Molecular Testing of Cancers: 'Let's Try to Get It Right'


September 21, 2016

This feature requires the newest version of Flash. You can download it here.

Hello and welcome. I'm Dr George Lundberg and this is At Large at Medscape.

Wilson Riles, the former director of public education for the state of California, once said at an event I attended (to paraphrase): "Never do anything just because you can." This dictum certainly applies to clinical laboratory and other diagnostic testing in medicine. Early imaging for low back pain, chest x-ray screening for lung cancer, mass screening for breast and prostate cancer, and blood testing for syphilis have all had their day—for better or for worse. Now we have advanced, sophisticated, and expensive next-generation sequencing for the genomic characterization of potentially lethal malignancies.

It is worthwhile to recall the brain-to-brain (B2B) loop in laboratory testing.[1] Some brain, usually that of the physician, decides to order a test and asks, "Why should I do this?" The loop then proceeds through specimen collection, identification, transportation, preparation, and analysis to reporting, interpretation, action, and outcome, which is now the tenth step.[2] What outcome resulted for the patient because of the decision to order the test? Did it make a difference? Was it beneficial? Was it worth the cost?

Molecular testing of human malignancies has been working its way through this B2B process. The costs have been high; the benefits thus far are mostly a better scientific understanding of the very nature of many cancers. But now, clinical data are accumulating that support molecular testing of many cancers. A simple diagnostic label of a cancer by organ location and histology is no longer enough. Many cancers now require molecular characterization in order to establish a proper actionable diagnosis.

Approximately 1 million Americans per year are cured of their potentially lethal cancers by surgery and radiation, but nearly 700,000 are not. Most of those patients are then given chemotherapy. Except for some hematologic and germ cell tumors, cure does not result from cytotoxic chemotherapy alone. A significant clinical response may be confined to some 12% of solid tumors. If a newer targeted therapy is provided based upon site and histology, only about 4% of these patients' cancers respond.[3] However, targeted therapy aimed at a previously defined molecular biomarker has shown a 30% response rate with approximately doubled progression-free survival.[4] These are big outcome differences that justify molecular testing. But these drugs can be very expensive. Molecular test results can prevent the waste of dollars on predictably ineffective targeted therapy.

The main decision still to be elucidated is when in the clinical life of a recognized cancer is the best time for molecular testing and how extensive should it be? At initial diagnosis, since the number of mutations may be fewer early in the course of the malignancy and effective intervention may be more likely? Or, after recurrence or spread, despite the likelihood of greater mutational load by that time? Or [should testing occur] sequentially, such as with liquid biopsies?[5]

An argument can be made for each approach. The best answers will get worked out by tumor response and outcome data—molecular cancer diagnosis by molecular cancer diagnosis.

The Association for Molecular Pathology (AMP) recently published a list of justifications for molecular testing of malignancies.[6] The AMP Framework for the Evidence Needed to Demonstrate Clinical Utility Task Force made specific recommendations in the following areas:

  • Adopting patient-centered definitions of clinical utility;

  • Recognizing that an accurate diagnosis has inherent clinical utility;

  • Recognizing that clinical utility for molecular diagnostics is context-dependent;

  • Supporting multiple modalities of evidence generation, including appropriate alternatives to randomized clinical trials;

  • Utilizing a modified ACCE (analytic validity, clinical validity, clinical utility, and ethical, legal, and social implications) model;

  • Recognizing the critical role of the molecular professional; and

  • Supporting professional organization-driven practice guidelines.

Upon receipt of a lab test result, the clinician wants to know: What does this mean? What should I do now? My activist approach tells me that the pathologist should advise at the front end (ie, why do this test?), determine how and when the specimen is collected, and where and how the test is performed. And at the distal end of the B2B loop, [advise on] what [the results] mean and what should now be done.

It's an exciting time in oncology and pathology. Let's try to get it right.

That is my opinion. I'm Dr George Lundberg, and my conflicts of interest include being a pathologist, belonging to the AMP, believing in this field, and serving as chief medical officer of CollabRx™, a San Francisco–based company that provides solutions generated by computer matching of oncologic genomic testing results with best subsequent clinical actions for each patient.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: