Pharma Strategy Blog

The growing importance of KRAS in ovarian cancer and its impact on pipelines

We've heard a lot about the impact of KRAS in colorectal cancer as a useful biomarker for determining whether or not to treat with EGFR therapy, depending on whether the mutation is wild-type or mutated, but now new evidence has emerged for it's possible role in ovarian cancer.

A paper by Ratner et al., from Yale University, looked at the relationship between an increased risk of OC and a KRAS variant.

Overall, the data demonstrated that a variant of the KRAS oncogene was present in 25% of all ovarian cancer patients. The same variant was also found in 61% of ovarian cancer patients with a family history of breast and ovarian cancer. Up until now though, it has proven impossible to tell which women with a strong family history of breast and ovarian cancers would go on to develop the disease. These results therefore offer some new clues in the puzzle.

Related research from another team from the UK, Denmark and US, published last year, looked at tagging single-nucleotide polymorphisms (SNPs) in candidate oncogenes and the susceptibility to ovarian cancer in approx. 1,800 women with invasive ovarian cancer compared with controls (n=3,000). The goal was to identify moderate/low-risk susceptibility alleles of the proto-oncogenes BRAF, ERBB2, KRAS, NMI, and PIK3CA, but no evidence of ovarian cancer association was found with these SNPs. When stratified by histologic subtype, however, one common variant allele have borderline evidence of association with epithelial ovarian cancer.

What was interesting to me in the latest research from Yale is that all the women in the study had a strong family history of cancer, but only half had known genetic markers of ovarian cancer risk such as BRCA1 or BRCA2 mutations. The results, albeit from a small sample (n=157), suggest that the KRAS variant may offer a more sensitive than currently available. 

The Yale group had previously established that the KRAS-variant is not somatic but germline, meaning it is identical in person's normal and tumour tissues, thereby enabling the researchers to collect primarily germline DNA from either blood or saliva, rather than from tumour biopsy samples.  This is huge from a patient perspective and physician point of view in terms of ease of use and convenience.

This study is important because ovarian cancer is the single most deadly form of women's cancer and is usually diagnosed in advanced stage disease.  Part of the reason is because of the lack of known risk factors or genetic markers of risk.  To this end, the authors concluded:

"Our findings strongly support the hypothesis that the KRAS-variant is a genetic marker for increased risk of developing ovarian cancer, and they suggest that the KRAS-variant may be a new genetic marker of cancer risk for hereditary breast and ovarian cancer families without other known genetic abnormalities."

What we now need is validation in large scale clinical trials to determine whether the genetic marker can be used commercially to determine prognostic risk for ovarian cancer earlier.  If the results ultimately prove useful in clinical trials, then this finding relating to the KRAS variant may well have important implications for future therapeutic strategies and pipeline development in ovarian cancer as well as for prognostic testing and earlier detection for improved outcomes. 

Ratner, E., Lu, L., Boeke, M., Barnett, R., Nallur, S., Chin, L., Pelletier, C., Blitzblau, R., Tassi, R., Paranjape, T., Hui, P., Godwin, A., Yu, H., Risch, H., Rutherford, T., Schwartz, P., Santin, A., Matloff, E., Zelterman, D., Slack, F., & Weidhaas, J. (2010). A KRAS-Variant in Ovarian Cancer Acts as a Genetic Marker of Cancer Risk Cancer Research DOI: 10.1158/0008-5472.CAN-10-0689 

Quaye, L., Song, H., Ramus, S., Gentry-Maharaj, A., Høgdall, E., DiCioccio, R., McGuire, V., Wu, A., Van Den Berg, D., Pike, M., Wozniak, E., Doherty, J., Rossing, M., Ness, R., Moysich, K., Høgdall, C., Blaakaer, J., Easton, D., Ponder, B., Jacobs, I., Menon, U., Whittemore, A., Krüger-Kjaer, S., Pearce, C., Pharoah, P., & Gayther, S. (2009). Tagging single-nucleotide polymorphisms in candidate oncogenes and susceptibility to ovarian cancer British Journal of Cancer, 100 (6), 993-1001 DOI: 10.1038/sj.bjc.6604947

Chin, L., Ratner, E., Leng, S., Zhai, R., Nallur, S., Babar, I., Muller, R., Straka, E., Su, L., Burki, E., Crowell, R., Patel, R., Kulkarni, T., Homer, R., Zelterman, D., Kidd, K., Zhu, Y., Christiani, D., Belinsky, S., Slack, F., & Weidhaas, J. (2008). A SNP in a let-7 microRNA Complementary Site in the KRAS 3' Untranslated Region Increases Non-Small Cell Lung Cancer Risk Cancer Research, 68 (20), 8535-8540 DOI: 10.1158/0008-5472.CAN-08-2129

Combined NOTCH and EGFR inhibition leads to synthetic lethality in breast cancer

This was a great paper I found while browsing My 6 Sense last night, a free but very cool app on my iPhone that takes my Google Reader rss feeds and intuitively tries to select the individual items (e.g. clinical journal articles on cancer) that are most relevant to me out of hundreds and thousands of items I have stored in there. 

Overall, the app does a pretty good job, despite the technical nature of many of the feeds I have in there (mostly cancer journals and science/tech articles), if this one is anything to go by.  My only beef is that when sharing or streaming scientific papers with long titles to Twitter, the 140 character limit often truncates the url, leaving some unhappy readers in the lurch as I found out very quickly from the @replies!

Not wishing to disappoint followers, I re-sent the link on Twitter but also thought it would be interesting to review here on Research Blogging as well.  

Now, basal-like breast cancers (BLBC) are not very nice, in fact they're highly aggressive and may confer a slightly poorer prognosis as a result, yet few if any current therapies are actually targeted at this particular subtype.  

The good news is that BLBC do over-express EGFR, but the bad news is that resistance to current EGFR targeted therapies has been well documented, suggesting that something else is at play that causes the resistance to develop.  Finding out what that might be may possibly offer a new therapeutic intervention strategy that could be explored, either in combination or in sequence.

It has also been observed in other studies that Notch signalling and activation is prevalent in BLBC.  The authors of this paper therefore decided to see what would happen with combined EGFR and Notch inhibition in animals studies:

"We show that although inhibition of either EGFR or Notch signaling alone is insufficient to suppress basal-like breast tumor cell survival and proliferation, simultaneous inhibition uncovers a synthetic lethal relationship between these two oncogenic pathways. This lethality is due in part to significant decreases in AKT activation caused by combined EGFR and Notch inhibition. 

Expression of the activated form of Notch1 restores AKT activity and enables cells to overcome cell death after dual-pathway blockade. Combined pathway inhibition is also dramatically more effective at suppressing tumor growth in mice than blocking EGFR or Notch signaling alone."

In breast cancer it would appear that phosphatidylinositol 3-kinase (PI3K)–AKT signaling is the best characterised mechanism to date of Notch-dependent resistance to cell death. AKT is also a recognized mediator of resistance to EGFR inhibition. Thus a much clearer picture begins to emerge whereby compensatory pathways of resistance also need to be considered in the treatment strategy.

Whether a combination of EGFR, AKT and PI3K inhibitors would work in women with basal like breast cancer isn't yet clear, but based on the preclinical evidence from this and other research, it's probably a strategy worth investigating.

You read it here first ;)

Dong, Y., Li, A., Wang, J., Weber, J., & Michel, L. (2010). Synthetic Lethality through Combined Notch-Epidermal Growth Factor Receptor Pathway Inhibition in Basal-Like Breast Cancer Cancer Research, 70 (13), 5465-5474 DOI: 10.1158/0008-5472.CAN-10-0173

Modified measles virus shows potential for treating childhood brain tumors

Whew, that headline caught my attention just now on the Nationwide Children's Hospital site while searching for information on medulloblastoma, a form of brain cancer that tends to affect children.

After all the recent brouhaha in the UK over Andrew Wakefield being struck off for professional misconduct relating to his role in the MMR vaccination scandal, tensions may be raised at the idea of using a modified form of the measles virus to treat a childhood brain cancer.

According to the hospital's press release:

Vaccine strains of measles virus have been used to kill tumor cells in a number of tumor types including one type of adult brain tumor. One vaccine strain of measles, the Edmonston strain, targets the cell surface receptor CD46 to gain entry into susceptible cells. “This preference most likely explains the efficacy of Edmonston strains in killing tumor cells, given the high level of expression of CD46 in multiple tumor types,” said Dr. Raffel. “It is also the reason we chose to explore a modified Edmonston’s strain of measles virus for use in medulloblastoma.”

You learn something new every day.

This news comes hot on the heels of some other research conducted in pediatric brain tumors at St Jude Hospital in Memphis, which reported that:

"Twelve percent of tumors in this study had extra copies of the gene PDGFRA. Researchers reported similar gene-expression patterns associated with high levels of PDGFRA were also found in childhood tumors without extra copies of the gene. Extra copies of PDGFRA were even more common in tumors from children who had received brain irradiation for treatment of earlier cancers."

As someone who experienced a malignant childhood cancer myself, it's always gratifying to see research ongoing in this area and improvement in 5-year survival rates.

Are we becoming immune to daily gene announcements?

The other day I clicked on a link someone shared on Twitter about new findings in breast cancer genes, which sounded really cool and interesting.  On clicking through it turned out to be a rather disappointing Reuters release on several fronts:

1. Where's the link to the original article (in these days of social sharing leaving it out is plain lazy)?
2. What/where are the "five common genetic factors" exactly?  No mention is made of them.
3. What does this finding actually mean?
4. What's the 'wow' factor here rather than the 'so what' factor?
   

I don't know about anyone else, but daily press releases and assaults in the media about identification of yet more gene blah blah that may or may not be relevant becomes anaesthetising after a while.

After a while, I finally tracked down the actual article in Nature Genetics myself to see what the data actually said (reference below with link to the paper for those interested).

What the researchers did was interesting; they conducted a large genome-wide study (GWAS) in around 8,000 people, approx. half of whom had a family history of breast cancer and half did not (controls). They then studied the DNA of another 24,000 women, with and without breast cancer.  By analysing over half a million SNP's, they were able to identify new 5 loci on chromosomes 9, 10 and 11 which if present, represent an increased risk of developing breast cancer.  This is in addition to 13 gene variations already identified in previous research.

Many of you will be aware of two high risk genes which are more likely to be defective in someone with breast cancer, known as BRCA1 and BRCA2.  These genes often confer higher risk of breast or ovarian cancers and can be tracked in families to try and pick up the cancer earlier. More recently, they have become targets for therapeutic intervention with PARP inhibitors such as AstraZeneca's olaparib in BRCA-positive breast and ovarian cancers.  More on this in another post. 

But what if these 5 new gene variations are also important and early identifiers of increased risk?

The reason that these findings are important is that as Prof Bert Vogelstein of Johns Hopkins never tires of pointing out at AACR meetings, most cancers, including breast cancer, are often diagnosed relatively late in their development and people die from the last 3 years of metastatic disease because the cancer went undetected for 20 odd years. 

GWAS as a field is also becoming more relevant, highlighting distinct subsets of patients who can then be evaluated for prognosis, biomarkers, efficacy and tolerability differences.

We still have very few methods of efficient and effective early detection, but what GWAS may allow us to do in future when we have more complete information (and gene testing is cheap, reliable and more widely available) is to potentially run screens to identify those most at risk for cancer.  These people could then be monitored more closely and watched diligently for early signs of cancer appearing.  Early stage cancer is potentially curable with surgery.  Inevitably, preventative studies could also be considered for high risk groups in future.

So these findings will have little immediate impact now, but in future if replicated, they could form the backbone of a comprehensive strategy to identify those people most at risk from developing cancer.  If we think about it, the enormous cost of end of life medical care massively dwarfs what the cost of earlier intervention and potential cure via surgery with or without neoadjuvant therapy would be.  Ultimately, we need this kind of data to move to a wellness model rather than a palliative model.

I would like to acknowledge my buddy Scott Hensley at NPR Health for prodding me into explaining the rationale behind my grump on Twitter about the flood of 'so what' rather than 'oh wow' gene data that constantly abounds daily.  If you're not following him on Twitter or listening to his awesome NPR Health podcasts on iTunes or the web, you should consider it!

Turnbull, C., Ahmed, S., Morrison, J., Pernet, D., Renwick, A., Maranian, M., Seal, S., Ghoussaini, M., Hines, S., Healey, C., Hughes, D., Warren-Perry, M., Tapper, W., Eccles, D., Evans, D., Hooning, M., Schutte, M., van den Ouweland, A., Houlston, R., Ross, G., Langford, C., Pharoah, P., Stratton, M., Dunning, A., Rahman, N., & Easton, D. (2010). Genome-wide association study identifies five new breast cancer susceptibility loci Nature Genetics DOI: 10.1038/ng.586

The biology of pancreatic cancer

The cause of pancreatic cancer is still unknown, although there are risk factors involved, such as increased exposure to tobacco smoke and a family history of the disease.  It is usually detected late when advanced stage disease has set in because most patients are asymptomatic for a long period and then it is too late.  This raises the question of how can we learn and understand more about the biology of the disease as well as develop early warning signs via biomarkers used in screening?  

It was therefore fascinating to be browsing through the latest edition of the New England Journal of Medicine over coffee this morning to discover an article on the medical progress in pancreatic cancer while simultaneously noticing an item on a new potential biomarker for the disease in Twitter!  The NEJM review hasn't been published online yet, so I'll add the reference later.

Essentially, the NEJM article reviewed the biology of what we know of pancreatic cancer so far:

1. Results from successive accumulation of gene mutations (average of 23 per tumour)
2. Most patients with malignant disease carry 4 or more mutations (eg KRAS, CDKN2A, TP53, DPC4)
3. It's extremely heterogeneous (more difficult to treat)
4. Formation of dense tumour stroma (dermoplastic reaction)
5. It is poorly vascularised (due to stellate cells and activation of TGFβ1, PDGF and FGFR)
6. Poor prognosis and treatment resistance has been conferred by the presence of cyclooxegenase-2 and PGFR, VEGF, SPARC and Hedgehog overexpression to name a few

The poor vascularisation of the tumour may well explain why anti-angiogenic therapies have not yielded promising results in the clinic to date.  

Treating advanced cancer is always more difficult than earlier cancers in the neoadjuvant and adjuvant settings.  The big challenge with pancreatic cancer to date, though, has been finding useful biomarkers for earlier detection, which would most likely impact outcomes for the disease.  According to the NEJM article:

"The overall 5-year survival rate among patients with pancreatic cancer is <5%."

That's a pitiful number that needs to change.

There are a number of agents in the clinic for the treatment of advanced pancreatic cancer including inhibitors of SPARC, MEK, Mucin-1, IGF-1R, Hedgehog, PDGF and FGFR, Src, RAS, JAK-STAT and TNFα, for example. 

Many of these agents are looking at single agent studies or in combination with the current standard of care, gemcitabine, itself with limited effectiveness. It is, therefore, quite hard to imagine that blocking one pathway alone is likely to be the panacea for such a hard to treat and highly complex cancer, as the graphic to the left shows (click to see the detail).

Source: Genway Bio

Meanwhile, progress may be possible on the detection front.  Goicoechea et al. reported in PLoSONE that a type of protein called "palladin" is produced in large amounts in groups of cells in the stroma surrounding pancreatic tumours, known as the 'tumour nest'.   By measuring the levels of palladin in patient samples, doctors could have an improved way to screen for the disease, potentially detecting it earlier than current tests. 

However, I say potentially, because these findings while interesting, need to be validated in large scale phase III clinical trials before we can really be certain of their long term validity.  Nevertheless, it is a promising start for a devastating disease.   

If there was one cancer that would hugely benefit from better earlier detection and diagnosis, it's pancreatic cancer.  We need more cowbell.  That is all.

Goicoechea, S., Bednarski, B., Stack, C., Cowan, D., Volmar, K., Thorne, L., Cukierman, E., Rustgi, A., Brentnall, T., Hwang, R., McCulloch, C., Yeh, J., Bentrem, D., Hochwald, S., Hingorani, S., Kim, H., & Otey, C. (2010). Isoform-Specific Upregulation of Palladin in Human and Murine Pancreas Tumors PLoS ONE, 5 (4) DOI: 10.1371/journal.pone.0010347

#AACR10: The Kinase Dilemma

There are a lot of clinical trials out there right with tyrosine kinase inhibitors; unfortunately many will fail because they were rushed into phase II or III trials without thinking through all the options.  There are, however, some smart companies out there who do think.

What was noticeable at AACR this year, was the surfeit of posters and presentations regarding logical combinations designed to eliminate escape routes and hence resistance.  For example, cross-talk is a common problem between ligands, eg IGF-1R and EGFR, so combining the two may reduce the problem but that isn't the whole story.

Feedback loops also exist, so targeting PI3-kinase alone is less likely to be effective than targeting both PI3-kinase and mTOR.  Neal Rosen from MSKCC showed some interesting data to this effect and argued cogently that oncogenes tend to lead to constitutive negative feedback.  He also noted that the BRAF mutation predicts for sensitivity to MEKi, for example.  Michael Korn also discussed the feedback activation loop between the RAS-ERK and PI3K pathways and how the inhibition of autophagy (where cells self digest themselves) can enhance apoptosis and the anti-tumour effect with smart combinations.

Targeting both MEK and AKT may therefore also have more effect than either alone, as you can see in the chart below: 

Source: Array Biopharma

In a recent trial reported at the the ASCO GI meeting in January, Merck described an elegant design where IGF-1R, EGFR and AKT inhibitors were all combined to target advanced pancreatic cancer, with promising early results.  I thought this was a prescient approach at the time, since it clearly sought to eliminate both cross-talk and feedback, so it was interesting to see numerous researchers advocating similar approaches in different tumour types based on the overexpression profiles at AACR last week. The design is based on rational biochemistry, which regular PSB readers will know I'm a big fan of, rather than randomly adding a kinase inhibitor to whatever is the standard chemotherapy of the day in a haphazard blunderbuss approach.

There are a number of MEKi and AKTi inhibitors out there (I counted nearly a dozen last time I checked), as well as a plethora of PI3-kinase and mTOR inhibitors, either alone or in combination.  Merck and AstraZeneca announced an interesting deal earlier this year to jointly pursue research with their AKT (MK-2206) and MEK (AZD6244) inhibitors.  This collaboration makes a lot of sense biochemically.  Novartis (a client) have one of a broadest kinase pipelines in the industry and just added to it prior to AACR in a deal with Array BioPharma to license their MEK inhibitors, of which ARRY-162 is the lead candidate. 

The compounds that ultimately win the race may not necessarily be the ones furthest ahead in clinical trials right now, but the ones with the smartest clinical trial designs to eliminate some of the issues associated with kinase inhibition - cross-talk, feedback, feed-forward loops and additional mutations. 

MEKi and AKTi are two of my favourite kinase approaches right now because they offer the flexibility to add to existing TKI's such as erlotinib, sorafenib or everolimus, for example, potentially improving the outcomes further in a variety of different cancers, never mind the future combination possibilities.  It's going to be a very interesting and hot area to watch in the near future, that's for sure.

If you have any thoughts or questions on this fascinating topic, please do add them in the comments below.

 

#AACR10: The Cancer Genome

One of the challenges of scientific conferences, at least the ones I attend, is that the investigators have a tendency to rush through a high volume of highly interesting slides at a fast pace, meaning ones notes end up looking like a chicken scratched them out and then a spider with wet feet ambled over them for good measure.  Teasing out the nuggets is sometimes a challenge, "what was I thinking when I scrawled that?"

And so it was with the plenaries at the American Association of Clinical Research meeting in DC this week. Looking at the notes from Charles Sawyers and Bert Vogelstein's plenary talks on the prostate cancer and the cancer genome, respectively, is particularly apt as you can see in the photo on the right!

Every time I listen to one of Prof Vogelstein's talks I learn something new and the last lecture makes more sense too.  He noted that 78 cancer exomes have been published to date.  The exome is the 1% of the human genome that is functionally relevant for phenotypic changes.  Exomes are comprised of short segments of DNA called exons, and provide the genetic blueprint for proteins.

The question is how many genes are altered in a typical solid tumour?Vogelstein showed that this varies by tumour type.  For example, in pancreatic cancer, there are an average of 44 non-silent mutations, while for more complex cancers such as melanoma and lung cancer, there are 100-200, 30-80 in breast, colon and brain cancers.  The lowest are seen in leukemias and medulloblastomas, at around 10.

Most mutations in a cancer are actually passengers though, so the big needle in the haystack question is working out which are the drivers? 

The other interesting fact I learned was that approx. 90% (263) of drivers are actually suppressors, as in tumour suppressor genes, and virtually all are found in 12 key pathways (eg NOTCH, PI3K/mTOR, RAS/RAF etc) and only 10% (33) are actually oncogenes.

However, the elephant in the room, as Vogelstein accurately alluded to is tumour heterogeneity.  It is the very diversity that makes it difficult to understand, describe and target appropriately with therapeutics to improve outcomes in humans.  After all, with respect to tumour suppressors, Vogelstein's pithy statement:

"You can't target something that isn't there!"

Rings very true, which is frustrating to say the least.  Much progress has been made in cancer research to date, but we still have a long way to go.  

The confluence of basic research, applied research, clinical trials, and therapeutics at this AACR meeting at least reassured me that the time from bench to bedside is improving, as are the technologies and strategic thinking based on a deeper understanding of the underlying biology. There was an urgency here in DC that I hadn't noticed before.  Long may it continue.

Zen and the art of Cancer Research at #AACR10

Yesterday afternoon I returned from DC and the annual American Association of Cancer Research (AACR) meeting tired but exhilarated.

Despite the Europeans unfortunately stranded by the Icelandic volcano, business continued as usual with many presenters doing their talks by teleconference to live slides in the room.  Quite a few researchers also FedExed their posters, so people could still peruse them and an envelope was provided for business cards or questions.

As we Brits say, "Keep calm and carry on."

There were approximately 18,000 delegates and 6,000 accepted posters this year, a sign of the times that the meeting is getting much bigger and cancer research is an active area of research.

The AACR staff did a fantastic job behind the scenes, were very responsive in whatever help or support was needed and also very active in social media, which was encouraged between delegates in order to share science and learnings with others in different sessions, those who were stranded and others who could not attend, but had an interest in the area.  

In general, I think most of us used common sense and tweeted interesting and general highlight points made, but did not tweet material that was clearly labelled as unpublished, in press or from private discussions in poster sessions.  If you want to find out more than the headlines, that's the point of travelling to the live meetings in person.

All this activity was very encouraging to see, because webcasts from key oral sessions and plenaries went up very quickly online for others to see.  The missing Europeans possibly speeded this up, but AACR have a history of sharing science, as past meeting podcasts and interviews will demonstrate on the website.  It also fosters comradeship and open science in the face of a common goal - curing cancer.  

Here are the webcasts to date and also the link to podcasts and media teleconferences.  Dr Ed Kim from MD Anderson did a superb job explaining the complexities of the BATTLE trial in simple, easy to understand language.  I'm actually going to be listening to some of the webcasts and podcasts myself over the weekend because there were so many sessions one couldn't attend those in parallel, so it's really useful to see them up there. 

If you are interested in cancer facts and figures, this short video from the AACR team may well surprise you: 

It's always easy to criticise what organisations are doing if things are not exactly to our liking, but in this case, I genuinely want to say publicly, "Nice job, AACR!"

Twitter had several impacts for me at the meeting.  I must confess that I dashed over to another session because Dr Naoto Ueno, an inflammatory breast cancer specialist at MD Anderson and a translational researcher, Dr Wafik El Deiry, at Hersey were both tweeting what looked to be a more interesting session than the one I was in (shhhh)!  Such is the power of persuasive tweets.  Twitter was also great for finding and meeting up with people one had been tweeting with, but not yet met.  Chatting in the cloud is great, but meeting in real life also solidifies those relationships for the future and I was honoured to meet many really awesome people at AACR.  I sincerely hope to catch those I missed at ASCO in June as it was a hectic meeting and trying to see everyone as well as attend sessions and posters was near nigh impossible.

It seemed particularly apt that a new Moleskine was christened at the beginning of the meeting for Dr Charles Sawyers opening plenary talk on prostate cancer and now the notebook is nearly a third full of chicken scratch notes already!

Still, many of you will be waiting with baited breath to hear what were the big picture emerging trends I noticed at this year's AACR? 

Here are some quick thoughts of the top of my head:

1. The cancer genome is now dominant and understanding the biology is very much a focus, even from manufacturers.
2. Free plasma DNA is going to be critical for biomarker research.
3. RNA seems appear in almost every other abstract (more later).
4. GWAS is becoming more relevant, highlighting distinct subsets of patients who can then be evaluated for prognosis, biomarkers, efficacy and tolerability differences.
5. Combination targeted therapy is the smart way to go to eliminate cross-talk, feedback and feedforward loops.  They may also reduce some acquired resistance, at least for a time.
6. Dr Bert Vogelstein noted that there are currently 12 critical pathways that impact cancer cell survival. Understanding these more will lead to new and better therapies.
7. Despite all the above, cancer is a heterogeneous disease and much work still needs to be done in unravelling the mysteries behind it.

Over the next few days, I'll be posting more in-depth thoughts about some of the key new TKI classes that are in the clinic, as well as some of the hot research topics that are emerging.  I have a lot of notes and over 500 poster handouts to condense in a Zen-like fashion before that can be done, not to mention some glorious blisters from miles of walking!  

Do feel free to add any thoughts, questions or immediate observations you may have from the meeting in the comments below in the meantime, it's always good to hear from others.

Photo Credit: Amazon

New developments in castration resistant prostate cancer - MDV3100

Last night I received an alert from Medivation announcing that an article would be published in today's online The Lancet regarding their prostate cancer agent, MDV3100.  Sure enough, here's the article:

Although the trial is an early phase I/II study, encouraging antitumour effects were reported in 140 men, of whom 65 were chemotherapy naïve.  Time to disease progression (TTP) was 47 weeks, which is decent considering the men in the study were all castration resistant and likely had advanced disease.  

Obviously, it is too early to tell yet quite what Medivation and Astellas have here, but to put things in context, docetaxel (Taxotere), was approved for hormone refractory prostate cancer (HRPC) with prednisone in androgen independent (hormone refractory) metastatic prostate cancer with a median overall survival of 18.9 months (approx. 85 weeks) in a large phase III trial. 

We must also remember that in another more recent trial, it was demonstrated that men who received cabazitaxel (sanofi-aventis' follow on taxane to docetaxel) lived a median of 15.1 months (68 weeks), compared with 12.7 months (57 weeks) for those who receive mitoxantrone, a difference that was statistically significant.

That said, the side effect profile of chemotherapy is certainly not benign, with significant risk of myelosuppression and infectious complications.

However, what was encouraging about the Medivation trial was early Kaplan Meier response curves, which demonstrated those men who had no previous chemotherapy generally did better than those who received it in terms of both PSA and radiological progression.

At the AACR Molecular Targets meeting last November, Charles Sawyers gave a keynote talk on prostate cancer and looked at the various opportunities for investigating new therapeutics in clinical trials for prostate cancer.  He suggested that circulating tumour cells (CTCs) might be a more useful marker for disease status than PSA.  In The Lancet study, the authors noted that:

"We recorded early post-treatment conversions from unfavourable to favourable in 75% of patients not exposed to chemotherapy and in 37% of those who were exposed. Decreases in PSA were generally associated with parallel falls (or no progression) in CTCs, but this finding was not consistent, suggesting that these measures assess different aspects of the malignant process. PSA decreases might in some cases be an indicator of the mechanism of action of MDV3100 as an androgen-receptor antagonist rather than an actual antitumour effect. However, the benefit of MDV3100 on several assessments, including CTCs and radiological time to progression, suggest that MDV3100 does have a true antitumour effect."

The other interesting observation that Sawyers made was that traditionally, early prostate disease is treated with oral hormonal therapies and once castration resistance sets in, the patient is referred to an oncologist for consideration of chemotherapeutic options.  With the advent of new therapies with different MOAs in development such as MDV3100, abiraterone (Cougar/J&J) and Provenge (Dendreon), the lives of men with prostate cancer can potentially be extended further before stage IV metastases sets in. 

At AACR last November, Sawyers also discussed the importance of androgen receptor signalling in the disease (previously discussed here with simple models) and the data from this trial seem to bear out his elegant theory.  I'd like to see more analysis though and learn whether those men who had AR amplified prostate cancer did better on MDV3100 than those who did not?

All in all, this is an exciting time for men with prostate cancer.  Dendreon's Provenge may well be the first new drug that fits between the hormonal therapies and chemotherapy, since the FDA PDUFA date is May 1st, although since that is a Saturday, we may well here the Go/No Go news on Friday 30th April.  

In the meantime, the annual AACR meeting in DC this weekend can't come soon enough!

Scher, H., Beer, T., Higano, C., Anand, A., Taplin, M., Efstathiou, E., Rathkopf, D., Shelkey, J., Yu, E., & Alumkal, J. (2010). Antitumour activity of MDV3100 in castration-resistant prostate cancer: a phase 1–2 study The Lancet DOI: 10.1016/S0140-6736(10)60172-9

The molecular genetics of aggressive B-cell lymphomas

As many readers here on PSB know, I've not been a big fan of genome-wide molecular profiling, preferring an oncogene addiction approach to drug development and targeted cancer therapies. However, every once in a while something comes along that stops you in your tracks and makes you think differently.

This morning I was reading the latest copy of the New England Journal of Medicine over coffee and was fascinated by a review article by Drs Lenz and Staudt at the NCI on the molecular genetics of diffuse large-B-cell lymphomas (DLBCL), which account for 30-40% of newly diagnosed lymphomas.

As the review article points out, it is well know that different subsets of diffuse large-b-cell lymphomas are associated with different overall survival rates after initial anthracycline based therapy.  For example, it is more favourable in people with PMBL and the GCB subytype but less favourable in those with the ABC subtype.  R-CHOP therapy has improved survival in people with ABC, but the cure rates are still lower than those with the GCB subtype.  Gene expression signatures can help identify the subtypes and predict survival rates:

Source: NEJM

Current therapeutic treatment with chemotherapy has made headway in improved survival, but in order to make further headway, new approaches are very much needed.  Targeted therapies have now begun to expand clinical trial options. 

The article talks about numerous pathways, but I particularly liked this one, which details the Nuclear Factor kB (NFkB) signalling pathways in normal and malignant lymphocytes:

Source: NEJM

Essentially, signalling is initiated when a SRC family kinase ( SFK) phosphorylates tyrosines in the immunoreceptor tyrosine-based activation motifs (ITAMs) on B-cell subunits.  SYK is then recruited to the ITAMs through the SH2 domains and becomes active.  Many of you will remember SYK inhibition from previous posts on Rigel's fostamatinib, a SYK inhibitor.  I think Celgene also mentioned a SYK inhibitor in early development at their recent R&D Day, although the Rigel-AZ is further ahead but more erratic in it's results, at least in immune disorders.  A more recent paper in Blood looked promising in NHL and CLL, though.  Companies are clearly starting to look at specific inhibitors in the downstream pathway for lymphomas and chronic lymphocytic leukemia.

What's interesting about the cartoon above is that you can also see that phosphatidylinositol-3-kinase (PI3-kinase) is activated in parallel, activating the mTOR pathway.  These two targets are getting a lot of attention from Pharma in the clinic, especially in leukemias and lymphomas, and we may well see more of their latest development at AACR next week and ASCO in June.  Exelixis and their partner, sanofi-aventis (a client), for example, have already announced 12 abstracts at ASCO, including 6 on their PI3K and mTOR inhibitors, but they are focusing on lung cancer, a much more difficult carcinoma, rather than NHL, where there is a strong rationale.

It's good see new treatment modalities being tested in leukemias and lymphomas and not just solid tumours, where most companies inevitably focus due to the larger population sizes.  That said, the challenge in lymphoma is going to be identifying rational combinations that kill lymphoma cells synergistically.  As we learn more about the underlying biology of the disease, targets and biomarkers, so more effective and less toxic solutions may evolve.

Lenz, G., & Staudt, L. (2010). Aggressive Lymphomas New England Journal of Medicine, 362 (15), 1417-1429 DOI: 10.1056/NEJMra0807082 

Friedberg, J., Sharman, J., Sweetenham, J., Johnston, P., Vose, J., LaCasce, A., Schaefer-Cutillo, J., De Vos, S., Sinha, R., Leonard, J., Cripe, L., Gregory, S., Sterba, M., Lowe, A., Levy, R., & Shipp, M. (2009). Inhibition of Syk with fostamatinib disodium has significant clinical activity in non-Hodgkin lymphoma and chronic lymphocytic leukemia Blood, 115 (13), 2578-2585 DOI: 10.1182/blood-2009-08-236471

The potential for PI3K pathway inhibitors in Lung Cancer

Life is about stories, events and even coming back full circle sometimes. I experienced that moment of déjà vu this morning on opening a link to an article that a colleague kindly sent from Avrum Spira's lab that was published in Science Translational Medicine, simply titled:

Airway PI3K Pathway Activation Is an Early and Reversible Event in Lung Cancer Development

Many moons ago I completed my doctoral thesis on the early detection of lung disease at King College London. Much of it was about the impact of smoking on the airways. Halfway through the 3 year MRC grant, while listening to some evening lectures at the Brompton Hospital by Prof Peter Barnes and Steven Holgate, I realised that the real answer lay in understanding the biochemical and molecular biology pathways… only we were working on the physiology of breathing patterns. An A-ha moment to be sure.

Still, I was able to pick up differences in breathing patterns between healthy non-smokers and those who were medium to heavy smokers.

The unaswered question, though, was why?

Prof Barnes loved to put up charts of pathways with a black background and different targets meticulously coded with coloured highlighter pens… PDE, leukotrienes, cholinergic mast cell activation and many others sprang to life as he walked students through the various possible mechanisms where pharmaceutic intervention might have a role to play.

Fourteen years on, I don’t think a cure for asthma has been found and many of those pathways have not yielded much in chronic obstructive pulmonary disorder (COPD) either, except perhaps anti-cholinergics such as tiotropium bromide (Spiriva) as maintenance therapy in controlling bronchospasm. My suspicion at the time was that the mechanisms underlying inflammation and aggravation of the airways by chemicals was key to unlocking why some people get lung cancer, others COPD and still others were unaffected.

Fast forward to the article.

Last week, Gustafson et al., published a paper in Science Translational Medicine suggesting that phosphatidlyinositol 3-kinase (PI3K) pathway activation is an early event in the development of lung cancer.

In particular, they found elevated PI3K in the airways of smokers with lung cancer compared to smokers without lung cancer:

Source: Science Translational Medicine 

The researchers validated their hypothesis that activation of PI3K is an early event in the development of lung cancer by comparing the cytology of airway epithelium from healthy smokers to those with mild to moderate airway dysplasia that reflects an early noncancerous abnormality, often considered a precancerous state.  The theory being that if smokers with dysplasia have higher PI3K pathway activation then this would suggest P13K pathway activation precedes the development of lung cancer. 

The results showed increased expression of genes induced by PI3K pathway in high risk smokers and those with lung cancer. This may well explain the dysplasia and inflammation seen, which you can see in the chart below:

Source: Science Translational Medicine 

One key question this research raises, but does not answer, is whether P13K pathway inhibitors have a role to play in the treatment of lung cancer?

Thinking about my doctoral research on early detection of lung disease, this statement made me sit up and pay attention:

“These results suggest that deregulation of the PI3K pathway in the bronchial airway epithelium of smokers is an early, measurable, and reversible event in the development of lung cancer and that genomic profiling of these relatively accessible airway cells may enable personalized approaches to chemoprevention and therapy.”

Whoa! Early detection of lung cancer? 

Forgive my excitement and enthusiasm, but it’s kind of cool to see read that scientists can actually find something biochemically active that may represent a new target for intervention fifteen years after you realise that early detection might be possible, if only you knew what the target might be.  That's the stuff of <goosebumps>.  

PI3-kinase is a fascinating target that quite a few biotech and pharma companies are looking at as a potential mechanism for the treatment of cancer. I’ve written and blogged extensively on this (see here, here, here, here and here) for some examples.

Interestingly, many of the PI3-kinase compounds in early clinical development are investigating the agents in leukemias, lymphomas, breast, ovarian and gastric cancer or basket solid tumour trials. Some of the interesting small biotech companies in this field include Calistoga, Semafore and Exelixis and big pharma is represented by Merck, Novartis, sanofi-aventis and Roche.  Only two agents seem to be specifically looking at non-small cell lung cancer (NSCLC) though - no prizes for guessing which ones!:

• Everolimus (Novartis), a combined mTOR-PI3 kinase inhibitor, in phase II development.
• GDC-0941 (Roche/Genentech), which is looking at whether the new agent adds anything to the current standard of care in NSCLC ie carboplatin + paclitaxel + bevacizumab and is currently recruiting patients.

Novartis (a client) have completed their trial according to the clinical trials database, so it will be interesting to see if there are any data at ASCO in June that will help us learn more about the disease.

If the data from the Novartis or Roche trials are positive, the companies may be very interested to read Gustafson et al.’s suggestion:

“Our work further suggests that additional lung cancer chemoprevention trials either targeting the PI3K pathway or measuring airway PI3K activation as an intermediate endpoint are warranted.”

Chemoprevention is a big word in cancer research. 

In two weeks time, I'll be at the AACR annual meeting in DC (do let me know if you are going and would like to meet up for a coffee and chat) and hopefully there will be more about this fascinating pathway then.  Watch this space!

Gustafson, A., Soldi, R., Anderlind, C., Scholand, M., Qian, J., Zhang, X., Cooper, K., Walker, D., McWilliams, A., Liu, G., Szabo, E., Brody, J., Massion, P., Lenburg, M., Lam, S., Bild, A., & Spira, A. (2010). Airway PI3K Pathway Activation Is an Early and Reversible Event in Lung Cancer Development Science Translational Medicine, 2 (26), 26-26 DOI: 10.1126/scitranslmed.3000251

Blocking POLQ gene could improve radiotherapy results

This was a very interesting article I came across in Cancer Research while looking for something else. Serendipity is a great thing sometimes.  A quote in the abstract caught my eye:

"We have conducted a small interfering RNA (siRNA) screen of 200 genes involved in DNA damage repair aimed at identifying genes whose knockdown increased tumor radiosensitivity."

Essentially, the researchers found that if they blocked the POLQ gene, which has a role in repairing damaged DNA, then radiotherapy was more effective.  What's more, POLQ knockdown has minimal effect on normal tissue, so blocking the gene did not have any effect on the sensitivity of normal tissue to radiation. This makes the gene an ideal target since it is more specific to, and abundant in, cancer cells.

The article stated that:

"Radiotherapy is a vital tool in the management of cancer patients. It is often given with curative intent either alone or with chemotherapy in patients with diseases as diverse as head and neck, cervix, bladder, and non–small cell lung cancer. The radiation dose that can safely be delivered to patients is limited by the dose tolerances of surrounding normal tissues. It is anticipated that the effectiveness of radiotherapy would be improved if tumor cells could be rendered more sensitive to ionizing radiation (IR) without altering the sensitivity of normal tissues."

It is therefore possible that POLQ inhibition might be used clinically to cause tumour-specific radiosensitisation.  In other words, if a new drug could be developed to block POLQ it might allow radiotherapy to be more effective without impacting normal surrounding tissues too much.

Higgins, G., Prevo, R., Lee, Y., Helleday, T., Muschel, R., Taylor, S., Yoshimura, M., Hickson, I., Bernhard, E., & McKenna, W. (2010). A Small Interfering RNA Screen of Genes Involved in DNA Repair Identifies Tumor-Specific Radiosensitization by POLQ Knockdown Cancer Research, 70 (7), 2984-2993 DOI: 10.1158/0008-5472.CAN-09-4040

Aeterna Zentaris and Keryx's Perifosine gets fast track designation for colorectal cancer

Last month at the ASCO GI Symposium, Keryx and Aeterna Zentaris reported statistically significant benefit in survival from updated results of a randomized, double-blind, placebo-controlled phase II study of KRX-0401 (perifosine) for the treatment of advanced metastatic colorectal cancer. 

The study was based on 35 evaluable, but heavily pre-treated patients, with relapsed or refractory metastatic colon cancer were randomized to receive capecitabine (Xeloda) from Roche at 825 mg/m2 BID on days 1 - 14 every 21 days plus either perifosine (P-CAP) or placebo at 50 mg daily (CAP).

The results showed that the P-CAP arm had 1 PR and 8 SD for an overall response rate of 64%, while the CAP arm had 0 PR and only 3 SD for a response rate of 27%.

In addition, looking at the overall survival, the data favoured the P-CAP arm a MOS of 15.3 weeks compared with 6.8 weeks for CAP alone.

Typical adverse events reported appeared to be relatively mild:

"The P-CAP combination was well-tolerated with Grade 3 and 4 adverse events of > 10% incidence for the P-CAP arm versus CAP arm as follows: anemia (15% vs. 0%), fatigue (0% vs. 11%), abdominal pain (5% vs. 11%), and hand-foot syndrome (30% vs. 0%)."

Perifosine is also being developed for the treatment of multiple myeloma, which is currently in phase III development.  The compound is an inhibitor of Akt and phosphoinositide 3-kinase (PI3K), as well as other pathways such as JNK and MAPK. These pathways are associated with programmed cell death (apoptosis), cell growth, cell differentiation and cell survival as shown in the diagram below:

Source: Aeterna Zentaris

Interestingly, this morning's news brought an announcement that the FDA have granted Fast Track Designation for the perifosine in advanced, refractory colorectal cancer.  Based on the phase II data, a randomised phase III trial is expected to begin this quarter based on a similar trial design.

Clearly, it will be a little while before the data is available, but definitely one to watch out for in the future given new therapy options are always needed in the refractory cancer setting.

Antisoma and Novartis lung cancer drug ASA404 flops in phase III

This morning I did a double take at an email alert that flashed through my email:

"ATTRACT-1 phase III trial of ASA404 halted following interim analysis"

Ugh, that's not good news considering the phase II data previously discussed in January looked very promising with a solid improvement (source data), ie median time to tumor progression was 5.5 months by investigator assessment and median survival was 14.9 months when used in combination with carboplatin-paclitaxel compared to the standard alone.

The Antisoma press release was very downbeat and clearly they were not expecting to receive any more funding from their partner, Novartis:

"The planned interim analysis of data from the ATTRACT-1 phase III trial of ASA404 in previously untreated non-small cell lung cancer (NSCLC) has shown that continuation of the trial would be futile, as there is little or no prospect of demonstrating a survival benefit with ASA404 in this setting. The ATTRACT-1 trial will therefore be halted."

Novartis (a client, but not on this particular agent) bought the global rights to ASA404 (vadimezan) from Antisoma, a British company, approximately three years ago. ASA404 was leading a new class of chemotherapy agents known as vascular targeting or disrupting agents, designed to inhibit angiogenesis and disrupt the flow of blood to tumours. The concept works in a different way to Roche and Genentech's bevacizumab (Avastin), which starves tumours from growing new blood vessels and proliferating the cancer mass.  ASA404 exerts direct and indirect effects on endothelial cells rather than the VEGF receptor, as this diagram shows:

Source: Novartis Oncology

Novartis and Antisoma were also developing ASA404 in breast cancer as well as non-small cell lung cancer, so we'll have to wait and see what happens to that program going forward.  For now, the negative news in lung cancer meant that investors cleared out their shares, sending the stock into a major freefall this morning.

It's been a tough year for cancer drugs in lung cancer in 2010 already, following several spectacular rounds of futile data from Pfizer's figitimumab, Novelos's NOV-002 and now ASA404, all of which had promising but early data in phase II, only to stumble in phase III.

An overview of Roche's oncology pipeline

At last weeks investor meeting held by Roche in downtown Wall Street, the Board reviewed the pipeline opportunities in a number of areas.  Earlier this week I wrote about the non-oncology pipeline and today will form an overview of the cancer drugs in development.

One of the things that Roche is renowned for is life cycle management.  They do this better than many in the industry in my opinion and it makes an enormous difference not only to continuity, but also long term revenues and performance. Too many companies take a short term view and do not think ahead to the future. This is a big mistake. Perhaps they get bogged down in classic silos or management do not see it as a priority, but it does make a difference. 

Why?

Well, for starters, think about the basics of marketing. It is much easier to sell new products to existing customers than it is to sell existing products to new customers and even harder to sell new products to completely new customers.  

Thus life cycle management is a smart strategy and done well, enhances the experience for everyone involved whether employees or investors.  I only wish more companies paid closer attention to this important aspect of Pharma marketing.

The other thing I like about Roche's approach to R&D is rigorous and strong proof of concept studies (usually in phase II). Between Roche and Genentech, they both do this particularly well in oncology, it seems to be their signature. This partly explains why they mostly end up with a continuous wall of data across several products including trastuzumab (Herceptin), rituximab (Rituxan) and bevacizumab (Avastin). Of course, negative trials do occur but overall, they seem to have more positive trials than not. This partly explains why they have fewer phase III flops than say, Pfizer, because they spend the time in phase II working things out rather than rushing aggressively ahead on the basis of early evidence.

So what did we learn from the pipeline presentations last week?

There are late stage oncology products in development that look promising.

One example is trastuzumab-DM1, which is basically modified Herceptin with a potent cell killing agent, DM1, bolted on. The goal is to improve the action of Herceptin in metastatic breast cancer, and at the same time invetigate whether the xenograft data in a variety of cancers (breast, ovarian, lymphoma and prostate) with an armed antibody is an effective strategy in people. Recent phase II data from the San Antonio Breast Cancer Symposium in heavily treated women with metastatic breast cancer look encouraging. Phase III trials have already begun and if all goes well, filing is currently anticipated by 2012.

Also potentially strengthening the breast cancer franchise is pertuzumab, a monoclonal antibody that targets HER2. Early phase I trials in several cancer types produced so-so results, but more recent phase II data in breast (combined with Herceptin) and ovarian cancer (in combination with gemcitabine) published this month in JCO look interesting.

Perhaps the most exciting compound though, is PLX4032/RG7204, a BRAF inhibitor being evaluated in malignant melanoma. Currently available data suggests survival is improved by 6 months so the big question is what causes resistance to develop and how this can be overcome. Data on this compound is expected at ASCO, where many are keen to see how it stacks up with ipilimumab (BMS).

In hematologic malignancies, GA101 or galiximab is being evaluated in non-Hodgkin's Lymphoma (NHL) and chronic lymphocytic leukemia (CLL). It appears to target a different part of the CD20 isotope than rituximab and this may increase it's efficacy. Phase III trials began in 4Q09 thus it will be a little while before we see some results. There is clearly an unmet medical need in the 3rd line refractory disease for more tolerable agents and rituximab is very much the bedrock of treatment for both across multiple lines of therapy either alone or in combination with chemotherapy. A similar agent will likely have good take up with the right approach.

Genentech are investigating various new and improved approaches to angiogenesis, but these are in much earlier development and the bar is very high with bevacizumab (Avastin), even for the company who manufacture it. The list of anti-angiogenesis compounds that didn't make it to market is very long indeed.

I've left the best to last, as hedgehog (Hh) signalling is one of my favourite pathways - it always reminds me of the cheerful cartoon character, Sonic the Hedgehog. RG3616, licensed from Curis, is currently in phase II trials for advanced basal cell carcinoma and trials are also underway for medulloblastoma. If interested, you can find out more about the pathway and the science here.

Overall, the oncology pipeline has a nice mix of follow on compounds to strengthen life cycle management with a raft of monoclonal antibodies with different targets in new cancer types.  Such a strategy should reduce risk and drive the future bottom line if the promise delivers in phase III trials.

Baselga, J., Gelmon, K., Verma, S., Wardley, A., Conte, P., Miles, D., Bianchi, G., Cortes, J., McNally, V., Ross, G., Fumoleau, P., & Gianni, L. (2010). Phase II Trial of Pertuzumab and Trastuzumab in Patients With Human Epidermal Growth Factor Receptor 2-Positive Metastatic Breast Cancer That Progressed During Prior Trastuzumab Therapy Journal of Clinical Oncology, 28 (7), 1138-1144 DOI: 10.1200/JCO.2009.24.2024

Makhija, S., Amler, L., Glenn, D., Ueland, F., Gold, M., Dizon, D., Paton, V., Lin, C., Januario, T., Ng, K., Strauss, A., Kelsey, S., Sliwkowski, M., & Matulonis, U. (2009). Clinical Activity of Gemcitabine Plus Pertuzumab in Platinum-Resistant Ovarian Cancer, Fallopian Tube Cancer, or Primary Peritoneal Cancer Journal of Clinical Oncology, 28 (7), 1215-1223 DOI: 10.1200/JCO.2009.22.3354