[ed note: Michael Jorrin, who I like to call “Doc Gumshoe”, is a longtime medical writer (not a doctor) who shares his thoughts with us from time to time, generally on non-financial topics in health and medicine (as today, though, he mentions a couple publicly traded companies). His words and opinions are his own.]
The needle in Doc Gumshoe’s handy Trend Indicator points unwaveringly towards Precision Medicine. This is possibly a public-relations-motivated bandwagon, and if you citizens of Gumshoe Republic have been paying attention, you know that recent jumpers-on have included the highest in the land. This is not to say that Precision Medicine has no merit other than its PR value, only that the nomenclature begs the question. Precision Medicine is obviously good, and that suggests that the rest of medicine – by implication imprecise? – is groveling in the mud.
There are other ways to describe this research avenue – individualized or personalized therapy, molecular targeted therapy, genomic-based therapy – which do not set up an automatic “good vs bad” distinction. The objective of these developing cancer therapies is to identify and target the vulnerabilities in the specific cancer in a specific patient. It does not take anything away from the immense potential of individualized therapy to point out the success of the therapies that have successfully targeted the behaviors that most types of cancers have in common. These were discussed in the first installment, “Cancer Care: Where Are We and Where Are We Heading,” which was posted a couple of weeks ago, and which you can see here. Among these are the characteristics that cancers grow more quickly and need somehow to induce new blood vessel growth to provide needed nourishment, and also that cancer cells will replicate endlessly. Drug therapies have been developed that target both of those behaviors, and these therapies have been hugely successful. They are not “Precision Medicine,” but neither are they some inferior, imprecise form of medicine, deserving of the scorn of the media and the public.
We’ll take a look at some of the approaches to individually-targeted therapy and some of the agents, some already on the market, but most at various stages of development. Many of these drugs look quite promising, both from the standpoint of the benefit that they may bring to cancer patients and also the profits that they may lavish on developers and investors. The clues about these agents’ potential come in the form of clinical trial results. But I need to warn you that in many cases, these clinical trial results appear a bit dim. They are expressed in terms of additional survival time in the patients taking the new drug, and quite often the additional survival is a matter of months. Critics, whose numbers are legion, point out that giving a few more months of misery to a terminal cancer patient is no great gift. However, there’s a reason why those results are often discouraging.
Why clinical trial results with new cancer drugs are often less than stellar
The ideal scenario for a drug developer is to find an agent that treats a disease or condition for which there are no known effective interventions. In such a case, the developer announces a clinical trial and is immediately overwhelmed with candidates, since even if the patient gets assigned to a placebo, it’s no worse for the patient than the current situation in which there is no treatment at all. And in most cases in such clinical trials, patients in the placebo arm are permitted to switch to the treatment arm after a certain period, at which point they get a drug which would not otherwise be available to them. For potential drugs that treat some diseases/conditions, there are also other options. In the case of a drug to treat hypertension, it may be possible to recruit newly-diagnosed patients to a clinical trial, even a placebo-controlled trial, because the effectiveness of a drug to lower blood pressure becomes apparent relatively quickly, so clinical trials are not over-long, and the patients in the placebo arm are not exposed to undue risk.
But with a cancer drug, the restrictions present a colossal hurdle. In the first place, it’s obviously impossible – or at least completely unethical – to put cancer patients on a placebo, so there’s no question of demonstrating efficacy in comparison to placebo. In the second place, there are many established forms of treatment for almost every form of cancer, and even if the treatment has mixed results, very few patients would agree to a completely experimental treatment for a disease that could kill them in fairly short order, when there are treatment forms available that have at least a reasonable chance of success. So there are very, very few clinical trials of new drugs that are able to enroll newly-diagnosed patients.
In most cases, cancer trials initially enroll patients in whom established forms of treatment have failed. Often, these are patients in whom a cancer that originated in one site has metastasized to another site or to multiple sites. The new drug will seldom be used by itself, but will be combined with another, more established drug.
Positive results in these cancer trials usually consist of only a few months longer survival than in patients treated with established drugs. Both the FDA and the EMA (European Medicines Agency, which is the FDA equivalent in Europe) recognize these as positive results and approve the drug, but that does not mean that the drug will be widely-used. An example is a promising drug, Zaltrap (ziv-aflibercept) from Regeneron. Zaltrap improved survival by only 1.4 months in patients with metastatic colorectal cancer as compared with Avastin (bevacizumab), from Genentech. Memorial Sloan-Kettering decided not to use it, because it costs about twice as much per month ($11,000) as Avastin. But that doesn’t mean that in recently diagnosed, early stage cancers, drugs like Zaltrap might not be a significant improvement. Currently, Avastin is approved for use in other cancer sites besides colorectal cancer – glioblastoma (a highly malignant form of brain tumors), non-small cell lung cancer, and metastatic kidney cancer – and is being studied in many other types of cancer. Zaltrap is also being studied in more than 80 clinical trials, in several types of cancer, but the way forward is not clear.
What cancer drug developers hope is that they come up with an agent that demonstrates better results than whatever else is out there, even if it’s only a few more months of overall survival in these exceedingly difficult-to-treat patients. Modest benefit in the most difficult patients can then provide justification for using that agent in patients with less advanced disease, frequently in combination with another drug, or after some other first-line therapy has been employed – what’s termed “adjuvant” therapy – or sometimes in advance of first-line therapy, such as in the attempt to shrink a tumor so that it can more easily be surgically removed. This last is called “neoadjuvant” therapy. The ultimate goal of drug developers is, of course, to have their drug recognized as the first-line therapy of choice. This does not happen overnight, but it has been known to happen.
What are the goals of “Precision Medicine”?
Obviously, the ultimate goal is to treat cancer more effectively – the question is, “How, exactly?” The wide-angle answer is that this approach proposes to base cancer treatment on cause rather than on location; meaning that instead of classifying cancers by the organ systems that they affect, Precision Medicine expects – ultimately! – to be able to pinpoint the genetic cause of many or most cancers and to develop treatments that target the specific genetic cause and, in some way block or inhibit its cancer-producing effects.
In promoting this approach, there’s a good deal of supercilious characterizing of location-based cancer treatment as imprecise and somehow behind the times. This needs to be viewed with skepticism. It will be a long, long time before oncologists ignore the location of a cancer in planning treatment: regardless of gene-sequencing, malignant melanomas will be treated differently than lung cancers.
But the Precision Medicine Initiative seeks to accomplish genuinely important objectives: essentially, by means of a proposed cohort of a million or more volunteers, statistically valid correlations will presumably emerge linking specific genomic sequences with specific cancers. This is something that Big Data can undeniably do, given enough time and enough money – and enough volunteers, of course. If robust correlations are found, it may be possible to make statistically valid assumptions about causality; this has already happened with several types of cancer, e.g., melanomas in patients with a mutated BRAF gene, and breast cancer in women with BRCA1 or BRCA2 genes or with particularly active HER2 receptors. In turn, those assumptions about causality have led to major treatment advances.
And what are they looking for?
Most of our genes don’t give a hoot about cancer, but among them there are heroes and villains. The heroes are the tumor suppressor genes, which identify cancerous cells early on and initiate processes that lead to their destruction, such as mobilizing the immune system. The villains are the driver oncogenes, which can signal cells to multiply and grow endlessly, until they overwhelm the entire organism, occupying space, stealing nourishment, interfering with necessary function, and destroying tissue. But the driver oncogenes can also initiate the cancer process by inhibiting the action of the tumor suppressor genes, or by interfering with the process that guides apoptosis – programmed cell death. Without apoptosis, normal cells can become cancerous.
A bit more specifically, here are some targets of genomic analysis:
- Abnormal gene mutations can be spotted in a total or partial gene sequence and statistical associations with cancers may emerge which permit forming a conclusion; i.e., this specific genetic mutation causes this specific cancer.
- The number of copies of specific genes in a tumor can be measured by means of a process called Fluorescence In Situ Hybridization (FISH); this provides information about gene duplication within a tumor.
- Gene translocation can be identified. This occurs when chromosomes break apart and rejoin, such that genes that are normally far apart are found in close proximity, with potentially malign consequences.
- Protein biomarkers in a tumor can be measured with approximate accuracy through immunohistochemistry, which can provide useful therapeutic information.
These analytic procedures can be used in the aid of finding the most effective drug as w