Principles of medical oncology
William N. Hait, MD, PhD James F. Holland, MD, ScD (hc) Emil Frei III MD (deceased) Donald W. Kufe, MD Robert C. Bast Jr. MD Waun Ki Hong, MD, DMSc (Hon)
Overview
The practice of medical oncology is evolving more rapidly than any field in medicine. The uncovering of the molecular basis of the numerous malignancies and their subtypes created a seemingly bewildering collection of new information that confronts the practicing physician. Since the principles that underlie the practice of this subspecialty, by definition, should remain constant, they can provide the scaffolding on which new information can be attached. To that end, this chapter attempts to define those immutable areas that apply to the care of the oncology patient, and through their understanding the massive amount of new information presented in this text can be organized.
Medical oncology evolved from a subspecialty of internal medicine whose hallmark was the ability to diagnose cancer and to safely administer dangerous drugs to the branch of medicine most directly linked to modern molecular biology; today, it heralds in the era of personalized medicine. A medical oncologist understands the biologic basis of malignant transformation and applies this knowledge to the prevention, early detection, and treatment of cancer patients. An individual with cancer should be viewed in the context of the etiology, pathogenesis, pathology, genetics, immunology, and biochemistry of the neoplastic process and humanistically as a person struggling with a terrifying disease.
Training of medical oncologists originated in cancer institutes, divisions of hematology, and departments of pharmacology. The American Board of Internal Medicine established the subspecialty as a separate discipline in 1971.1 Medical oncologists and hematologists have overlapping interests in neoplastic diseases of the hematopoietic tissues and share historic interests in training subspecialists in both fields. However, because each discipline became more complex, individual training programs have become the preferred approach.
In this chapter, we introduce the subspecialty by attempting to define a set of principles that underlie the practice of medical oncology.
Principles
Certain principles underlie the practice of medical oncology and are listed in Table 1. These tenets, although not mathematically derived or subject to rigorous validation, should nonetheless serve the purpose of providing both the uninitiated young practitioner and the seasoned veteran with a perspective gleaned from the several vantage points of the authors.
The treatment of cancer is multidisciplinary, requiring consultation with knowledgeable colleagues in related subspecialties
The practice of medical oncology is dependent on highly productive interactions with cognate disciplines, particularly surgical oncology, radiation oncology, urology, orthopedic surgery, radiology, and pathology. Multiple other interactions occur with nursing oncology, psycho-oncology, neuro-oncology, gynecologic oncology, rehabilitation medicine, and, for young patients, pediatric oncology. Infectious diseases and more recently autoimmune diseases are common complications of cancers and their treatments, forming a natural alliance with specialists in infectious diseases and rheumatology. Today, with the gratifying number of long-term survivors, the medical oncologist must work closely with primary care physicians for follow-up surveillance and psychologists and psychiatrists, who may be required to address the existential complexities of survivorship and the sequelae of treatment.
The medical oncologist is often involved in the final decisions concerning management and is frequently the final common pathway through which decisions are implemented. The timing of surgery and radiotherapy, the decision whether to take a curative or a palliative approach, and the decision whether watchful waiting is the appropriate course, or whether vigorous action is necessary are often entrusted to the medical oncologist.
Most patients have a relationship with other physicians before being diagnosed with cancer; a family physician or internist who referred the patient to a medical oncologist and other specialists may already have been involved with the patient before the recognition of a neoplastic disease. Medical oncologists must recognize their interest and continuing role in the management of patients with multisystem disease and communicate with them effectively. In the absence of such consultants, however, the medical oncologist must also attend to all aspects of internal medicine. This book contains detailed descriptions of various diseases; the modalities used in their treatment; the pharmacologic, immunologic, neurologic, psychological, biochemical, epidemiologic, and the modern emphasis on molecular biologic aspects of cancers; and the complications that cancers cause. Oncologic emergencies, rehabilitation, and the oncologist’s relationship to medical informatics and to government are also presented. Familiarity with these topics constitutes a foundation for medical oncology from which the principles derive.
Table 1 Principles of medical oncology
The treatment of cancer is multidisciplinary, requiring consultation with knowledgeable colleagues in related subspecialties |
The suspicion of cancer is based on clinical acumen and the diagnosis on examination of tissue |
Prevention is more effective than treatment |
The medical treatment of cancer patients is based on a clear understanding of the mechanism of drug action, potential for harmful side effects, mechanisms of drug resistance, and the principles of therapeutics |
Early stage cancers are more curable than late-stage cancers; the first treatment is more effective than the next |
The best treatment is often found through participation in clinical trials |
Cancer surveillance must be based on validated assumptions |
Oncologic care is for life |
The suspicion of cancer is based on clinical acumen; the diagnosis on examination of tissue
Oncologists must be highly competent internists to diagnose, exclude, and treat cancer. Many diseases can mimic the signs and symptoms of cancer, but the medical oncologist must not miss a nonmalignant cause. Conversely, the internist and the medical oncologist must remember that cancer, like syphilis in the past, is the “great masquerader” and, thus, must be considered in every differential diagnosis.2 A medical oncologist must understand the pathophysiology of cancer, the genetic predispositions, the existence of molecular subtypes, and the basic pharmacology that is the bedrock of effective cancer treatment. All cancers are not the same, and all patients with a given type of cancer do not behave in a similar manner. Increasingly, cancers are being segmented into a variety of distinct molecular entities, each with subtle and not so subtle differences in prognosis and treatment. As the more precise classification of leukemias and lymphomas led to improvements in treatment in the past, the use of powerful diagnostic tests is identifying subsets of solid tumors that respond differently to treatment are leading to improvements in treatment today.
When new syndromes appear in a patient who once had cancer, such as pulmonary insufficiency, meningoencephalopathy, or inexplicable pain, it is indispensable to establish by objective criteria that cancer is the proximate cause. Cancer patients are not protected from other symptomatic noncancerous diseases, such as pulmonary fibrosis and central nervous system disorders, or painful conditions, such as a herniated disk. No symptom should be attributed to cancer without persuasive evidence. Yet cancer must be suspected every time.
To ascribe a finding to cancer requires proof. Modern medical oncologists will need to both understand the histologic appearance of malignant and premalignant lesions and be able to understand the interpretation of complex profiles of nucleic acid sequences (DNA and RNA), gene transcription and protein translation, as well as chromosomal abnormalities to properly diagnose and manage their patients. These advances will lead to an even greater number of individuals who know that they are at risk of developing cancer, living with a diagnosis of cancer, or putatively cured of cancer but are concerned about recurrence due to the presence of minimal residual disease. It is an exceptional case when a medical oncologist can consider treatment without a firm diagnosis. Certain oncologic emergencies, such as spinal cord compression or superior vena cava syndrome, were once considered exceptions, but with modern imaging and biopsy techniques, a tissue diagnosis can usually be made rapidly and safely. Cytologic diagnoses may provide sufficient information in the presence of unambiguous clinical syndromes, but cytology of the bronchus, stomach, cervix, and body fluids has produced sufficient numbers of false-positive identifications to show that corroborating evidence is essential. The use of circulating tumor cells and free tumor DNA can aid in diagnosis and monitoring of patients, but as yet these surrogates have not been fully validated. Highly specific biochemical and molecular markers may be helpful when clinical and radiologic findings are characteristic and when a major intracavitary (cranial, thoracic, or abdominal) biopsy procedure would constitute an ominously serious event for a particular patient. It is always preferable to have histologic evidence whenever possible.
The management of cancer patients requires an understanding of the genetics, biology, and pharmacology of neoplasia and the psychosocial impact of a cancer predisposition or diagnosis on patient and physician
The practice of medical oncology requires wisdom, sensitivity, and resourcefulness. For the patient in whom relatively asymptomatic findings lead to a diagnosis of cancer, it is useful to consider that the day before the discovery, the patient was living with cancer. It is a source of some encouragement to patients to know that a diagnosis of cancer does not lead immediately or inevitably to death. The medical oncologist may be able to stress the long-term evolution of a cancer, the several stages that intervene between the carcinogenic stimulus, genetic mutations, selection of cells with a survival advantage, and the appearance of an autonomous neoplasm. Because this process usually takes years and often decades, it is of value to place the neoplastic process in perspective (vide infra). This is particularly important when a patient is being urged to make an immediate decision regarding surgery, radiation, or chemotherapy. When one considers that, on average, it will take more than 5 years for a malignant cell to undergo the requisite 30 doublings to achieve a 1 cm mass (one billion cells) required for most diagnoses, an additional week or two to gather information and opinions is unlikely to have a negative impact on the ultimate course of the disease. In contrast, the situation with certain malignancies such as acute myelogenous leukemia requires immediate action as the diagnosis is made when the near-fatal 1012 malignant cells exist.
Increasingly, individuals are recognized to harbor genetic predispositions to cancer owing to subtle genomic and epigenomic alterations. The oncologist must combine an understanding of the impact of these changes on the probability of being diagnosed with cancer and weigh the risks and benefits of medical intervention while being sensitive to the emotional impact that this knowledge has on individuals and their families. In some, cancer or the prevention of cancer may become a chronic process requiring ongoing treatment analogous to the management of diabetes or hypertension. Similarly, oncologists must recognize environmental exposures that increase the risk of cancer and attempt intervention to mitigate that risk.
The medical oncologist must distinguish between a neoplasm in which a chance for cure exists versus one that is currently incurable (precurable). Most cancers are curable if detected early, and most cancers are incurable if detected late. Whereas the former is likely to be driven by few genomic alterations and confined to one or two critical pathways, the latter is likely to harbor multiple changes in several pathways that make successful treatment less likely. The medical oncologist must therefore appreciate the importance of prevention, screening, and early detection and must actively educate colleagues and patients.
Advances in our understanding of cancer biology encourage the belief that one day all cancers will be prevented or cured. It is axiomatic that the day before the first metastatic choriocarcinoma was cured with high-dose methotrexate,3, 4 metastatic cancer, in general, was considered incurable by most observers. Similar considerations apply to every neoplastic disease that is now curable (Table 2). Other neoplasms are currently not cured by medicines alone but require participation of surgery or radiotherapy as an intrinsic part of the therapeutic process (Table 3).
Table 2 Cancers curable with chemotherapy
Choriocarcinoma |
Acute lymphocytic leukemia of childhood |
Burkitt’s lymphoma |
Hodgkin’s disease |
Acute promyelocytic leukemia |
Large follicular center cell lymphoma |
Embryonal carcinoma of testis |
Hairy cell leukemia |
Table 3 Cancers subcurable with chemotherapya
With regional therapy |
Wilms’ tumor |
Osteosarcoma |
Ewing’s sarcoma |
Embryonal rhabdomyosarcoma |
Adenocarcinoma of breast |
Small cell carcinoma of lung |
Squamous cell carcinoma of upper aerodigestive tract1 |
Adenocarcinoma of ovary |
Thymoma without regional therapy |
Acute lymphocytic leukemia of adulthood |
Acute myeloid leukemia |
Lymphomas, some subsets |
a By definition, <50% curable with chemotherapy alone; cure rates obtained with chemotherapy plus regional therapy are significantly superior to those with regional therapy alone (i.e., chemotherapeutic cure of micrometastatic disease only).
b Cure rates <50% in most series.
Patients are often influenced by their present state of subjective well-being. It is the responsibility of an oncologist to recognize the often pernicious behavior of cancer in its potential for recurrence and metastasis. In this context, the medical oncologist must interact directly with the patient and family, as well as with the medical record, films, slides, and other critical raw data. It is important to understand how patients make choices so as to present information in a way that does not bias an eventual decision. In addition, there is an opportunity to alter behaviors of patients and their families at the time of a cancer evaluation.5 For example, following a “scare” that a lung nodule could have been malignant may be the best time to treat nicotine addiction in the patient and in members of the family. The diagnosis of cancer constitutes a serious emotional burden that may distort ordinary reason. By firsthand intimacy with the diagnosis, the extent of the disease, and the patient’s attitudes and infirmities, the medical oncologist can make rational recommendations to the patient and to the other physicians involved.
Explanations of disease, anticipated therapies, randomized protocols, and unknowns must be tailored to the intellectual and emotional levels of the particular patient. It is never permissible to lie, but it may be prudent not to deposit all of the truth at once on a patient who cannot accept the full details and ramifications of diagnosis and management. “Your patient has no more right to all the truth you know than to all the medicine in your saddle-bags” was a humane and ethical tenet advanced by Oliver Wendell Holmes more than a century ago.6 It is dishonest to twist facts or to deny specific features, such as the existence of metastases. It is also wrong to deny a patient an opportunity to make final dispositions with respect to self, family, religion, the law, and business by falsely stating that a disease is benign or cured. Families who assert that the patient must not know because the patient could not stand it are usually twice wrong: the patient often knows already or may be more distraught by being excluded from knowing, and the patient ordinarily incorporates the information into his or her life equation indistinguishably from other patients. A reading of Tolstoy’s masterful The Death of Ivan Ilyich should convince any doubting oncologist about the terror of uncertainty and the value of direct and honest, yet humane, interactions with the patient. When a patient asks, “There is hope, isn’t there?” the oncologist can always be positive. Hope is a uniquely human characteristic that sustains the will to continue, and all oncologists and all patients do hope for a better outcome.
The treatment of patients with life-threatening disease can take its toll on providers of care. A sense of frustration can affect anyone who encounters barriers to successful completion of an important task. This is particularly true of intellectual tasks and invisible barriers. When the barrier is a lethal disease about which the oncologist can do little that is effective, the frustration can be all consuming. Oncologists who encounter several instances of recrudescent or refractory disease in a short time (especially if punctuated by the deaths of young or favorite patients, uninterrupted by counterbalancing compensatory successes) may well experience frustration, a sense of inadequacy, and depression. Frequent repetition of this cyclic phenomenon may lead to burnout.
The medical oncologist knows that many of today’s cancers are not yet curable. To the extent that the medical oncologist can be involved, actually and conceptually, in the solution to these complex mysteries, the frustration is lessened. Cancer research, whether at the basic or the clinical level, is held in high esteem by our fellow citizens. Group identity, “being one of the team,” helps offset the self-deprecation when human tragedies mount despite one’s best efforts. The camaraderie of other oncologists helps because they battle the same enemy with the same primitive weapons. Another oncologist can understand the trauma and the distress; it is an encounter on familiar terrain.
The appreciation that the horizon is distant, and that oncologists are all working intently to see beyond it, puts present frustration in a more appropriate perspective. Involvement in the systematized academic pursuit, whether in an academic setting, a medical school outreach, an oncology society, or a local collaborative group, provides the security of collegial support. A sound mind in a sound body implies rest, exercise, nutrition, and enjoyment. To ensure the last, the first three are prerequisites. Avocation and vacation are a portion of good mental health, included in the terms rest and exercise.
Finally, all oncologists can benefit from the vast array of continuing medical education programs given at academic medical centers, on the World Wide Web, at national meetings such as the American Association for Cancer Research and the American Society of Clinical Oncology, the American Society of Hematology, and through textbooks such as Cancer Medicine.
Prevention is more effective than treatment
The seminal work of Dr. Bert Vogelstein of John Hopkins Medical School demonstrated that as a normal epithelium transforms through dysplasia, anaplasia, and, eventually, neoplasia, that there is a progressive increase in genomic alterations, a process that may take many years.7 Similarly, as chronic myelogenous leukemia progresses from the chronic phase, through the accelerated phase to blast crisis, more chromosomal and genetic abnormalities are acquired. Exposure to environmental carcinogens, most commonly through tobacco use owing to nicotine addiction, accelerates the oncogenic process most prominently in the bronchial mucosa. These nucleic acid changes can activate complex signaling pathways that favor cell viability and inactivate pathways that normally balance uncontrolled growth through mechanisms of cell death. Not surprisingly, many of these molecular changes impart resistance to cancer treatment. It follows that the earlier a cancer is diagnosed and treated, the more effective the treatment. Furthermore, the oncogenic process provides substantial time for interventions that could prevent the formation of cancer in susceptible individuals, so-called disease interception.
Inhibition of cyclooxygenases within the intestinal epithelium with nonsteroidal anti-inflammatory drugs prevented many cases of colorectal cancer.8, 9 Hepatitis B vaccines dramatically reduced the incidence of hepatoma, once the Eastern world’s most common malignancy, by preventing the chronic inflammation and cellular damage produced by chronic hepatitis.10 New curative treatments for hepatitis C should also decrease the burden of hepatocellular carcinoma.11 Retinoic acid can decrease the appearance of second malignancies in heavy smokers12; tamoxifen and raloxifene can prevent breast cancer in patients at substantial risk.13 Minor surgical interventions, such as colonic polypectomy14 and loop electrosurgical procedure for carcinoma in situ of the cervix,15 can prevent colon and cervical cancer, respectively.
Increasingly, the medical oncologist will be called on to decide whether medical or surgical prevention is indicated in individuals without cancer but for whom cancer is a significant risk. This will ultimately extend to subtle genetic abnormalities, such as single nucleotide polymorphisms and circulating biomarkers. In addition, the use of preventive drugs will require long-term monitoring if used each day for years. Currently, the medical oncologist is asked to decide who should receive tamoxifen for prevention of breast cancer, retinoids for prevention of head and neck cancer, and oral contraceptives for prevention of ovarian cancer. Medical oncologists frequently advise patients regarding the benefits of prophylactic oophorectomy and mastectomy in patients with BRCA1 and BRCA2 mutations. Similarly, oncologists must be aware that the hepatitis B vaccine will prevent hepatoma, and a vaccine against human papillomavirus will prevent many cases of cancer of the uterine cervix.16 The oncologist must take a leading role in treating addiction to nicotine in all of its manifestations and must educate physicians and the public on the importance of early detection through appropriate screening.
Medical oncologists should counsel patients and families about good nutrition and healthy sexual practices as well as screening tests available for some cancer types. Several chapters of this treatise deal with prevention and early detection, and numerous publications that deal with these topics are available for distribution to patients and families from the National Cancer Institute (NCI) and the American Cancer Society (ACS). The NCI Cancer Information Service (800-4-CANCER) and the ACS National Cancer Information Center (800-ACS-2345) will send available publications free of charge and publish this information on their web sites.
Cancer therapeutics is evolving
The origins of cancer chemotherapy began with the observation that mustard derivatives such as mechlorethamine could treat patients with lymphoma, evolved through targeting biochemical pathways involved in DNA synthesis (antimetabolites) and microtubule function (vinca alkaloids and taxanes), through combination chemotherapy to the era of targeted therapies. These new medications are likely to be less toxic than traditional cancer chemotherapy drugs that most commonly target DNA or microtubules. The proper use of the new-targeted therapies will require an understanding of the molecular pathways responsible for malignancy in general and possibly in each individual’s tumor. Recent examples include imatinib to inhibit the transforming oncogenic protein in chronic-phase chronic myelogenous leukemia (Bcr-Abl)17; gefitinib to effectively treat patients with nonsmall cell lung cancer by targeting those individuals whose tumors harbor activating mutations in the epidermal growth factor receptor (EGFR)18; erlotinib to extend survival in patients with nonsmall cell lung cancer19; trastuzumab to treat metastatic breast cancer in patients whose cancers have amplification of the HER2/neu oncogene20; cetuximab to treat colorectal cancers by targeting the EGFR21; bevacizumab to treat colorectal cancer by targeting the vascular endothelial growth factor receptor and blocking angiogenesis,22 and venurafinib to inhibit mutant bRAF in patients with melanoma.23
Today, we are in the modern era of immuno-oncology, which began with the pioneering work of Dr. Steve Rosenberg at the NCI who demonstrated that certain T-cell subsets could be harnessed to treat refractory melanoma and renal cell carcinoma.24 Today, with the advent of immune check point inhibitors (e.g., ipilimumab), the medical oncologist is faced with a new set of therapeutic challenges based on autoimmune side effects that require additional skills to manage.
In addition, one must appreciate the importance of drug resistance and recognize it when it occurs. When choosing treatment for recurrent disease, one should consider the likely cross-resistance to several natural products owing to the multidrug resistance phenotype mediated by the adenosine triphosphate–binding cassette family transporters [(e.g., P-glycoprotein, MRPs (multidrug resistance proteins)], the low probability of response to drugs of the same drug class (e.g., taxanes), or sequential hormonal therapies when selecting drugs with the same mechanism of action (e.g., aromatase inhibitors). Furthermore, the medical oncologist be familiar with mutations within drug targets that lead to either sensitivity or resistance to a given class of drug, for example, EGFR inhibitors, bRAF inhibitors, abl inhibitors, and inhibitors of BTK (Bruton’s tyrosine kinase) to name a few.
Treatment of cancer patients includes the use of drugs for host support. The effects of the tumor and its products on the structure and function of the patient’s normal tissues, as well as the mind and emotions, define an understanding of the disease process and of the patient in whom it takes place. It is not sufficient to order a therapy with the appropriate dose and schedule. There must be a broad understanding of and attention to potential toxicities, which represent the drug’s effects on normal tissues. A medical oncologist should understand the interaction of the administered drug with target molecules present in normal tissues and the potential for drug–drug interactions to avoid unnecessary toxicities.
The availability of effective antibiotics and the use of platelet transfusions were intrinsic to early cures of the acute leukemias. Colony-stimulating factors (filgrastim and sargramostim) have significantly altered the prospect of drug-induced granulocytopenia, and recombinant erythropoietin can diminish drug-induced anemia. Means to ameliorate thrombocytopenia, other than with platelet transfusions, are now available.
The use of cytokines to collect circulating hematopoietic progenitor (CD34) cells enabled the convenient collection of marrow-repopulating precursors, allowing autologous stem cell transfusions to substitute for autologous marrow transfusion. New antibiotics make granulocytopenia less ominous, and oral prophylaxis with antibiotics and antifungal agents has decreased hospital admissions. These assets allow chemotherapy to be given more safely at the intended dose and schedule without delay or dose reduction.
The availability of highly effective antiemetics makes cancer chemotherapy less dreaded than in the past. The emergence of psycho-oncology as a widely appreciated discipline has also made it possible for patients to strengthen their resolve to undertake approaches aimed at cure or to accept the unlikelihood of cure with greater serenity.
Intravenous medications that may be toxic to the venous wall, and vesicant if extravasated, require the use of central venous access. When venipuncture is difficult because of anatomy or obesity, repeated needle sticks and much time are wasted in attempting peripheral cannulation. Needle phobia is a perverse part of being under treatment; establishing permanent venous access can largely obviate it.
Most patients with cancer are over the age of 60 and are receiving treatment for comorbid conditions. Therefore, care must be taken when prescribing treatments that could have dire drug interactions, leading to untoward side effects. No modern physician should be without ready access to the Internet for a compendium of drug interactions, drug descriptions, and appropriate methods of monitoring for efficacy and toxicity. A medical oncologist benefits from using an electronic medical record system that saves time and prevents errors through online calculation of dosing and electronic ordering. These systems also allow the creation of databases that can be queried for identification of trends and unexpected results.
Therapies that are totally appropriate for someone whose disease might be cured by judicious application of surgery, radiotherapy, and/or chemotherapy might be totally inappropriate when applied to another person with widely metastatic disease for whom no known cure exists. Therapy with curative intent may be toxic but of relatively short duration. On the one hand, conservatism aims at saving a life, not avoiding toxicity; on the other hand, treatment for palliative purposes would not ordinarily condone similar risks and iatrogenic effects that diminish the quality of life, even temporarily. The same is true for therapies aimed at cancer prevention because these may be taken by asymptomatic individuals for prolonged periods of time. The recent experiences with high-dose cyclooxygenase II inhibitors over long periods, which caused increased cardiovascular side effects in patients participating in one of two cancer prevention studies, resulted in serious consequences for the pharmaceutical manufacturers.25
Certain principles govern the application of therapies, no matter what the disease. These were enunciated more than a half-century ago by Robert F. Loeb, Bard Professor of Medicine at Columbia University’s College of Physicians and Surgeons. These simple rules have profundity and nearly universal applicability but must be tempered, however, by an understanding of the neoplastic process.
The first law is if what you are doing is “doing good,” keep doing it. It is implicit that a physician measures the effects of any intervention on both the tumor and the host. However, the lessons learned from the treatment of acute lymphocytic leukemia of childhood are noteworthy. For example, vincristine plus prednisone is an excellent induction treatment, but in 1968, a question was raised: why not keep administering this highly active induction regimen rather than shifting to antimetabolite management? A cohort of children who were induced into remission by vincristine and prednisone were randomized to continue the induction treatment or shift to the antimetabolite. They rapidly became resistant and relapsed, whereas those who were shifted to the antimetabolite experienced long-term sustained remissions and cures (Cancer and Leukemia Group B, unpublished data). Thus, the first law of therapeutics does not always apply to cancer for which sequential treatment regimens may have special importance. Much of curative oncology relates to the biology of the unseen tumor (minimal residual disease), for which the current clinical status may not be informative. The first law seems more applicable to clinically recognizable disease.
The second law of therapeutics states that if what you are doing is not “doing good,” stop doing it. Most therapeutic regimens have little chance of success if after 8 weeks of treatment they have failed to elicit therapeutic benefit. It is, nonetheless, advantageous to undertake a second month of treatment in most instances because a well-documented early increase in tumor diameter on radiographic examinations or increased pain can, indeed, be followed by tumor regression, notably for certain forms of hormonal therapy as well as immunotherapy. Before stopping treatment, corroborating information should be sought by direct measurements, biomarkers that include circulating tumor cells and circulating tumor DNA may aid in the decision to stop an ineffective treatment. Increased bony uptake of radionuclides can be a sign of bone healing, even of a previously unsuspected lesion, and is not a suitable end point. The appearance of a new metastatic deposit or the continued growth of a previously documented tumor despite treatment speaks against continuing that regimen.
Hippocrates’ admonition, Primum non nocere (first do no harm), is also subject to careful cautious consideration in oncology.26 Thus, the second law of therapeutics does not extend to toxic effects unless they are life-threatening or profoundly disabling. With many of the therapeutic agents available today, complete avoidance of toxicity would doom many patients to death from their neoplasm. Some patients can obtain cure, and more can achieve meaningful remission by accepting the transient effect of intensive therapy that kills tumor cells and normal cells alike. The patient almost always recovers, but the less resilient tumor may not. To treat a population of patients at so low a dose that it would avoid toxic harm to any patient would surely exact a higher price in depriving others of adequate doses to achieve maximum benefit. Curative and subcurative cancer chemotherapy as well as newer immunotherapies, as we know it today, are often toxic but rarely fatal. Attempts to abrogate toxicity for all by reducing the dose of an established regimen might compromise benefit for the majority.27, 28 Dose adjustment for an individual may be necessary and prudent but must always be considered with respect to other means of mitigating toxicity without dose reduction.
The third law of therapeutics—“if you do not know what to do, do nothing”—counsels against uninformed action. In many circumstances, a rush to judgment or, worse, a rush to do something, anything, can be disastrous. Aside from oncologic emergencies and certain forms of leukemia, there is rarely an occasion when observing the evolution of symptoms and findings or seeking consultation with another individual for a fresh viewpoint is contraindicated because of time pressure. In the presence of pain, one should not delay pain relief, but other therapy may be delayed to build an informed formulation. In the presence of a differential diagnosis that includes diseases other than cancer, particularly infections, one must be certain that delay does not risk mortality or morbidity from the other possible disorders. Thus, the time invested for observation and consultation should not be extravagant. In circumstances where the benefits of existing treatment are poor or unknown, clinical trials should be a first consideration.
The fourth law of therapeutics is “never make the treatment worse than the disease.” This relates to total life equation: the price the oncologist knows the patient may be obliged to pay in present side effects to attain future real effects. Often, the patient’s vision is foreshortened because today’s symptoms caused by drug toxicity can be more severe than the original complaints related to the cancer. The medical oncologist must ascertain the patient’s attitude toward quality of life versus duration of life. It is a medical oncologist’s responsibility to counsel the patient concerning this weighty topic. With the rapid appearance of new oncology products and numerous new types of treatments available through clinical trials, it is becoming increasingly difficult to distinguish therapy with curative intent from a palliative orientation, except in the extreme situations of newly diagnosed, low-stage and grade disease versus palliative care for the terminally ill. In all cases, the proper goal is maximal life at maximal quality. For some patients, the toxic effects of treatment outweigh the value of possible extension of life. This perception is often related directly to age. Pain and disability from cancer may temper the desirability of certain therapies, which offer only temporary and partial relief. It is not a kindness to defer death only transiently by rescuing a dying patient back to a raft of suffering. Heroic efforts are justified only when a meaningful therapeutic option exists.
It is inappropriate for the medical oncologist to substitute professional judgment for a patient’s ardent wishes when the patient strives to accomplish something that is a reasonable therapeutic goal. The medical oncologist must serve as a bastion of reality, however, advising the patient of what is possible and of what is likely. In the course of doing this, the laws of therapeutics and of humanity always include hope.
Finally, the medical oncologist must remember that patients do not fail treatments; rather, treatments fail patients. The careless use of the former, inadvertently implies a disrespect for the patient and a lack understanding of the extremely fragile state of patients whose disease is unresponsive to treatment.
Early stage cancers are more curable than late-stage cancers; the first treatment is more effective than the next
As no method of prevention is likely to be completely effective, medical oncologists must recommend appropriate tests to detect cancer at its earliest possible stage. They include mammography, colonoscopy, occult blood in the feces, digital rectal examination of the prostate, Papanicolaou smears, and examination of the skin and body orifices for signs of premalignancy, such as leukoplakia and dysplastic nevi. Soon, a variety of validated biomarkers will likely be available to identify early cancer or the presence of premalignancies, for example, ductal carcinoma in situ, prostatic intraepithelial neoplasia, colonic polyps, and cervical dysplasia. As described earlier, progression of cancer leads to the acquisition of changes in the host and the cancer genome that select for cell survival over cell death. Many of these changes allow tumor cells to exist in harsh environments characterized by hypoxia and a low pH. These same changes produce tumor cells that are increasingly resistant to cancer therapies.
Cancer cells that emerge following initial treatment often represent clones of cells with intrinsic resistance or cells in which the treatment has selected for drug resistance mechanisms. For example, treatment of patients with adenocarcinoma of the lung who progress after initial response to inhibitors of EGFR may acquire a mutation that renders these kinase inhibitors ineffective.29 This reality places a premium on getting it right the first time so that medical oncologists and their team must be certain that they have all of the requisite information at hand to allow the best choice for initial treatment. For example, treating a breast cancer patient without knowledge of the status of estrogen and progesterone receptors and the HER2/neu oncogene correctly measured, or subtyping nonsmall cell lung cancers in terms of EGFR, alk, and other genomic alterations would likely compromise initial treatment. Similarly, embarking on the treatment of acute myelogenous leukemia without full genotyping and phenotyping risks making the wrong choice for initial treatment. Furthermore, resistance to imatinib is associated with both P-glycoprotein and specific mutations in the BCR : ABL gene. These changes are defined, and new drugs are available to overcome these forms of resistance. Finally, the proper approach to patients with lung cancer and melanoma requires molecular characterization to select the proper treatment.
The advantage of treating early stage patients has proved so persuasive that the profession and patients have accepted the technique of postsurgical or adjuvant chemotherapy, acknowledging that this entails treating some if not the majority of patients whose risk of relapse is already zero. Adjuvant therapy after surgery has been demonstrated to be curative in several diseases for which surgery alone has low cure rates and where chemotherapy alone cannot cure metastatic disease. Breast cancer, Wilms tumor, and osteosarcoma are the prime examples. In many diseases, there is evidence of prolonged disease-free survival and of longer survival, such as stage II and III breast cancer,30, 31 stage III ovarian cancer,32 and stage III colon cancer.33 Recent evidence suggests that transcriptional profiling of breast cancer specimens may help identify patients for whom adjuvant therapy is unnecessary.34
Because adjuvant treatment is aimed at micrometastatic disease remote from the primary tumor, exploration of chemotherapy before surgery (neoadjuvant) has been undertaken in a few types of cancer. In addition to earlier exposure of the micrometastases, this approach has two other beneficial characteristics. First, regression of the primary lesion predicts that the micrometastases will also likely be sensitive.35 Second, regression of the primary tumor may make primary surgery unnecessary or less debilitating or disfiguring, allowing curative radiotherapy, as in some head and neck cancers and as shown in a large series of patients with breast cancer.36 In other instances, surgery after chemotherapy may be technically easier, although not always less radical, because there is no certainty that every cell has been eradicated at the original boundaries. Induction or concurrent chemotherapy may also significantly enhance the effectiveness of radiotherapy for other tumors, such as anal carcinoma, thereby decreasing the need for surgery.
In the past, treatment of recurrent disease was too often empiric. The response to these subsequent treatments predictably diminishes over time. For example, a breast cancer patient whose tumor strongly expresses the estrogen and progesterone receptor has more than a 50% chance of responding to first-line hormonal therapy, with a likely duration of response of approximately 18 months.37 On relapse, that same patient’s probability of response to second-line therapy is cut in half, as is the likely duration of response.38 Today, the use of dynamic monitoring of circulating tumor cells, circulating tumor DNA, and re-biopsy with genomic analysis promises to help select the most appropriate treatment for the relapsed patient.
The best treatment is often found through participation in clinical trials
The most common and lethal cancers, including metastatic lung, colon, and breast, are often incurable by approved treatments. Furthermore, most approved treatments have significant risks of side effects and produce limited prolongation of life. Yet, under these circumstances, most patients receive standard rather than experimental therapies. For many, the best choice of treatment may be available only through clinical trials.
Patients are increasingly involved in decisions regarding the choice of therapy, having been empowered by information widely available through the Internet. Patients with cancer are often apprehensive that they may not receive the best treatment and will challenge us with appropriately tough questions before moving ahead. The medical oncologist can speak with greater authority when a deliberate comparison is being made because the goal of such studies is toward improvement on the standard, not toward finding treatments that are equally good.
A number of ethical issues are abrogated by the certainty that a specific patient’s disease is not currently curable. For metastatic disease for which no cure is known, it is not only ethical but also important that a systematically designed investigation of new treatments through participation in clinical trials be undertaken early in the course of the patient’s disease. This allows assessment of a drug’s activity before toxicity arises from conventional therapies that might limit dosing. Conventional therapies might also elicit resistance of one or another kind or immune system depression, which might foreclose the opportunity to recognize the activity of the candidate compound.
For diseases with an especially unfavorable outlook and rare therapeutic success, delays in introducing candidate compounds to ensure that they carry little or no risk of toxicity are an unwise investment of resources and time, let alone the patient’s short-lived opportunity for possible benefit. The outcome of unsuccessfully treated cancer is more ominous than the hazards of clinical investigation.
Patients should be made aware of clinical trials as part of the initial discussions regarding treatment. Too often patients are not aware of clinical trials and discover this option through friends or the Internet rather than through their physicians. A trial of candidate phase II agents prior to conventional chemotherapy for metastatic breast cancer has been conducted without significant compromise in response to the established regimen.39 Today, therapies aimed at validated molecular targets may enter the clinic once proof-of-principle is obtained in cell lines and safety and efficacy are confirmed in animal models. Major differences in pharmacokinetics exist between mouse and man, and the ultimate test remains the carefully designed, meticulously conducted clinical trial. The testing of new, targeted therapies is likely to change how we view early trials of anticancer drugs. For example, the phase I trial of imatinib did not reach a maximum tolerated dose. Rather, the dose was determined based on “maximum biological dose,” that is, the dose in which the target enzyme was maximally inhibited.40 Similarly, dynamic imaging techniques, such as positron emission tomography, predict the responsiveness to imatinib in patients with gastrointestinal stromal tumors well before changes occur in tumor size.41 Finally, several newer agents are showing promising activity in very early phase studies, further compelling patients and their physicians to participate in properly designed studies.
Therefore, until the day arrives when all cancers are preventable or curable, enrollment of patients in clinical trials will be the hallmark of the practice of the best medical oncologists. No cancer is so well treated that an improvement in outcome or therapeutic approach cannot readily be imagined. Thus, research is imperative.
An individual in practice cannot devote the same time and energy to clinical research as one who serves full time on the faculty of a university, research institute, or hospital. However, the private practitioner has an opportunity to participate in clinical trials through cooperative groups or in collaboration with cancer centers; this opportunity should not be wasted. Every oncologist should participate in clinical research during his or her training, and the oncologist’s office should include the capacity to perform clinical research. It is the responsibility of the medical oncologist to discuss clinical trials with patients in a way that does not bias an ultimate decision. There is much reason to anticipate that progress would be more rapid if clinical research were accepted as an integral part of the practice of medical oncology, as it is in pediatric oncology. The technology exists in medical informatics for community oncologists to ally themselves with their alma mater or other academic centers to participate in diagnostic, preventive, and therapeutic research trials using the computer, electronic mail, and fax as expedient tools. Those oncologists who have so heavy a workload that it prevents their devoting the necessary time to participate in clinical research risk depriving their patients of access to research advances. Clinical investigation should serve as the bridge to fundamental science and the excitement generated by the new molecular biologic understanding of cancer as a malignant tissue.42
Cancer surveillance must be based on validated assumptions
Following completion of curative therapy and attainment of a complete remission or following adjuvant therapy, surveillance for early signs of recurrence is based on the logical assumption that the earlier a recurrence is detected, the better the outcome of a therapeutic intervention. Although true in principle, in practice, this is rarely the case. There are several reasons for this illogical result, including the possibilities that surveillance tests are insensitive or nonspecific or that further treatment options are ineffective. Over testing of patients at low risk of recurrence increases the probability of a false-positive result and the attendant morbidity, psychological or physical, associated with attempting to make a definitive diagnosis. In contrast, under testing in patients at risk of recurrence of a form of cancer for which effective salvage therapy is available (e.g., germ cell tumors, large cell lymphoma, Hodgkin disease, osteosarcoma, breast cancer, prostate cancer) is inexcusable. Thus, the oncologist must be aware of the predictive value of surveillance tests (tumor markers and imaging studies) in the context of specific malignancies and apply them accordingly.
Standards for quantifying diagnostic tests used frequently in surveillance have been adopted in most parts of the world [Système Internationale (SI) units], except the United States. It is impossible to read an international medical journal without being thoroughly familiar with SI units. They are presented in Table 4 so that readers can have ready access to a source for translation from the old nomenclature, characteristically American, which pervades this treatise.
Table 4 Representative Système International units for laboratory tests of importance in oncology
Component | Present reference interval | Present unit | Conversion factor | Intervals | Unit symbols |
Albumin | 4.0–6.0 | g/dL | 10.0 | 40–60 | g/L |
α-Fetoprotein, radioimmunoassay | 0–20 | ng/mL | 1.00 | 0–20 | g/L |
Bilirubin | |||||
Total | 0.1–1.0 | mg/dL | 17.10 | 2–18 | µmol/L |
Conjugated | 0–0.2 | mg/dL | 17.10 | 0–4 | mol/L |
Calcium | 8.8–10.3 | mg/dL | 0.2495 | 2.20–2.58 | µmmol/L |
Cholesterol | <200+ | mg/dL | 0.02586 | <5.20 | µmmol/L |
Cortisol | 4–19 | g/dL | 27.59 | 110–520 | nmol/L |
Creatinine | 0.6–1.2 | mg/dL | 88.40 | 50–110 | mol/L |
Fibrinogen | 200–400 | mg/dL | 0.01 | 2.0–4.0 | g/L |
Glucose | 70–110 | mg/dL | 0.05551 | 3.9*6.1 | µmmol/L |
Hemoglobin | |||||
Male | 14.0–18.0 | g/dL | 10.0 | 140–180 | g/L |
Female | 11.5–15.5 | g/dL | 10.0 | 115–155 | g/L |
Immunoglobulins | |||||
IgG | 500–1200 | mg/dL | 0.01 | 5.00–12.00 | g/L |
IgA | 50–350 | mg/dL | 0.01 | 0.50–3.50 | g/L |
IgM | 30–230 | mg/dL | 0.01 | 0.30–2.30 | g/L |
IgD | <6 | mg/dL | 10 | <360 | mg/L |
IgE | 20–1000 | ng/mL | 1.00 | 20–1000 | g/L |
Iron | 80–180 | g/dL | 0.1791 | 14–32 | µmol/L |
Iron-binding capacity | 250–460 | g/dL | 0.1791 | 45–82 | µmol/L |
Lipoproteins | |||||
Low-density lipoproteins (LDL), as cholesterol | 50–190 | mg/dL | 0.02586 | 1.30–490 | mmol/L |
High-density lipoproteins (HDL), as cholesterol | 30–70 | mg/dL | 0.02586 | 0.80–1.80 | mmol/L |
Magnesium | 1.8–3.0 | mg/dL | 0.4114 | 0.80–1.20 | mmol/L |
1.6–2.4 | mEq/L | 0.500 | — | — | |
Metanephrines (as normetanephrine) | 0–2.0 | mg/24 h | 5.458 | 0–11.0 | µmol/day |
Osmolality | 280–300 | mOsm/kg | 1.00 | 280–300 | nmol/kg |
Phosphate (as inorganic P) | 2.5–5.0 | mg/dL | 0.3229 | 0.80–1.60 | mmol/L |
Potassium | 3.5–5.0 | mEq/L | 1.00 | 3.5–5.0 | mmol/L |
Protein, total | 6–8 | g/dL | 10.0 | 60–80 | g/L |
Serotonin | 8–21 | g/dL | 0.05675 | 0.45–1.20 | mol/L |
Thyroxine, free (T4) | 0.8–2.8 | ng/dL | 12.87 | 10–36 | pmol/L |
Triiodothyronine (T3) | 75–220 | ng/dL | 0.01536 | 1.2–3.4 | nmol/L |
Urate (as uric acid) | 2.0–6.0 | mg/dL | 59.48 | 120–360 | µmol/L |
Urea nitrogen | 8–18 | mg/dL | 0.3570 | 3.0–6.5 | mmol/L |
Vanillylmandelic acid | <6.8 | mg/24 h | 5.046 | <35 | µmol/day |
Oncologic care is for life
The relationship of a medical oncologist with patients is intimate and should not end once therapies aimed at controlling the spread of cancer are no longer effective. The medical oncologist must be skilled in the principles and practice of palliative care and collaborate actively with specialists in symptom control, for example, neurologists, psychiatrists, and hospice staff. No greater feeling of abandonment can occur than when a patient is abruptly released from care by an oncologist who fails to recognize this lifelong responsibility.
It is, however, also the responsibility of the medical oncologist to address end-of-life planning with the patient and the family. Advice regarding living wills, power of attorney, and resuscitation falls squarely within the purview of the medical oncologist. This responsibility is highlighted in states that require that do not resuscitate (DNR) orders be written on patient charts prior to death. In the absence of such orders, when a nurse finds a patient apparently dead, she must, by law, initiate emergency calls for resuscitative efforts.
When death comes from cancer as the expected final event of a gradual deterioration of vital forces, resuscitative efforts do not succeed. When we are unable to keep someone alive, the likelihood of bringing him or her back to meaningful life is infinitesimal. Resuscitative efforts should be applied to patients with cancer who were not expected to die because reversible phenomena, such as pulmonary emboli, cardiac arrhythmias, aspiration, and similar events, can provoke unexpected death in a patient with a neoplasm, just as in any other hospitalized or ambulatory patient. Many patients, particularly the elderly and those apprised of the progress of their disease, can discuss the decision not to resuscitate with equanimity and, indeed, with a certain personal satisfaction of avoiding the fruitless anguish that such a procedure entails for the surviving family. Most patients sign living wills or appoint a health care proxy if these possibilities are presented to them.
Because of the legal implications involved, where particular religious scruples obtain or where families have emotionally uncontrolled members who cannot accept the anticipated death of a loved one, the medical oncologist should spend considerable time planning for an eventual death. DNR forms are a technique of documentation and constitute further evidence that society has moved medicine onto a new plateau of accountability.
The medical oncologist should make known his or her intentions concerning the advisability of resuscitative efforts for each particular patient in advance to forestall unnecessary trauma to the patient, family, and staff; to forestall litigation; and to settle in advance any serious disagreements with the patient or family. An impasse might occasion a medical oncologist to find a suitable substitute physician if there is irresolvable conflict concerning the plans surrounding an anticipated death.
DNR orders do not imply that there be diminution of effort to control or palliate the disease before death. However, if good judgment indicates that continued efforts are fruitless and can only inflict suffering, with no prospect of benefit, discontinuation of active therapy should always be accompanied by DNR orders.
Summary
The medical oncologist stands at the crossroads of modern molecular biology and medical practice and often serves as the final common pathway for the application of cancer research to patients. A complex corpus of information is available that expands rapidly, both deeper into the nature of the cancer process and wider into new approaches that provide demonstrated effectiveness in therapy, prevention, or support.
The increasing appreciation that oncogenes and tumor suppressors act through usurpation of normal autocrine and paracrine signaling pathways and that cancer is in reality not merely a disease of cancer cells but of a cast of supporting characters that form a malignant tissue (blood vessels, fibroblasts, smooth muscle cells, macrophages, lymphocytes, etc.) provides a variety of new targets for therapy. The realization that malignant transformation is a multistep process of accumulation of genetic abnormalities over long periods of time gives impetus for the design of rational preventive strategies and guidance to target patients at highest risk. These pathways have already led to effective targeted therapies that are less toxic than traditional cancer chemotherapy. Thus, the tide of fundamental discovery is washing away many of the unknowns. It is axiomatic that certain cancers can be cured today without knowing the intimate nature of neoplasia. How better the day, perhaps soon upon us, when we are more confident that we know what we are doing!
Clinical accomplishments have similarly been exceptionally productive in the 60 years since the first cancer was cured with drugs.3 A large assortment of drugs has since been provided. A new array of genetically engineered drugs support host function, and others that are still early in their development are on the way. Imaging technologies will continue to revolutionize the ability to detect, stage, treat, and monitor cancers. Biochemical markers of tumor behavior will provide increasing diagnostic and monitoring capacity and may offer new targets for therapy.
There is probably no cancer in which some progress in diagnosis or therapy has not been achieved in the last decade. Similar achievements for cancer prevention are materializing. Oncologists must assume greater responsibility for health preservation. Much could be accomplished by applying what is already known about lifestyle, diet, and exercise. Medical facts without political action were slow to change the tax on health that tobacco levies. A concerted effort within most states has begun, but a federal role in managing the tobacco plague has been thwarted.
The horizon has never been closer. Although still distant, there are enough promising paths to follow that one of them may prove considerably faster than even reasonable optimism would suppose. The information that serves as our foundation, its rate of accrual, its revelations, and the demonstrated success of translating science to clinical applications augur well for the future of medical oncology and for cancer patients.