What does great science look like?

Roche continues to be a leader in cancer research and development, as evidenced by our extensive pipeline and extensive set of data being presented at this year’s ASCO annual meeting.

Inevitably, with so much going on there will occasionally be disappointing results from our clinical trials. While this is obviously bad news for us, more importantly it means that a potentially medicine won’t get to patients.

We asked three of our senior R&D leaders for their thoughts on clinical development at Roche, and how they see the company continuing to drive progress forward despite setbacks.


From left to right: John Reed, Head of Roche pRED; Sandra Horning, Chief Medical Officer and Head of Global Product Development, Roche; William Pao, Global Head of Oncology, Roche pRED

Roche has experienced a number of disappointments in Phase 3 recently. What went wrong?

[SH] Negative study results come with the territory of being at the forefront of developing new medicines. Not everything is going to work, and if it does, it probably means that we were just lucky in our endeavours to develop ground-breaking medicines for patients who need new options. While it is disappointing when potential new medicines don’t produce the outcomes we hoped they would, they are highly valuable because what we learn from each study shapes future research. Making good decisions about our clinical trials doesn’t mean the outcome will be positive. It means making the best decisions possible based on sound scientific rationale and available data.

Even though the results of our METLung trial for NSCLC weren’t what we hoped for, we learned a great deal about the development of a companion diagnostic and we are conducting a thorough “lessons learned” analysis of the clinical data.

Of course, we need to remember that Roche has an enviable record of multiple Phase 3 successes in oncology, including those that led to the approval of five new molecular entities from 2011 to 2013, as well a number of line extensions.

How do you define a “failed” study?

[WP] Clinical studies are designed to test a hypothesis. Despite being negative, the study can help us to move forward with new questions to address. As long as we know the drug hit target, the study may also teach us about biology that was previously unappreciated.

What is Roche doing to try to prevent this happening again in late-stage clinical development in the future?

[WP] A fundamental understanding of the biology of our targets and diseases is key to our future success.

Can you give us an insight into what goes into the design of a clinical study?

[WP] There are many factors that go into the design of a clinical study. For example, we have to take into account the target of the drug and the importance of that target to a specific cancer type. We also have to consider the number of patients that may have that target and the current treatment options for those patients. We want to develop new drugs that have a significant impact with potentially fewer side effects for those patients.

Has the emergence of personalized medicine changed how the industry approaches clinical development? If so, how and why?

[WP] One of the biggest changes is that we can now understand the molecular basis for why only some patients may benefit significantly from a certain drug. This can facilitate drug development, because we can then target precisely those patients most likely to be helped. Another benefit is that we can potentially minimize side effects by hitting targets specific to cancer cells and not normal cells. A downside, however, is that some of these subsets of patients are small. Identification of those patients using diagnostics will be critically important.

How is Roche going to stay at the forefront of clinical development in what appears to be an ever-more crowded oncology space?

[JR] Roche is differentiated from other companies in several important ways. First, we have the most extensive portfolio of oncology drugs and drug candidates in the industry, with over 30 molecules marketed or in development between Roche, Genentech and Chugai. This abundance of targeted therapeutics provides Roche with rich opportunities to explore synergistic drug combinations – and it is well recognized that combinations of target therapeutics is the key to improving outcomes for cancer patients.

Second, Roche is the industry leader in innovative antibody drug design. For example, last year we launched two milestone therapeutic agents: an antibody-drug conjugate, and a glycol-enhanced antibody. In addition, our research teams at the Roche Group are actively pursuing many exciting new antibody drug platforms for oncology.

Third, our commitment to Personalised Healthcare and the marriage of diagnostics with therapeutics provides a foundation for linking companion diagnostics to every cancer therapeutic. This is a truly differentiated strength of Roche, which enables us to identify the right patient subpopulations for our medicines, thus increasing our chances of success in both early and especially late development.

Test your clinical trials IQ

Clinically challenged?

correct out of . Yikes; you’re certainly below the median. You’d best start swotting now. Our guide to Understanding Clinical Trials should help make sure you get to grips with the basics.

New to this?

correct out of . While you are not at home with your Kaplan-Meier curves, we are certain you will get there. Where to start? Well we think our guide to Understanding Clinical Trials might be a pretty good place.

Getting there… slowly

correct out of . So, you’re pretty… average to say the least! Don’t live life in the middle lane; learn more about by downloading our Understanding Clinical Trials guide now.

Almost excellent

correct out of . Your hazard ratio from this quiz was well under 1, so you are certainly on the right track! Make sure you don’t slip up by downloading our guide to Understanding Clinical Trials.

Statistical supremo

correct out of . Woo hoo - we’re impressed! There is no risk of you being bamboozled by the trial data at this year’s conference.