Personalized medicine has revolutionized the management of lung cancer. It is estimated that oncogenic driver mutations can be found in approximately 64% of patients with lung adenocarcinoma, while a relentless search is ongoing for new drivers.1 Therapy targeted towards specific genetic mutations has resulted in improved response rate, progression free survival, and decreased adverse effects.2–8
A significant proportion of lung cancer patients are diagnosed at a late stage when surgical resection is not possible and diagnosis is established on cytologic specimens obtained by minimally invasive procedures.9,10 These specimens are increasingly utilized for assessing genetic oncologic drivers.9,11,12 Furthermore, tissue acquisition for genotyping through invasive biopsy can carry a nontrivial risk of complications and may not provide sufficient material for biomarker analysis.13
Pleural fluid can be used to diagnose malignancy as well as assess genetic mutations. Current studies have looked at testing for individual mutations in pleural fluid cytology specimens, like mutations in epidermal growth factor receptor (EGFR), Kirsten rat sarcoma virus (KRAS), and v-Raf murine sarcoma viral oncogene homolog B (BRAF) genes and anaplastic lymphoma kinase (ALK) rearrangement.10,14–16 Most of these studies reported laboratory experience on a limited cohort of specimens, and it is not clear how many patients were not accounted for.9,14 Therefore, these data do not inform about the diagnostic yield of pleural fluid for mutations in all-comers in a clinical setting. Testing for rearrangement of gene coding for proto-oncogene receptor tyrosine kinase (ROS1) has not been reported in pleural cytology specimens. However, current guidelines recommend routine testing for EGFR, KRAS, ALK, and ROS1 mutations.17 The volume of pleural fluid necessary to assess genetic mutations has not been established. In addition, more studies are needed to assess the utility of pleural fluid for mutation analysis in the US population, as there are significant mutation prevalence differences across various ethnic groups.18
In this study, we report our experience with simultaneous testing for multiple oncogenic driver mutations including EGFR, KRAS, BRAF, ALK, and ROS1 in all patients who presented with malignant pleural effusions related to lung adenocarcinoma. We also investigated the amount of pleural fluid needed to optimize the yield of molecular markers.
MATERIALS AND METHODS
We retrospectively reviewed data of all patients with malignant pleural effusions secondary to metastatic adenocarcinoma of the lung from May 1, 2012 to August 30, 2016. Only patients with pleural fluid cytology diagnostic of bronchogenic adenocarcinoma were included. The study was approved by Institutional Review Board at Duke University Medical Center (Protocol no. Pro00069486, Durham, NC). Testing was performed for oncogenic drivers, EGFR, KRAS, and BRAF mutations, and ALK and ROS1 rearrangements on pleural fluid. The pleural cytology yield was compared with independent histopathology specimens including surgical resection, core needle, and transbronchial or endobronchial biopsies.
Analysis for genetic mutations was performed on genomic DNA extracted from macroscopically dissected tumor from formalin-fixed, paraffin embedded (FFPE) sections created from cell blocks of pleural fluid. In histopathology specimens, testing was performed on formalin-fixed, paraffin-embedded tissue. Cytology or histopathology specimens were categorized as adequate if mutation analysis could be performed for all the ordered oncogenic drivers, including EGFR, KRAS, BRAF, ALK, and ROS1. If analysis could not be performed for any one or more of these mutations, the specimens were categorized as inadequate. However, yield of testing for individual mutations was recorded.
Cell Block Processing
Cell blocks were prepared using thrombin to congeal cellular material. A 15 mL centrifuge tube was used to perform centrifugation at 2800 rpm to generate a pellet. The supernatant was discarded and the pellet was resuspended in Cytolyt (Hologic, Marlborough, MA), followed by another round of centrifugation with supernatant discarded. Drops of thrombin (at a concentration of 5000 units/mL of deionized water) were added until the cellular sediment clotted. The clot was placed in a bag and cassette and placed into 10% neutral buffered formalin for routine histologic processing. An alternative method was used if cellular material was difficult to identify using 95% ethanol. If the pellet was insufficient, more cell blocks might be requested by the pathologist using additional volumes of sample. For most of the patients, only 1 pellet was used for cytologic analysis. If the pellet showed cells consistent with lung adenocarcinoma, further molecular testing for genetic mutations was performed.
Testing for EGFR and KRAS mutations was performed using PCR amplification. A hematoxylin and eosin stained slide from the specimen was first evaluated to identify the regions of greatest tumor content. These regions were then macrodissected and genomic DNA was extracted. The EGFR assay examined mutations in exons 18 to 21 in EGFR gene, including codons 719, 768, 790, 858 or 861, and deletion or complex mutations in exon 19. The assay was not designed to detect an acquired mutation that was present below the 5% detection limit (ie, neoplastic cell population of <10% of total cells in the sample).
The KRAS assay examined codons 12, 13, and 61 within exons 2 and 3 of the KRAS oncogene for activating mutations. The assay used PCR with Locked Nucleic Acid (LNA)-mediated PCR clamping followed by Sanger DNA sequencing to detect activating mutations. This assay would not detect an acquired mutation present below the 2.5% detection limit (ie, neoplastic cell population comprising <5% of total cells).
Over the timeframe of the study, the genetic analysis approach changed to next generation sequencing (NGS) of selected regions in 50 target genes, including EGFR, KRAS, and BRAF, which was used in more recent cases (n=13). This assay used genomic DNA from FFPE sections from cell block to perform massively parallel or NGS performed on the Ion Torrent platform (Thermo Fisher Scientific, Waltham, MA). The lower limit of detection for this assay was 5%, that is, a mutant allele could be detected in a 95% wild type allele background (or 10% neoplastic cell content). Sensitivity was impacted by the depth of sequence coverage, mutation type and allelic frequency of the mutation. Coverage of at least ×250 was required to detect a mutation at the 5% lower limit of detection. Indeterminate coverage was considered to be ×100 to ×250 and might be adequate to detect mutations in samples with a greater neoplastic cell content. Inadequate coverage was below ×100.
Testing for ALK and ROS1 gene rearrangements was performed using fluorescence in situ hybridization (FISH). Specifically, ALK assay tested for ALK-EML4 fusion by assessing for chromosome 2p23 rearrangement. A dual color break-apart interphase FISH assay was performed on FFPE sections of the tumor using the Vysis ALK Break-Apart FISH Probe Kit (Abbott Molecular, Des Plaines, IL). Signals were enumerated in 50 nuclei from an area of the tumor located by comparison with the hematoxylin and eosin stained adjacent section. Tumors were considered FISH-positive if ≥15% of scored tumor cells had split ALK5′ and 3′ probe signals or isolated 3′ signals.
FISH using break-apart probe for the ROS1 gene region of chromosome 6q22 was performed on FFPE sections from cell block. A dual color break-apart interphase FISH assay was performed using the Vysis ROS1 Break-Apart FISH Probe Set (Abbott Molecular, Des Plaines, IL). Signals were enumerated in 100 nuclei from an area of the tumor located by comparison with the hematoxylin and eosin stained adjacent section. Tumors were considered FISH-positive if ≥15% of scored tumor cells had split ROS1 5′ and 3′ probe signals or had isolated 3′ probe signals.
Descriptive statistics for patient characteristics are presented as mean±SD when normally distributed and median (interquartile range, IQR), when not normally distributed. χ2 test was used to compare categorical variables. Two-sided P-value of <0.05 was considered statistically significant. R version 3.3.1 (R Core Team, 2016, Vienna, Austria) was used for statistical analysis.
A total of 134 patients with malignant pleural effusion secondary to lung adenocarcinoma were included in the study. Out of these, 56 patients had oncogenic driver mutations testing ordered and performed on pleural fluid. Other patients did not have mutation testing on pleural fluid as this information was obtained from histopathology specimens. The patient characteristics are shown in Table 1. Pleural fluid cytology was adequate for complete mutation analysis including EGFR, KRAS, and BRAF mutations, and ALK and ROS1 gene rearrangements in 40 (71.4%) patients, as shown in Table 2. This was further stratified by individual oncogenic drivers. EGFR testing was adequate in 38 of 49 (77.6%), KRAS 22 of 28 (78.6%), BRAF 10 of 13 (76.9%), ALK gene rearrangement 42 of 51 (82.4%), and ROS1 gene rearrangement in 21 of 28 (75%) pleural fluid specimens. Since PCR amplification was performed to detect EGFR and KRAS mutations in the earlier cases, and NGS was adopted later to assess EGFR, KRAS, and BRAF mutations, we compared the diagnostic yield of both methods. PCR amplification was adequate in 30 of 39 tested pleural fluid samples, whereas NGS was adequate in 10 of 13 samples (0.76 vs. 0.76; P=1).
In comparison, the histopathology specimens were able to provide complete mutation analysis in 72 of 84 samples (85.7%), as shown in Table 2. The yield of histopathology specimens for EGFR testing was adequate for 74 of 82 (90.2%), KRAS 44 of 48 (91.7%), BRAF 11 of 12 (91.6%), ALK rearrangement 71 of 74 (95.9%), and ROS1 rearrangement in 30 of 34 (88.2%) histopathology specimens. As patients had testing done on either pleural cytology or histopathologic specimens, but not both, a paired comparison is not possible. Unpaired comparison analysis of the yield between the two specimen types is shown in Table 2.
We found the following distribution of genetic mutations in all the pleural fluid samples, which were sufficient for testing: EGFR 5 of 38 (13.1%), KRAS 6 of 22 (27.2%), BRAF 0 of 10 (0), ALK 7 of 42 (16.6%) and ROS1 0 of 21 (0). Further details of EGFR and KRAS mutation location and frequency is shown in Table 3.
Volume of pleural fluid tested for mutation analysis was compared with its diagnostic yield, as shown in Table 4. The mutation analysis was possible in 13 of 19 samples (68.4%; median, 90 mL; range, 5 to 100 mL) with volume ≤100 mL versus 27 of 37 samples (72.9%; median, 1000 mL; range, 110 to 2200 mL) with >100 mL of pleural fluid sent to the laboratory (χ2 test P-value=0.7).
This study shows that malignant pleural effusions related to pulmonary adenocarcinoma can be used to assess oncogenic driver mutations. About 100 mL of pleural effusion tested for mutations yields results comparable to larger volumes.
Previous studies have looked at pleural fluid testing for individual genetic mutations in lung adenocarcinoma. Guan et al14 assessed EGFR mutation in pleural fluid and lung histologic specimens in 50 patients with lung adenocarcinoma and found a similar prevalence of 30% versus 34%, respectively. However, they did not comment on the yield of pleural fluid for detection of the mutation. Billah et al9 looked at the PCR yield of EGFR and KRAS in 27 body fluid specimens from nonsmall and small cell lung cancer patients and found that cytology was sufficient to perform testing in 96.3% patients. They showed that both cytology smears and cell blocks can be used for molecular testing. The yield was lower with specimens with low cellularity (<300 tumor cells). In another study, Betz and colleagues reported their experience of assessing ALK rearrangement in 13 pleural fluid smear and cell block specimens. FISH testing was performed only on the specimens with sufficient cellularity and was successful in all cases. In our study, the main aim was to assess the adequacy rate of pleural fluid cytology for molecular markers. It could be taken as synonymous for diagnostic yield which included both positive and negative results for molecular markers (positive and negative results/all samples tested). We expanded the previously reported literature by showing that simultaneous testing ordered for all currently targetable genetic mutations and KRAS is possible on pleural cytology in 71% samples. We also report the experience of FISH testing for ROS1 rearrangement and NGS testing for BRAF mutation in pleural cytologic specimens. In addition, we describe a real clinical yield of the pleural fluid and included all consecutive patients who presented with lung adenocarcinoma-related pleural effusions, decreasing the selection bias potentially present in previous studies.
The yield of surgical biopsies was better compared to pleural fluid cytology for comprehensive as well as individual mutation analysis, especially EGFR and ALK rearrangement where the difference was statistically significant. However, all the targetable mutations and KRAS could be assessed in 70% to 80% of the patients with pleural fluid cytology. Since pleural fluid drainage is a minimally invasive procedure, testing for genetic mutations in pleural fluid can obviate the need for more invasive biopsies. We recommend starting with pleural fluid to assess genetic mutations in patients with lung adenocarcinoma. If the pleural cytology is inadequate, further invasive or surgical biopsies should be considered as histopathology specimens have a better yield to assess the mutation status.
We also showed that yield of molecular markers is comparable whether ≤100 mL or larger volume of pleural fluid is tested. In a previous study, 50-mL of pleural fluid has been shown to have similar yield for diagnosis of malignancy compared to larger volumes.19 Carter et al16 recently reported their experience of molecular testing in 50 patients with nonsmall cell lung cancer associated pleural effusions. They found a yield of 54% for EGFR, KRAS, BRAF, and ALK testing, regardless of the pleural fluid volume analyzed. We confirm their findings with our cohort, showing a better pleural fluid yield for molecular markers independent of the pleural fluid volume tested.
Role of cytologic specimens for assessment of PD-L1 expression is under investigation. Heymann and colleagues reported their findings of PD-L1 testing in a cohort of 214 surgical, small biopsy, and cytology specimens. Twelve patients with pleural effusion were included in this group. They found that PD-L1 testing was feasible in patients with cytology specimens. However, future robust studies are needed to explore this further.20
Strengths of our study include demonstration of a pragmatic and minimally invasive way of assessing genetic mutations in patients with lung adenocarcinoma-related pleural effusions. It also shows that currently targetable mutations and KRAS can be simultaneously tested on pleural fluid cytology with fluid volume as little as 100 mL. Finally, this is the first report of ROS1 FISH testing in pleural fluid. On the basis of our results, we recommend a minimum of 100 mL of pleural fluid for mutation analysis and encourage more definitive studies to explore this question.
Our study is limited by its retrospective nature and relatively small sample size. Since patients did not have testing done on both cytologic and histopathology specimens, we cannot comment on the concordance of these specimens. We were able to successfully assess for BRAF and ROS1 mutations, but no patient in our cohort had these mutations. Given our relatively small sample size and the previously estimated prevalence of 2% to 4% for BRAF mutation and 1% for ROS1 rearrangement in lung adenocarcinoma,21,22 this is an expected result. Similarly, we did not have any patients with T790M EGFR mutation, but we expect that it could be potentially diagnosed using pleural fluid cytology. Lastly, detection of EGFR and KRAS mutations was performed by PCR amplification in the earlier cases, whereas the process was switched to NGS later. However, we did not find a significant difference in the yield between the 2 methods in this limited sample.
Cytologic specimens from malignant pleural effusions can prove valuable for assessment of oncogenic driver mutations. Volume of malignant pleural effusion analyzed for molecular testing does not correlate with its diagnostic yield. We recommend a minimum of 100 mL of pleural fluid for molecular marker testing.
1. Kris MG, Johnson BE, Berry LD, et al. Using multiplexed assays of oncogenic drivers in lung cancers to select targeted drugs. JAMA. 2014;311:1998–2006.
2. Mok TS, Wu YL, Thongprasert S, et al. Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma. N Engl J Med. 2009;361:947–957.
3. Maemondo M, Inoue A, Kobayashi K, et al. Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR. N Engl J Med. 2010;362:2380–2388.
4. Mitsudomi T, Morita S, Yatabe Y, et al. Gefitinib versus cisplatin plus docetaxel in patients with non-small-cell lung cancer harbouring mutations of the epidermal growth factor receptor (WJTOG3405): an open label, randomised phase 3 trial. Lancet Oncol. 2010;11:121–128.
5. Zhou C, Wu YL, Chen G, et al. Erlotinib versus chemotherapy as first-line treatment for patients with advanced EGFR mutation
-positive non-small-cell lung cancer (OPTIMAL, CTONG-0802): a multicentre, open-label, randomised, phase 3 study. Lancet Oncol. 2011;12:735–742.
6. Rosell R, Carcereny E, Gervais R, et al. Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation
-positive non-small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial. Lancet Oncol. 2012;13:239–246.
7. Sequist LV, Yang JC, Yamamoto N, et al. Phase III study of afatinib or cisplatin plus pemetrexed in patients with metastatic lung adenocarcinoma with EGFR mutations. J Clin Oncol. 2013;31:3327–3334.
8. Wu YL, Zhou C, Hu CP, et al. Afatinib versus cisplatin plus gemcitabine for first-line treatment of Asian patients with advanced non-small-cell lung cancer harbouring EGFR mutations (LUX-Lung 6): an open-label, randomised phase 3 trial. Lancet Oncol. 2014;15:213–222.
9. Billah S, Stewart J, Staerkel G, et al. EGFR and KRAS mutations in lung carcinoma: molecular testing by using cytology
specimens. Cancer Cytopathol. 2011;119:111–117.
10. Betz BL, Dixon CA, Weigelin HC, et al. The use of stained cytologic direct smears for ALK gene rearrangement analysis of lung adenocarcinoma. Cancer Cytopathol. 2013;121:489–499.
11. Hopkins E, Moffat D, Parkinson I, et al. Cell block samples from endobronchial ultrasound transbronchial needle aspiration provide sufficient material for ancillary testing in lung cancer-a quaternary referral centre experience. J Thorac Dis. 2016;8:2544–2550.
12. Yarmus L, Akulian J, Gilbert C, et al. Optimizing endobronchial ultrasound for molecular analysis. How many passes are needed? Ann Am Thorac Soc. 2013;10:636–643.
13. Tam AL, Kim ES, Lee JJ, et al. Feasibility of image-guided transthoracic core-needle biopsy in the BATTLE lung trial. J Thorac Oncol. 2013;8:436–442.
14. Guan Y, Wang ZJ, Wang LQ, et al. Comparison of EGFR mutation
rates in lung adenocarcinoma tissue and pleural effusion
samples. Genet Mol Res. 2016;15:1–7.
15. Rosenblum F, Hutchinson LM, Garver J, et al. Cytology
specimens offer an effective alternative to formalin-fixed tissue as demonstrated by novel automated detection for ALK break-apart FISH testing and immunohistochemistry in lung adenocarcinoma. Cancer Cytopathol. 2014;122:810–821.
16. Carter J, Miller JA, Feller-Kopman D, et al. Molecular profiling of malignant pleural effusion
in metastatic non-small-cell lung carcinoma. The effect of preanalytical factors. Ann Am Thorac Soc. 2017;14:1169–1176.
17. Ettinger DS, Wood DE, Akerley W, et al. NCCN guidelines: non-small cell lung cancer, version 5.2017. J Natl Compr Canc Netw. 2017;15:504–535.
18. John T, Liu G, Tsao MS. Overview of molecular testing in non-small-cell lung cancer: mutational analysis, gene copy number, protein expression and other biomarkers of EGFR for the prediction of response to tyrosine kinase inhibitors. Oncogene. 2009;28(suppl 1):S14–S23.
19. Abouzgheib W, Bartter T, Dagher H, et al. A prospective study of the volume of pleural fluid required for accurate diagnosis of malignant pleural effusion
. Chest. 2009;135:999–1001.
20. Heymann JJ, Bulman WA, Swinarski D, et al. PD-L1 expression in non-small cell lung carcinoma: Comparison among cytology
, small biopsy, and surgical resection specimens. Cancer Cytopathol. 2017;125:896–907.
21. Nguyen-Ngoc T, Bouchaab H, Adjei AA, et al. BRAF alterations as therapeutic targets in non-small-cell lung cancer. J Thorac Oncol. 2015;10:1396–1403.
22. Shaw AT, Ou SH, Bang YJ, et al. Crizotinib in ROS1-rearranged non-small-cell lung cancer. N Engl J Med. 2014;371:1963–1971.