|
| |
Cancer: New Technologies, New Directions
From: Columbia University
| By:
Karen Antman |
EDITOR'S INTRODUCTION |
While scientists and physicians can now cure many of the acute infectious diseases that strike humans, chronic disease has proved far more difficult to treat. The mechanisms of cancer development, in particular, are varied; no single pathway, no single event reproducibly causes cancer,  so no one intervention provides effective prevention or treatment. We have adapted standard treatments--surgery, chemotherapy and radiation--to many of the common cancers with growing success. However, new treatments, from gene therapy to diagnostic gene chips, promise to target cancer cells much more specifically along their distinct pathways.Karen Antman, director of Columbia University's Herbert Irving Comprehensive Care Center, addresses the progress made in treating cancer to date and the new therapies that may help prevent or eradicate the disease. |
 ancer is certainly a major public-health problem. Cancer is currently the second-leading cause of death in the world. Seven million people develop cancer every year, and there are 4 million deaths. It is the major killer of American women aged 34 to 74 and is as lethal as heart disease in middle-aged men. |
People ask me, "When we are going to cure cancer?" as if it is a single disease, but they never ask, "When we are going to cure infectious diseases?" Just as we have many infectious diseases--tuberculosis, AIDS, E. coli--we have different kinds of cancers. Over the past 50 years, we have improved survival from one type of cancer at a time. Before we can explore the new treatments for cancer, we need to better define the current status of cancer therapy. |
A history of the development of effective <br>treatments for cancer
Cancer is not a new disease, and it's not an industrial disease. Non-industrial countries have relatively similar rates of cancer, but they have different cancers. The risk of esophageal and stomach cancer in Asia is much higher than in the United States, but Americans are more likely to get breast, prostate and colon cancers. |
Dinosaurs were affected by cancer--we can see evidence of metastatic cancer in their bones. Sharks get cancer. Cancer has been found in Egyptian mummies, and papyri from 1500 BCE mention non-healing ulcers that were almost certainly sun-related skin tumors. Hippocrates described stomach and uterus cancers, and attributed them to "an imbalance of the humors." |
In the Middle Ages, the belief was that cancer spread locally, so surgery and radiation began to develop. A mastectomy carving from the 1500s describes giving the patient wine, making incisions under the breast and cauterizing the wound. Surgery became much safer after antiseptics were introduced and much more humane when anesthesiology was developed. A microscope was essential for telling normal cells from cancerous ones. |
The first X-ray was taken by Roentgen in 1895, showing his wife's hand, and soon after he won the Nobel Prize in physics. X-ray was recognized immediately as a major achievement, and remissions from radiation of lymphomas, breast and other cancers were described within a year. The first caveat appeared a decade later, when radiation-associated cancers began to develop. From 1900 to 2000, increasing energy and better technology made radiation safer and more effective. Now we can deliver high doses deep into the tissue with excellent cosmetic results, thus avoiding amputations and disfiguring surgeries that would have been necessary to achieve the same local control. |
As local control of cancers improved with surgery and radiation, physicians were disappointed to find that tumors in many patients later spread to other organs, despite being controlled at the primary site. Thus, the tumor must have already spread to the other organs before their original surgery. Cancer investigators began to conceptualize cancer as a systemic disease, even at the time of diagnosis. Systemic treatment (drugs) would be needed to successfully treat occult systemic disease. |
Some anti-tumor drugs were developed along with sulfa antibiotics starting in the 1950s. As physicians learned to treat tuberculosis with several drugs at once, to keep the bacteria from developing resistance, anti-tumor drugs began to be given in combinations for the same theoretical reasons. (We see the same strategy more recently with AIDS treatments given in combinations.) Thus, drug combinations are essential in both infectious diseases and cancer. |
Surgery became less extensive as randomized trials showed similar survival rates for patients treated with extensive surgery alone compared with a combination of conservative surgery and radiation or drugs. Patients with cancer in the arms or legs could be treated with limb-sparing surgery and radiation instead of amputation, for example, and lumpectomy and radiation was as good as mastectomy for breast cancer. Surgery was refined and caused fewer side effects. |
Current understanding of cancer
The current model for cancer development is based on both genetics and environmental exposures. Cancer is now considered a genetic disease. Abnormal genes can be inherited, and normal genes can be damaged by radiation, chemicals or infections after birth. There are about 1015 cells in the body--that's one with 15 zeroes--and the DNA in one of these cells may become damaged. A single mutation may persist, which in and of itself rarely causes cancer. A variety of cell pathways detect and repair DNA damage. Damaged DNA, which is not repaired, may accumulate additional mutations. However, if second, third and fourth mutations occur, and if these mutations allow unregulated cell growth and/or extend cell life, cancer can result. |
Most Americans who develop cancer will do so in their fifties, sixties or seventies. Patients who inherited a mutation from a parent who carries a germ-line mutation, meaning that first mutation is in every cell of the body, often develop cancer decades earlier, as young adults or even children. |
Technically, "cure" means living until death from another cause without relapse. Waiting that long isn't very practical in evaluating new treatments, so doctors look instead at five-year survival, which is a very clean end point for comparing the effectiveness of different treatments. Most patients who have not had disease recur in the first five years are cured, although some cancers are exceptions. For example, the risk of a recurrence of breast cancer does not decrease to a normal baseline for 15 years after the original surgery. |
"Complete response" means having no tumor detectable with any of our best tests, but a complete response does not necessarily result in cure, because some patients who have a complete response can still relapse. However, if the tumor doesn't go away completely with treatment, a "partial response," the patient will not be cured. |
Getting cancer is much more common than dying from it--i.e., the incidence of cancer is higher than its mortality. About 59 percent of American patients with cancer are currently cured, so less than half of Americans who are diagnosed this year with a cancer will ultimately die from it. |
The risk of having had a cancer before the age of 40 is under 2 percent for both men and women. Between 40 and 60, the risk of ever having had cancer rises to about 10 percent. Between 60 and 79, the risk for men is almost 40 percent and for women it is 22 percent. Men at age 85 have a 48 percent risk, and a woman has 38 percent. It makes sense, as we get older, to screen for tumors for which we have an effective treatment. |
The four most common cancers
The four most common cancers for Americans--breast, prostate, lung and colorectal cancers--together account for more than 50 percent of the incidence of cancer and cancer deaths each year. (The next most common malignancy is lymphoma, followed by pancreatic cancer and leukemia.) Although a discussion of the treatment of each of these diseases would take more than the hour allotted, a discussion of what we know about who gets these cancers might lead to prevention strategies. |
Breast cancer
Breast cancer develops in one in eight women if she lives to age 85. About a third of those who develop breast cancer die of it. Breast cancer is exceedingly rare before age 30, and is by far the most common cancer in women until age 80, when colon cancer becomes more common. |
As published in the New York Post, the incidence of breast cancer in New York City varies by zip code, with the highest incidences on the Upper East Side of Manhattan. Although the Post quotes women who are "afraid to go outside because of the risk," the higher incidence of breast cancer in the Upper East Side is reasonably well understood. Breast cancer is more common in women of Northern European extraction, and in those in high socioeconomic status (SES) groups--in other words, the better you do financially, the higher your risk. The Upper East Side population is by and large white and well off. Latin areas in New York have a lower risk of breast cancer, while risk in Harlem is intermediate. |
The incidence of breast cancer is high in the whole Northeast US compared with the West and the South. Women in the Northeast tend to be well nourished; thus they tend to start menstruating at 13 or younger, and menopause comes later, over 50. They postpone having their first child until a later age, and they may not breast-feed. These factors account for about 70 percent of the increased risk. The other 30 percent we don't yet understand. Columbia University has had a leadership role in a large epidemiology study of breast cancer on Long Island. |
Internationally, breast cancer mortality is falling in Northern European countries and in Canada and the US. The US does not have the highest incidence of breast cancer--we rank 15th. The risk in Japan is about a fourth of the risk in the US. The lower risk cannot simply be genetics. First, the risk in Japan is now rapidly increasing. Second, when a Japanese family moves to Los Angeles, the risk increases to that of other Americans within one or two generations. |
Recently several genes have been associated with a high risk of breast and ovarian cancer. About half of women with abnormal BRCA1 or BRCA2 gene will develop breast cancer (strikingly, even in women in their twenties and thirties). Some of these mutations are found in about 2 percent of Jewish families from Eastern Europe (Ashkenazi Jews). This is the first ethnic group closely examined, and other populations may have equivalent rates of specific mutations. |
Scientists rank the factors that they believe increase breast cancer risk differently than the lay public does. Delayed childbearing, obesity after menopause, estrogen use and less breast-feeding increase the risk of breast cancer in epidemiologic studies, but chemical use, pesticides and electromagnetic fields do not. (Farm chemicals may be associated with lymphomas and sarcomas.) The lay public ranks chemical use, pesticides and electromagnetic fields highest. |
Prostate cancer
The incidence of prostate cancer is quite high. About one-fourth of those who develop prostate cancer die from it. Prostate cancer is more common in the Northeast and is particularly common in African-American men. The percentage of men with prostatic pre-cancers in Japan and in the US is about the same, but the number of clinical cancers in American men is much higher. The incidence has been increasing slowly over decades, but mortality has been increasing even more slowly. |
When screening for prostate-specific antigen (PSA) became available as a test for prostate cancer, the known incidence initially increased substantially and then abruptly decreased. The same peak-and-fall was seen when mammography screening for breast cancer was introduced. PSA screening is controversial. Many older men harbor early prostate cancers that would never be life-threatening even if left untreated. The side effects of surgery and radiation for prostate cancer are impotence and incontinence. Ideally we want to diagnose only those cancers that would eventually be life-threatening and would like to avoid detecting and treating those that never would have produced clinical symptoms. |
Lung cancer
Lung cancer is the major cause of cancer death for both men and women. Ninety-five percent of patients with a diagnosis of lung cancer die from the disease. The risk of dying from lung cancer is decreasing among men, probably because many men quit smoking about 15 years ago. For women, lung cancer risk is increasing. |
Areas where smoking is still common, like Kentucky, where 40 percent of the population smokes, are predominant locations of lung cancer in the US. Along the coasts, near shipyards with asbestos, the risk is even higher. More than nine of 10 patients with lung cancer are or were smokers. After quitting, an ex-smoker's risk of lung cancer remains higher than in never-smokers for about 20 years. A new screening test, called spiral CT, is under evaluation in patients with a history of smoking. Although spiral CT can detect small lung cancers, it is not established whether its use can improve survival. |
Colon cancer
Colon cancer risk is highest in the Northeast for both men and women. The lifetime risk of developing colon cancer for an American is 6 percent. We know that four or five genetic changes are required to produce a colon cancer. One or more mutations leads to development of a benign polyp. Several more genetic changes lead to cells behaving more aggressively, finally leading to a colon cancer capable of invading through normal tissue. |
This long process of cancer development allows time for screening. Polyps can be detected with a stool blood test, sigmoidoscopy or colonoscopy. The risk of colon cancer can be decreased by removal of the polyps before they become frankly malignant. All Americans over the age of 50 should undergo at least the simplest screen, checking stool for digested blood. Screening sigmoidoscopy and colonoscopy involves checking the lower colon or the whole colon, respectively, for lesions with a fiber-optic tube. |
A high percentage of patients with colon cancer also have other family members with colon cancer. About 5 to 10 percent of cases involve abnormal genes for which we can screen. One abnormal gene, which causes familial polyposis, causes thousands of polyps to develop in the colon by the time an affected family member is an adolescent. The usual treatment in such patients is to prophylactically remove the colon, since virtually 100 percent of these patients would otherwise develop colon cancer. If you're going to lose the battle, remove the battlefield. |
A second gene that can cause familial colon cancer is caused by damage in a "mismatched repair" gene. Cells cannot repair DNA damage. Mutations in other essential genes accumulate, and affected individuals have a high risk of colon and endometrial cancers. |
Current status of cancer treatment
How have we done over the decades in the war against cancer? Back in 1935, 20 percent of Americans diagnosed with cancer survived. By 1965, about one-third survived, by 1990, about half, and today in the year 2000 about 59 percent survive. Despite the improved percentage cured until 1992, the number of Americans dying of cancer was increasing, mostly because the incidence of cancer, particularly of lung cancer, was rapidly increasing. |
Fortunately, cancer mortality per 100,000 has been falling since 1992. Because the US population has been increasing, the absolute number of cancers first evened off, but then in the last few years even the absolute number of cancer deaths has now begun to drop. The chance of an individual getting or dying from cancer has decreased about 7 percent since 1992. |
What has worked? Prevention, particularly decreased smoking. Perhaps better nutrition has contributed to the decreasing risks of dying from colon and stomach cancer. The benefit of screening for cervix, breast and colon cancers are established, but screening for prostate and lung cancer remains controversial. For those who develop cancer, survival can be achieved despite less extensive surgery. Drug and radiation treatment at the time of diagnosis can decrease the chance of relapses. |
Better imaging for diagnosis and staging
One of the most important developments is better imaging. First computerized technology (CT) scans, and then magnetic resonance imaging (MRIs), and now PET and SPECT scans allow us to detect smaller tumors without invasive tests or exploratory surgery. A PET (positron emission tomography) scan may identify abnormalities missed even by CT scans. |
MRIs of the breast, which can detect cancers missed by a mammogram, may be useful for young women from families with abnormal breast cancer genes. (Mammograms are less helpful for screening young women who have dense breasts.) MRI is more expensive and has not been used as a general screen as mammography has, but it may prove cost-effective for these younger women with a high risk of cancer. MRI also uses no radiation, something physicians would like to avoid in women already at high risk for breast cancer. |
Better diagnostics
Currently a diagnosis of cancer is made by a pathologist examining a piece of the tumor under a microscope. A new tool called a gene chip can distinguish between two kinds of leukemia: acute lymphocytic leukemia (ALL), common in childhood, and acute myelocytic leukemia (AML), which is more common in adults. Although genetic differences cannot be seen through a microscope, the genes expressed in ALL and AML can be distinguished by a gene chip, which can be designed for many kinds of tumors. |
Targeted drugs
A new leukemia drug has been developed that targets a protein produced only in cancer cells, by the broken chromosome associated with leukemia, and not in normal cells. In an early clinical trial, 30 out of 31 leukemia patients treated had a complete remission. In 30 percent of these people, no abnormal cells could even be detected using DNA technology. This is one of the first examples of a drug targeted only against tumor cells. |
Several gene therapies target p53, a gene that is mutated in half of all cancers. Normal p53 acts as a tumor suppressor, repairing damaged DNA or signaling cells with DNA damaged beyond repair to die. If p53 is mutated or lost, DNA damage accumulates and cancer can develop. A gene therapy technique uses a modified virus that can infect only cells that carry an abnormal p53 gene. The virus multiplies only in cancer cells, and kills them. |
Another gene therapy uses a virus that replicates only in cells that make prostate-specific antigen (PSA). Men who have already had a prostatectomy do not produce PSA unless they have a recurrent tumor. In a man whose PSA levels climb after a prostatectomy, such a virus could seek out and destroy only cells expressing PSA. |
The drugs Angiostatin and Endostatin caused excitement in 1998 when the New York Times reported that they "eradicate any type of cancer with no obvious side effects, no drug resistance in mice. The drugs... work by interfering with the blood supply tumors need. Taken together they make tumors disappear and never return." Many drugs good enough to get into clinical trials can cure mice or other laboratory models. Nevertheless, these drugs represent a very important concept in dealing with tumors. A tumor larger than one millimeter cannot grow any bigger unless it attracts a new blood supply. If it manages to mutate a gene or two to get a blood supply and some nutrients, it can grow and become life-threatening. The cells that form the vessels of a tumor's blood supply are not cancerous and thus are a very stable target compared with the tumor itself. One problem with the currently available drugs to treat tumors is that tumor cells continue to accumulate mutations, which can confer resistance to the drugs. The cells of the blood vessels are not accumulating mutations and thus should be less likely to mutate to resistance to anti-angiogenic drugs. |
Another interesting new drug target is telomerase. Telomeres exist at the ends of chromosomes, and they get shorter over time as DNA replicates. Cells stop dividing and die when the telomeres get too short. This is probably why human beings don't live to be 200. Germ cells (sperm and egg cells) have the enzyme telomerase, which can lengthen their telomeres and make them immortal. Cancer cells also make telomerase. A vaccine against telomerase might target only cancer cells and germ cells. Although loss of fertility would be a predictable toxicity, many cancer patients who have already had their families might find this side effect acceptable. |
Vaccines are already being used to prevent cancers caused by infectious agents. We already have a hepatitis vaccine; hepatitis causes liver cancer. Epstein-Barr virus is known to cause certain kinds of lymphomas. Papilloma viruses cause cervical cancer. Vaccines against cancer-causing papilloma virus are now being tested. We now have a diagnostic test for the herpes virus that causes Kaposi's sarcoma and hope to eventually have a vaccine against it. |
Twenty-first-century cancer detection and treatments
Within 10 years, we will probably have much better cancer screening. We are now evaluating spiral CT to screen smokers for lung cancer. It costs about the same as a mammogram. We can't cure advanced lung cancer, but maybe we could cure early lung cancer if we could detect it with better imaging. |
We may soon be able to diagnose cancer not only with a microscope but also with gene chips. We would know not only what a tumor looks like but what genes it expresses and what specific drugs might target that particular tumor. Treatment could be made specific not only for the genetic profile of the tumor but also for the genetic profile of the patient. |
It doesn't take much imagination to think that in another 20 years we could do DNA screening at or before birth to figure out who will be at risk for what tumors. We could intervene throughout a person's life to prevent cancer, rather than treating it. |
This is an exciting time in cancer research not because of the ideas out there--there were always good ideas--but because of the number of good ideas. We used to be testing one or two new drugs in any given year, and now we are testing 10 to 20. The information is going by so quickly now it's hard to even keep up, and that's good. |
Caveats
The first and most important reality check to this speculation--and the only thing I'm really sure of--is that events will not unfold as we expect. Advances in science often come from unexpected directions, and it's difficult to predict them. For every 10 drugs that come to clinical trial, only one is found effective. |
Historically, effective treatments have come from unexpected places. The first remissions of cancer were from Mustargen, originally used in chemical warfare. Bone marrow transplant research, now used to cure leukemias, was funded by the United States Defense Department, which was primarily interested in treating potential radiation exposures from nuclear war. Cisplatin, the drug that in combination cures testis cancer, was originally developed as an antibiotic. Tamoxifen, one of our best anti-estrogens for breast cancer, was originally developed as a birth-control pill. A gene mutated in colon-cancer families and described by investigators in Baltimore sounded to a researcher in Boston like a gene he was working on in yeasts. He used the yeast gene to identify the human gene, and now we have a diagnostic test to identify those members of colon-cancer families that inherited the risk of colon cancer. |
Another caveat: 25 to 50 percent of current US cancer deaths are preventable with what we already know. Many patients with Medicare (who have full health insurance) don't get mammograms or colon cancer screening. If you're 50 years old, you should be screened for breast and colon cancer, and perhaps prostate cancer. Thirty percent--not even half--of Medicare patients with colon cancer get adjuvant chemotherapy that's known to be curative. |
In summary, the gains in fighting cancer are real, if modest. The decrease in mortality is 7 to 8 percent since 1992, but mortality continues to fall and the rate of decline appears to be accelerating. The path from the lab to the clinic is longer than the press and the public appreciates. Twenty years elapsed between the laboratory discovery of monoclonal antibodies and the creation of one that was clinically useful. We've known about the p53 gene for about 23 years, and we're just evaluating it as a target of therapy now. |
Molecular diagnostics are already available; therapeutics are coming. But certainly don't discount serendipity. |
|
| |
