Product Description

The National Cancer Act of 1971 propelled the War on Cancer, mainly by tapping the vast resources of the Federal government to confront the growing cancer challenge. As a result, all cancer initiatives funded by Federal dollars were channeled through a restructured National Cancer Institute and some predicted the conquest of cancer by the nation’s bicentennial. Yet, over the ensuing three decades progress in the clinical setting has been slow, and cancer remains a largely incurable disease.

After reviewing the history of cancer and its impact on the population, Dr. Faguet exposes the antiquated notions that have driven cancer drug development, documents the stagnation in treatment outcomes despite major advances in cancer genomics and growing NCI budgets, and identifies the multiple factors that sustain the status quo. He shows that, contrary to frequent announcements of breakthroughs, our current cancer control model cannot eradicate most cancers and the reasons why. Significantly, this book also delineates a way forward via a shift from the discredited cell-kill approach of the past to an integrated, evidence-driven cancer control paradigm based on prevention, early diagnosis, and pharmacogenomics. The author's views are based on data published in mainstream scientific journals and other reliable references, 432 of which are cited.

Product Details

• Amazon Sales Rank: #1103311 in Books
• Published on: 2008-05-29
• Original language: English
• Number of items: 1
• Binding: Paperback
• 227 pages

Review

From the reviews:

"I am quite impressed with the scope of the effort, the excellent writing, and the sharp and accurate critique over a broad area of modern oncology practice and research. Frankly, I am not aware of many scientists who could synthesize both the basic and clinical material, and lucidly depict the implications in such a compelling manner". (Neil Caporaso, M.D., Chief, Pharmacogenetics Section, Genetic Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute, MD.)

From the reviews:

"Dr Faguet's approach is not unlike an analysis of the so-called industrial military complex.…a complex consisting of the federal government, pharmaceutical drug companies and academia—all have a part of the pie and therefore a reason for maintaining the status quo. Insurance companies, regulatory agencies such as the FDA, and patients themselves add further layers of complexity. Faguet walks through all of these areas in a critical but unemotional manner... His book should be of interest to policy-makers, cancer-care workers, and the public at large". (Gerald E. Marti, M.D., Ph.D. Chief, Flow and Image Cytometry Section, Laboratory of Stem Cell Biology, National Institutes of Health MD.)

From the reviews:

"I think your book will be a valuable corrective to much of the hype that has been splashed around in greater and greater quantities by the cancer establishment as it tries to stifle the message that it has failed to make much of an impact" (John C. Bailar, M.D., Ph.D., Former Deputy Associate Director for Cancer Control, National Cancer Institute; Former Editor-in-Chief of the Journal of the National Cancer Institute; Professor Emeritus, University of Chicago, IL.)

From the reviews:

"The War on cancer is a trenchant and provocative book that offers a critical view of the last four decades in the field of oncology and a vision for the future. To the mainstream oncology community, I would commend it as a challenging treatise" (Randall Hughes ,Journal of the American Medical Association, June 2006)

Customer Reviews

An excellent critique
A prominent cancer researcher has published a trenchant critique of conventional oncology. Guy B. Faguet, MD, Professor, Department of Medicine, Section of Hematology and Oncology, Medical College of Georgia, has written The War on Cancer: An Anatomy of Failure, A Blueprint for the Future (2005).

Dr. Faguet is no stranger to cancer research. After receiving his MD degree in Bogota, Colombia, he did postgraduate work at the University of Texas and at Ohio State University. He then conducted cancer research in Augusta, GA, for 28 years, funded mainly by the National Cancer Institute (NCI) and the Department of Veterans Affairs (VA). His output has included 140 peer-reviewed articles, 7 book chapters, and two previous scientific books on cancer. He is an expert on chronic lymphocytic leukemia (CLL).

Starting about fifteen years ago, the doctor told me in a recent telephone interview, he began to develop serious misgivings concerning the lack of progress in the war on cancer. At the urging of departmental colleagues, he began writing The War on Cancer about five years ago and it was finally published by the German medical publisher, Springer, in late 2005.

Faguet has a keen sense of medical history. He points out that after the development of effective treatments for Hodgkin's lymphoma, in the 1970s, many researchers thought that the cure for the more common cancers was just around the corner. But the successful conquest of Hodgkin's disease has remained an isolated victory. "Indeed," Faguet notes, "little additional progress has been made towards the cure of most invasive cancers. In fact, in the last 20 years, only testicular cancer has been added to the short list of malignancies routinely curable using chemotherapy" (p. xiv).

How then does he explain the much-vaunted decline in the death rates for some kinds of cancer? Stomach cancer is a good example: in the early years of the 20th century, this was the most common form of internal cancer in the US, but is now relatively rare. Faguet attributes the improvement not to any dramatic advance in cancer therapy, but to "prevention and early-stage detection, to food refrigeration, to improved infection control and transfusion therapy, to enhanced nursing, social, and rehabilitation services, and to better general medical support, rather than to advances in cancer treatment" (ibid.).

In other words, general health and sanitation measures have improved survival, while there has been no change in the effectiveness of the treatment itself - a type of progress that many promoters of conventional therapies conveniently overlook.

Dr. Faguet also takes on the misleading nature of five-year survival statistics. Improvements in five-year survival are frequently cited as proof that cancer treatment is increasingly effective. One need only look at how such improvements are showcased by the American Cancer Society (Cancer Facts & Figures 2006: 17-18). For instance, for all cancers, five-year survival rose from 51 percent in 1974-76 (the beginning of the war on cancer) to 66 percent in a more recent period (1996-2001). This is the basis of claims that whereas only half of all patients survived at the start of the war on cancer, today two-thirds of patients survive their disease - an improvement that is usually ascribed to steady progress in the realm of cancer treatment, especially chemotherapy.

But, as Faguet shows, this is a gross over-simplification. "While improvements in five-year survival are frequently presented to the public and to policymakers as evidence of success in the War on Cancer, they should not be," he asserts. "This is because while survival is a valid measure of treatment outcome within a clinical trial, it is misleading when applied over long periods of time. Indeed, factors other than therapy affect survival favorably. They include improvements in supportive medical care and better screening and diagnostic tools."

An additional reason for observing improved cancer survival over the years is that, as cancer detection tools improve, cancer is diagnosed in incrementally earlier stages leading to a phenomenon called "lead time bias". Simply stated, the earlier the diagnosis the longer will patients live with their disease, giving the false impression of increased survival that can and has often been attributed to newer treatments.

Faguet also shows that cancer incidence and death rates present a very mixed picture, but in general are not falling, as we have been led to believe. In fact, if the age and size of the US (and world) population continue to increase at current rates, so too will the overall number of cancer patients. Cancer is a primarily a disease of aging populations (the average age of diagnosis for adults is 67 years), and so the graying of the baby boomers will in all probability herald a new spike in cancer incidence and mortality figures.

Faguet ascribes the general failure of the war on cancer to the application of the "cancer cell kill paradigm" that was fostered by the application of the microbial model to cancer treatment. Scientists in the late 19th century generally believed that one or more microorganisms also caused cancer. Even after this "cancer microbe" theory was broadly rejected, however, drug development and patient management continued to be based on the premise that cancer is in essence some sort of foreign invader that must be eradicated at all costs. But while it is true that in some limited cases cancer is indeed caused by a virus (such as the human papilloma virus that causes cervical cancer), in general cancer is essentially a runaway product of the human host. To paraphrase the humorist Walt Kelly, "We have met the enemy and he is us."

The cell kill paradigm holds that these "foreign" cancer cells must be eradicated like swarming germs before they overwhelm the host (p. 63). For various technical reasons, this aggressive approach has worked sufficiently well in the case of Hodgkin's disease, where a combination of four drugs (the so-called MOPP protocol) is curative in many cases, albeit at a significant cost in toxicity and second cancers. However, as Faguet points out, "this early success was seldom replicated despite a myriad of subsequent clinical trials launched to test a variety of intermittent combination chemotherapy regimens in many types of cancer over the ensuing four decades."

Chemotherapy has also cured acute lymphocytic leukemia (ALL) in children, choriocarcinoma, germ cell tumors, and a few other rare types of cancer in pediatric and young adult patients. Additionally, as an adjuvant, it modestly improves survival after surgery in a number of adult cancers. But by and large it has been a failure in treating advanced disease. As both Faguet (2005) and the Australian researchers Graeme Morgan et al. (2004) have shown, chemotherapy is responsible for curing only approximately 2 percent of those who receive it for advanced cancer.

Faguet is also critical of immunotherapy, which dominated cancer research in the 1970s and part of the 1980s. Few now remember the initial hype that greeted the emergence of interferon, interleukin, and the other so-called 'biological response modifiers' (BRMs).

"Each immune enhancer rode a wave of enthusiasm within the medical community and in the press," says the author, who himself performed scientific work on the immunological dimensions of cancer. For example, the drug interferon-alpha, a cytokine, was greeted with a deluge of media coverage, mainly thanks to its astute promoters. "It was touted as a 'magic bullet,' a 'miracle cure,' 'like the genie in a fairy tale,' that was equally good to cure the common cold or cancer," he writes. He recalls that when an article in the New York Times finally called its efficacy into question, four scientists from Sloan-Kettering Institute, New York, wrote a letter to that newspaper "expressing dismay that such reporting might undermine public support for interferon research" (p. 65).

Eventually, however, the media and the public caught on to the hype, and interferon has disappeared from the treatment of most kinds of cancer therapy (although it still has a limited role in treating a number of malignancies). This overselling of interferon should have been a cautionary tale for all concerned, yet in fact it became the model for future hype campaigns. Just a few years later came an even more extreme promotion of another cytokine drug, interleukin.

"Despite two decades of intense studies," Faguet writes, "immune stimulants have had little impact on cancer management" (p. 66). This section of the book is perhaps a bit ungenerous towards the immunological concept, which I believe still holds promise as an adjunctive treatment for cancer. In his eagerness to make his argument, Faguet gives short shrift to the great progress that has been made in the field of immunogenetics and in the development of various kinds of cancer vaccines.


Fallacies of Chemotherapy


Faguet saves his most trenchant criticism for the fallacies involved in the application of cytotoxic chemotherapy, especially high-dose treatment. His background as a researcher into these very treatments makes this part of his analysis truly compelling. He writes here with a sure hand that will certainly have the effect of increasing the ranks of those who are doubtful about chemotherapy's effectiveness.

The essential fallacy of chemotherapy, says Faguet, is "that while most patients achieve some degree of tumor response few survive longer as a result." This is certainly the essential point - the general lack of any correlation between tumor responses (especially partial responses) and overall prolongation of life. He treats the reader to a fascinating overview of the history of chemotherapy, leading up to passage of the National Cancer Act of 1971. Faguet has a clear understanding of the various forces that led to passage of that Act. But while that multi-billion legislation (NCI's budget is now $4.8 billion per year) has funded astonishing progress in the basic sciences, it has been an almost total failure in finding actual cures for common cancers.

"Three decades later," says Faguet, "the process of anti-cancer drug development remains mostly anchored on this century old, conceptually antiquated, technically inefficient, labor intensive, costly, and low-yield 'hit and miss' (mostly miss) screening approach engineered and sponsored by the National Cancer Institute (NCI)."

He shows how the National Cancer Institute's approach has been largely empirical. For example, although NCI has doggedly screened tens of thousands of potential cancer treatments, natural as well as synthetic, it has found only a handful of useful agents, almost all of which are cellular poisons. "No existing laboratory method," says Faguet, "can accurately predict the anticancer efficacy of a particular chemical..." (p. 73).

He is also strong at describing the competing and interlocking theories of cancer cell dynamics that underlie the application of toxic drugs. For many years, the dominant concepts were those of Howard Skipper, MD. These were based on test tube models that did not mimic what happened in actual human patients. Skipper's "laws" were succeeded by Mendelsohn's concept of growth fraction and the hypothesis of Goldie and Coldman on drug resistance. Oceans of ink have been spilled in disputing these conflicting theories. But, as Faguet points out, the fact is that "none of these hypotheses led to more efficacious cancer management and today the outcome of most cancer patients remains grim" (p. 78). The bottom line for any cancer theory is always its actual effect on patients' survival.

Faguet also explains how the limited progress made in treating a few cancers has been used to obscure the failure to cure more common forms of the disease. Thus, he writes, "true prolongation of survival has been achieved over the last decade or so in subsets of patients afflicted by some cancers including breast, prostate, and colon. On the other hand, favorable survival trends in many cancers observed over several decades relate to factors other than cancer treatment" (p. 89).

Honorable intentions, good effort, questionable solution
1. Scope of the Problem

This book offers a detailed critique of the "War on Cancer" by Guy Faguet, MD, an oncologist and cancer researcher with about 30 years of experience. The fact that only three people have previously reviewed this book reinforces my impression that most cancer clinicians and researchers are too mired in the orthodox cancer establishment to adequately question it. That's deeply unfortunate because, as Faguet demonstrates with abundant and tragic statistics (mostly from the NCI), we've been losing the War and, if anything, the situation is getting worse rather than better.

To put things in perspective, if "cure" is defined as 5 continuous years of disease-free survival (a questionably timid definition), only about 11 out of the more than 200 types of advanced (invasive and metastatic) cancer are currently routinely curable, with meaningful survival prolongation for only a few more. This corresponds to about a 2% cure rate overall. To look at it another way, about 1,500 people in the US alone die every day from cancer, or one per minute around the clock, day in and day out; that's the equivalent of a 9/11 death toll every two days, or 180 losses of 9/11 size every year.

Considering the magnitude and complexity of the cancer problem, and our continued inability to solve it, a cursory analysis of the situation won't suffice. Accordingly, my review of this book is quite long and detailed, so bear with me ...


2. Biological Factors

From a biological standpoint, Faguet identifies several key contributors to the problem:

* Cancer cells differ from normal cells by (at least) having a mutated genome, especially mutations of oncogenes, tumor suppressor genes, and DNA repair genes. These mutations confer a survival and growth advantage to tumor cells. Oddly, Faguet doesn't mention epigenetics.

* Patient heterogeneity results in a wide range of responses to a given drug. Inability to adequately address this heterogeneity has resulted in resorting to randomized controlled trials which are analyzed statistically. The outcome for each patient thus becomes a largely random matter of "odds" with few patients "beating the odds." Though Faguet doesn't make the distinction, there are at least three types of heterogeneity to consider: (a) genetic and phenotypic differences between patients which are unrelated to their tumors, (b) differences in the tumors of patients despite having the same visually-based histopathologic diagnosis (intertumoral heterogeneity), and (c) genetic and phenotypic heterogeneity among the tumor cells within a particular patient's tumor (intratumoral heterogeneity), along with related "ecological" aspects involving the interactions among tumor cells and with normal tissues.

* Drug development has similarly been largely a matter of trial and error. As a result, only a tiny fraction of compounds screened in vitro have wound up being used clinically. For example, according to Faguet, between 1990 and 1998, the NCI screened 70,702 compounds, of which only 10 were eligible for toxicity trials in animals and humans.

* Tumor shrinkage isn't a sufficient condition for long-term survival, since shrinkage can be transient and recurrence can be rapid. Similarly, tumor shrinkage isn't a necessary condition for long-term survival either, since one can survive long-term if tumor size is stable (zero or minimal growth rate). Short-term tumor response (eg, PFS-6) is thus not strongly correlated with long-term survival and isn't an adequate surrogate.

* Minimal residual disease (tumor) is very difficult to detect using currently available methods.

* Besides being mostly (but not entirely) ineffective, chemotherapy drugs generally have a narrow (or zero) therapeutic window. Moreover, switching chemotherapy drugs after treatment failure is usually (but not always) futile.

* Genomic instability is a hallmark of cancer. Though Faguet doesn't note it, this means that most tumors are likely to evolve, both naturally and in response to selective pressures of treatments. This is a major reason for the heterogeneities noted above.

* What were once thought to be simple one-way biochemical pathways, especially signaling pathways, are in fact "multidirectional, redundant, and recurrent." What Faguet doesn't note is that our mapping of these complex networks of pathways is far from complete (and probably impossible), some of it is probably wrong, and almost all of it is more qualitative than quantitative. We can also add that homeostasis is a pervasive concept in physiology, yet it's rarely mentioned in the cancer literature; it shouldn't be a surprise that tumors and tumor cells employ a suite of redundant mechanisms to robustly preserve their integrity in the face of changing environments, including changes induced by treatments. The fact that a tumor continues to develop over years and decades, and persists in the face of a barrage of toxic treatments, is self-evidence for the robustness of the tumor.

* Tying together many of the above points, even the lauded targeted treatment Gleevec (imatinib) doesn't work universally for chronic myelogeneous leukemia (CML). In some CML patients, Gleevec doesn't inhibit tyrosine kinase activity because of increased drug efflux, bcr-abl amplification, other kinase mutations, and other factors. In other patients, even if the kinase is inhibited, Gleevec is ineffective or relapse occurs due to mutations other than bcr-abl.


3. Questionable Biological Considerations

Faguet also discusses several other biological considerations, and here I find his analysis questionable:

* He claims that the cell-kill paradigm (modeled on germ control) is flawed. I agree that this approach has been mostly ineffective so far, but there have been some successes, and the concept of killing cancer cells while sparing normal cells is rational. We might be able to find better ways to selectively kill cancer cells in the future.

* He claims that most solid tumors exhibit no chromosomal abnormalities. My understanding is that chromosomal instability (CIN) and aneuploidy are quite common.

* He correctly notes that strategies to overcome chemotherapy drug resistance and increase efficacy, such as combination therapy and alternative dosing schedules, haven't been very effective so far. I agree but, again, there are important exceptions, so these strategies shouldn't be summarily dismissed.

* He calls chronotherapy "whimsical" but provides no evidence or arguments for this dismissal. My understanding is that oncologists such as Dr. Keith Block have had some success with this approach.

* He has a rather negative view of immunotherapy, arguing that cancer cells aren't fundamentally different from normal cells (thus not recognizable as nonself by the immune system), and that two decades of immunotherapy failure argues for abandoning this approach. At best, this conclusion is premature. For example, certain vaccine immunotherapies (eg, DC-Vax and CDX-110) are already proving to be among the best options available for malignant brain tumors, with survival times and toxicities which are much better than traditional chemotherapies.

* He likewise dismisses chemopreventive drugs intended to prevent incidence or recurrence of cancer, on the grounds that they have uncertain mechanisms of action and their side effects may offset any benefits. First of all, basic measures like good diet, exercise, and appropriate use of supplements can provide substantial chemopreventive protection. Secondly, once again, inability to develop chemopreventive drugs so far doesn't preclude their development in the future. If successful, such an effort could have a huge payoff, especially if we continue to lack effective cancer treatments.


4. Institutional Factors

Faguet describes several institutional reasons why we're losing the War:

* The cancer establishment, funded by the NCI, drug companies, and ultimately the public, is heavily shaped and perpetuated by inappropriate financial incentives which frequently go against the public interest. Researchers are constrained in the types of ideas they can pursue because of the need for funding, drug companies are motivated by profit streams rather than low-cost solutions, clinical trials are lucrative for investigators, and oncologists in private practice have a serious conflict of interest because they often sell chemotherapy and it can produce 2/3 of their personal annual income, which ranges from about $250K to over $1 million.

* Clinical trials are cumbersome, slow, and expensive, with hundreds of millions of dollars per drug being typical. Faguet notes that Phase I trials take about 1.5 years and enroll 20 to 100 patients, Phase II takes about 2 years and enrolls 100 to 500 patients, and Phase III takes about 3.5 years and enrolls about 1,000 to 5,000 patients, so the total is about 7 years. Moreover, trials with unfavorable outcome are often not reported. We can add that clinical trials are far from a perfect tool, since they have limitations related to the way they're designed, exclusion/inclusion criteria for patients, and the ways their results are analyzed and interpreted. These limitations can lead to falsely positive and falsely negative conclusions, as well as uncertainties in applying study findings to both particular patients and broader patient populations.

* The media, in sync with researchers and drug companies, and pandering to public hopes, often reports alleged "breakthroughs" in treatment which are usually preclinical and don't pan out clinically, but they don't follow up and report these failures. This gives a false impression of progress and a false hope that effective treatments are right around the corner.

* The "standard of care" establishes a norm which becomes difficult to deviate from in either direction (doing less or doing more), even though it may not be the best option for a given patient, nor even for the overall patient population.

* Faguet notes that off-label use of cancer drugs is a widespread practice, based largely on anecdotal data, but it seldom works. I agree that it has seldom worked so far, but I think it's sensible to keep this option open, since careful clinical and theoretical investigation can suggest genuinely promising combination treatments using off-label drugs, and the evidence strongly suggests that combination treatments have a better chance of success than monotherapies (as Faguet himself notes). Moreover, off-label drugs can be effective yet inexpensive (eg, chloroquine for treatment of malignant gliomas).

* Although Faguet doesn't note it, we can add that the cancer establishment is incredibly fragmented, with rather poor coordination of efforts and poor integration of findings from different lines of research. This makes for high cost and meager results. We can also add that profit motives prevent the development of treatments which may be effective but unprofitable (eg, supplements and drugs with expired patents).


5. A Way Forward

Finally, Faguet proposes a way forward, hoping that it will lead to winning the War:

* He notes that most breakthroughs in medicine come from creative independent researchers rather than organized centrally-directed research. This argues strongly for a more decentralized and interdisciplinary approach to cancer research, with a substantial number of smaller grants being awarded to support exploration of novel ideas, and with a forum (probably web-based) for efficient and effective collaboration. The NCI has taken some steps in this direction in recent years, but it's too little and too slow.

* He argues that we need more effort to prevent cancer. I agree. Consider that smoking is responsible for about a third of cancer deaths in the US, and kills as many Americans as the next 10 causes combined. Also, viruses are estimated to cause at least 15% of all cancers worldwide, so vaccination may significantly reduce cancer risk. We can add that we need more effort to identify and control the environmental factors which contribute to development of cancer.

* He argues that we need early diagnosis via screening. Again, I agree, although we need to be careful about the many adverse consequences of false positives and false negatives, especially given that current screening methods are still somewhat primitive.

* He disagrees with Sidney Farber's statement that cancer can be cured without understanding mechanisms of action. I see Faguet's point, but I mostly side with Farber on this. First of all, biochemical pathways and mechanisms of action may involve complex networks of interactions which will remain beyond our understanding for many years, if not forever, so an effective "bottom-up" reductionistic solution based on molecular biology may not achievable. Secondly, a proper "top-down" analysis of the vast experimental and clinical database already at hand, looking at the details of patterns of success and failure (including outliers), may suggest effective treatments without requiring an understanding of mechanisms of action. The concept of "undiscovered public knowledge" pioneered by Donald Swanson is very relevant here. Remember that our true need is results, not knowledge. Cancer is ultimately a human problem, not a scientific question. If we can find a solution to the problem, it doesn't really matter if we don't know why the solution works.

Most fundamentally, Faguet argues that, instead of the cell-kill paradigm, we need treatments which target the molecular and genetic defects of cancer cells, aiming to normalize the behavior of cancer cells rather than necessarily killing them. Such treatments would be developed using a bottom-up pharmacogenomics approach, may use multiple agents to target most or all molecular and genetic defects, and may use dosages less than the maximum tolerated dosages (MTD) typically used with traditional chemotherapies. He cites Gleevec (for CML) as a "proof of concept" for this approach.

Unfortunately, the argument for targeted treatment runs into several problems. First of all, CML is "simpler" than most cancers, so it's more likely to be the exception than the rule, and the intratumoral heterogeneity and complexity of most cancers may put them beyond the reach of targeted treatments. Secondly, Faguet himself acknowledges that even Gleevec doesn't work for all CML patients and eventually it stops working (see above). Thirdly, adequately hitting the correct target while avoiding off-target effects is much easier said than done. And perhaps most distressing, many targeted agents have actually been tried clinically in recent years and they've been mostly ineffective. Most of these trials used single agents, though a few trials have combined two or three agents, still without much efficacy. It's possible that targeted agents might still be effective if we combine a relatively large number of them (~5 to 15) and select them on a strategically rational basis, but cumulative and interactive toxicities would be a concern with such large combinations (especially related to off-target effects). Using dosages much lower than the MTDs of each agent alone may substantially alleviate this risk, but it's unclear how much therapeutic window would be available (if any). And it's also possible that combining many targeted agents simply won't work, for the reasons already noted (heterogeneity, evolution, robustness, etc.).


6. Other Ideas

Overall, I think Faguet's book is important and I recommend reading it because it at least offers a detailed critique of the War, raises important issues, and makes some valid points. But obviously I don't fully agree with his analysis and recommendations. I see more promise in the following approaches not mentioned by Faguet:

* Develop a "philosophy of oncology" in order to better understand the existing cancer paradigm, its problems, and how it can be improved (or totally revamped via a paradigm shift). This would be analagous to the established fields of philosophy of physics, philosophy of biology, philosophy of medicine, etc. There's even a philosophy of chemistry, yet no philosophy of oncology -- simply incredible when one consider the human toll taken by cancer.

* Use an adaptive Bayesian approach in order to make clinical trials more efficient and effective (see especially the work of Donald Berry in this regard).

* When reporting results of clinical trials, focus on survival outcomes and don't rely on other surrogates unless they've been shown to strongly correlate with survival. And don't report only median survival, which is inadequate because (a) it's insensitive to the shape of the survival distribution and (b) it represents only the "middle of the pack" and thus gives no indication of the variability of outcomes. Instead, routinely report the 15th, 50th, 85th, and 98th percentile outcomes. The first three of these approximately represent the median values of the bottom, middle, and top third groups of outcomes (with each group thus being equally likely), and the 98th percentile gives an indication of the best-case outcome that could reasonably be expected. Since the treatment benefits associated with 15th and 50th percentile outcomes are usually relatively small (on the order of months), the 85th and 98th percentile outcomes would give a better indication of how much benefit a treatment would provide if a patient is lucky enough that the treatment is effective. In short, this approach would provide a much better basis for making rational treatment decisions.

* Continue development of immunotherapies, and also develop other novel cytotoxic treatments which are selective for cancer cells, based on principles different from traditional cytotoxic chemotherapies.

* Consider alternative chemotherapy schedules, such as metronomic chemotherapy.

* Individualize treatment (including dosing amounts and schedules) to each patient, including adapting the treatment as the patient's particular clinical course unfolds. In other words, get away from "one size fits all" and "time-invariant" treatment. Tumor genetic and molecular profiling may help in this regard, but this approach faces many challenges, including the genetic (intratumoral) heterogeneity and instability of tumors. A more direct and reliable approach may be chemosensitivity testing, in which particular drugs are directly tested against patient tumor specimens. Improved imaging methods and biomarkers may also help in monitoring tumor behavior and treatment efficacy.

* If treatment has been effective, control growth of minimal residual tumor by continuing treatment long after a tumor is no longer clinically detectable (provided that toxicity is acceptable), and also continue chemopreventive measures (at least CAM).

* Target tumor stem (or stem-like) cells.

* Target the tumor microenvironment (eg, neovasculature), in addition to (or instead of) targeting tumor cells directly.

* Inhibit tumor invasion (this is particularly important if tumor cells can't be killed).

* Inhibit tumor evolution, or steer tumor evolution in a direction which is more benign or makes the tumor more vulnerable to treatment. This could involve using anti-mutagenic agents (with consideration for possible reduction in efficacy of radiotherapy and conventional cytotoxic chemotherapy), as well as particular treatment sequences, including rotating treatment schedules.

* Target tumor metabolism, especially aerobic glycolysis, by exploiting the increased glucose appetite of cancer cells, as well as other metabolic processes to which tumor cells may be "addicted" (eg, autophagy).

* Apply findings from developmental biology to better understand tumor development and behavior, and related treatment possibilities.

* While I agree with Faguet on use of combination treatments, don't limit such treatments to targeted treatments.

* Use mathematical modeling to optimize treatments based on the quantitative relationship between treatments and tumor growth rate. Why bother with a target which looks qualitatively attractive (eg, based on position within a signaling network), if even perfectly hitting the target may have little quantitative benefit in terms of tumor growth rate and patient survival? Only a quantitative model can answer this question.

* Apply complexity science to better understand the complex nature of cancer and to identify better treatment strategies. Systems biology could play a role here.

* Apply information-theoretic approaches as an alternative (and relatively abstract) way to potentially better understand and control cancer. In this regard, it may be noteworthy that carcinogenesis, tumor growth, and tumor evolution appear to be processes in which entropy increases. Thus, models from statistical mechanics and thermodynamics may have some relevance here (such models are being attempted even in economics, in the field of "econophysics").

* Use CAM measures such as diet, exercise, supplements, and mind-body approaches, supported by meaningful NCI funding for clinical trials of such measures.

Thoughtful Critique
Written by an experienced oncologist and cancer researcher, this book is a searching critique of the general direction of cancer research and treatment over the past generation. Faguet aims at reaching a broad audience and covers cancer epidemiology, basic genetics and cell biology, treatment approaches, policy issues, and recommendations for the future in a series of concise chapters. This is a well written book and Faguet does a very nice job of being broadly informative without minimizing the difficulty of the problems involved.

In the early 1970s, Federal support for cancer research and treatment was increased enormously; the so-called 'War on Cancer' aimed a producing cures for cancers. Faguet devotes the meat of the book to a critique of this effort. Thirty plus years later, cancer mortality is falling but not because of marked improvements in medical treatments. The decline in cigarette smoking and improved screening for some common cancers - breast, colon, prostate - is largely responsible for the decline in cancer mortality. While powerful treatments have been discovered for some rare to uncommon cancers, medical treatment for common cancers remains highly unsatisfactory. Faguet is not nihilistic about developments in basic research related to cancer; quite the opposite, like everyone who is familiar with this field, he is enthusiastic about the potential of more recent work in the basic biology of cancer. The problem, as Faguet argues well, is that most clinical cancer research and drug development has been dominated by an inadequate scientific model - the concept of eradicating malignant cells via cytotoxic therapies. Even when alternative theories have been pursued, such the the immune surveillance idea, they have not been scientifically well justified. The basic flaw of building the treatment development effort on a shaky scientific foundation resulted in huge dispersion of resources without commensurate yields. Faguet does not spare clinical practice. The intellectual success of the cytotoxic concept, plus the culture of standard of care practice and some perverse financial incentives, resulted in the very widespread use of cytotoxic therapies, even when benefits are likely to be modest.

Faguet's prescriptions for change are sensible. He suggests better efforts at controlling cancer risk factors, which in practice means much more aggressive tobacco control efforts. Since screening has clear benefits, he suggests more effort be devoted to developing better screening methods and programs. Finally, and this is not particularly novel, he urges abandonment of the cytotoxic approach in favor more biological sensible treatments. To a considerable extent, this means abandoning the cure idea for the model of managing chronic disease.

Both Faguet's diagnosis and prescriptions are cogent. The War on Cancer is a textbook example of failed dirigiste science policy. One area where I think Faguet may be wrong is his enthusiasm for new therapies based on our now much better knowledge of cancer biology. Many of these cellular defects are quantitative, as opposed to qualitative, differences in crucial cellular functions. That presents a real challenge in developing treatments with a high ratio between efficacy and toxicity.