by Azra Raza
There are several problems with the approach. First, it is extremely rare to have one gene driving one cancer. Second, even if such a mutation is identified, there aren’t many effective, approved, targeted therapies with which to treat the patient. Third, when a genetic mutation is matched to a drug, response is not guaranteed; in fact, the response rate is 30 percent at best. And finally, if everything works as planned and the patient even responds to the targeted therapy, the response offers no more than six months of improvement in survival over unmatched therapies. And this is the fundamental problem with most of the approaches; cancer treatments either don’t improve survival or the improvement is measurable in weeks or a few months, at tremendous physical, financial, and emotional burden.
How many thousands of tumors will need to be sequenced to find these rare patients, at what immense cost, and for what little benefit? Vinay Prasad argued in Nature in 2016 that the numbers will be very high; a sequencing program at MD Anderson Cancer Center was able to match only 6.4 percent of 2,600 patients with a targeted drug for identified mutations. A National Cancer Institute trial of 795 people who have relapsed solid tumors and lymphoma was only able (as of May 2016) to pair 2 percent of patients with a targeted therapy. Even then, Prasad reminds us, “being assigned such a therapy is not proof of benefit.” Only a third of patients respond to drugs given based on biological markers, and median progression-free survival is less than six months. Prasad estimated that precision oncology would benefit only 1.5 percent of patients, such as those in the NCI trial.
At the 2018 meeting of the American Association of Cancer Research, David Hyman from Memorial Sloan Kettering Cancer Center presented data on tumors in more than 25,000 patients. Of them, 15 percent matched with an FDA-approved drug and 10 percent with a drug in clinical trials. Prasad found similar proportions in his latest analysis, where 15 percent of 610,000 US patients with metastatic cancer were eligible for an FDA-approved, genome-guided drug. But once matched, just 6.6 percent likely benefited. Similar finds emerge from a study in Europe. From 2009 to 2013, the European Medicines Agency approved the use of forty-eight cancer drugs for sixty-eight indications. Only for twenty-six, or 38 percent, of those indications was there an improvement in survival, with a median benefit of only 2.7 months.
When I have questioned the practical feasibility of conducting such trials at the cost of hundreds of millions of dollars, one answer I regularly receive is a rather self-righteous one: “Well, Azra, for those 6.6 percent of patients, the extra five to six months or more mattered.” Of course the time mattered, although—since we’re talking about medians—it’s important to remember that half the patients would get less than the median benefit. And what about all the toxicity caused to the 93.4 percent who derived zero benefit from it? And all the wasted resources of sequencing thousands of tumors?
Take the example of the latest arrival in this area of precision oncology. In November 2018, the FDA approved the drug larotrectinib for the treatment of adult and pediatric solid tumors that express a neurotrophic receptor tyrosine kinase fusion gene (TRK). The trial of this small molecule, which led to approval of the drug, included a total of 55 patients, 22 percent showing complete and 53 percent a partial response. How long did the response last? Six months for two-thirds of the patients and a year for 40 percent. The test alone costs thousands of dollars per patient to find a very rare case. Treatment is likely to run in hundreds of thousands of dollars. For some two dozen patients who benefited for at least a year with this treatment, the approval of the drug is fantastic news. But bear in mind how small that number is in the face of the 1,735,350 new cases of cancer that will be diagnosed in the United States and the 609,640 who will die from the disease. This cannot be the most cost-effective way for us to move forward, yet such approvals are greeted as the new horizon, the game changer, the paradigm shift. It is my contention that these rare cases would be identified anyway through routine genetic profiling if we shift our focus to employing the genomic technology toward early detection. Instead of declaring victory, this approval by the FDA should serve as the impetus to envision better strategies for the future that can help a majority of cases.
Precision oncology ultimately fails because it ignores the evolutionary nature of cancer. As Theodosius Dobzhansky observed, “Nothing in biology makes sense except in the light of evolution.” In 1837, Charles Darwin sketched a tree trunk in his notebook with radiating branches representing the evolution of species from a common ancestor. Today, a graphic representation of cancer with all its genetic diversity and presence of multiple competing subpopulations of cancer cells emanating from the primary tumor is superimposable on Darwin’s tree of evolution. It has taken oncologists a long time to reach this understanding, thanks to a strange cleavage appearing among researchers early on. As molecular biology took off in earnest in the 1970s, investigators became convinced they would crack the cancer enigma. Reductionists, devoted to studying molecular genetic happenings in the cell, awash in overconfidence, drowned out the pluralists tracking the behavior of tumors as a whole. There was almost no cross talk between the two groups.
It is therefore no surprise that Peter Nowell’s clairvoyance about cancer being an evolving entity, with all its attendant therapeutic implications, an idea hailed as truly revolutionary today, remained largely ignored when first published in 1976. Fortunately, my husband, Harvey, was an exception. He immediately saw the genius behind Nowell’s paradigm, and he asked me to present the paper. Harvey was merciless in shredding one to pieces for the tiniest error during these weekly lab meetings. It was one’s familiarity with details that impressed him; any misstep, no matter how small, would discredit the entire presentation. I had to study the paper very carefully and read up on all sorts of background material in order to present the ideas coherently. That single paper helped me develop a radically different view of cancer very early on in my career. At the risk of testing the patience of readers not initiated into the specialized, dense, telegraphic language of science, it is worthwhile reproducing Nowell’s 138-word abstract from the classic paper “The Clonal Evolution of Tumor Cell Populations”:
It is proposed that most neoplasms arise from a single cell of origin, and tumor progression results from acquired genetic variability within the original clone allowing sequential selection of more aggressive sublines. Tumor cell populations are apparently more genetically unstable than normal cells, perhaps from activation of specific gene loci in the neoplasm, continued presence of carcinogen, or even nutritional deficiencies within the tumor. The acquired genetic instability and associated selection process, most readily recognized cytogenetically, results in advanced human malignancies being highly individual karyotypically and biologically. Hence, each patient’s cancer may require individual specific therapy, and even this may be thwarted by emergence of a genetically variant subline resistant to the treatment. More research should be directed toward understanding and controlling the evolutionary process in tumors before it reaches the late stage usually seen in clinical cancer.
That was more than forty years ago. Today, mapping mutational profiles in hundreds of individual tumors, combined with an astounding failure to develop any meaningful therapies for cancer in the interim, have confirmed the veracity of every word Nowell wrote. Simply stated, tumors also evolve by the Darwinian process of natural selection. Cancer begins in a single cell with one or more genetic mutations driving its release from growth-controlling signals. As the cell starts unchecked proliferation, its daughters pick up additional mutations, giving rise to multiple branches emanating from the tree. Each branch of cells carrying the driver mutation of the founder cell and the novel passenger mutations acquires novel metabolic and physiologic properties. Cells whose genotype matches the microenvironment develop a growth advantage, selectively expanding their population. Others wait their turn silently. No patient has one cancer.
There are countless cancers within each cancer. Since chemotherapy cannot kill every canc
er cell, the surviving cells are selected to adapt and regrow. This is the reason why even the most successful targeted therapies fail; they only kill off the cells with peculiar characteristics susceptible to the treatment, selecting the outgrowth of others with biologic diversity.
Every cancer is unique, yet some common principles apply to all. First, the malignant process begins in a single cell for practically all known cancers. Mutations accumulate in key genes related to proliferation, cell growth, and cell death, eventually giving rise to a cell with a growth advantage. This cell divides rapidly to produce clones of itself. All the daughters will share the same foundational genetic mutations, but in addition, some of the daughters will sustain additional mutations that give them biologic characteristics that are distinct from the parent. Formation of such subclones happens constantly in a tumor, but usually, a few clones dominate at any given time while others remain on the sidelines, waiting for sequential recruitment. Of course, malignant cells also leave their natural habitats and wander off to form metastases.
The presence of innumerable, biologically distinct daughter cells with additional mutations, chromosomal changes, and altered nutritional and metabolic requirements is the reason why even the best of targeted therapies are of transient benefit. Treatment to which one clone is sensitive leads to the selection of refractory, resistant subclones and a more invasive disease. A biologically new cancer results with an entirely different natural history, novel rules of proliferation and differentiation, newfound invasive potential, unpredictable responsiveness to therapies. These frightening, abrupt transformations of the disease are a spectacle to watch through the clinical prism of changing blood counts, paraneoplastic syndromes, and immune reactions. As clinicians, we regularly witness this kaleidoscopic, repetitive dance of motley populations within populations of cancer cells unfolding in real time in vivo.
Competing groups of cells take turns expanding and shrinking; changing places, honeycombing, crumbling, only to be reignited into action by newly acquired copying errors in the reeling, replicating DNA strands, seeking comfort in uninhabited beds, forging alliances with cooperative bedfellows in marrow niches and safe havens of supportive organs. Occasionally, a leukemia arises in the background of MDS with such malignant ferocity that all we can do is watch the vertiginous descent into entropy, spellbound, helpless in front of so anarchic a rebellion.
The microenvironment of tumors plays a critical role in clonal selection and promotion. Properties of the soil vary in different areas of the body. When studying ovarian cancer, a tumor spreading by direct physical invasion in the abdominal cavity rather than traveling through blood or lymphatics, researchers found that subsets of cancer cells thrived in a site-specific manner. Characteristics of the microenvironment were differentially suited to promote the growth of one clone over another. One seed, one soil; change the properties of a seed through a mutation and it would have to find a new home. It is an important reason why preclinical cancer platforms employing cell lines and patient-derived xenografts are likely to remain wholly inadequate as models for drug development; they are devoid of the in vivo microenvironment.
There is no sickness worse for me than words that to be kind must lie.
—AESCHYLUS
The secret to success in life is relationships. The secret of relationships is trust. The secret of trust is acknowledgment of pure and simple truth. The problem in oncology, as in life, is that truth is rarely pure and never simple.
One historic incident that has stayed with me since I first read about it as a teenager growing up in Karachi involves Mr. M. A. Jinnah, the founder of Pakistan. He spoke to a crowd of approximately ten thousand at a public gathering in Agra, India, in the early 1940s, years before the partition of the subcontinent into India and Pakistan. Probably five hundred people in that crowd had a passing knowledge of English, and about fifty of the elite among them understood it well. Trained as a barrister at Lincoln’s Inn in London, Mr. Jinnah spoke in chaste English with a British accent for forty minutes, and only in the last few did he address the commoners through a broken hybrid version of Urdu-Hindi-English. Shockingly enough, the crowd sat mesmerized throughout despite a complete lack of understanding. When asked afterward about what captivated them to such a degree, one man’s answer was, “Look, it’s true that I did not understand a word of what Mr. Jinnah said in English, but I have full trust that whatever he said was for my good and meant to protect me.”
Was the man’s blind trust justified? Trust is not just the sugarcoating glaze; it is indispensable, essential, vital. Too much willingness to trust is naive—a leap of faith that can earn deception. Yet a deeply meaningful blind trust is justified as long as the trustworthiness of the individual is already established. The man’s trust was based on an intelligent and experiential assessment of Mr. Jinnah’s previous actions, competence, reliability, integrity, and his demonstration of benevolence and empathy for the common man. Trust is not a static entity; it must be continually won.
Patients have the right to trust their physicians the same way Mr. Jinnah was trusted by his constituency. Do we deserve the trust?
In 1986, I had gone to Pakistan for a brief visit. One of the elderly female relatives at a family gathering, delighted to see me after several years, asked a curious question. “I don’t care how many degrees a doctor has, even if they are known to cure cancer, if they don’t have the reputation of shifa in their hands, I stay miles away from them. What I want to know is if you have been graced with shifa in your hands yet?” Shifa is an Urdu word loosely translated as “the healing power.” It is the equivalent of blind trust in one’s doctor—a powerful, intangible confidence that no matter how deadly their health challenges, and especially when medical knowledge is stumped, the physician alone possesses the wisdom to remain sensitive, to proceed in caring, empathic ways, always exclusively focused on the patient’s interest.
Lady N. thought I possessed shifa. She expressed her confidence at least half a dozen times during every clinic visit. She trusted me with her life. I obsessively tally the number of ways in which I let her down, this terrified, trusting, vulnerable woman, sitting in the consultation room, her mind and body besieged from within and without, desperately seeking a lifeline I had no power to conjure. Lady N. and I both knew that she had a fatal illness, that it was simply a matter of time before she would enter the bedlam surrounding end-of-life issues. Obviously, I had no cure to offer, no magic bullet to eliminate the coming leukemia, and each time she expressed her implicit trust in my power of shifa, I reminded her gently of what was expected. She scoffed, she laughed it off, she changed the topic, sometimes she became agitated, abruptly walked out. I broached the subject of involving our palliative care team, which she dismissed out of hand. What about a psychiatrist? “I have been on antidepressants practically all my life. My mother thought I was hyper when I was two! I don’t need more doping, thank you.” Lady N. simply refused to accept that the end could come for her. She demanded therapy for her cancer, not her mind, willing to be a guinea pig for any experimental approach I could concoct.
Things spiraled out of control in her case with frightening speed. Within days, she was admitted to the hospital with a high fever. I sat on her bed early one morning as she struggled to breathe. We were, for once, without the usual team of nurses, oncology fellows, and medical students crowding the bed, craning their necks to catch snippets of our conversation. Strangely enough, the intimacy of solitude had a distancing effect, introducing a formality in our communication, an uncharacteristic courtesy with which to speak about unspeakable things.
“We have to intubate you now and place you on a respirator. You can refuse.”
She caught her breath as the color drained from her face, then rallied and shot back, “Refuse and do what? Dr. Raza, I will not give up. Do whatever you can to keep me alive. For God’s sake, my mother is alive at a hundred. I have good genes. Freeze my body if I die. I want you to clone me when you have the technique
s worked out. I know you can. You are the only one I have full confidence in.”
I kissed her and called the anesthesiology team. We wheeled her down to the MICU. Within minutes, she was intubated, placed on a respirator.
What followed was less about supporting life and more about prolonging death. There was zero chance that Lady N. would ever be able to breathe on her own again since her fundamental issue was not the rapidly progressive pneumonia but her untreatable, fatal cancer. Infections were flaring up precisely because all of a sudden, pathogens had a free pass in her leukemia-riddled body. The immune system was fast approaching a state of total collapse as the bone marrow failed to produce the most critical first line of defense, white blood cells. I knew precisely what frightful days lay ahead. She did not.
I am a clinician first, and my medical, moral, and ethical obligation is to relieve distress and suffering caused by disease. I should enable my patients to benefit from the best that science and technology has to offer, not be hurt by it. Yet by offering to intubate her and connect her to artificial life support, as if death were an option, did I fail to protect Lady N.? What forces compelled me to offer her a choice of intubation, inviting her to accept unspeakable horrors that she had no clue about? The law, of course. What had I done to help Lady N. accept mortality? Did I do enough to explain the hopeless nature of her leukemia, the pointlessness of placing her on artificial life-support systems? Was it a failure on my part as her treating oncologist that somehow I transmitted a false sense of hope to Lady N.? Did I use language that made sense to Lady N. instead of confusing her? Or was it Lady N. whose nature dictated revolt, who would never take things lying down, no matter how much I tried to explain the hopelessness of her prognosis? What I know beyond a shadow of doubt is that to intubate her and attach her to a ventilator was the worst possible thing to do to her, and yet, against my better judgment, I was forced to give her the choice. So was I morally wrong in knowingly letting her enter the hellish nightmare of the next week? Where does medical and individual responsibility end and societal responsibility take over?
