitivity to a therapeutic strategy. In this case, existence of the matched model allowed us to confirm the biological relevance of the

  • STAG2 mutation by showing sensitivity of the model to a DNA cross-linking agent.

Therefore, annotated patient-derived models provide a substrate upon which to functionally dissect the significance of novel and potentially actionable genetic events that occur within a tumor.

Another challenge of genomics-driven personalized medicine is

  • assessing the effect of specific molecular aberrations on therapeutic response in the context of complex genetic changes present in individual tumors.
  • KRAS has been proposed to modify therapeutic dependency to EZH2 inhibitors (Kim et al., 2015), and in the models tested, responses to this class of drugs were not uniformly present in cases harboring mutations in chromatin-remodeling genes.

This finding suggests that, although tumors acquire genetic alterations in specific genes, the implicated pathway may not be functionally inactive or therapeutically actionable. Therefore, annotated patient derived models provide a unique test bed for interrogating specific therapeutic dependencies in a genetically tractable system.

Empirical Definition of Therapeutic Sensitivities and Clinical Relevance

Cell lines offer the advantage of the ability to conduct high throughput approaches to interrogate many therapeutic agents. A large number of failed clinical trials have demonstrated the difficulty in treating PDAC. Based on the data herein, the paucity of clinical success is, most probably, due to the diverse therapeutic sensitivity of individual PDAC cases, suggesting that, with an unselected patient population, it will be veritably impossible to demonstrate clinical benefit. Additionally,

  •  very few models exhibited an exceptional response to single agents across the breadth of a library encompassing 305 agents.
  • We could identify only one tumor that was particularly sensitive to MEK inhibition and another model that was sensitive to
  • EGFR and
  • tyrosine kinase inhibitors.

In contrast to the limited activity of single agents, combination screens yielded responses at low-dose concentrations in the majority of models. Specific combinations were effective across several models, indicating that, by potentially screening more models, therapeutic sensitivity clades of PDAC will emerge. In the pharmacological screens performed in this study,

  • MEK inhibition, coupled with MTOR, docetaxel, or tyrosine kinase inhibitors, was effective in _30% of models tested.
  • Resistance to MEK inhibitors occurs through several mechanisms, including
  • Upregulation of oncogenic bypass signaling pathways such as AKT, tyrosine kinase, or MTOR (mammalian target of rapamycin) signaling.

In the clinic, the MEK and MTOR inhibitors (e.g., NCT02583542) are being tested. An intriguing finding from the drug screen was

  • sensitivity of a subset of models to combined MEK and docetaxel inhibition. This combination has been observed to synergistically enhance apoptosis and inhibit tumor growth in human xenograft tumor models (Balko et al., 2012; McDaid et al., 2005) and is currently being tested in a phase III study in patients with KRAS-mutated, advanced non-small-cell lung adenocarcinoma (Ja¨ nne et al., 2016).

Interestingly, in the models tested herein, there was limited sensitivity imparted through

  •  the combination of gemcitabine and MEK inhibition.

This potentially explains why the combination of MEK inhibitor and gemcitabine tested in the clinic did not show improved efficacy over gemcitabine alone (Infante et al., 2014).

Another promising strategy that emerged from this study involves using

  • CHK or BCL2 inhibitors as agents that drive enhanced sensitivity to chemotherapy.

Together, the data suggest that the majority of PDAC tumors have intrinsic therapeutic sensitivities, but the challenge is to prospectively identify effective treatment.

Patient-Derived Model-Based Approach to Precision Medicine

This study supports a path for guiding patient treatment based on the integration of genetic and empirically determined sensitivities of the patient’s tumor (Figure S7). In reference to defined genetic susceptibilities, the models provide a means to interrogate the voracity of specific drug targets. Parallel unbiased screening enables the discovery of sensitivities that could be exploited in the clinic. The model-guided treatment must be optimized, allowing for the generation of data in a time frame compatible with clinical decision making and appropriate validation.

In the present study, the majority of models were developed, cell lines were drug screened, and select hits were validated in PDX models within a 10- to 12-month window (Figure S7). This chronology would allow time to inform frontline therapy for recurrent disease for most patients who were surgically resected and treated with a standard of care where the median time to recurrence is approximately 14 months (Saif, 2013).

Although most models were generated from surgically resected specimens, two of the models (EMC3226 and EMC62) were established from primary tumor biopsies, indicating that this approach could be used with only a limited amount of tumor tissue available.

In the context of inoperable pancreatic cancer, application of data from a cell-line screen without in vivo validation in PDX would permit the generation of sensitivity data in the time frame compatible with treatment.

[We] acknowledge that model-guided treatment is also not without significant logistical hurdles, including the availability of drugs for patient treatment, clinically relevant time frames, patient-performance status, toxicity of combination regiments, and quality metrics related to model development and therapeutic response evaluation.

Additionally, it will be very important to monitor ex vivo genetic and phenotypic divergence with passage and try to understand the features of tumor heterogeneity that could undermine the efficacy of using models to direct treatment. As shown here, drug sensitivities remained stable with passage in cell culture and, importantly, were confirmed in PDX models, suggesting that the dominant genetic drivers and related therapeutic sensitivities are conserved.

In spite of these challenges, progressively more effort is going into the development of patient-derived models for guidance of disease treatment (Aparicio et al., 2015; Boj et al., 2015; Crystal et al., 2014; van de Wetering et al., 2015).

Several ongoing trials use PDX models to direct a limited repertoire of agents (e.g., NCT02312245, NCT02720796, and ERCAVATAR2015). Given the experience here, PDAC cell lines would provide the opportunity to rapidly interrogate a larger portfolio of combinations that could be used to guide patient care and provide a novel approach to precision medicine.

Validation of this approach would require the establishment of challenging multi-arm or N-of-1 clinical trials. However, considering the dire outcome for PDAC patients and the long-lasting difficulty in developing effective treatments, this non-canonical approach might be particularly impactful in pancreatic cancer.

SOURCE

Witkiewicz et al., 2016, Cell Reports 16, 1–15

August 16, 2016 ª 2016 The Author(s).

http://dx.doi.org/10.1016/j.celrep.2016.07.023

Agnieszka K. WitkiewiczPress enter key for correspondence information
Uthra Balaji
Cody Eslinger
Elizabeth McMillan
William Conway
Bruce Posner
Gordon B. Mills
Eileen M. O’Reilly
Erik S. KnudsencorrespondencePress enter key for correspondence information
Publication stage: In Press Corrected Proof
Open Access

Resource Integrated Patient-Derived Models Delineate Individualized Therapeutic Vulnerabilities of Pancreatic Cancer

Correspondence

awitki@email.arizona.edu (A.K.W.),

eknudsen@email.arizona.edu (E.S.K.)

Accession Numbers: GSE84023

Other related articles on this topic published in this Open Access Online Scientific Journal include the following:

Pancreatic Cancer: Articles of Note @PharmaceuticalIntelligence.com

Curator: Aviva Lev-Ari, PhD, RN

https://pharmaceuticalintelligence.com/2016/05/26/pancreatic-cancer-articles-of-note-pharmaceuticalintelligence-com/

Advertisements

Read Full Post »