Objective. Epidermal growth factor receptor (EGFR) and K-ras mutations guide treatment selection in nonsmall cell lung cancer (NSCLC) patients. Although mutation status is routinely assessed in biopsies, cytological specimens are frequently the only samples available. We determined EGFR and K-ras mutations in cytological samples.Methods. DNA was extracted from 150 consecutive samples, including 120 Papanicolau smears (80%), 10 cell blocks (7%), nine fresh samples (6%), six ThinPrep tests (4%), and five body cavity fluids (3.3%). Papanicolau smears were analyzed when they had >50% malignant cells. Polymerase chain reaction and direct sequencing of exons 18 -21 of EGFR and exon 2 of K-ras were performed. EGFR mutations were simultaneously determined in biopsies and cytological samples from 20 patients. Activity of EGFR tyrosine kinase inhibitors (TKIs) was assessed.Results. The cytological diagnosis was adenocarcinoma in 110 samples (73%) and nonadenocarcinoma in 40 (27%) samples. EGFR mutations were identified in 26 samples (17%) and K-ras mutations were identified in 18 (12%) samples. EGFR and K-ras mutations were mutually exclusive. In EGFR-mutated cases, DNA was obtained from stained smears in 24 cases (92%), pleural fluid in one case (4%), and cell block in one case (4%). The response rate to EGFR TKIs in patients harboring mutations was 75%. The mutation status was identical in patients who had both biopsies and cytological samples analyzed.Conclusion. Assessment of EGFR and K-ras mutations in cytological samples is feasible and comparable with biopsy results, making individualized treatment selection possible for NSCLC patients from whom tumor biopsies are not available. The Oncologist 2011;16: 877-885

Thirty consecutive surgical patients with glioblastoma, were operated upon using fluorescence induced by 5-aminolevulinic acid as guidance. The fluorescent quality of the tissue was used to take biopsies from the tumor center, from the invasive area around it and from adjacent normal-looking tissue. These samples were analyzed with HE, Ki-67 and nestin. Nestin expression in tissue surrounding glioblastoma cases was compared to tissue surrounding vascular lesions, metastasis and hippocampal sclerosis. The rate of gross total resection assessed by volumetric MRI was 83%. Using HE examination as the gold standard, fluorescence identified solid tumor with 100% positive predictive value, invasive areas with 97%, and normal tissue with 67% negative predictive value. Ki67 stained some cells in 69% of the non-fluorescent samples around the tumor. There was always strong nestin expression around the tumor but it was similar to control cases in non-glioma lesions with subacute expansion. 5-aminolevulinic acid fluorescence guidance is very reliable and can help to study the tumor-brain interface. Nestin expression is strong and constant in the tissue around the tumor, but is mostly an acute glial reaction, not specific of the neoplasm. Nestin staining is not recommended as a tumor stem cell marker.

- Molecular testing can be performed on cytologic specimens, especially on direct smears. Rapid on-site evaluation by cytopathologists has improved the adequacy and the management of cytology samples for molecular testing. Mutational profiling of NSCLC using next-generation sequencing can be performed on cytology samples from very small amounts of DNA. Fluorescence in situ hybridization assays on cytology specimens, including stained direct smear, offer some distinct advantages over their histologic counterpart, and are used to detect ALK and ROS1 rearrangements in NSCLC. Cytology specimens allow assessment of the entire tumor cell nucleus, avoiding signal loss from truncation artifacts. The use of cytology samples for assessing programmed death ligand-1 protein expression is currently being developed. Protocols for bisulfite conversion and DNA droplet digital polymerase chain reaction assays have been optimized for cytology smear to investigate aberrant DNA methylation of several NSCLC-related genes.

Sunitinib is one of the most widely used targeted therapeutics for renal cell carcinoma (RCC), but acquired resistance against targeted therapies remains a major clinical challenge. To dissect mechanisms of acquired resistance and unravel reliable predictive biomarkers for sunitinib in RCC, we sequenced the exons of 409 tumor‐suppressor genes and oncogenes in paired tumor samples from an RCC patient, obtained at baseline and after development of acquired resistance to sunitinib. From newly arising mutations, we selected, using in silico prediction models, six predicted to be deleterious, located in G6PD, LRP1B, SETD2, TET2, SYNE1, and DCC. Consistently, immunoblotting analysis of lysates derived from sunitinib‐desensitized RCC cells and their parental counterparts showed marked differences in the levels and expression pattern of the proteins encoded by these genes. Our further analysis demonstrates essential roles for these proteins in mediating sunitinib cytotoxicity and shows that their loss of function renders tumor cells resistant to sunitinib in vitro and in vivo. Finally, sunitinib resistance induced by continuous exposure or by inhibition of the six proteins was overcome by treatment with cabozantinib or a low‐dose combination of lenvatinib and everolimus. Collectively, our results unravel novel markers of acquired resistance to sunitinib and clinically relevant approaches for overcoming this resistance in RCC.