Archives

  • 2019-10
  • 2019-11
  • 2020-03
  • 2020-07
  • 2020-08
  • ref br cMET cellular Mesenchymal Epithelial Transition facto

    2020-08-28

    ref.
    cMET: cellular Mesenchymal Epithelial Transition factor, ALK: Anaplastic Lymphoma Kinase, LCLC: Large-Cell Lung Cancer, NOS: Not Otherwise Specified, HR:
    Hazard Ratio, Ref: Reference, CI: Confidence Interval.
    4. Discussion
    The role of c-MET protein overexpression as a prognostic and pre-dictive biomarker in early stage NSCLC was the primary objective of our study. It is still unclear how c-MET protein expression, MET am-plification and MET exon 14 mutation correlate with each other. The Cancer Genome Atlas study of lung adenocarcinoma and a Japanese study, both done in surgically resected tumors (early stage disease), showed that MET exon 14 mutation and MET amplification are mu-tually exclusive [19,20]. A recently published study with 28 MET exon 14 mutated cases showed a statistically significant co-existence of MET exon 14 skipping mutation and both MET gene amplification/c-MET protein overexpression in stage IV NSCLC [9]. In this study there was a significant association between MET amplification and stage IV disease, something which was not verified in a larger cohort with 298 MET exon 14 skipped patients [15]. Dziadziuszko et al showed that MET gene copy number determined by silver in situ hybridization (SISH) and protein expression evaluated by IHC correlated significantly in a cohort of 189 patients with surgically resected NSCLC [21]. This finding could not be confirmed in another surgical cohort (n = 222), where no sig-nificant association was observed between c-MET expression and gene copy number, but these two studies used different IHC scoring systems [8]. Bubendorf et al found a statistically significant correlation between MET gene amplification and c-MET protein overexpression in a large cohort of surgically resected NSCLC patients [6].
    In the present surgical cohort, we aimed to evaluate c-MET as a prognostic and/or predictive biomarker and identify a c-MET H-score cut-off value with a prognostic or predictive clinical utility. The uni-variate analysis done in the whole study population failed to show any significant benefit for any of the c-MET quartiles, but this could be due to the crossover of the Kaplan-Meier curves after some years of follow-up. Because of this we used specific Cox regression analyses, taking time-dependent co-variates into consideration. These multivariate models showed a positive independent prognostic value for c-MET H-score ≥20 and for higher c-MET H-score, when c-MET was included as a continuous variable, which is in MG-132 with existing trial results [3–5,10], albeit in concordance with the Asian study discussed in the Introduction [2]. This generated the hypothesis that there may be an effect-modifying influence of adjuvant therapy in our models. The fact that the HR of c-MET H-score ≥20 was lower in the multivariate analysis done in the subgroup of patients who received adjuvant therapy, compared to the multivariate analysis including adjuvant treatment as a co-variate strengthened the hypothesis that c-MET H-score ≥20 may indicate a predictive value for patients receiving ad-juvant chemotherapy. This finding is strengthened by the Korean study of Kim et al [2] which showed that adjuvant chemotherapy was a po-sitive independent prognostic factor in c-MET positive but not in c-MET negative patients. In our analyses, it was not meaningful to use a con-trol group with patients not receiving adjuvant therapy. It was due to comorbidity that some patients did not receive adjuvant therapy, though indicated. The OS in this group of patients is strongly influenced by comorbidity, rendering it a suboptimal control group in order to test the effect modifying hypothesis. There is no indication from the lit-erature which cut-off value of c-MET best reflects the tumour biology and we have chosen the median value which is a usual cut-off used for different biomarkers in IHC studies. Furthermore, we tested different cut-off values with no further impact on prognosis. It is difficult to correlate expression level of a biomarker with a biological effect and it is possible that even low levels of c-MET may exert a biological effect.
    An optimal cut-off value for c-MET protein expression with clinical impact on prognosis or prediction has not yet been defined. In our study we analysed different quartile cut-off values using the H-scoring system from 0 to 300 and identified H-score 20, being the 50-quartile, to be a clinically significant cut-off value. We believe that the H-scoring system is a good method taking into account both the intensity and extent of protein expression. Other studies have identified different clinically  Lung Cancer 133 (2019) 69–74
    significant cut-off values, including H-score 60 in stage IV NSCLC pa-tients [22] and ≥50% of cells with moderate or strong staining in the MetMAb trial [14]. These differences may be due to several factors that differ between the studies, including the use of different antibodies, scoring methods, disease stage and treatments. Regarding different antibodies, patient selection through immunohistochemical evaluation of c-MET expression in formalin-fixed and paraffin-embedded (FFPE) samples is problematic, because only a limited number of validated antibodies to c-MET that work in FFPE are available. Most studies in different types of cancer demonstrating a correlation between Met and disease progression have used antibodies directed against the COOH-terminus of Met. However, in one study on lymph node negative breast carcinoma, antibodies against the intracellular domain of Met were predictive of worse outcome whereas antibodies against the extra-cellular domain were not [23]. Furthermore, the cellular distribution of Met may vary and influence the IHC results; it has been shown that Met may be expressed not only in the cytoplasmic and membranous com-partments but also in the nucleus, especially in the invasive front of tumors [24]. A harmonization between different studies would be needed to decide on an optimal cut-off value.