Virtual Library
Start Your Search
Steffen Dietz
Author of
-
+
P3.02 - Biology/Pathology (ID 620)
- Event: WCLC 2017
- Type: Poster Session with Presenters Present
- Track: Biology/Pathology
- Presentations: 1
- Moderators:
- Coordinates: 10/18/2017, 09:30 - 16:00, Exhibit Hall (Hall B + C)
-
+
P3.02-032 - Spatial Heterogeneity of EGFR and KRAS Variant Allele Frequencies Correlates with Histological Patterns of Lung Adenocarcinomas (ID 8401)
09:30 - 09:30 | Presenting Author(s): Steffen Dietz
- Abstract
Background:
Analysis of spatially distinct regions revealed a considerable histological and molecular heterogeneity of lung adenocarcinomas (ADCs). The predominant histological growth pattern and driver gene alterations (e.g. in EGFR and KRAS) have been shown to be of high prognostic and therapeutic relevance. This project aimed to examine the spatial distributions of EGFR and KRAS mutation frequencies in the various histological growth patterns of ADCs.
Method:
Central tumor sections from 19 ADCs were subdivided into 467 tumor segments of 5x5 mm. The segments were evaluated separately in order to enable a systematic analysis of histological and molecular markers within and between patients. We determined the predominant histological growth patterns and the variant allele frequencies (VAFs) of EGFR and KRAS in each segment by digital PCR. We further quantified the absolute cell counts and proportions of tumor and non-neoplastic cells in all segments in order to examine the cellular fractions and allow a precise normalization of VAFs.
Result:
Histopathological classification revealed morphological intratumor heterogeneity with more than one histological growth pattern in 16 of the 19 cases. The 467 malignant segments exhibited a mean tumor cell fraction of 28% with an extensive variability within and between the individual cases. The predominant solid pattern revealed the significantly lowest fraction of tumor cells. Furthermore, EGFR and KRAS VAFs were measured by digital PCR and normalized to the cellular fractions of the respective segment. While driver gene mutations were detected in > 99% of malignant segments, we found a heterogeneous spatial distribution of normalized VAFs: Some cases showed ubiquitously low or high mutant allele frequencies, others revealed regions with focally elevated frequencies and driver gene amplifications. In addition, we found a significant correlation between mutated allele frequencies and histological patterns. Micropapillary and solid patterns harboured higher mutant allele frequencies compared to lepidic and papillary histologies. Correlation analysis did not show an association between the tumor size and the normalized VAF or driver gene amplification.
Conclusion:
Our findings indicate that driver gene mutations are present with high levels of inter- and intratumor heterogeneity throughout all ADC samples tested. Allele frequencies correlated with histological growth patterns, but not with tumor size.
