Organ or Pathway?

Organ or Pathway?

George W. Sledge, MD, FASCO

Jun 27, 2011
At meetings these days, a growing chorus of investigators sings a particular tune: organ-based therapy is dead; it is all about the pathways. But then I go home to my clinic and treat my patients in a very organ-based fashion. My patient has a screening X-ray read by an organ-specific radiologist, then has a biopsy, frequently by an organ-based specialist, and (if my patient is fortunate) has the biopsy interpreted by a good organ-based pathologist. I always get nervous when my organ-specific pathologist is out of town and a neuropathologist subs for him. And then my patient has definitive local-regional therapy by an organ-based team of surgeons and radiation oncologists before finding her way (as I am a breast cancer specialist) to an organ-based medical oncologist. We call this a “multidisciplinary team,” and it is generally acknowledged to offer the best cancer care.

Wait a minute, I hear you saying, we weren't actually talking about taking care of a patient, we were talking about drug development and systemic therapy. And here, of course, the point is well taken. What caught my eye at the ASCO Annual Meeting this year was a fascinating phase I experience from M. D. Anderson in which investigational drugs were assigned by predicted mode of action and measurable biological target. This makes sense in so many ways. If the drug is a home run drug, an imatinib for CML, you see the ball leaving the park in a hurry, you save a bunch of money in development costs, and (greater ethical clarity) you do not treat patients who will not benefit and may experience unwarranted toxicity. In theory, if you know the target and have a drug that precisely hits that target, you can design trials that recruit patients irrespective of the organ of origin. The fact that one has a Src or ALK mutation should matter more than whether the tumor developed in the lung or the brain. Pathways, not organs, right?

Well, yes and no. Take the saga of medullary thyroid cancer, one of our recent targeted therapy successes. Medullary thyroid carcinoma is rare, accounting for no more than 5% of thyroid malignancy. Mutations in the RET oncogene drive all of hereditary and 50% of sporadic MTC (pathways, hurrah!). It is sufficiently common in MTC, in fact, that the study that led to the approval of vandetanib for MTC used the pathologist's diagnosis of MTC as the "biomarker" for study entry (organ-based pathology, hurrah!).

But wait, there's more, as they say on Saturday late-night commercials for obscure kitchen implements. Vandetanib is kind of a messy drug. It was developed as a VEGFR inhibitor, and as such was tried in non-small cell lung cancer, a disease where VEGF inhibition had established benefit. But it was a dud there, with a negative phase III trial. Fortunately, it was a somewhat promiscuous kinase inhibitor, which allowed its resurrection via MTC, with a trial that accrued patients—you guessed it—in clinics specializing in thyroid cancer from around the world. As with many orphan tumors, the rarity of the disease requires trials to recruit patients in highly specialized, organ-specific clinics. Maybe there will be RET-based clinics someday, but I doubt it.

So the “pathways versus organs” argument is a false dichotomy. No, we do not want to try crizotinib in every patient with lung cancer, but at the same time virtually every lung cancer patient with an ALK mutation has an adenocarcinoma. An H&E stain can probably save a good deal of expensive testing. Kinase targets are not randomly distributed; they clump in specific organ-based sites, and in specific tumor types within those sites. What is more, and I find this fascinating for what it implies, resistance to kinase inhibitors can be quite organ-specific. Resistance to EGFR inhibition in lung cancer is a matter of mutations in the kinase domain, but in colorectal cancer resistance is mediated by KRAS. Why is that? And is it also true for other kinase-targeting agents?

So should the FDA start approving drugs for a mutated BRAF indication rather than for a metastatic melanoma with mutated BRAF indication? This would be a fascinating approach, and certainly a sea change in drug development. But let me suggest that the experience to date, as mentioned above, suggests that we are not quite there yet, either at a general drug development level or at a specific patient-treatment level. The argument in favor of this approach makes a few too many assumptions: that drugs are pristine in their targeting of specific kinases, that we can always measure the target or be confident that it drives a tumor, and that a mutant kinase gene means the same thing biologically in the context of the kidney microenvironment as it does in the colon. A testable hypothesis, yes; but one well tested, no. There are now studies that are beginning to investigate this approach, and trial designs (such as the randomized discontinuation design) well suited to pathway-based clinical trials.

Tossing out the false choice between pathway and organ makes sound biological sense. What we see under the microscope at the tissue level is the product of the molecular biology underlying it, both benign and malignant. It is the acting out at a macro level of innumerable molecular events. So at least for the moment it makes sense not to fire all the pathologists and hire molecular biologists in their place. We can use them both, at least until the time when deep sequencing of a patient's tumor is sufficiently cheap to become routine, and the meaning of those tortured three billion base pairs sufficiently transparent to trust for routine care.


The ideas and opinions expressed on the ASCO Connection Blogs do not necessarily reflect those of ASCO. None of the information posted on is intended as medical, legal, or business advice, or advice about reimbursement for health care services. The mention of any product, service, company, therapy or physician practice on does not constitute an endorsement of any kind by ASCO. ASCO assumes no responsibility for any injury or damage to persons or property arising out of or related to any use of the material contained in, posted on, or linked to this site, or any errors or omissions.

Back to Top