Molecular markers for targeted therapy in lung cancer: USCAP 2010

I attended the Special Course at USCAP 2010, "Basic Principles and Practice of Molecular Pathology in Cancer."  One of the talks was given by Dr. Marc Ladanyi from MSKCC and one of the leaders in the field of molecular pathology of lung cancer.  The title of his presentation was "Molecular Predictive Markers for Targeted Therapies in Lung Cancer."  Here's my "take-home" points from the talk:

• Term: "lung adenocarcinoma oncogenome"--(I like this) and he presented a pie chart illustrating the different (and (generally) mutually exclusive) mutations found so far in lung ACa
• •  KRAS
  • EGFR
  • BRAF
  • ERBB2 (HER2)
  • ALK fusions (especially, ALK-EML4)
  • MEK1
  • NF1
  • but still about one-third of lung ACa have an unknown mutation
• Although there are morphologic and clinical features associated with EGFR mutations, their predictive value are insufficient to exclude EGFR mutation testing.  Thus, every adenocarcinoma should be tested for EGFR mutation (with the exception of mucinous ACa/BAC which are always KRAS mutants).  Although uncommon or even rare, one does find EGFR-mutant ACa in former and current smokers and in men who could benefit from targeted therapy.  This is important for testing algorithms (see below).
• EGFR mutation versus EGFR amplication is difficult to tease out because they frequently occur together but--
• • The response rate to EGFR-TKIs is high in EGFR-mutated cases regardless of the EGFR copy number.
  • EGFR amplified cases lacking one of the mutations have a very low response rate to EGFR-TKIs.
• Why are some patients with EGFR-mutant tumors non-responders to EGFR-TKIs?
• •  Not all EGFR mutations confer sensitivity (dependent on type of mutation detection method, i.e., direct sequencing versus targeted approach to known sensitive mutations)
  • There may be concurrent "downstream" mutations, e.g. BRAF, PTEN loss, PIK3CA, etc.)
  • Not a well-studied group.
• Why do some patients with EGFR-wild type tumors respond to EGFR-TKIs?
• • Undetected sensitive mutations due to targeted detection method.
  • Poor sensitivity of direct sequencing methods
  • Not a well-studied group.
• KRAS mutations function better as a negative predictor of response to EGFR-TKIs than EGFR mutations do as a positive predictor.
• Dr. Ladanyi presented data from a MSKCC study of patients with EGFR-mutant tumors who initially responded to EGFR-TKI therapy but relapsed while on therapy and underwent re-biopsy.  Two-thirds of these patients developed acquired resistance due to the development of the EGFR T790M mutation; some also were shown to have MET amplification.
• BRAF mutations (most commonly exon 15 V600E) is also a negative predictor of response to EGFR-TKIs but does predict benefit from MEK-selective inhibition: PLX-4032, an experimental agent from Plexxikon which targets mutated BRAF (currently in trials for melanoma in which >50% of tumors show the V600E mutation).
• Very helpful point in discerning which type of EGFR/KRAS mutation detection testing will be serve your patients: Dr. Ladanyi distinguished between "diagnostic sensitivity" and "technical sensitivity."
• • Diagnostic sensitivity (the "comprehensiveness" of an assay): the percentage of all mutations that will be found with a given assay
  • Technical sensitivity: the percentage of tumor cells that need to be present in the specimen in order for a mutation being tested for to be identified
  • There is a trade-off between low technical sensitivity and the risk of false negatives and increasing comprehensiveness or high diagnostic sensitivity and the risk of false positives due to detection of rare mutations of unknown or no clinical significance.
• There are mutation-specific rabbit MoAbs that are commercially available for the detection of mutant EGFR in lung ACa and these could facilitate detection in an algorithm and perhaps eliminate the need for mutation detection testing in certain situations: Antibody 3197 against EGFR exon 21 L858R mutation and Antibody 2085 against the EGFR exon 19 E746-A750del--both from Cell Signaling Technologies.
• Dr. Ladanyi showed 95% sensitivity and 99% specificity for EGFR L858R Ab and 85% sensitivity and 99% specificity for EGFR exon 19 del Ab. (paper in press)
• EML4-ALK detection issues: RT-PCR requires multiple assays because of multiple possible isoforms, other ALK and other kinase fusion partners have been reported, new IHC antibody clone to detect ALK fusion (rabbit mAb #3633 D5F3, Cell Signaling Technologies) may be useful in screening for EML4-ALK detection.
• Dr. Ladanyi presented the testing algorithm used at MSKCC and contrasted with one published recently in J Clin Oncol 2009;27:4232-4235.
• A future post will include my merged algorithm, combining these and incorporating new IHC.