See related article on pages 344–53, this issue.
Standardizing preanalytical variables for molecular cytopathology
Article first published online: 27 MAR 2013
© 2013 American Cancer Society
Volume 121, Issue 7, pages 341–343, July 2013
How to Cite
da Cunha Santos, G. (2013), Standardizing preanalytical variables for molecular cytopathology. Cancer Cytopathology, 121: 341–343. doi: 10.1002/cncy.21290
- Issue published online: 12 JUL 2013
- Article first published online: 27 MAR 2013
- Manuscript Accepted: 22 FEB 2013
- Manuscript Received: 10 FEB 2013
Biomarker discovery studies that rely on the use of molecular assays for the selection of patients who may benefit from therapeutic agents that specifically target molecules/genes have emerged within the past few years. The development of new technologies for the detection of multiple alterations and advances in molecular technologies that require low DNA quantity have enabled the use of minimal material, including cytological samples. In patients with non-small cell lung cancer, many reports consisting of large series of cases have demonstrated the suitability of cytological specimens for the detection of epidermal growth factor receptor (EGFR) mutations and the selection of patients for treatment with EGFR tyrosine kinase inhibitors.[2-7] However, those studies have shown variations in the type of specimens and cytological preparations, types of fixation, and staining techniques as well as the molecular platforms used for the analysis. Although these variations might in fact provide a multitude of options for the use of cytological specimens, they also impose challenges in standardizing and optimizing preanalytical and analytical requirements. Reliable, consistent, robust, and accurate results from molecular tests using these samples will depend on standardized protocols for maximizing DNA yield and quality and preanalytical variables will have a direct impact on the analysis. Therefore, studies regarding understanding the effect of different preanalytical and analytical variables on molecular results are urgently needed.
As demonstrated in the study by Dejmek et al, reliable EGFR genotyping can be achieved with different cytological preparations, fixation techniques, mounting media, and staining. Using human lung carcinoma cell lines, the authors performed a series of comparisons of some preanalytical variables currently used in routine clinical cytology practice. To the best of my knowledge, this is one of the few systematic studies to compare different cytological preparations and fixation techniques with regard to their effect on DNA yield and quality. More importantly, the study by Dejmek et al also evaluated different mounting media and staining methods. Although the results were mainly obtained from cell lines, it has to be emphasized that the study was funded by a pharmaceutical company that produces one of the EGFR tyrosine kinase inhibitors, which may indicate the industry is possibly paying attention to the increasing use and value of cytological material for clinical studies.
The authors compared spray, ethanol, and liquid-based cytology and air-dried slides as variants for fixation techniques and found that spray or ethanol-fixed Papanicolaou-stained slides provided the best results in terms of yield and fragment length. However, another publication using archival smears from clinical samples reported better results with Romanowsky staining compared with Papanicolaou-stained slides, with superior DNA preservation and integrity. Dejmek et al acknowledged that washing the cells with phosphate-buffered saline (PBS) before producing the Romanowsky-stained air-dried smears, which is an artificial condition and not performed in routine clinical practice, might have introduced a potential bias in the study and therefore have limited their conclusions. Although they diligently limited the time of PBS exposure to minimize hydration of the cells and did not detect morphological changes with PBS treatment, it is still possible that it might have some effect on DNA integrity. Differences in staining and fixation (reagents) protocols for the same staining technique at different institutions also might have some impact on the results. It is interesting to note that the authors have commented that it cannot be concluded whether the poor results noted with air-dried May-Grunwald-Giemsa–stained specimens were the result of the fixation or the staining or even a combined effect of both. In addition, the trend toward the use of molecular-friendly, nonformalin fixatives with the possibility of improved results for molecular analysis coincides with the need for rapid fixation and processing of tissues and therefore further reinforces the use of cytological samples that are usually fixed immediately after collection with fixatives other than just formalin. The study from Dejmek et al clearly corroborates those aspects.
When comparing liquid-based cytology preparations, CytoLyt solution (Hologic, Inc, Bedford, Mass) provided a 5-fold higher DNA yield than CytoRich Red collection fluid (Thermo Fisher Scientific Inc, Waltham, Mass). The authors highlighted that both the specific fixative and the liquid-based preparation technique used might have affected DNA retrieval. Another study using clinical samples had already reported the successful use of DNA extracted from samples fixed in CytoLyt solution for EGFR mutation analysis. The poor results observed with CytoRich Red collection fluid were attributed to its formaldehyde content. In fact, the effects of formalin on DNA integrity due to chemical crosslinking of proteins and nucleic acids are well known.
A significantly higher DNA yield was obtained with slides mounted with EcoMount (BioCare Medical LLC, Concord, Calif) when compared with the xylene-based mounting medium Pertex (CellPath Ltd, Newtown, Powys, UK). As explained in the article by Dejmek et al, EcoMount is a low-hazard, organic, polymer-based mounting medium used as a substitute for xylene-based media and therefore is preferred for reasons of environmental and workplace safety.
Although the minimal amount of clinical data provided might be viewed as a limitation, a significant and important contribution of the study by Dejmek et al, as pointed out in their discussion, is that by using cell line experiments instead of actual specimens, they were able to compensate for the considerable differences in the cell number and percentage of neoplastic cells observed in clinical samples. This strategy allowed for the strict control of variables such as fixation and staining, thereby leading to stronger conclusions with regard to their effect on DNA yield and quality.
For molecular studies, especially advanced genomic high-throughput platforms, DNA quality/integrity is perhaps even more important than the yield to obtain reliable results. As we witness an expanded role for cytological samples in providing critical molecular data for targeted therapies, the evidence shown in the literature and further corroborated by the study from Dejmek et al indicates that, regardless of differences in sample preparation, the different fixation methods for cytological material provide superior results in terms of DNA quality when compared with the formalin-fixed paraffin-embedded technique. In this setting, perhaps the time has finally come that we as cytopathologists can confidently say: “tissue is NO longer the issue.”
No specific funding was disclosed.
CONFLICT OF INTEREST DISCLOSURES
The author made no disclosures.