by Ashutosh Jogalekar
A few days ago I finished watching a new documentary on Bill Gates’ life and work. One of the episodes narrated the sad story of the death of his mother in the mid 1990s from late stage breast cancer. She was a great philanthropist and a doting parent who managed to see Bill get married just before she died. At that point her son was already one of the most successful and wealthiest individuals in the world, but with all his resources and wealth, her life could not be saved. Steve Jobs, another person who had access to every medical treatment that money can buy, died early from cancer. These two stories tell us how great the leveling effect of cancer is, taking poor and rich alike without discrimination. Like war, cancer is the father of us all.
Today breast cancer can be treated much better than it was in the 1990s. There are better drugs and better radiation treatment options available, but for resistant late stage breast cancer the prognosis isn’t much better. In fact, as Dr. Azra Raza who is a distinguished oncologist at Columbia University tells us in this eloquent, thought-provoking and immensely sobering book, what’s true for breast cancer is true for most other kinds of cancer except for a few rare exceptions. The hard hitting truth is that in spite of tens of billions of dollars fueled into research around the world done by some of the smartest people in the field, the truly relevant endpoint for cancer – the increase in someone’s life span – has not changed much even after thirty years. For instance, a study of FDA-approved drugs from 2002 to 2014 showed that these drugs extended people’s lives by an average of only 2 months. Dozens of Nobel Prizes have been given out for basic cancer discoveries, cancer ‘moonshots’ have been promoted by politicians, startups and hospitals working on cancer continue to spend countless dollars and hours on a cure for the disease, but the two things that matter most for patients and their loved ones – extension and quality of life – haven’t changed much.
To know why this depressing scenario persists, Raza offers a simple reason with a hard answer: we are focusing too much on late stage cancer treatment, when the disease has already progressed and spread throughout the body, and much less on early stage detection and prevention. In spite of purported cancer breakthroughs in the media, the treatment is essentially the same as it has been for decades – slash (surgery), poison (chemotherapy) and burn (radiation), a triad of interventions sounding like they have been imported from the Stone Age, used because we can’t use anything better.
Perhaps there’s no better metaphor for the paucity of humane sophistication in cancer treatment than the realization that some of the early cancer drugs were derived from mustard gas and other chemical agents used in World War I. Today when you get malaria or the flu or even AIDS, at least in the developed world, a cure or a lifelong treatment is close to certain. When you get cancer, you instead do a mad dance of trial-and-error, trying out your best guess and hoping that it works.
The focus on cancer treatment rather than prevention comes from a complicated legacy of academic and industrial research in which cancer was treated the same way other diseases were, as a problem that could be solved by giving someone a drug. In Raza’s view this state of affairs needs to be fundamentally tweaked for a better future. Her book is a clarion call for changing the very face of cancer research, and it should be read not just by physicians and scientists who are working on cancer but by patients and their families as well as by politicians. The urgency of changing the nature of cancer treatment is driven home in stark, painful terms by Raza’s stories of some of her patients; the most painful and sobering story is of the untimely passing of her own husband Harvey, himself an accomplished oncologist. Many of her patients died deaths that were traumatic for both themselves and their families; many under circumstances where the purported cure was almost as bad as the disease. But while these case studies are sometimes harrowing to read, they underscore the fervent need to change the dialogue and the mechanics of talking about cancer, and they also reveal the luminous hope and humor that somehow tear their way through in her patients in the depths of despair. The book also has a very literary bent, and the erudite Dr. Raza quotes liberally from Urdu, English and Greek poets. She belongs to a select cadre of physician-scientist-writers who combine technical excellence and compassionate storytelling with literary eloquence. This is one of those books you keep thinking about long after you turn the last page.
Cancer is a hard nut to crack because unlike many other diseases, it’s a natural disease, a rebellion of the body against itself. It’s hard to kill because you are killing your own body in a sense. It’s only in the last twenty years that our view has changed from looking at cancer as a single disease to looking at it as a heterogeneous, emergent, complex set of diseases. However the basic cause is the same. As our cells divide from the time we are born to the time we die, there are errors in their imperfect division. The vast majority of errors are corrected by some very clever error-correcting enzymes, but some persist. The possibility of cancer arises when some of these errors lead to mutations affecting proteins whose job is to promote or curtail cell division or to modulate the cell’s metabolism. As the mutations build up, they hobble these protein ‘switches’, turning them perpetually on or off and leading to uncontrolled cell growth. The goal of most cancer research is to find drugs that block or kill cells with mutated proteins.
The problem, as Raza starkly and comprehensively describes, is that by the time cancer is actually discovered and treatment started, there are so many mutations in so many cell populations that treatment becomes extremely challenging, if not impossible. It’s like the quintessential hydra which grows a new head every time an existing one is chopped off. Yet modern cancer research continues to use the same outdated paradigm for treating this very complex set of affairs. For instance, almost all of basic cancer research is done using cells grown on petri dishes or in animals like mice. In some cases these cells are taken out of a body and grown in or on mice, while in some cases mice are given cancer by deleting genes for proteins that act as brakes on the cell cycle. Not surprisingly, there can be enormous differences in the genetic and biochemical makeup of cells in the complex, constantly changing milieu of a human body and the relatively simple, static milieu of a petri dish or even a mouse model (as a distinguished scientist I know once said in a talk, “Whatever else you may think of me I am not a large, hairless mouse.”). As just one basic difference, cells in the constrained environment of a capillary can be ‘hypoxic’ – that is, they have evolved to thrive in a low oxygen environment – whereas those in a petri dish are awash in oxygen. Studies have demonstrated that cells in a mouse or in a lab accumulate or eliminate mutations that ones in the human body harbor, making them quite different.
No wonder that drugs tested on cancer growing in such an artificial environment have a dismal rate of success when translated to human beings; for instance as Raza notes, therapies tested in a particular kind of mouse model of cancer called CDX have only a 5 to 7% success rate when tested in patients. A mouse model called patient-derived-xenografts (PDX) where human cancer cells are directly grafted on mice is more promising, but even there success rates are wildly uncertain. And as she astutely points out, even the drugs that do translate between mice and humans probably work because they are global cell killers or cytotoxic agents rather than because they target a specific mutated protein; because they target both cancer and normal cells, these drugs typically have very bad side effects like nausea, hair loss and ulceration of the mouth. In addition, conducting mouse cancer experiments is expensive and is a long process, so the cost-to-benefit ratio becomes even higher.
The side effect problem is in fact a universal one since cancer cells often don’t produce different proteins, just more quantities of the same proteins normal cells need. Many cancer treatments, including the much-publicized recent CAR T-cell therapy, therefore work by targeting cells indiscriminately. CAR T therapy can occasionally lead to striking remission rates, but it’s intrinsically dangerous because it works by suddenly killing billions of cancer cells, triggering an overwhelming immune response by the body that has to be managed by drugs as an emergency. Often whether a patient lives or dies therefore comes down to how much they can tolerate horrific side effects and not whether the drug works. This is a dismal situation and makes one feel that we are still battling cancer with the equivalent of sticks and stones. As human beings with claims to high intellects and humane aspirations, we must put a stop to these crude practices.
Raza also takes a blunt scalpel to the fashionable buzzword of ‘precision medicine’. The concept is not fundamentally misguided, but we simply don’t have a good way to turn it into a predictive tool. The basic idea of precision medicine is to tailor individual treatments to patients’ individual genetic makeups. But the problem with precision medicine is that it still relies on an overly simple, reductionist treatment of the problem. Most precision medicine still targets single mutations or proteins in a particular cancer and patient, but as we saw, cancer is an emergent, complex disease caused by several mutated proteins. Sometimes when one mutation is dominant and detected early enough the treatment can work, but such cases are rare; unfortunately these also happen to be cases widely publicized by the media and offering unrealistic hope to patients.
For instance, a drug called Gleevec for chronic myeloid leukemia (CML) became wildly successful and much talked about in the early 2000s when it spectacularly cut the mortality rate by almost 90%. It was the kind of medical miracle that doctors ardently await their entire career. But CML is a very special disease in which progression is driven dominantly by a single mutation, and Gleevec works effectively only in the chronic phase of the disease, when there are few complicating mutations, rather than the later acute phase. Thus, the story of Gleevec is actually a poster child for early intervention. This makes Gleevec a one of a kind drug, one that would be ineffective for other cancers in adults. Unfortunately as Raza points out, many of the miracle stories reported in the press are similar, involving a handful of drugs that work for a very small patient population under specialized conditions. Even the much-touted field of immunotherapy which received a Nobel Prize last year can have success rates of only 30%. What’s worse is that it’s often impossible to predict beforehand if the patient you are treating would be one of those 30%, so you simply have to try out a few treatments, each with their own side effects.
There are a few fundamental lessons to take away from these widely publicized occasional successes: One is that even immensely exciting, Nobel-caliber basic research does not automatically translate to actual extension of lives of cancer patients, the variable that should matter the most. Secondly, even when drugs work, we are pretty bad at predicting which patients they would work on. Thirdly, stop thinking that a drug is great because the FDA has approved it. Fourthly and perpetually, don’t believe everything you read in the papers.
So what’s the way out? One obvious solution is to combine different therapies that target many different mutant proteins; not only can this solution target cancer on multiple fronts, through multiple mechanisms, but it can also have a bigger probability of success because of the low likelihood of all proteins becoming resistant. This solution has worked successfully for AIDS, but unfortunately for cancer the drugs typically have so many side effects that combining them is often intolerable. Raza offers several other exciting suggestions. One of the transformative changes in our view of cancer is to see it as an evolutionary disease, with cancer cells constantly shifting on the mutational landscape. But this view cannot come from mouse models or cells in plastic containers; it has to come from patient samples. Along these lines, Raza has initiated a biobank effort in which she stores patient samples over time and examines their changing mutational profile. This kind of temporal view of cancer is enormously valuable and also directly relevant to patients. Unfortunately it’s harder to get funding for this important initiative than it is for Yet Another Mouse Model Study, and that is where our dialogue about cancer research needs to fundamentally change. This attitude of directly looking at patient samples – going to the source, so to speak – also needs to be baked into the education of young physicians and scientists, otherwise the flawed animal-and-petri dish cycle will continue.
Perhaps the most fascinating paradigm as far as I am concerned is to look to other, non-traditional fields for early detection and understanding of cancer. New improved methods of imaging, especially at the single cell level, could be transformative. Raza describes some very interesting insights she has had from the field of dynamical systems to describe cancer. In this view, cancer is like a sandpile which topples with the addition of that last grain of sand (this phenomenon is called ‘self-organized criticality’ a neologism from the physicist Per Bak who happened to be one of Raza’s patients and who unfortunately also died prematurely), but what we ignore is the entire foundation on which the pile rests and which has been building up over time. If we can modulate the growth of that foundation itself, we will be able to control the avalanche much better. There is no doubt in my mind that insights from fields like graph theory which can look at cancer protein networks and biomedical engineering which can deliver existing drugs better will make dents in cancer therapy in the future.
There are several promising directions for detecting that foundation and intervening in its growth. Although Raza notes some of these I would have appreciated some more detail, and although I share her optimism about the potential of some of these directions, I am not as optimistic as she is about how fast we can make progress in these fields. For instance she notes both AI and intensive sequencing of blood samples to detect even a handful of cancer cells floating around, but there are a lot of fundamental statistical sampling challenges in doing this; detecting a few cancer cells in many liters of blood is like trying to find a needle in a haystack of needles. AI technology for cancer is much hyped and little substantiated, and the cases where it works well involve diagnostic applications like pathology image analysis rather than early detection. There are also some out-of-the-box approaches that Raza notes; one particularly amusing experiment involves ‘yelling’ at a cancer, essentially making a mass of cells vibrate using ultrasound and looking at the cells that are shed.
Personally I also think that looking into ancient Indian and Chinese medicine – the discovery of a cure for malaria from the latter won a Nobel Prize a few years ago – can be very promising, since these systems have intrinsically focused on a holistic and emergent rather than a reductionist view of disease for centuries (the problem with these systems is that the data is of uncertain quality and hard to get hold of). I am also very excited about the potential of wearable and continuous monitoring technology embedded in our houses and on our person. We are already seeing the benefits of watches and mobile phones monitoring basic signs of health, and I think it’s not going to be too long before we see things like smart toilets or nano-sensing pills that monitor more sophisticated measures of health.
Whatever the merits and timelines of these approaches, the overwhelmingly important, urgent message in Raza’s book is that we need to focus on detection and prevention with the same fervor that we continue to investigate drugs against mutated proteins and mouse models. Doctors, scientists, the FDA, the NIH, patients and politicians all need to change the conversation from treatment at a later stage to detection at an earlier one and need to read this book. Cancer is a disease which in some sense we carry with us all the time, emerging as it does from holistic, natural processes of division and growth. As Dr. Raza so effectively, poignantly and eloquently describes in this volume, only a holistic, natural approach can help us conquer this emperor of all maladies. In the words of President Kennedy, let us begin.