Pharma Strategy Blog

There were a number of fascinating talks over the last four
days covering the latest information on biomarkers and personalized
therapy.  The summary that follows
is highly incomplete, but hopefully reflects some key points I’ve learned at
this excellent AACR meeting.  Unpublished
data reported by the presenters, will not, however be discussed and thus
commentary will be limited to general observations based on published data or
hypotheses evolving from them.

The small intimate event of 300 people with a common focus
and a well organized format contributes significantly to the learning and
offline discussions.  That is a
tribute to not only the organizing committee, but also the co-chairs, Drs Roy
Herbst from MD Anderson and David Carbone from Vanderbilt.  Both of these gentlemen, along with Dr
Paul Bunn from the Colorado Cancer Center, were particularly giving of their
time in helping me understand some of the more complex issues and I’m extremely
grateful to their patience with my many questions.  Any inaccuracies that may result in this report are entirely
mine.

First a little bit a history. 

Not so long ago, patients with non-small cell lung cancer
(NSCLC) would be treated with series of chemotherapy doublets or singlets, with
little regard for the underlying biology. 
As such, there is very little (or small) difference reported in the
literature between the various combinations such carboplatin plus paclitaxel,
gemcitaibine or docetaxel with cisplatin, or vinorelbine or pemetrexed  combinations, for example.  It was very much an empirical way of
looking at the options based on physicians choice depending upon patient
characteristics (stage, age, performance status etc) and physician preference
or experience with managing toxicities.

More recently, it was noticed that patients taking erlotinib
(Tarceva), an EGFR inhibitor, tended to do better if they had an EGFR mutation,
were young, female, non-smoker and with an adenocarcinoma histology.  Meanwhile, pemetrexed (Alimta), was
found to have a greater efficacy benefit in non-squamous compared to squamous
patients.

This represents a small advancement in identifying which
patients are most likely to respond to therapy, thereby sparing the exposure to
a drug that is most unlikely to have any benefit.  Four presentations were particularly intriguing to me and deserve some mention in separate blog posts.

In the first one, Dr Herbst presented the BATTLE (biomarker
based targeted therapy for lung cancer elimination) concept being developed and
evolved at MD Anderson across a large multi-functional team involving
researchers, physicians, and nurses. 
Aside from oncologists, intervention radiologists, diagnostic and
molecular pathologists were also heavily involved, as were the research groups
involved with molecular biology and biomarkers. 

The basic concept was that not all patients are the same and
tumour biopsies could be critical in determining the optimal clinical approach
to therapy in a highly heterogeneous disease.  Having biomarker driven clinical trials and understanding
molecular mechanisms of response will help improve knowledge of the underlying
disease for future studies.

BATTLE1 began in November 2006 and enrolled 341 patients
with NSCLC who had received at least one prior chemotherapy, had a good
performance status and biopsy amenable disease.  Eligible patients were randomized to receive one of four
regimens, depending upon their initial biopsy result, i.e. using adaptive
randomization.  Eleven core
biomarkers were used to help stratify patients and also determine the effect of
therapy before and after treatment, with the primary end point being 8 week
disease control rate: 

• EGFR mutation – Erlotinib
  
• Ras/Raf – Sorafenib
  
• VEGF – vandetanib
  
• Cyclin D1/RXR – Erlotinib plus bexarotene
  

The trial was closed in October 2009 and 227 patients had
evaluable tumour biopsies.

The concept appears to have proven successful and BATTLE2 is
now underway, with the goal of enrolling over 300 patients being evaluated with
different therapies in four new regimens, ie: 

• Erlotinib
• Erlotinib + IGF-1Ri
• Erlotinib + AKTi
• MEKi + AKTi

However,
Herbst noted that the IGF-1R combination may need to be reconsidered following the
recent announcement of the figitumumab futility data.

What I like about this approach is that treatment is based upon a logical design with a strong rationale from the underlying biology, rather than just relying on patient characteristics and physicians (sometimes biased) opinion.  

I do think we will see more of the work from translational medicine and research start to impact patient selection in the future.  Right now, we're just scratching the surface of what's possible.