Targeting genetic markers in premalignancy is an emerging concept. In speaking at PMO Live 2015, Scott M. Lippman, MD, said that genetic drivers can identify premalignant conditions and even certain benign conditions, and genetic drivers can aid in identifying higher-risk populations and populations most likely to respond to targeted agents.
“Finding genetic drivers early identifies potential new targets for intervention in earlier phases of carcinogenesis, such as high-risk premalignancy,” said Lippman, Director, University of California, San Diego (UCSD), Moores Cancer Center.
Evidence supports the existence of the same genetic drivers in premalignancy that occur in advanced disease and metastasis. The question then becomes whether these important drivers can also be targeted early with intervention. “It would seem that that would be an effective way, maybe more effective, than targeting advanced disease,” he said.
Mutations in Lung Adenocarcinomas
Combinations of genetic drivers involving oncogenes and tumor suppressor genes can lead to cancer. The earliest targetable genetic driver of lung adenocarcinoma, the epidermal growth factor receptor (EGFR), was discovered and reported in 2004. “A number of other genetic drivers have been discovered such that even in a disease like lung cancer, it is becoming
a large group of rare diseases, and of course, that’s
the concept of precision medicine,” said Lippman.
Patients who have a sensitizing mutation do much better with EGFR inhibitors/tyrosine kinase inhibitors than with chemotherapy. The hazard ratio (HR) for progression-free survival with treatment with erlotinib was 0.16 compared with gemcitabine plus carboplatin in patients with lung adenocarcinomas with EGFR mutations (Lancet Oncol. 2011;12:735-742).
In patients with advanced lung cancer and known EGFR mutation, the same gradient effect of the same EGFR-activating mutation was observed in the normal epithelium within the tumor or surrounding the tumor (Cancer Res. 2005;65:7568-7572). The EGFR mutation is absent in the normal epithelium in patients without EGFR-mutant tumors, “which suggests that in this clonal path of normal cells that acquire EGFR mutation early, the tumor will arise,” he said.
A number of potential driver mutations in lung adenocarcinoma and premalignancy/adenomatous hyperplasia overlap, “but what’s fascinating is there are certain mutations that seem to occur only in the premalignancy, and then you don’t find them in the cancer,” said Lippman. “So they may be important early, at initiation, and then are lost.”
The genetic driver, 3q26, that has been identified in lung squamous cell carcinoma also drives lung premalignancy. In airway basal cells, evidence of overexpression of 3q26 was found from premalignancy to tumor development in squamous cell carcinoma cases (Cancer Prev Res [Phila]. 2014;7:487-495).
Aberrations in Benign Disease
Aberrations in the expression of BRAF, HER2, FGFR3, PIK3CA, TP53, and NOTCH1 have all been documented in both benign/premalignant conditions and in malignant conditions.
“The same BRAF mutation seen in advanced melanoma is seen in dysplastic nevi, and actually benign nevi, at a very similar rate,” he said. “HER2 is overexpressed to a greater degree in ductal carcinoma in situ premalignancy than in basic breast cancer. FGFR mutations, which can occur in cancers, can be found in seborrheic keratosis. TP53 mutations are broadly found in many cancers and have actually been reported in synovial tissue in rheumatoid arthritis.”
The Story in Head and Neck
A series of randomized clinical trials of retinoic acid (isotretinoin) was performed in patients with oral premalignant lesions. There was no difference in cancer-free survival in patients despite differences in the response rates between treated and untreated patients, which suggests that unselected use of a targeted therapy in an unselected population is ineffective as chemoprevention.
One of the first examples of genetic driver detection in a premalignant setting was the discovery of cyclin D1 expression/gene amplification in high-risk premalignant head and neck lesions. Cyclin D1 protein expression is resistant to ubiquitization by various drugs, including retinoids, in individuals who carry the A allele.
Investigators at the MD Anderson Cancer Center found that the cyclin D1 A allele was associated with increased chromosome instability, decreased favorable modulation of cyclin D1 expression, and less likelihood of response to biochemopreventive intervention in patients with preinvasive upper aerodigestive tract cancer compared with the G allele (J Natl Cancer Inst. 2003;95:198-205).
“Individuals that had the GG genotype that was sensitive to retinoid-induced proteolysis did well [ie, a lower cancer rate], while those that were resistant and had no downregulation of cyclin D1 had a higher cancer rate,” said Lippman. “That’s evidence of a targeted genetic driver. It seems that if the genetic driver is sensitive, like in the GG genotype, that would be beneficial.”
Another genetic driver in oral premalignancy is loss of heterozygosity (LOH) profiles, or allelic imbalance. Individuals with LOH at certain sites of tumor suppressor genes have a higher rate of cancer development than those who have retention (Cancer Prev Res [Phila]. 2012;5:1081-1089).
In the National Cancer Institute–funded EPOC trial, nearly 400 patients with oral intraepithelial neoplasia were selected and screened for LOH. Ultimately, 150 patients were randomized to erlotinib 150 mg/day or placebo for 1 year, and then followed for an additional 2.5 years, for a median follow-up of 35.4 months. During the intervention, there was a trend toward improved survival in the treated patients that did not achieve significance, but this trend was lost when the study drug was terminated.
In an exploratory finding, patients who developed a rash within the first month of erlotinib therapy had a much lower rate of developing cancer than those who did not experience the rash (HR, 0.28; P = .009). “This has been observed in a number of therapy trials in lung and head and neck,” said Lippman. “We’re not sure what the mechanism is, but it does have implications, potentially for future designs.”
In 2014, The Cancer Genome Atlas (TCGA) group identified candidate therapeutic targets and driver oncogenic events in head and neck squamous cell carcinomas, from which the “two-hit” event of TP53 mutations and loss of chromosome 3p was found to confer a markedly worse prognosis than either event alone or neither event (Nat Genet. 2014;46:939-943). Research at UCSD then discovered that in the subgroup of patients positive for human papilloma virus (HPV), having the 3p LOH alone worsened survival. “As we know, HPV inactivates p53; a finding relevant to premalignancy,” he said.
Genes frequently mutated in head and neck cancer are NOTCH1 and PIK3CA, the TCGA found. Within the EPOC trial, PIK3CA was the important mutation in the premalignancy setting in head and neck, said Lippman, and may therefore represent a genetic driver to target early.
The NOTCH1 mutation is an inactivating mutation in Caucasians, making it difficult to target. In China, “the same premalignant lesion is associated with a high rate of NOTCH1 mutations, but they’re oncogenic…they’re activating,” he said. “This changes everything…they could be targetable with NOTCH1 inhibitors [used as preventive agents].”
