[M_NotableLabs]

Hi I'm Matt, one of the founders of Notable Labs. Here are more specifics on our process:

Step 1: The patient undergoes surgery to remove their brain tumor, which is sent to our lab where the cells are grown. Step 2: We use lab robotics to apply thousands of combinations of FDA-approved drugs directly to the cancer cells. Step 3: Notable Labs analyzes the test results and provides them to the patient’s doctor, who can prescribe the drugs immediately without a clinical trial. Step 4: Each treatment result allows us to algorithmically improve the drug selection process for future patients.

Glad to answer any questions!

[refurb]

What is in this set of "thousands of FDA-approved drugs"? Just oncolytics? Or more broad than that?

A few potential challenges:

1. Just because a combination of drugs works on a cell culture, doesn't mean it works in a human body; there are just so many more variables in vivo

2. You're just testing for efficacy, not safety; what happens if one drug potentiates the other and results in the patients getting very sick or dying? Likely a small risk since most drugs list contraindications, but a few problems might slip through

3. How are you planning on dealing with reimbursement? Insurance companies (Medicare/Medicaid included), won't pay for cancer drugs unless there is proof that they work; that proof is usually a clinical trial or at least some published data supporting it; there may be some pushback if insurance companies either don't believe your data or just don't know how to interpret it

Not trying to shit on your company, just curious if you've looked at these issues yet. It's definitely a cool idea!

[M_NotableLabs]

Thanks for your questions! We are testing both cancer and non-cancer drugs.

1. We agree, the reason we are using cell culture is because it is very fast and allows us to test many combinations at once. We have improved standard cell culture conditions and continue to do so to mimic the patient's brain as much as possible.

We confirm our results using in vivo mouse models where we have injected a patient's brain tumor cells into the mouse's brain.

2. Another great question, we prioritize the results based on safety first, then efficacy. Since these are exclusively approved therapies there is existing software that can predict toxicity interactions. We will also have a group of pharmacists that will validate the safety of the combinations.

3. This was one of the first hurdles we encountered after we started the company. The drugs we are testing are mostly off-patent generic drugs that are cheap for patients to buy directly without reimbursement from their insurance company.

If it is a more expensive drug, we can help the patient and their doctor appeal to get reimbursement on the rationale that the standard of care for grade 4 brain cancer is ineffective.

We will be starting a feasibility clinical trial as soon as we can for the purpose of publishing data on the safety and feasibility of this approach.

[tgokh]

If you're using mouse models to confirm cell culture results, roughly how long does it take to go from receiving the tissue sample to sending out a report to the patient's physician?

What type of responses to your reports have you gotten from practicing oncologists?

[vobios]

> You're just testing for efficacy, not safety

The drugs being tested are all already approved, so safety would not be really a concern.

> Insurance companies (Medicare/Medicaid included), won't pay for cancer drugs unless there is proof

After some time to collect the data, you could fairly easily compare outcomes of patients who undergo this test versus those who do not.

[refurb]

The drugs being tested are all already approved, so safety would not be really a concern.

Not true at all. Two drugs could be very safe on their own, but in combination could cause serious adverse events or even lead to death (combining some drugs with grapefruit juice can be bad). When drugs are approved by the FDA, the obvious combinations are sometimes tested for (or theoretical interactions are called out). New combinations are basically an unknown.

After some time to collect the data, you could fairly easily compare outcomes of patients who undergo this test versus those who do not.

Unless you are running a controlled clinical trial, I'm not sure anyone would trust the data. There is too much variability among patients, that if not controlled for, could seriously skew results. Sure it might indicate a promising lead, but I doubt that's enough for insurance companies to start paying for it.

[vobios]

> Two drugs could be very safe on their own, but in combination could cause serious adverse events

That is true. My assumption was that only single compounds or previously tested combinations are suggested.

> Unless you are running a controlled clinical trial, I'm not sure anyone would trust the data.

I agree. I just meant it would be relatively easy to collect enough preliminary data to have a reasonable idea of how a proper clinical trial would go. It's not like going from an animal model to human trials or from a tiny cohort of late-stage patients to a larger more heterogenous group.

[refurb]

It's not like going from an animal model to human trials or from a tiny cohort of late-stage patients to a larger more heterogenous group.

Fair point. A retrospective analysis would provide some data. The problem is insurance companies have pretty high standards (at least for cancer drug that cost a lot)!

[ejstronge]

> The drugs being tested are all already approved, so safety would not be really a concern

I think the problem is that Notable Labs would be testing combinations of FDA-approved drugs:

  Step 2: We use lab robotics to apply thousands of combinations of FDA-approved drugs...

> After some time to collect the data, you could fairly easily compare outcomes of patients who undergo this test versus those who do not

Until then, this won't be broadly available. I do think such a comparison would be interesting.

I guess you'd need to compare the expected benefit of using the Notable Labs-suggested combination to using the standard of care. You'd need to have cells from patients who did not use the Notable Labs-suggestion. This requirement could turn an 'easy comparison' into a lengthy clinical trial.

[M_NotableLabs]

Thanks for your questions!

Yes we'll be testing combinations, as this brain cancer has not responded historically to single agent therapy for decades.

With combinations of existing drugs it is possible to affect multiple driver pathways in the tumor at once. Similar to the dramatic improvement of AIDS therapy and infectious disease treatment with cocktails, tumors could respond much differently to combinations instead of a "one at a time approach" that has resulted in very small survival gains in this disease.

Please see this paper for more detail on the scientific approach to mutli-agent combinations to target many pathways at once. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4226667/

Patients have been taking the CUSP-9 protocol without any serious adverse affects on a compassionate use basis. This has now moved into a prospective clinical trial for recurrent glioblastoma patients in Germany http://www.anticancerfund.org/projects/cusp9-a-combination-o...

Personalized medicine services like this could allow clinicians to move beyond traditional randomized controlled trials. Commercial services such as Foundation Medicine provide test results to oncologists who then use their discretion to treat the patient outside of a clinical trial with scientific rationale based on the test itself as opposed to a standard of care.

In a severe disease like Glioblastoma where average survival is 15 months, if the majority of patients using a system like Notable Labs live substantially longer, is a control arm necessary? The HAART cocktail for AIDS treatment never went through a phase 3 trial, so there is precedent in high need diseases.

I also would like provide a link to a recent documentary that highlights the combination approach in brain cancer by telling the story of a 20 year survivor of Glioblastoma, Ben Williams. Other patients who safely used combinations of existing drugs are featured as well. http://www.survivingterminalcancer.com/

[newman8r]

I have not looked into your company - but what's your opinion on crowdsourcing some of this? Some of these devices are remarkably similar to 3d printers and not so tough for the hobbyist to assemble.

If you do that then the community can start banding together to test formulations and then send the promising ones to you for further analysis.

Good luck. I love the idea and I personally have a few ideas of my own in the field that I want to start in 2016/2017

Edit * of course everyday people won't necessarily have access to approved therapies - but perhaps they can test other compounds and environmental conditions (or get the FDA to somehow approve tiny little test pellets for different drugs that would be legal for hobbyists to use in testing)

[vobios]

> I think the problem is that Notable Labs would be testing combinations of FDA-approved drugs

I assume that only previously used combinations are suggested.

> This requirement could turn an 'easy comparison' into a lengthy clinical trial.

But at least you would have a fairly reasonable idea of how a clinical trial would go. It's not like going from an animal model to human trials. I assume they can get a large enough sample size as they can probably recruit patients easily for what can be viewed as a second opinion.

[untilHellbanned]

Kudos to your efforts to help cancer patients and their families.

1) These types of experiments have long been go on in labs all around the world. Why did YC think your approach was unique? Why didn't they just fund one of the dozens of academic/industry labs in e.g., Boston, already doing the same?

2) Many people have doubts about the utility of the results obtained from treatments on ex vivo tumor cells relating to the in vivo situation. Was YC aware of this?

I'm all for YC going into new areas, but those two points above make me concerned that they are considerably out of their leagues when it comes to the biomedical sciences.

[M_NotableLabs]

Thanks!

1. There are indeed similar efforts in academic labs around the world. The constraint they face is lack of funding to expand their experiments beyond testing drugs on their small sized tumor banks.

Industry labs don't generally have access to primary (from patients) tumors so they have to partner with academics or contract research organizations (CROs). No one institution or CRO has a Glioblastoma tumor bank beyond around 100 patient samples.

These smaller tumor banks don't capture the heterogeniety of the patient population as a whole, so one of the advantages of being a company not tied to an institution is the ability to develop a substantially larger tumor bank that accomplishes this. We accomplish this by acquiring primary tissue through our testing service around the world.

Academics also often rely on standardized assay conditions and commercially available cell lines because their major incentive is to publish their results. Standard methods and cell lines are often preferred by reviewers and working with primary lines is difficult and expensive. Translation to response in patients isn't the primary objective of the screens.

2. Yes, Elizabeth Iorns, a former cancer biologist and the CEO of Science Exchange, is one of the part time YC partners.

[neurotech1]

My thoughts;

1) There is a large amount of "guesswork" on the part of oncologists. I can well imagine YC would fund a startup that provides more data.

2) That is a valid point, although it may be that this technology may reduce the use of ineffective combinations, as much as find more effective ones.

Disclaimer: I'm not physician etc. and this isn't medical advice.

[srunni]

Hey, this is an area of interest to me, as I'm building an early-stage neuro-oncology imaging startup. A few questions:

1. How do you plan to deal with mutations/regrowths? Would the tumor have to be repeatedly resected for subsequent in vitro analyses?

2. What is your plan for dealing with impending LDT regulation?

The thing I'm most interested in is how you're tying the in vitro results to in vivo efficacy, as well as your reimbursement strategy, but refurb already asked those 2 questions.

[M_NotableLabs]

Very cool! I'd love to follow up with you if you'd be interested, imaging is a major problem in this disease.

1. Great question, this is why we'll be focused on newly diagnosed glioblastoma patients. If they do recur and have a surgery to remove the recurrent (and hypothetically mutated) tumor, we would test that tissue as well and compare its drug response profile as well as any biomarker changes.

We'll also be using in vivo xenografts to try to model mutations/regrowths, by continuously observing treated mice to see if we get a regrowth. The mice have faster metabolisms than humans so tumors can sometimes regrow faster in them than in patients.

If we have a regrowth in a mouse, we could then culture that tumor in vitro for a screen to develop a plan to treat a patient's recurrence before it occurs. This has been proven out already as a strategy by another company using xenografts, Champions Oncology.

2. Since this LDT regulation change is still in motion we are observing it very closely. In discussions we've had with experts on the topic they have highlighted that any changes will be implemented gradually over time.

[srunni]

> Very cool! I'd love to follow up with you if you'd be interested, imaging is a major problem in this disease.

Sure, just shoot me an email at one of the addresses in my profile.

It'll be very interesting if you can robustly translate murine xenograft data into successful clinical outcomes. Two potential hurdles I can see are (1) high cost per patient despite relying on generics, from all the animal models (though maybe you can get an adventuresome ACO on board: http://www.hhs.gov/news/press/2015pres/03/20150310b.html) and (2) insufficient characterization of the mutations/drug interactions/interpatient variability state space, due to the relatively low GBM incidence rate (compared to other cancers).

[jforman]

- What's your approach on how to eventually apply the process to cells that don't normally grow in suspension?

- How do you guard against harmful drug interactions in vivo, given that most of these drug combinations probably haven't been seen in a person at the same time? Does this fall under compassionate use?

- Are you sequencing tumors and looking for predictive markers for future applications?

Wonderful project and best of luck.

[M_NotableLabs]

- This is one of the reasons we have started with Glioblastoma Multiforme (grade 4 brain cancer), as it has been grown in serum free suspension cultures (the neurosphere assay) by many labs for the past decade or so. When we eventually look to move into other cancers we will need to re-engineer the process.

-Since all of the drugs are FDA approved we don't have to apply to get compassionate use as you would if you were obtaining an experimental therapy outside of a clinical trial. The approved drugs can be used at a doctor's discretion as an off label prescription within our current system. As for safety please see my response to the first comment.

-Yes we are going to be sequencing all of our tumors. The difference between our approach to this versus existing personalized medicine sequencing is that we're more interested in using biomarkers in the future once we've identified a patient population that responds to a certain drug combination.

This is our long-term goal, to use a patient's biomarkers to predict drug response that we can then validate in vitro and in vivo rather than just relying on empirically physically testing drugs on a patient's cells.

[srunni]

> Are you sequencing tumors and looking for predictive markers for future applications?

For anyone interested in predictive cancer biomarkers, here are some good (open access) reviews of some of the challenges in creating them:

* http://www.biomedcentral.com/1741-7015/12/156

* http://cancerres.aacrjournals.org/content/72/23/6097.long

* http://www.biomedcentral.com/1741-7015/10/87

There's also a good Institute of Medicine report on the Evolution of Translational Omics: Lessons Learned and the Path Forward (free PDF): http://www.nap.edu/catalog/13297/evolution-of-translational-...

[M_NotableLabs]

This is a really great point srunni thank you for bringing it up. We agree that the current approach to identifying predictive cancer biomarkers has many challenges. That's why we believe in a different approach.

Glioblastoma was actually the first tumor profiled for the Cancer Genome Atlas Project. http://cancergenome.nih.gov/cancersselected/glioblastomamult...

There have been four subtypes found in the disease, but all four are still treated ineffectively with the same standard of care chemotherapy (Temozolomide). To date none of these predictive biomarkers have led to a successful targeted therapy.

This was a personal frustration for me, as I had my dad's tumor sequenced for genetic mutations, but there were no targeted therapies that reach the brain in high enough concentrations to be effective.

This is an important distinction in personalized medicine specific to brain cancer, as many drugs do not cross the blood brain barrier. For example the most prevalent mutation in Glioblastoma is a mutation in the Epidermal Growth Factor Receptor (EGFR) pathway. This is also a driver mutation in lung cancer that has approved targeted therapies like Afatinib.

Unfortunately none of these drugs have worked in trials for brain cancer, in large part because they do not cross the blood brain barrier well. For example here is a study recently published on the poor results with Afatinib in brain tumor patients. http://www.ncbi.nlm.nih.gov/pubmed/25140039

There are currently very few targeted therapies in development for brain cancer as it is a small disease for pharma companies to market drugs.

This is why our approach is to correlate tumor response to existing therapies with empirically identified biomarkers. In contrast using tumors alone to predict cancer biomarkers is challenging because it requires a hypothesis or target to be tested one at a time with a novel therapy in a trial.

For underserved diseases like glioblastoma, this could result in many years passing by as each hypothesis driven therapy fails in clinical trials.

[quadrature]

stupid question, is there no way to administer drugs directly where they are needed ?

[M_NotableLabs]

Definitely not a stupid question it is really important!

There are several novel ways to deliver drugs to tumors in the brain in trials. Some use specially designed pumps, nanotechnology, and intra-arterial delivery (where a catheter is snaked to the tumor through blood vessels).

The problem is that all of these methods are experimental so they can't be used in combination. We're very hopeful that they are successful though!

[dguaraglia]

Hi Matt, first I wanted to thank you for doing this. Cancer, and unfortunately the treatment for it, can be devastating. Anything we can do to improve that, is money well spent.

I had a few questions:

1) Does NotableLabs produce a standard pathological report for the oncologist before starting the tests? (I ask because in my experience that process can take weeks in a normal lab, so I wonder what's the feasibility of performing tests on a specimen after that long.)

2) Are there any challenges/differences when applying the drugs directly to the cells? Do the cells in a Petri dish behave differently from cells in the patient's body?

3) Out of curiosity, why did you choose brain tumors?

[M_NotableLabs]

1. Could you clarify your question on this please?

We won't be replacing the pathologist report for the oncologist. We also don't obtain the tumor after their analysis and report. Our testing will be running in parallel.

We produce our report using test results from the actual patient's living tumor that we obtain from the operating room. Pathologists embed the tumor cells in paraffin wax, which kills the tissue.

2. Yes there are challenges with applying the drugs because the cells are in an artificial environment outside of the body. We have improved the standard conditions the cells are tested in and will continue to develop methods to make the petri dish more like the patient's brain.

3. Initially brain cancer chose us, my dad passed away from the disease two weeks ago.

[dguaraglia]

I'm really for your loss Matt. As someone in remission (from a punier kind of cancer) myself, I'm thankful for people like you, looking for a cure. Thanks for your reply. My heart goes out to you and your family.

[vobios]

That's a big leap to go from in vitro to in vivo. That would usually not even be sufficient to enter clinical trials.

Besides, given GBM's heterogeneity, I am not sure it's clear which cells you are growing and how representative they are.

Is this FDA-approved?

[M_NotableLabs]

We arrived at the approach for Notable Labs based on the needs of patients, unfortunately through personal experience. We are not trying to replicate the traditional target, in vitro screening, in vivo studies, and then clinical trials that academics and pharma are pursing.

We started with the basis that all brain cancer patients receive a standard of care that will do little to extend their lives. Only about 10% enroll in clinical trials in the US, which only test one drug at a time. Most of the clinical trials are performed on patients who have a regrowth of their tumor, which has little chance of responding to treatment.

309 Currently Open Trials for Glioblastoma, but only 75 for newly diagnosed patients. https://clinicaltrials.gov/ct2/results?term=glioblastoma&rec...

With 90% of the patients not enrolled in a clinical trial we wanted to design a commercial solution that could help patients today rather then relying on the standard of care.

We have collaborations with academic labs that are studying the heterogeniety of neurospheres cultured from multiple sites from the same patient's tumor. Ideally to get a more representative sample you would be able to obtain multiple samples from the same patient and derive a separate culture from each one.

We will not be seeking FDA approval for our service as it is not a drug or diagnostic for a disease. The pathway for us is as a CLIA certified service similar to other personalized genomics companies in cancer like Foundation Medicine.

[vobios]

> Only about 10% enroll in clinical trials in the US, which only test one drug at a time.

Clinical trials in the US test drug combinations and not just single drugs.

> With 90% of the patients not enrolled in a clinical trial we wanted to design a commercial solution that could help patients today rather then relying on the standard of care.

In other words, you are enrolling them in your own clinical trial without calling it such. Clinical trials are not perfect, but it's the best way we have to guarantee safety and efficacy.

> The pathway for us is as a CLIA certified service similar to other personalized genomics companies in cancer like Foundation Medicine.

Actually, it's very different. Foundation Medicine looks for known genetic variants and offers suggestions based on those. It is not offering novel or untested solutions. Their report is actually very poor if you are looking to make new discoveries.