Chemosensitivity Testing – What It Is and What It Isn’t

Several weeks ago I was consulted by a young man regarding the management of his heavily pre-treated, widely metastatic rectal carcinoma. Upon review of his records, it was evident that under the care of both community and academic oncologists he had already received most of the active drugs for his diagnosis. Although his liver involvement could easily provide tissue for analysis, I discouraged his pursuit of an ex vivo analysis of programmed cell death (EVA-PCD) assay. Despite this, he and his wife continued to pursue the option.

As I sat across from the patient, with his complicated treatment history in hand, I was forced to admit that he looked the picture of health. Wearing a pork pie hat rakishly tilted over his forehead, I could see few outward signs of the disease that ravaged his body. After a lengthy give and take, I offered to submit his CT scans to our gastrointestinal surgeon for his opinion on the ease with which a biopsy could be obtained. I then dropped a note to the patient’s local oncologist, an accomplished physician who I respected and admired for his practicality and patient advocacy.

A week later, I received a call from the patient’s physician. Though cordial, he was puzzled by my willingness to pursue a biopsy on this heavily treated individual. I explained to him that I was actually not highly motivated to pursue this biopsy, but instead had responded to the patient’s urging me to consider the option of performing an EVA-PCD assay. I agreed with the physician that the conventional therapy options were limited but noted that several available drugs might yet have a role in his management including signal transduction inhibitors.

I further explained that some patients develop a process of collateral sensitivity, whereby resistance to one class of drugs (platins, for example) can enhance the efficacy of other class of drugs (such as, antimetabolite) Furthermore, patients may fail a drug, then be treated with several other classes of agents, only then a year of two later, manifest sensitivity to the original drug.

Our conversation then took a surprising turn. First, he told me of his attendance at a dinner meeting, some 25 years earlier, where Dan Von Hoff, MD, had described his experiences with the clonogenic assay. He went on to tell me how that technique had been proven unsuccessful finding a very limited role in the elimination of “inactive” drugs with no capacity to identify “active” drugs. He finished by explaining that these shortcomings were the reason why our studies would be unlikely to provide useful information.

I found myself grasping for a handle on the moment. Here was a colleague, and collaborator, who had heard me speak on the topic a dozen times. I had personally intervened and identified active treatments for several of his patients, treatments that he would have never considered without me. He had invited me to speak at his medical center and spoke glowingly of my skills. And yet, he had no real understanding of what I do. It made me pause and wonder whether the patients and physicians with whom I interact on a daily basis understand the principles of our work. For clarity, in particular for those who may be new to my work, I provide a brief overview.

1.    Cancer patients are highly individual in their response to chemotherapies. This is why each patient must be tested to select the most effective drug regimen.
2.    Today we realize that cancer doesn’t grow too much it dies too little. This is why older growth-based assays didn’t work and why cell-death-based assays do.
3.    Cancer must be tested in their native state with the stromal, vascular and inflammatory elements intact. This is why we use microspheroids isolated directly from patients and do not grow or subculture our specimens.
4.    Predictions of response are not based on arbitrary drug concentrations but instead reflect the careful calibration of in vitro findings against patient outcomes – the all-important clinical database.
5.    We do not conduct drug resistance assays. We conduct drug sensitivity assays. These drug sensitivity assays have been shown statistically significantly to correlate with response, time to progression and survival.
6.    We do not conduct genomic analyses for there are no genomic platforms available today that are capable of reproducing the complexity, cross-talk, redundancy or promiscuity of human tumor biology.
7.    Tumors manifest plasticity that requires iterative studies. Large biopsies and sometimes multiple biopsies must be done to construct effective treatment programs.
8.    With chemotherapy, very often more is not better.
9.    New drugs are not always better drugs.
10.  And finally, cancer drugs do not know what diseases they were invented for.

While we could continue to enumerate the principles that guide our practice, one of the more important principles is humility. Medicine is a humbling experience and cancer medicine even more so. Patients often know more than their doctors give them credit for. Failing to incorporate a patient’s input, experience and wishes into the treatment programs that we design, limits our capacity to provide them the best outcome.

With regard to my colleague who seemed so utterly unfamiliar with these concepts, indeed for a large swath of the oncologic community as a whole, I am reminded of the saying “There’s none so blind as those who will not see.”

Reposted from March 23, 2012

Tumor Ecology, Not Tumor Biology

During the first years of this millennium as the newly discovered field of anti-angiogenesis was reaching a fevered pitch, I had the opportunity to attend an AACR Special Symposium, held at Whistler Resort in British Columbia. While there I attended a symposium by Dr. Rakesh Jain. Dr. Jain a long-time colleague of Judah Folkman, MD,  at Harvard University presented his observations on tumor vascularity and its implications for therapy. Despite the prevailing belief that tumor angiogenesis was a linear phenomenon, from cessation of blood supply – and thereby nutrients and oxygen – to the death of cancer, Dr. Jain provided compelling evidence to the contrary.

Every so often I read an article, hear a lecture, or attend a symposium that changes the way I think. Dr. Jain’s presentation that year was just that type of lecture. In the span of an hour he described the dynamics of blood flow through the network of disorganized tumor blood vessels. He showed that anti-angiogenic factors actually “pruned” the blood supply and returned normal flow. He went on to point out that most of the experiments being reported at that time by other investigators had short windows of observation during which the effects of Bevacizumab could be captured, photomicrographed and published to great acclaim in the most prestigious journals. But there was a fly in the ointment. Bevacizumab by itself had a miniscule response rate. Indeed, in the absence of chemotherapy, it was single digits.

Jain, an engineer by training, developed a novel tissue “window” method that enabled him to explore the temporal sequence of cellular response to VEGF therapy. He found that it all wasn’t as simple or tidy as it had seemed. The short-term control of vasculature was followed by revascularization. Cells deprived of oxygen and nutrients devolved into more stem cell-like phenotypes. Therapies based on an incomplete understanding of angiogenesis might, in his opinion, be adding to the problem.

As the years have gone on I’ve carried the insights from that lecture with me. At a subsequent AACR presentation by Napoleon Ferrara, PhD, many years later, Dr. Ferrara, who developed Bevacizumab, reminded his audience that VEGF was originally known as VPF (vascular permeability factor). Perhaps this aspect of the VEGF effects were responsible for its minimal single agent activity, yet profound combinatorial effect.

With this as a backdrop, I sat among 15,000 medical oncologists at the plenary session lecture where Dr Jain presented his work and I delighted in the possibility (however slight) that his message of experimental analysis and systems biology would sink in.

Cancer is not a cell, it is a system. Tumor cells are but a small portion of the process. Carcinogenesis may represent a response to cellular stress, some of which, we as “therapists” may inflict. The indiscriminate use of cytotoxic agents and antivascular drugs may, in some circumstances, be more harmful than helpful to our patients.

What is the appropriate dose of Bevacizumab? How should it be given? In what sequence with radiation or chemotherapy? With what drugs or targeted agents? Are low doses better than high doses? Is the effect of VGEF inhibition a driver of response or an epiphenomenon? What about the fibroblast matrix, lymphatic vessels, infiltrating monocytes, T-cells, B-cells and neutrophils? Dr. Jain elegantly outlined the complexities of the human tumor microenvironment.

It was with more than a small amount of satisfaction, that I realized how quite correct our approach to this disease has been over the years. It is not just the cancer cell that is important, but the tumor as a whole. Cancer cells are just part of the problem. Using native state microspheroids replete with vasculature, cytokines, stromal elements and tumor cells; we feel that we are now poised to advance the growing use of effective targeted therapies in ever-expanding ways.

Chemosensitivity Testing – What It Is and What It Isn’t

Several weeks ago I was consulted by a young man regarding the management of his heavily pre-treated, widely metastatic rectal carcinoma. Upon review of his records, it was evident that under the care of both community and academic oncologists he had already received most of the active drugs for his diagnosis. Although his liver involvement could easily provide tissue for analysis, I discouraged his pursuit of an assay. Despite this, he and his wife continued to pursue the option.

As I sat across from the patient, with his complicated treatment history in hand, I was forced to admit that he looked the picture of health. Wearing a pork pie hat rakishly tilted over his forehead, I could see few outward signs of the disease that ravaged his body. After a lengthy give and take, I offered to submit his CT scans to our gastrointestinal surgeon for his opinion on the ease with which a biopsy could be obtained. I then dropped a note to the patient’s local oncologist, an accomplished physician who I respected and admired for his practicality and patient advocacy.

A week later, I received a call from the patient’s physician. Though cordial, he was puzzled by my willingness to pursue a biopsy on this heavily treated individual. I explained to him that I was actually not highly motivated to pursue this biopsy, but instead had responded to the patient’s urging me to consider the option. I agreed with the physician that the conventional therapy options were limited but noted that several available drugs might yet have a role in his management including signal transduction inhibitors.

I further explained that some patients develop a process of collateral sensitivity, whereby resistance to one class of drugs (platins, for example) can enhance the efficacy of other class of drugs (such as, antimetabolite) Furthermore, patients may fail a drug, then be treated with several other classes of agents, only then a year of two later, manifest sensitivity to the original drug.

Our conversation then took a surprising turn. First, he told me of his attendance at a dinner meeting, some 25 years earlier, where Dan Von Hoff, MD, had described his experiences with the clonogenic assay. He went on to tell me how that technique had been proven unsuccessful finding a very limited role in the elimination of “inactive” drugs with no capacity to identify “active “drugs. He finished by explaining that these shortcomings were the reason why our studies would be unlikely to provide useful information.

I found myself grasping for a handle on the moment. Here was a colleague, and collaborator, who had heard me speak on the topic a dozen times. I had personally intervened and identified active treatments for several of his patients, treatments that he would have never considered without me. He had invited me to speak at his medical center and spoke glowingly of my skills. And yet, he had no real understanding of what I do. It made me pause and wonder whether the patients and physicians with whom I interact on a daily basis understand the principles of our work. For clarity, in particular for those who may be new to my work, I provide a brief overview.

1.    Cancer patients are highly individual in their response to chemotherapies. This is why each patient must be tested to select the most effective drug regimen.

2.    Today we realize that cancer doesn’t grow too much it dies too little. This is why older growth-based assays didn’t work and why cell-death-based assays do.

3.    Cancer must be tested in their native state with the stromal, vascular and inflammatory elements intact. This is why we use microspheroids isolated directly from patients and do not grow or subculture our specimens.

4.    Predictions of response are not based on arbitrary drug concentrations but instead reflect the careful calibration of in vitro findings against patient outcomes – the all-important clinical database.

5.    We do not conduct drug resistance assays. We conduct drug sensitivity assays. These drug sensitivity assays have been shown statistically significantly to correlate with response, time to progression and survival.

6.    We do not conduct genomic analyses for there are no genomic platforms available today that are capable of reproducing the complexity, cross-talk, redundancy or promiscuity of human tumor biology.

7.    Tumors manifest plasticity that requires iterative studies. Large biopsies and sometimes multiple biopsies must be done to construct effective treatment programs.

8.    With chemotherapy, very often more is not better.

9.    New drugs are not always better drugs.

10.   And finally, cancer drugs do not know what diseases they were invented for.
While we could continue to enumerate the principles that guide our practice, one of the more important principles is humility. Medicine is a humbling experience and cancer medicine even more so. Patients often know more than their doctors give them credit for. Failing to incorporate a patient’s input, experience and wishes into the treatment programs that we design, limits our capacity to provide them the best outcome.

With regard to my colleague who seemed so utterly unfamiliar with these concepts, indeed for a large swath of the oncologic community as a whole, I am reminded of the saying “There’s none so blind as those who will not see.”

Cancer Research Becomes “Curiouser and Curiouser”

Following the Gina Kolata New York Times article on July 8, 2011, which described the failure of the Duke University gene profile program in lung cancer, a second New York Times article popped up on the radar screen.  “Cancer’s Secrets Come into Sharper Focus” by George Johnson, examined the growing complexity of cancer research.

This article explored the growing realization that human biology is not linear. Included were references to work that we have previously described in this blog, including the groundbreaking work of Pier Paolo Pandolfi. It also described the interaction between the human body and its microbial flora. We have long recognized that human health is, in part, associated with our interaction with microbes in our environment. The gastrointestinal tract has numerous species that are increasingly believed to contribute to our health. The growing field of probiotics, wherein people consume “healthy organisms,” has gone from quackery to community standard in less than a decade.

What is interesting over the past years is the growing recognition that many cancers are related to infections. Viral infections are known to be oncogenic, with the Epstein-Barr virus, HPV and other viruses now known to be causative of lymphomas, cervical, head and neck, and other cancers. The association between helicobacter and ulcers, gastric lymphoma, and esophageal malignancies are of interest both epidemiologically and therapeutically.

What is most interesting of all is the growing recognition that the cancer cell is but a small component of the cancer.

Here at Rational Therapeutics we recognized the interplay between cells, stroma, vascular elements, cytokines, macrophages, lymphocytes and other environmental factors. This lead to our focus on the human tumor primary culture microspheroid, which contains all of these elements. In our earlier work, we endeavored to isolate tumor cells from their benign constituents so as to study “pure” tumor cells. As time went on, however, we found that these disaggregated cells were artificially sensitized to the effects of chemotherapy and provided false positive results in vitro.

Early work by Beverly Teicher and Robert Kerbel that examined cells alone and in 3-dimensional structures, lead to the realization that cancer cells inhabit a microenvironment. Our lab now studies cancer response to drugs within this microenvironment, enabling us to provide clinically relevant predictions to our patients.

It is our capacity to study human tumor microenvironments that distinguishes us from other platforms in the field. And, it is this capacity that enables us to conduct discovery work on the most sophisticated classes of compounds that influence cell signaling at the level of notch, hedgehog and WNT, among other (Gonsalves, F, et al. (2011). An RNAi-based chemical genetic screen identifies three small-molecule inhibitors of WNT/wingless signaling pathway. PNAS vol. 108, no. 15, pp. 5954-5963).  With this clinically validated platform we are now positioned to streamline drug development and advance experimental therapeutics.

The EVA-PCD™ Platform

At the recent meeting of the American Association for Cancer Research (AACR) held in Washington D.C (April 16-21, 2010), the theme remained consistent with the ground swell of interest in personalized care. Many of the sessions reflected the changing paradigm of clinical trials with a growing focus on biomarker analysis and patient selection predicated on genomic and proteomic features. Among the most compelling presentations were those that examined the manifest complexities of human signaling circuits. One presentation by Dr. Neal Rosen from Memorial Sloan Kettering in New York examined redundancy and feedback as principal determinants of clinical response to signal inhibitors. That session, chaired by Dr. Engleman from Harvard Medical School, examined the cross talk between EGFr and PI3K pathways. Using cell line systems, these investigators drilled down onto RNA and DNA expression profiles to examine how inhibitors acting for one pathway might up or down regulate parallel pathways.

This work dovetailed perfectly with our presentation on Monday, April 19, 2010 (Nagourney, R. et. al, Horizontal and vertical signal pathway inhibition in human tumor primary culture micro-spheroids. Abstract 1764, proceedings AACR 2010). In this analysis, we used small molecules tyrosine and serine/threonine kinase inhibitors to examine the points of commonality and disparity in these two crucial signaling pathways to assess how future drug combinations might provide response with these novel classes of agents.

The most exciting aspect of our work is the capacity of the human tumor micro-spheroid platform (EVA-PCD™) to capture all of the operative mechanisms of response and resistance. This, more closely than other platforms, recapitulates the complexity of human tumors and provides insight into these  complex and redundant biological pathways. No genomic or proteomic tool can approximate the clinical relevance of the EVA-PCD™ platforms’ predictions.

The Primacy of Microspheroids

After incorporating the realization that cancer biology was predicated on cell survival and not cell growth into our laboratory platform, we moved away from proliferative end points to cell death measures, and then redoubled our efforts to recreate the human tumor micro-environment in tissue culture. We immediately recognized that this required the preservation of cell-cell interactions found normally in the body as cellular clusters.

These cellular clusters better known as microspheroids, represent cohesive populations that interact directly with stroma, vasculature, inflammatory cells, and other tumor cells. Thus, the microspheroid recapitulates the human tumor environment. By applying cell death endpoints (the most rigorous of predictive measures) to these microspheroids, we have overcome most of the pitfalls encountered by earlier technologies. And, for the first time, a truly predictive human tumor model has been developed.

Of the two fundamental changes that we as a laboratory have brought to the field of chemosensitivity-resistance testing, the maintenance of cancerous tumor cells in their “native state” as microspheroids has been fundamental to our success.

Despite these important advances, many physicians have not grasped their significance. Falling back on their out-dated understanding of chemosensitivity studies that used growth-based endpoints (clonogenic, growth-to-confluence, and H₃* thymidine incorporation, etc.) many physicians have failed to incorporate the use of these highly validated methodologies into their clinical practices.

A review of the published literature, correlating these more rigorous predictive methodologies with clinical outcomes, clearly establishes the validity of cell death in microspheroids as an important breakthrough in cancer treatment.