Improving the yield of circulating tumour cells facilitates molecular characterisation and recognition (original) (raw)

Improving the yield of circulating tumour cells facilitates molecular characterisation and recognition of discordant HER2 amplification in breast cancer.

Author(s): Flores LM, Kindelberger DW, Ligon AH, Capelletti M, Fiorentino M, Loda M, Cibas ES, Jänne PA, Krop IE

Publication: Br J Cancer, 2010, Vol. 102, Page 1495-502

PubMed ID: 20461092PubMed Review Paper? No

Purpose of Paper

Two methods of circulating tumor cell (CTC) isolation, the CellSearch Epithelial kit (CEK) and the CellSearch Profile kit (CPK), were compared in terms of CTC recovery, purity, and molecular characterization in clinical specimens from patients diagnosed with metastatic cancer as well as peripheral blood specimens spiked with cancer cell lines. The paper also examined whether CTC recovery is compromised by the method of cell counting (manual, semi-automatic), the type of collection tube used, or a delay to processing, and whether CTCs isolated using the CPK method are viable and suitable for analysis by in situ hybridization and immunofluorescence.

Conclusion of Paper

CTC isolation using the CPK method resulted in significantly higher median cell counts and superior purity than the CEK method when blood from patients diagnosed with metastatic breast cancer or NSCLC were analyzed. Differences in cell yield between the CEK and CPK methods were not a result of manual versus semi-automatic cell counting. A higher percentage of apoptotic cells were isolated using the CPK method in comparison to the CEK method, when specimens were pretreated with apoptosis-inducing agents, suggesting that the method of CTC isolation adversely affects the recovery of more fragile cell types. Further, the CPK method could reproducibly isolate pure, viable CTCs from EDTA blood specimens after a processing delay at room temperature of up to 72 hrs. Detectable levels of total and phosphorylated HER2 and EGFR was also stable among blood specimens collected in CellSave tubes that experienced a processing delay at room temperature for up to 144 h.

  1. Study Purpose

This study compared the sensitivity of two methods for the enumeration of circulating tumor cells (CTCs) in peripheral blood from 71 patients with metastatic NSCLC and 75 patients with metastatic breast cancer and assessed the purity of CTCs isolated by the CPK method. Blood specimens were collected into CellSave blood collection tubes and stored at room temperature until processed (≤ 72 h). For the CEK method, specimens were processed with the CellSearch system according to the manufacturer’s instructions and analyzed by semi-automated counting of labelled CTCs. For the CPK method, specimens were collected in a dilution buffer and immediately prepared as cytospin preparations by centrifugation at 500 x g for 5 min with a cytology filter. Slides were fixed with methanol for analysis by immunofluorescence (cytokeratin, CD45, DAPI, HER2, Phospho-HER2, EGFR, and Phospho-EGFR) or fixed with methanol:acetic acid (3:1) for fluorescent in situ hybridization (FISH) analysis of cytokeratin, CD45, DAPI, MET, EGFR, pEGFR, HER2, and pHER and manual counting using a standard fluorescence microscope.

Summary of Findings:

Median cell counts from the CPK method were significantly higher than those obtained using the CEK method for specimens from breast cancer patients (116.5/7.5 ml of blood compared to 4/7.5 ml of blood, 29-fold difference, P<0.0001) and patients with NSCLC (145/7.5 ml of blood compared to 4/7.5 ml of blood, 36-fold difference, P<0.0001). Lower yields of cells isolated by CEK were not due to method of cell counting (manual vs semi-automatic) as differences in median cell counts were not statistically significant between cells counted using the semi-automated CEK method and those counted manually (4 CTCs versus 5 CTCs/7.5 ml of blood) in specimens from breast cancer and NSCLC patients. Further, a comparison between cells counts on a per patient basis revealed a strong correlation between those obtained using the semi-automated CEK method and the manual CEK method in breast cancer specimens (R2=0.9930) and lung cancer specimens (R2=0.7200).
The purity of cells isolated by the CPK method from 30 NSCLC and 30 breast cancer patients revealed that 60% of cells were identified as tumor cells (DAPI+, cytokeratin+, CD45-; range: 21–95%), 6% were white blood cells (DAPI+, cytokeratin-, CD45+; range: 1–22%), and 32% could not be classified as either tumor cells or white blood cells (DAPI+ only, range: 5–68%) which corresponds to a median absolute number of cytokeratin positive cells of 280 per specimen (range 20–20,000) and a median of 34 (range 20–476) CD45 positive cells per specimen. Fluorescence-activated cell sorting on eight NSCLC and three breast cancer patients revealed similar results for percentage of cytokeratin positive cells (61%; range 35–77%) but a lower percentage of CD45 positive cells (1%; range 0–7%). FISH analysis of the CPK-isolated CTCs from 24 patients with HER2 positive breast cancer resulted in 72% of cells with a HER2 gene copy number ≥4 and only 3% of the cells had two copies of HER2. In contrast, only 5% of CPK-isolated cells from the six HER2-negative breast cancer patients had ≥4 copies of HER2 and 95% of cells had two copies of HER2. Similarly, 53% of blood specimens from 30 NSCLC patients had ≥4 copies of EGFR, 13% had ≥4 copies CEP 7, and only 5% had two copies of EGFR and CEP 7. Analysis of peripheral blood specimens from 40 healthy volunteers with either the CEK semi-automated enumeration or with the CPK manual count resulted in no CTCs (CD45 negative/cytokeratin positive) and a median of five CD45 positive/cytokeratin negative cells (indicative of white blood cells) per 7.5 ml of blood (range: 0–50) with only two gene copies of CEP7/EGFR or HER2 as detected by FISH analysis were isolated by the CPK method. 2. ###### Study Purpose
This study investigated the reproducibility and effects of a pre-processing delay at room temperature and effects of tube type on the number of CTCs isolated from peripheral blood using the CPK method. Blood specimens from seven patients with metastatic NSCLC were collected into CellSave blood collection tubes and EDTA tubes, stored at room temperature for up to 144 h before processing by the CPK method. Slides were prepared as cytospin preparations by centrifugation at 500 x g for 5 min with a cytology filter and fixed with methanol for analysis by immunofluorescence (cytokeratin, CD45, and DAPI).

Summary of Findings:

Peripheral blood specimens from seven NSCLC patients processed in triplicate by the CPK method resulted in a similar number of isolated cells with a low coefficient of variability (9.7%). When blood specimens were collected from NSCLC patients, CellSave and EDTA tubes that were stored at room temperature for 24–72 h before processing by the CPK method did not differ from one another in cell count (CV 9.1% for CellSave tubes and CV 6.5% for EDTA tubes). Cell count was not significantly altered when blood specimens collected in CellSave tubes were subjected to a pre-processing delay at room temperature for up to 144 h; however, cell count did decline significantly when blood specimens collected in EDTA tubes were subjected to a pre-processing delay of ≥96 h. 3. ###### Study Purpose
This study investigated whether CTC isolation methods (CEK and CPK) were biased in their recovery of proliferating, non-proliferating, and apoptotic cells from peripheral blood specimens. Peripheral blood from healthy volunteers was spiked with cells from the breast cancer cell line SK-BR-3 or NSCLC cell line HCC827, then pretreated with apoptosis-inducers (EGFR/HER2 inhibitor lapatinib for SKBR3 cells and EGFR inhibitor gefitinib for HCC827) or left untreated as a control. Specimens processed using the CEK method according to the manufacturer’s instructions were analyzed by semi-automated counting of labelled CTCs. Specimens processed using the CPK method were collected in a dilution buffer and immediately prepared as cytospin preparations by centrifugation at 500 x g for 5 min with a cytology filter. Slides were fixed with methanol for analysis by immunohistochemistry (IHC) or Apoptosis Detection Using Terminal Transferase and Biotin-16-dUTP (TUNEL).

Summary of Findings:

CTC isolation methods differed from one another in the recovery of apoptotic and proliferating cells from peripheral blood specimens. While the number of apoptotic cells (approximately 30-70%) was highest when cell lines were treated with an apoptosis inducer and analyzed without processing (smeared directly on the slide), when cells from cell lines were spiked into peripheral blood and recovered by CTC isolation, a higher percentage of apoptotic cells were recovered with the CPK method than the CEK method, whether treated with an apoptosis inducer (approximately 10% vs. 0%) or not treated (approximately 5% vs. 0%). A small number of apoptotic cells was also recoverable from seven peripheral blood specimens collected from NSCLC patients with the CPK method (0-6, 4%) while no apoptotic cells were recoverable with the CEK method (0%).The authors conclude that recovery of apoptotic and non-proliferating cells is disproportionately lower in processed samples, particularly with the CEK method, due to their fragile state. 4. ###### Study Purpose
This study investigated if immunofluorescence and FISH are possible with CPK-isolated CTCs, and whether cancer biomarker results using isolated CTCs are representative of the primary tumor or metastasis in patients diagnosed with metastatic breast cancer or NSCLC. Membrane expression of total and phosphorylated HER2 and EGFR was evaluated by immunofluorescence in cells isolated using the CPK method and HER2 status was assessed by FISH in the same cells. Peripheral blood specimens from seven NSCLC patients, 20 HER2-positive breast cancer patients, and from healthy volunteers (spiked with cells from the breast cancer cell line SK-BR-3 or NSCLC cell line HCC827) were processed with the CPK method. Briefly, slides were prepared as cytospin preparations by centrifugation at 500 x g for 5 min with a cytology filter and fixed with methanol for analysis by immunofluorescence (HER2, Phospho-HER2, EGFR, Phospho-EGFR, and MET). FISH results were compared to those obtained in tissue biopsies from the primary tumor or a metastasis in the same cohort of patients diagnosed with breast cancer.

Summary of Findings:

While immunoflourescent analysis detected EGFR and pEGFR membrane expression on CTCs isolated by the CPK method in all specimens from NSCLC patients and EGFR, pEGFR, HER2, and pHER2 expression on CTCs from HER2-positive breast cancer specimens, the percentage of stained cells was lower than what was observed when spiked cells from a cancer cell line were recovered by the same processing method from healthy peripheral blood specimens.
The HER2 status of CTCs isolated from peripheral blood of patients diagnosed with HER2 positive metastatic breast cancer were strongly concordant with the biopsy of the primary tumor and metastasis (95% of patients). However, concordance among all three specimen types was only 66% in HER2 negative breast cancer patients. Of the 10 patients that presented discordant results, the primary tumor was HER2 negative while both the metastasis and CTC status was positive for 9 patients. 5. ###### Study Purpose
This study evaluated the viability of CTCs collected in EDTA tubes and isolated using the CPK method. Peripheral blood from healthy volunteers collected in EDTA tubes was spiked with a range (5–1000) of SKBR3 or HCC827 cells, processed using the CPK method, and then grown in culture on laminin-coated plates.

Summary of Findings:

Viable SKBR3 and HCC827 cells were isolated from EDTA blood spiked with a cancer cell line. Isolated cancer cells successfully propagated in vitro with a minimum of 50 cells and optimal growth occurred in the presence of enriched medium and a cell adhesion matrix (data not provided).

You Recently Viewed

News and Announcements

More...