The Daily Sign-Out

Incidentalomas revisited

Dr. Bruce Friedman posted a thoughtful blog on Lab Soft News recently concerning "incidentalomas" ("How to Avoid the Risks of a CT Incidentaloma").  This post happened to coincide with a new working committee at my hospital commissioned with developing a virtual clinic for following patients who have incidental lung nodules discovered.

Perhaps unlike other sites, there are widely disseminated radiographic recommendations for following lung nodules (Fleishner criteria).  Of interest, a recent article found that despite widespread awareness of these criteria, actual incorporation of these criteria into practice has been found wanting--only around 50% of radiology practices actually used these recommendations in reports.  One issue that we will be challenged with is how to track these patients across multiple different providers for appropriate follow-up.  One model we are considering is already in place: using a breast "navigator" to follow-up patients with abnormal mammograms.  This issue arose recently because of a series of patients presented at our Tumor Board who had incidental lung nodules identified but were not followed appropriately and then were re-discovered with inoperable advanced stage disease months later.

This idea also came up just today in a conversation I was having with a urologist colleague regarding renal incidentalomas treated with cryoablation.  I recently diagnosed an incidentally discovered 2.5-cm mass in an 80-year-old woman on CT-guided needle biopsy as an muscle-predominant angiomyolipoma (mini-case report on its way!).  For better or worse, the tumor was ablated probable before it was even grossed!  But, to return to Dr. Friedman's post, what if the patient was simply observed with serial imaging?  The urologist was unaware of any recommendations similar to the Fleishner criteria for lung nodules.  If you, dear reader, know of any please post them in a comment.

Finally, today's email brought a "Published Ahead of Print" article from Journal of Thoracic Oncology on-- mediastinal incidentalomas.  This is a prospective study of 88 patients in which mediastinal lymphadenopathy was found as an incidental finding on CT scans for indications other than staging.  The title is a bit deceptive but the scenario is common.  The finding of incidental mediastinal lymph nodes on CT were characterized by multiple Naruke stations involved, relative small sizes of nodes (median ranging from 6 to 14 mm for all nodes), and coexistence with hilar adenopathy.  In this study, sarcoidosis was diagnosed by EBUS-FNA in 22% of patients and 2 patients developed lung cancer 2 years after initial finding.

While I concur with the basic premise of the post (i.e. that one shouldn't jump into a full-blown diagnostic work-up of an incidental finding), I think we should also insist on developing evidence-based criteria for managing these types of findings such as, Fleishner criteria for lung nodules.

Systems biology for studying cancer

This month's issue of Future Oncology has a well-written editorial from Dr. John McDonald from Georgia Tech regarding systems biology in cancer research.  This is a free download and I highly recommend it as a nice overview contrasting the traditional "reductionist" approach to cancer cell biology research versus the emerging application of a systems biology approach.

First, it is appreciated that the traditional approach has grudgingly yielded some valuable insight into cancer cell biology which has led to the development of "rationally designed" targeted therapies.  But while many of these therapies have had some initial impact for patients enrolled in clinical trials (and even some dramatic responses), I think it is fair to say that they also have not radically shifted overall survival or long-term outlook of the most common lethal cancers.

However, we are beginning to realize the potential for combining high-throughput molecular technologies with computational strategies and capabilities for combining multidimensional datasets to find inferred network relationships.  Dr. McDonald refers to this as the "top-down" approach.  He contrasts this with the traditional reductionist "bottom-up" approach which links human oncogenes or tumor suppressor genes with known interacting genes previously characterized in experimental models using cancer cell lines or animal models to identify signaling pathways implicated with cancer.  He stresses that these approaches are by no means mutally exclusive but indeed are complementary.

The systems approach starts with establishing a list of component parts.  Using new technologies, we can now speak of the genome, transcriptome, proteome, metabolome, etc.  The revolutionary step is using sophisticated computer models to detemine how these various components interact and which of these interactions are relevant for cancer initiation and progression and which represent potential druggable targets.

Again, this is a nice appetizer for additional reading on this exciting area!

DDR2 Kinase Mutations in Lung Squamous Cell Carcinoma

Yes--you read that right!

The journal Cancer Discovery reports in a publish ahead-of-print article (abstract)the discovery of mutations in the discoidin domain receptor 2 tyrosine kinase (DDR2) gene in a series of lung squamous cell carcinomas (SQC) that may identify a  potential therapeutic target similar to those described for lung adenocarcinomas.  Interestingly, DDR2 is a receptor kinase that normally binds collagen as its ligand and has been shown to promote cell migration, proliferation, and survival when activated by ligand binding and phosphorylation.

While subtypes of lung adenocarcinoma (ADC) have been increasingly better defined and characterized on multiple levels with respect to identifying therapeutic targets/responses, minimal progress has been made with squamous cell carcinoma.

The authors of this study used conventional Sanger sequencing of 201 genes, including the entire tyrosine kinome, in an initial set of 20 primary lung SQC and matched normal controls and identified 2 samples with DDR2 mutations; a secondary screen of 48 lung SQC samples identified as additional 4 DDR2 mutations.  Following this, the authors sequenced DDR2 in a validation cohort of 222 primary lung SQC samples and identified 5 additional samples with mutations.  Excluding samples from lung SQC cell lines, there was an overall frequency of 3.2% of DDR2 mutations in primary lung SQC samples--about the same frequency reported for ALK fusion mutations in lung ADC.

The authors analyzed the TKIs imatinib and dasatinib, previously reported to inhibit DDR2, to evaluate whether targeting DDR2 mutated in SQC might be a potential therapeutic strategy.  Dasatinib showed particular efficacy in inhibiting proliferation and promoting cell death through apoptosis in SQC cell ines with DDR2 mutations.  In addition, they report one of 7 subjects from an early-phase trial of dasatinib or combination dasatinib and erlotinib with advanced stage lung SQC cancer who exhibited significant shrinkage in tumor size while receiving the dasatinib/erlotinib combination.  The authors were able to perform direct sequencing of DDR2 in a pretreatment tumor specimen from this individual and identified a novel DDR2 kinase mutation.

This is an intriguing paper that you might well keep in mind.  Hopefully, this may stimulate other investigators to not only look for other potential genomic alterations that either drive SQC and/or represent targetable proteins but also go back and scrutinize histological and other clinicopathological characteristics of lung SQC that may have therapeutic or prognostic importance.