KNDU/NBC TV “Living Well” Show: Dr. Lawenda Discusses How Oncologists Choose Cancer Drugs (June 3, 2017)
Most of us in oncology like to say that we practice “personalized” medicine.
What we mean by this is that we take into account many patient and cancer-specific variables when we formulate our treatment recommendations. These include such things as:
- Cancer type
- Pathologic subtype
- Numerous pathological features (i.e. hormone receptor status, perineural or angiolymphatic invasion, etc.)
- Molecular features
- DNA mutations
- Patient age, overall health status, other medical conditions, etc.
Standard Treatment Recommendations (One-size-fits-all):
Using all the above information, we follow treatment guidelines developed by multi-disciplinary groups of subspecialty experts (i.e. medical oncologists, radiation oncologists, surgeons, radiologists, pathologists, etc.)
Most of the time, we rely on guidelines such as these to make sure that we are managing our patients according the latest standard of care. In the United States, the most commonly employed guidelines (for non-pediatric cancers) come from the National Comprehensive Cancer Network (NCCN.)
Your Cancer Is Unlike Anyone Else’s:
One major problem with following a consensus-based guideline is that none of them are able to answer the million dollar question: ‘will this person’s cancer die when we do X, Y, or Z to it?’ What I mean by this is, ‘how do we know that this person’s cancer is best treated with any particular therapy or combination of therapies?’ The answer is…we don’t know until we treat them and then wait to see what happens.
- When it comes to surgery and other interventional therapies, the choices of which procedure(s) to use is/are often established by studies that have demonstrated superiority of one over another. As technologies evolve, we continue to analyze the pros and cons of which (i.e. open surgery, robotic surgery, radiofrequency ablation, etc.) are the most effective and offer a quicker surgical recovery.
- When it comes to radiation therapy, we have expert, consensus-based guidelines and studies that help us select what doses and technologies to use. For many cancers, radiation therapy is able to completely eradicate tumors from the treated area. One day we will hopefully have an assay that will allow us to more precisely pick the best radiation dose for every person’s cancer (i.e. higher doses vs. lower doses.) Researchers are getting closer to developing a personalized radiation sensitivity assay, but they are not ready for prime time yet.(Moffitt Cancer Center, 2015)
- When it comes to chemotherapy and targeted biological drug therapy, we also use expert, consensus-based guidelines and studies to pick the drug(s) to use against your cancer. For some cancers (i.e. lymphomas, leukemias, testicular cancers) we have seen remarkable effectiveness with drug therapies. Unfortunately, not all cancers (i.e. gastric cancer, pancreatic cancer, lung cancer, melanoma, glioblastoma multiforme, ovarian cancer) are as exquisitely sensitive to the standard drug regimens recommended by the consensus guidelines. This is highlighted by a controversial study that found that chemotherapy only improves 5-year survival rates by a mere 2%.(Morgan, Ward, & Barton, 2004) Many oncologists have criticized this study for ‘cherry picking’ the least chemosensitive cancers to include in their analysis, while excluding some of the more chemotherapy sensitive cancers. Although I agree with the critics on this point, one thing is very clear…we still have a long way to go.
What about experimental drug trials?
If you have an advanced cancer and standard therapies are not effective, your oncologist may recommend that you enroll in an experimental drug trial. Typically, after researchers have discovered that an experimental drug shows some effectiveness in animal studies, the most promising of these agents will be tested in a “phase 1” trial in humans to see if they are safe. The main endpoint of phase 1 trials is not actually to find an effective drug. In fact, studies have shown that phase 1 trial drugs only have a 10% probability of having any efficacy.(Horstmann et al., 2005)
“Phase 2” trials take the least toxic drugs from the phase 1 trials and test them in patients with a particular cancer to see if they have any efficacy (i.e. slower cancer growth, cancer regression, reduction in metastases, etc.)
Finally, “phase 3” trials take the most effective drugs from the phase 2 trials and compare them to other standard drugs. Only 4-38% of the experimental drugs are found to be superior to the standard drugs in phase 3 trials.(Gan, You, Pond, & Chen, 2012; Maitland, Hudoba, Snider, & Ratain, 2010)
Every cancer is different & Why Molecular Pathway and Genetic Mutation Assays Are Flawed:
Your cancer is not same as anyone else’s cancer on the planet, so why do we continue to treat cancers ‘similar’ to yours with standard drug regimens or a ‘shot in the dark’ experimental drug? Oncologists have started to use “next generation sequencing” (NGS) genetic assays to detect genetic abnormalities in certain cancers, and depending on the results of these assays, they may prescribe a drug targeting the detected abnormality. These type of assays don’t tell us whether the drug will actually be effective in killing your cancer, however they do give us a hint that it might work.
Studies have found that the latest NGS assays can detect targetable genetic abnormalities in the majority of cancers, through tumor tissue or blood samples (“liquid biopsies”.) Unfortunately, the effectiveness of this approach has not been dramatic in most cases, with response rates of 2-6 months on average. Occasionally, the responses can be much greater, but these have been far a few between.
Cancer is much more complex than its genetic blueprint, and simple genetic abnormalities often do not fully account for why a cancer grows out of control. Merely targeting one genetic pathway mutation or abnormality typically does not stop cancers from progressing, as there are many growth pathways the cancer can use simulataneously to bypass a roadblock. The genetic abnormalities identified in the NGS assays also may not be expressed in the cells, as genetics (‘genotype’) are simply a blueprint and not the final product (‘phenotype’.)
Furthermore, cancer is extremely heterogeneous, so a small sample of tissue from a needle biopsy or liquid biopsy may not tell us what is happening throughout the cancer and its metastatic tumors. Studies have found that if you biopsy multiple locations in the same tumor, the genetics of each biopsy specimen will be more distinct than similar. To make matters even more complex, the metastatic lesions often have different genetics than the primary tumor. As time goes on in the course of the disease and successive treatments are employed, the cancer will continue to evolve and develop new genetic aberrations that make it resistant to targeted treatments.
If you really want to know if a drug(s) will kill a tumor, you must observe the effects of that treatment on both the tumor and its microenvironment (i.e. tissue fibroblasts, blood vessels and immune cells.) The interplay amongst all of these cellular components determines how sensitive the tumor is to any drug or drugs thrown at it. NGS assays only look at tumor genetics and do not take into account any of the other factors that are essential in cancer growth.
Instead of treating everyone the same way or experimenting on you with a new study drug, is there another way?
- Can we do a better job of predicting which cancer drug(s) will have the best chance of killing your specific cancer?
- Can we do this using a test that does not subject you to their toxic side effects?
- Can we get the results from this test quickly?
In my opinion, the answer to all of these questions is a resounding ‘YES.’
Does drug(s) XYZ kill your cancer?
What if we were able to remove cancer cells from you and test to see which drug or combination of drugs was the most effective in killing those cells? This is called “chemosensitivity testing.” This offers the benefit of testing for therapeutic efficacy without subjecting patients to the toxicity of the drugs tested, and the results take a few days-to-months to see which drug(s) actually work.
Labs have been developing personalized assays in either cell (“in-vitro”) or in animal (“in-vivo”) models to see how sensitive a patient’s cancer is to the drugs tested. These are not perfect assays, as they cannot take into account the innumerable physiological variables in an actual person.
That said, studies comparing outcomes from patients who received personalized drug regimens using these tests have found significantly improved cancer outcomes (i.e. better cancer response rates, survival, etc.) compared to patients who were not treated based on these tests.(Christian, 2013; Nagourney et al., 2012) (Bosanquet et al., 2007; Tian et al., 2014.) In a review of drug sensitivity assays, the clinical response rates were significantly improved for numerous cancer types when patients were treated with drugs based on the assay results (see the table below.)(Nagourney, 2006)
For example, patients with cancers that typically only have a 10% clinical response rate (slowed tumor growth or tumor regression) with standard drug therapy, will have a 46% chance of having a clinical response if the chemosensitivity assay predicts that the drug(s) are effective and only a 2.6% clinical response if the assay predicts the drug is not effective.
Wouldn’t you want to know this information?
Your cancer may be more sensitive to drugs typically used for treating other cancer types:
These chemosensitivity assays may find, for example, that your lung cancer is actually much more sensitive to a drug regimen typically recommended for gastric cancer. Or, your cancer might, be much more sensitive to 1 specific drug combination out of the 10-20 NCCN recommended drug regimens for your cancer. There would be absolutely no way for your oncologist to know this predictive information without a chemosensitivity assay performed on your cancer. Oncologists are often confronted with choosing any one of the numerous ‘approved’ drug options in these consensus-based guidelines (see example below from a 2017 NCCN guideline.) If there was a better way to predict which one of these regimens is more likely to work on your cancer, wouldn’t you want that information to help direct your treatment?
Why are these assays not covered by insurance nor included in the consensus-based guidelines?
Unfortunately, chemosensitivity testing is still considered “experimental” and is not typically covered by medical insurances. This is a travesty, in my opinion.
What will it take for these assays to be approved by the insurance companies and to be included in the consensus guidelines? The answer is: randomized clinical trial data. Academic institutions need to conduct randomized controlled trials, where 50% of the patients will be randomized and treated based on the assay results and the other 50% will be treated with standard drug regimens. If the assay-directed therapy patients consistently do better in these studies, the experts and insurance companies will likely change their approach to cancer treatment guidelines. Although we currently have retrospective and prospective data that demonstrate these results, without randomized controlled trials to confirm it the experts will remain skeptical and the insurers will deny coverage.
What is the process?
The details will vary depending on the assay, but in general, to get these tests done, you will need:
- To take about 2-to-4-weeks off from your treatment (cancer drug therapy or radiation therapy) to coordinate getting a sample of your tumor. Depending on the cancer type, you will need to have your doctors collect either: blood (leukemia), bone marrow (leukemia, myeloma), tumor tissue (solid tumors) or malignant effusions (pleural fluid or ascites fluid.) NOTE: Getting a sample of your tumor to submit for testing is not a risk-free procedure. You will have to discuss with your oncologist if the potential risks outweigh the potential benefits.
- Submit the sample to the lab. This is done through FedEx or other delivery services in special packaging.
- Wait for the results. Depending on the test, this may take only 1 to 2-weeks to get your results. These will be sent directly to your oncologist.
- Talk with your oncologist about which regimen makes the most sense for you based on your assay results and other factors.
My Favorite Chemosensitivity Test:
Numerous top academic cancer centers (i.e. Harvard, Fred Hutchinson Cancer Research Center, Weill Cornell Medicine, Sunnybrook Research Institute, etc.) are developing their own versions of these state of the art assays, but most are all still in the research and development stage. Read the articles below to learn why these assays are the future of where we are heading in precision oncology.
The best chemosensitivity assay that is currently available for patients to use now is the EVA-PCD Functional Profile (Nagourney Cancer Institute, Long Beach, CA), also known as “Ex-Vivo Analysis of Programmed Cell Death.”
I have referred numerous patients, for EVA-PCD assay testing, whose cancers have recurred after multiple drug regimens. Often these patients come to me for a second opinion, having been told that there are no good treatments left for them.
This assay requires a freshly obtained sample (preferably, at least 1 cubic centimeter) of your tumor, which is mailed out (same day day service) to the Nagourney Cancer Institute. Once it is received by their lab, they process specimen into tiny 3-dimensional clumps of tissue called “microspheroids,” which are comprised of your tumor cells, tissue stromal cells (the cells that make up a tumor microenvironment) and immune cells. These microspheroids mimic the biology of your cancer, and therefore serve as a good model to test numerous drugs for anticancer activity. Here are some of the drugs used in the EVA-PCD assay:
It is important to distinguish genetic “NGS” and molecular pathway assays from the EVA-PCD assay. NGS assays look for genetic abnormalities in your cancer for which oncologists have drugs that might be effective anticancer targets. Unfortunately, a cancer cell is much more complicated than just its DNA, RNA or proteins. Just because you have a drug that targets any of these does not mean you will kill that cancer cell.
In contrast, the EVA-PCD assay doesn’t care about the mechanisms in which the cancer cell is defective. All this test looks for is whether a drug or combination of drugs actually kills it or not. This is a much more relevant result.
The first generation (and much less accurate) chemosensitivity assays used colonies of “clonogenic” cancer cells or cancer stem cells instead of whole tissue biopsy specimens (which include: tumor cells, stromal cells and immune cells), and subjected these to chemotherapy drugs. Unfortunately, there were many flaws in this approach:
- Clonogenic cells acquire mutations that make them behave very differently than native tumor cells. (Some labs still use clonogenic assays.)
- Since the patient’s tissue stromal and immune cells were not incorporated into these first-generation assays, real-world tumor response was not accurate. We now know that stromal and immune cells interact with tumor cells to modify the response to drug therapies. (Most labs do not take into account stromal and immune cells in their assay.)
- Circulating tumor cells (“CTC”) assays collect free-floating cancer cells from the blood, but these CTC’s may not represent the same behavior (i.e. cell death) in the presence of drugs as in whole tissue assays.
It is not surprising that these flawed, first-generation chemosensitivity assays were a flop in many early clinical trials, as they did a poor job of predicting the most effective drugs for the tumors treated. In 2004, ASCO reviewers concluded that these assays were not useful. Unfortunately, many oncologists have not yet embraced the latest, second-generation chemosensitivity assays (i.e EVA-PCD assay), believing that all chemosensitivity assays are equally flawed. The current, next-generation chemosensitivity assays use fresh, live tumor tissue (or ascites/pleural fluid) which gives a much more representative picture of the physiologic response to drug therapies.
I wish more patients had access to the EVA-PCD assay prior to receiving the vast majority of oncologic drug treatment regimens. The results of this assay potentially could be lifesaving.
Unfortunately, this test is not typically covered by insurance. The cost depends on the drugs tested, but one of my patients was recently quoted a price of approximately $4,000; this does not include the procedure cost of the sample collection (i.e. biopsy, etc.) nor the shipping.
Personally, if I or any of my loved ones had a cancer that required a potentially toxic (and possibly ineffective) oncologic drug therapy, I’d contact Rational Therapeutics to see if the EVA-PCD assay would be appropriate.
Here is an example report:
Here’s a brochure you can give to your oncologist:
Dr. Robert Nagourney is the Director of the Nagourney Cancer Institute. He is one of the most brilliant medical oncologists I have ever met and had the pleasure of working with.
I highly recommend you watch his excellent Tedx Talk:
You can contact Nagourney Cancer Institute here:
When should you get a chemosensitivity assay?
- It can be done at any time point before you start your cancer treatment. The earlier in the course of treating your cancer, the less likely it will have become resistant to multiple therapies.
- It can be done during your cancer treatment. You will need to come off your drug therapy for a short period of time to get the biopsy.
- It can be done if you have a recurrence of your cancer. This may be especially helpful if metastatic lesions are no longer responding to your current drug therapy.
- Did you know that as cancers grow and metastasize they acquire new mutations that may make them sensitive to drug therapies that initially they were not sensitive to? For example, a breast cancer that was estrogen receptor negative or HER-2 negative can develop metastatic disease that develops a sensitivity to anti-estrogen therapy or Herceptin therapy (reference.) This is why doing a chemosensitivity assay on a freshly obtained biopsy specimen of a metastatic lesion could be very helpful, as this may uncover these new drug sensitivities.
I do not receive any compensation for my endorsement of the Nagourney Cancer Institute.
Bosanquet, A., Kaspers, G., Larsson, R., Nagourney, R., Nygren, P., Pieters, R., . . . Weisenthal, L. (2007). Individualized Tumor Response (ITR) Profiling for Drug Selection in Tailored Therapy: Meta-Analysis of 1929 Cases of Leukemia and Lymphoma. Paper presented at the ASH Annual Meeting.
Christian, A., Souza, K., Hornuss, C., Weisenthal, L., Nagourney, RA.,. (2013). Accuracy and clinical utility of in vitro cytometric profiling to personalize chemotherapy: Preliminary findings of a systematic review and meta-analysis. Paper presented at the ASCO Annual Meeting.
Gan, H. K., You, B., Pond, G. R., & Chen, E. X. (2012). Assumptions of expected benefits in randomized phase III trials evaluating systemic treatments for cancer. J Natl Cancer Inst, 104(8), 590-598. doi: 10.1093/jnci/djs141
Horstmann, E., McCabe, M. S., Grochow, L., Yamamoto, S., Rubinstein, L., Budd, T., . . . Grady, C. (2005). Risks and benefits of phase 1 oncology trials, 1991 through 2002. N Engl J Med, 352(9), 895-904. doi: 10.1056/NEJMsa042220
Maitland, M. L., Hudoba, C., Snider, K. L., & Ratain, M. J. (2010). Analysis of the yield of phase II combination therapy trials in medical oncology. Clin Cancer Res, 16(21), 5296-5302. doi: 10.1158/1078-0432.CCR-10-0669
Moffitt Cancer Center. (2015). Moffitt Cancer Center Researchers Develop First Genetic Test to Predict Tumor Sensitivity to Radiation Therapy. from https://moffitt.org/newsroom/press-release-archive/2015/moffitt-cancer-center-researchers-develop-first-genetic-test-to-predict-tumor-sensitivity-to-radiation-therapy/
Morgan, G., Ward, R., & Barton, M. (2004). The contribution of cytotoxic chemotherapy to 5-year survival in adult malignancies. Clin Oncol (R Coll Radiol), 16(8), 549-560.
Nagourney, R. A. (2006). Ex vivo programmed cell death and the prediction of response to chemotherapy. Curr Treat Options Oncol, 7(2), 103-110.
Nagourney, R. A., Blitzer, J. B., Shuman, R. L., Asciuto, T. J., Deo, E. A., Paulsen, M., . . . Evans, S. S. (2012). Functional profiling to select chemotherapy in untreated, advanced or metastatic non-small cell lung cancer. Anticancer Res, 32(10), 4453-4460.
Tian, C., Sargent, D. J., Krivak, T. C., Powell, M. A., Gabrin, M. J., Brower, S. L., & Coleman, R. L. (2014). Evaluation of a chemoresponse assay as a predictive marker in the treatment of recurrent ovarian cancer: further analysis of a prospective study. Br J Cancer, 111(5), 843-850. doi: 10.1038/bjc.2014.375