Personalized Medicine was our Health Future of the Year 2006. For 2007, the award goes to what — for the moment — amounts almost to the same thing: the War on Cancer.

When FDA Commissioner Andrew von Eschenbach predicted, a few years ago as head of the National Cancer Institute, that cancer would effectively be licked by 2015, he was accused of over-exuberance. Given the advances in the War on Cancer in 2007, come 2015 we might wonder why Dr. Eschenbach was so timid, and how cancer center boards and administrators could have been so shortsighted as to expect a financial return on such long-term investments as 125 million dollar proton beam therapy machines.

Our confidence is bolstered by data unavailable to Dr. Eschenbach when he made his prediction, including the finding that in 2003-4 (the latest years for which figures are currently available) the number of cancer deaths in the United States dropped for the second straight year. Death from childhood cancer in the United States also declined, by 1.7 percent from 1990 to 2004, even though the number of reported cancers increased by 0.6 percent annually.

These successes were the result of more screening, less smoking, better detection and treatment of colorectal, breast, and prostate cancers. and “a revolution in treatment between 1998 and 2000, and revolution is a mild word,” as one expert put it. “We went from having one drug to having six or seven good drugs. The cure and survival rates have increased dramatically as a result. The cost of care has also gone up, but you get what you pay for.”

But that was then, when the revolution largely involved refinements to the “slash, burn, and poison” (surgical, radiological, and chemical) approaches to cancer treatment. Those rates and those approaches are starting to seem grossly under-exuberant in light of the personalized, tightly targeted therapies now beginning their domination of the cancer battlefield.

Personalized Cancer Diagnosis

The beginning of the end of cancer can be seen in the accelerating discovery of cancer and cancer-risk genetic biomarkers and the development of biomarker test technologies such as spectrometry and gene chips. Certainly, there are minefields to be negotiated: the FDA does not regulate the tests, many lack clinical evidence of effectiveness and are not reimbursed by payers, and patients may fear job and insurance discrimination if they test positive for a disease or disease risk.[*]

The minefields will cause delays and casualties along the way, but we have begun to pick our way through. At least two biomarker tests are gaining acceptance in clinical practice and are being reimbursed by several major payers (the US$3,200 Oncotype DX and the similarly priced MammaPrint, both of which test for risk of recurrence in breast cancer. MammaPrint is even FDA-approved.) FDA regulation of such tests would probably slow their proliferation but the payoff would be better test validity and reliability. In the meantime, genetic tests will at least be used clinically to supplement existing diagnostics and therapies.

More biomarkers – fodder for the test-makers — were unearthed in 2007. For example, the finding that about fifty percent of a sample of 20 ovarian cancer patients did not benefit from the drug paclitaxel because they had low levels of a protein called TGFBI identifies TGFBI as a biomarker for selecting patients likely to respond to this class of drug.

Personalized Cancer Therapy

That study also found that TGFBI sends messages to the cancer cells, sensitizing them (making them vulnerable) to paclitaxel. This finding could lead to new treatments that simulate such messages, leading to a significant improvement in paclitaxel response in ovarian cancer patients normally unresponsive to it, and also for other available paclitaxel-like (taxane) drug treatments for lung and breast cancer patients.

Coupling biomarker discoveries and resultant tests — personalized diagnostics — to drugs that precisely target cancer cells is the essence of personalized cancer medicine. Both are advancing in lock step. Fifteen targeted cancer drugs were FDA-approved by February 2006 (the last date for which we could find a coherent list), and more than 100 are in development. But while you might get what you pay for, you might not be able to pay at all: targeted cancer drugs can cost as much as $10,000 for one month’s treatment. This is a problem that, in our view, will only be solved when in silico trials replace human clinical trials. Dr. David Eddy has launched a spirited offensive on that front, with his Archimedes program.

Computational methods (of which Archimedes represents one type) are already making their mark in personalized cancer medicine. For example, last year, a mathematical model found that if the HMMR gene mutates, it increases a woman’s risk of breast cancer by more than a third.

Identifying cancer risk genes enables earlier detection of risk and earlier, more effective, treatment. One treatment introduced in 2007 is truly a Big Gun in the War on Cancer: the cancer vaccine.

Cancer Vaccines

The cervical cancer (strictly speaking, HPV) vaccine Gardasil was the first shot fired by that big gun, and with an anticipated US$2+ billion in annual worldwide sales, it is a shot being heard around the world. New shells in the form of vaccines for melanoma, breast, and lung cancers are already being loaded into the gun. About two thirds are therapeutic vaccines; the rest, prophylactic.

Provenge, a therapeutic vaccine for asymptomatic hormone-refractory prostate cancer patients, had encouraging early clinical trials and could be on the market by the end of 2008. GVAX, another vaccine for hormone-refractory prostate cancer patients, is about a year behind Provenge. (So much for one expert’s prediction last year that the incidence of prostate cancer is likely “to absolutely go through the roof in the next decade or two” as the population ages.) MyVax, a patient-specific vaccine for previously untreated patients with stage III and IV follicular non-Hodgkin’s lymphoma, was granted FDA Fast-Track Status in 2006 and is in Phase III clinical trial.

Another therapeutic vaccine has triggered a strong immune response to cancer cells in mice. The results were described as “astounding.” The vaccine produced levels of a key immune system antibody sufficient to kill mouse-derived cultured epithelial cells (commonly involved in solid tumors) and to stimulate an immune response in healthy mice. Human phase I clinical trials are expected to begin in 2008.

In a similar but more developed approach, a therapeutic vaccine for epithelial ovarian cancer has already produced phase I clinical trials results described as “encouraging.” The vaccine stimulates production of specialized T cells to target a protein produced in a high proportion of ovarian cancer cells, but not in healthy cells. The T cells were detected in patients up to 12 months after immunization, suggesting the vaccine may be prophylactic as well as therapeutic.

The ultimate cancer prophylaxis might lie in the genes. A breed of mouse engineered in 2007 to carry a gene called Par-4 was apparently invulnerable to cancer, and it lived longer as well. Whether giving humans the Par-4 gene would work the same way as in the mice remains to be seen, but it seems possible in principle.

Other Promising Weapons

Here are some examples of other banner advances on the cancer battlefield in 2007:

Make Cancer Cells Mortal

The discovery of the three-dimensional structure of a part of the telomerase molecule may lead to the development of direct inhibitors of the enzyme, which is associated with the uncontrollable proliferation of cells seen in as many as 90 percent of all of human cancers. Knowing the physical structure will enable pharmaceutical companies to design drug molecules to disrupt telomerase self-assembly in cancers. Although telomerase is essential for normal cell division, it is switched off after about 50 divisions, to prevent runaway cell proliferation. Many cancer forms deactivate the “off” switch, permitting the cancer cells to replicate indefinitely.

Turbocharge the Immune System

By copying the immune system of patients with the rare neurological disorder PCD (paraneoplastic cerebellar degeneration) it is possible to turn normal immune cells barely able to detect breast and ovarian tumors into ones that destroy them. The immune system kills cancerous cells in all of us all the time. However, in PCD patients it does not stop there, but goes on to kill a group of neurons in the brain that express a protein, cdr2, that cancer cells also express. When examined because of their neurological symptoms, PCD patients are often found to have small gynecological tumors that have failed to metastasize – they are practically immune to the cancers. It turns out they have a unique type of immune cell, called CD8+, that specifically targets cdr2 and destroys tumors.

To make breast and ovarian cancer patients’ immune cells like those of PCD patients, researchers endowed normal human immune cells (in experimental mice) with CD8+ characteristics, turning the normal cells into cancer killers. This points to a potential therapy for gynecological cancers, though to this lay mind it also suggests that patients who receive the therapy would then be at risk for PCD. Nevertheless, the research provides deeper understanding of tumor immunity in all cancers and holds out promise of an end to the “slash, burn, and poison” approach of surgery, radiation, and chemotherapy.

Sock ‘em with SIR Spheres

Radioactive plastic microspheres called “SIR-Spheres” infused directly into the liver through the hepatic artery halted the growth of tumors in 71 percent of patients tested in a 20-patient clinical trial during ten months of follow-up, with minimal side effects. Liver function tests in the responding patients have become normal or have stabilized.

SIR-Spheres were first approved for use by the US Food and Drug Administration in 2002. They now appear to offer a treatment option for patients who develop multiple liver tumors and have no other effective treatment options, and to work best in patients who have good blood flow to their tumors.

Burst their Bubble

Bursting bubbles non-invasively generated at the site of a tumor by High Intensity Focused Ultrasound (HIFU) release heat energy that kills the cancer cells. Clinical trials of the technique in terminal kidney and liver cancer patients are under way in the UK. However, it takes up to five hours to treat a 10cm tumor, compared with 45 minutes or so for surgical resection, and the results are not immediately observable. To overcome these limitations the researchers are developing a sensor that can “hear” the tiny bubbles bursting which will help deliver the treatment more accurately, deeper in the body, and to monitor the treatment in real time.

The therapy is not suitable for cancers that have metastasized. Candidates for the procedure would generally be patients with isolated solid tumors in the kidneys or liver.

Conclusion

Hospital and Health Networks editor Mary Grayson has lamented the paucity of prognostications this New Year just past, leaving healthcare institutions and practitioners to stumble blindly into 2008. “No one,” she suspects, “can come up with anything particularly new or surprising to say about health care in 2008.” She may be referring more to near-term policy and reimbursement predictions than to longer term auguries of the future of medicine, but even there, she would be quite right.

It’s neither new nor surprising that cancer will be effectively licked by 2015, not to Dr. Eschenbach, not to us, and hopefully not to you. But if you are surprised, perhaps we can claim that this edition of the Digest has helped in some small way to set the universe back on course, by predicting that in 2008, we will see an exponential advance in the prosecution of the War on Cancer.