Novel Therapeutics Target Gene Mutations, Tumor Suppressors

David Sidransky, MD

The genes that populate the squamous cell carcinoma of the head and neck (SCCHN) neighborhood have some new residents in the form of previously unknown genetic methylation patterns and new mutations.

These genetic and epigenetic alterations could prove valuable for diagnosing SCCHN, tracking its progress, and developing targeted therapies, according to David Sidransky, MD, who chaired an educational session at the ASCO meeting on developments in SCCHN.

Sidransky is a professor of oncology, otolaryngology, pathology, genetics, urology, and cellular and molecular medicine at Johns Hopkins University in Baltimore, Maryland. Other speakers at the session were Christine Chung, MD, who outlined new strategies for targeting SCCHN mutations, and Ezra Cohen, MD, who highlighted pivotal trials for disease therapy.

Sidransky’s detailed talk offered 2 major takehome points:

• At least 2 genes, KIF1A and EDNRB, appear to be highly methylated in SCCHN patients when compared with healthy people who are at risk for SCCHN from smoking and other environmental exposures. The methylation of these genes might aid in disease diagnosis and monitoring.
• New mutations have cropped up in genes that are not usually associated with SCCHN, including NOTCH1, which is a known genetic mutation in other types of cancer. However, unlike other cancers, NOTCH1 seems to be inactivated in SCCHN.

Table. Top 12 Genes With Mutations in Squamous Cell Carcinoma

Gene

Detection

TP53

43%

CDKN2a (p14ARF)

39%

FGFR3

22%

CDKN2A (p16)

18%

MET

14%

SMAD4

13%

HRAS

10%

PIK3CA

9%

EGFR

3%

PTEN

3%

KRAS

2%

BRAF

1%

The mutation data were obtained from the Sanger Institute Catalogue of Somatic Mutations In Cancer Website, http://www.sanger.ac.uk/ cosmic. Bamford et al (2004) The COSMIC (Catalogue of Somatic Mutations in Cancer) database and Website. Br J Cancer. 2004. 91(2):355-358.

Targeting Tumor Suppressors

Tumor suppressors dominate in SCCHN, and the conventional wisdom has been that they are difficult to target with therapeutic agents. But that mindset needs to change, said Chung, associate professor of oncology at the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins Hospital in Baltimore, Maryland.

Chung acknowledged that when it comes to targeting tumor suppressors, “it’s much more difficult to restore lost activity than to inhibit increased activity [as with oncogenes]. You have to circumvent the loss by targeting downstream effectors,” she explained. “I think this dogma [that] tumor suppressors cannot be targeted is a bit dated as we know more about these mutations. Knowing more of the [mutation] characteristics...may actually provide novel therapeutic targets.”

For instance, p53 is one of the most common genes with mutations in SCCHN, Chung said. But simply eliminating p53 function will not work, as the gene is linked to multiple pathways; knocking out p53 can actually increase activity in these other pathways.

But there are therapies in development that will target p53 by pinpointing the dominant downstream pathway, thereby inhibiting the mutation’s loss of function. These include PRIMA-1 and CP-31398, both of which are stabilizing agents of p53 mutations. Also, RITA and Nutlin-3 prevent ubiquination of the dominant downstream pathway, Chung said.

Christine Chung, MD

Additionally, research has shown that the p53 mutation has its own activity, opening up the possibility for approaching the mutation as if it were an oncogene. “Some people argue that mutant p53 can function as a pro-oncogenic gain of function mutation, and you can actually target it,” Chung said.

Another key to developing targeted agents and changing clinical practice is predictive (rather than prognostic) biomarkers. Again, unlike with gain-of-function mutations that occur in a hot-spot, loss-of-function mutations occur in multiple spots, requiring multiple assays to detect all possible abnormalities.

Human papillomavirus (HPV) status and smoking history are great prognostic biomarkers but not predictive ones, Chung said. But combining the HPV profile with multiplex biomarkers that characterize and catalog all mutations would serve as a predictive marker. “That justifies getting more tumor biopsies and blood samples,” Chung said.

Chung cited 2 current studies that will use HPV as predictive biomarkers for cetuximab (Erbitux) sensitivity. The phase II E1308 trial will study induction therapy followed by concurrent intensity-modulated radiation therapy (IMRT) and cetuximab treatment. The phase III RTOG 1016 trial will pit IMRT and cisplatin therapy against an IMRT and cetuximab protocol.

“

I think this dogma [that] tumor suppressors cannot be targeted is a bit dated as we know more about these mutations. Knowing more of the [mutation] characteristics... may actually provide novel therapeutic targets. ”

—–Christine Chung, MD

Targeted Therapy Trials

Based on published data, single-agent therapy has demonstrated modest results in SCCHN. For instance, single-agent response rates to epidermal growth factor receptor- (EGFR) targeted therapies and tyrosine kinase inhibitors (TKIs) have ranged from 13% with cetuximab to 2% with gefitinib (Iressa).

So clinicians need to maintain realistic expectations when it comes to how well singleagent therapy will work in SCCHN. “Only a minority of patients are going to benefit from a specific agent,” said Cohen, assistant professor of medicine at the University of Chicago Medical Center. “There is a great deal of heterogeneity in the biology of these tumors.”

Ezra Cohen, MD

But current clinical trials aim to show that traditional targeted therapy in SCCHN should not be written off yet. “Let’s not throw everything out and say ‘There are too many mutations and there’s too much heterogeneity. This is a disease that is not treatable,’” Cohen said.

Trials are underway in EGFR-refractory patients that focus on the PI3K/AKT/ mTOR signaling pathway. Cohen is involved in a randomized phase II trial that will test everolimus (Afinitor) versus placebo as adjuvant therapy in patients with locally advanced SCCHN. Patients will first undergo curative-intent therapy, and once there is no evidence of disease, will be randomized to everolimus or placebo for 1 year. The trial also will assess activity in the PI3K/mTOR activity pathways fr om tumor tissue samples, Cohen said.

Investigations also will look at PANHER blockade because all HER/ErbB family members are expressed at high levels in SCCHN, Cohen said. A completed phase II trial by Cohen’s group compared treatment with a TKI, afatinib, with cetuximab. The RECIST partial response rates with afatinib came in at 14.5% versus 3.2% with cetuximab (P = .0331). However, progression-free survival was essentially identical for the 2 agents.

“Although afatinib may be more effective at inducing response in these patients, the benefi t may be comparative to cetuximab,” Cohen said. “Nonetheless, there is now a TKI that does appear to be as eff ective as the approved monoclonal antibody.”

Finally, there are angiogenic inhibitors. Single-agent activity has been modest for drugs such as sunitinib (Sutent) and sorafenib (Nexavar) with RR less than 5%. Although vascular toxicities have been comparable with these agents in SCCHN to other disease sites, recurrent SCCHN tumors are often close to major blood vessels, which increases the potential for serious bleeding events, Cohen explained.

Instead, angiogenic inhibitors may have more value in the chemoprevention setting, and with an agent that has the ability to inhibit EGFR and vascular EGFR-2. Cohen described an upcoming randomized phase II study that will compare the dual-inhibitor vandetanib with placebo in patients at high risk for SCCHN.

“When we find these patients and match them with the right agent, the potential for benefit can be substantial,” Cohen stated.