Ixchel (YC S15) Predicts How Cancer Drugs Will Behave in Humans | Hacker News Reader

Ixchel (YC S15) Predicts How Cancer Drugs Will Behave in Humans

techcrunch.com

21 pointsjcr20 comments

Hey everyone,

Please let us know if we can answer any questions about what we do, the current state of cancer treatment, or anything about cancer, in general.

Julia (CEO, Ixchel Scientific; www.ixchelsci.com)

Isn't biology / medical research too expensive to do enough meaningful work on a YC grant?

The sums YC provides work well for a couple of programmers with laptops - can keep 'em going for a while, but you folks need labs and stuff.

Yes, biological research requires a lot of infrastructure and laboratory set-up, so it is indeed expensive. However, YC is an amazing platform to get your technology off the ground and in front of customers. In addition to the funding, advice from YC members has been invaluable in getting a strong business strategy.

Hi Julia, thanks for answering questions.

I noticed something odd in the techcrunch article; the author, Sarah Buhr (@sarahbuhr), capitalized your company name incorrectly as "IxChel" in the article title. I fixed the capitalization after submitting this to HN, and dang (Dan) or another HN admin most likely fixed the "YC-backed" to the usual "(YC S15)". You might want to ask techcrunch to fix it?

I have not (yet) read the linked papers in your site:

http://ixchelsci.com/#publications

Which I hope will provide more details, but at the moment, I'm curious how your whole system/service works?

Our platform is a cell-based laboratory tool where we reconstruct the environment of human organs in a dish. When cells are placed in these physiological conditions, they behave the way they would behave in the human body (unlike the conventional models where cells are grown on a surface of plastic dishes). Such set-up allows us to evaluate how cancer cells will respond to drugs (or drug combinations) under native conditions of drug-resistance. I.e. when cancer cells interact with their environment in the human organs, this environment protects them against chemotherapy and other drugs, so when cells are taken out of their physiological environment, they become more sensitive to drugs. This results in a false assumption that the drug has anti-cancer activity, but when it's introduced into a patient, the drug doesn't work.

What we provide, is a system where drugs can be tested under the conditions similar to those seen by the cancer cells in the human.

What we are offering right now, are services to test drugs in our system, and we are gearing-up to provide all-inclusive kits that can be used by researchers to run these assays in their own labs.

Please let us know if you would like know things in more detail.

The link on your site to the SPIE Journal of Biomedical Optics paper is paywalled [1], but the paper is freely available from NIH [2]. You might want to change the link on your site, or just add NIH link(s) as backups (I also found [3] on NIH).

So far, I've only read through the abstract, but the use of Raman spectroscopy [4] for identification looks fascinating. I had no clue that it's possible to differentiate cancer cells with a spectrograph, and no, I really have no idea how or why the analysis works. On the other hand, automating test equipment in a lab is a whole lot of fun!

At the moment, the closest thing I can manage to a useful question is: Can/Will/Does your service automate this analysis method?

[1] http://biomedicaloptics.spiedigitallibrary.org/mobile/articl...

[2] http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3832300/

[3] http://www.ncbi.nlm.nih.gov/pmc/articles/PMC141028/

[4] https://en.wikipedia.org/wiki/Raman_spectroscopy

Thanks for the suggestion, we will add the links to the NIH repository for the papers that are available.

Regarding automation, the Raman spectroscopy studies was done in collaboration with a colleague at Purdue University, so at the moment, we don't have plans to extend these studies.

However, we are working on automating our workflow to streamline the set-up and analysis of our cultures. So, if you know anyone working on bio lab-related automation, we'd love an intro.

Thanks! I asked TechCrunch to fix the company name.

Hey everyone,

I'm the co-founder & CTO of Ixchel Scientific - would love to answer any questions you may have about our company!

Is it just me or does the S15 batch seem to have a crazy number of medical/health startups compared to previous batches? There were at least 2 on the front page yesterday. Maybe they're just all launching on TC early?

Not saying it's a bad thing, all these companies look awesome, just curious.

There are 11 or 12 healthcare/medical companies in this batch -- which isn't more than W15. They all just seem to be hitting TC at the same time.

Would your product be a suitable replacement for mouse models? https://edge.org/response-detail/25429

The short answer is, ideally yes, but not completely. The animal models of cancer are poor at predicting how the drug will work in a human, but where they have value is in evaluation of drug safety. So, we envision that the toxicity studies would still be done in animals, but near-human, organ-specific models would ultimately replace animal studies of drug efficacy.

Personalized medicines are a frontier for the next major advances in medicine. We have multiple medications for the same disease, each working differently in different individuals.

I did some work in the area a decade ago, and data was the hardest thing to come by.

FYI Ixchel is the Mayan goddess of Medicine.

what are good ways of educating people about cancer? I have a relative who was diagnosed in the later stages and it was a very traumatic exp for them and the family. Are there some tests that can be integrated with the regular body check up that can help people with earlier diagnosis?

I've been thinking that I would like to get CT and MRI scan done for my whole body, maybe every 4-6 months. Then doctors could keep track of any suspicious lumps, and perform biopsies if they are growing at an unusual rate.

I've read many articles that say MRIs are overused and lead to overdiagnosis. They say that many people have benign nodules and cysts that you don't need to worry about, and you might even end up worrying about shadows or other scanning errors. You'll end up having too many biopsies, which can be invasive and have negative side-effects.

But I think this might be a different story if you could compare two full-body scans, and take note of the differences. In fact, I'm starting to wonder if rich people get this done all the time, and it's just too expensive to provide this level of medical care to the rest of the world.

It's just a fact that a tumour could be growing inside me right now, and I wouldn't know about it until it starts causing side effects. I hope we see some breakthroughs for the early detection of cancer, and I'm really glad to see companies like this being funded by YC.

American Association for Cancer Research (AACR), American Cancer Society (ACS), and National Cancer Institute (NCI) are good sources for reliable information on cancer. There are many sites that talk about cancer, but if you want information based on research (both lab and clinical) these three provide the most trustworthy information.

Regarding tests, some are already available, such as mammography, and many people are working on creating molecular diagnostics that can identify cancer early from a blood sample. A few blood tests already exist, such as CA125 for ovarian cancer, PSA for prostate, CEA for colon.

Did you get significant improvements over the 2D cell lines? or 3D cell lines?

Yes, absolutely. Our platform provides physiologically relevant 3D environment that recapitulates the native conditions seen in human tissues. Moreover, unlike the models with immortalized cells lines, which for many cancers are not representative of the disease, our platform sustains primary cells from patients. Our studies demonstrate that our platform has high correlation with clinical response (i.e. how cells respond to drugs in our system matches how the patient responds).

Reply on news.ycombinator.com