The advancement of molecular pathological techniques and the identification of additional targeted therapy sites for lung adenocarcinoma necessitate a refined molecular subtypes of non small cell lung cancer(NSCLC) patients as an essential prerequisite for personalized treatment (10). The molecular testing results derived from cytology specimens play a crucial role in formulating clinical treatment plans, particularly for patients with advanced diseases where surgical specimens are unavailable. Some previous studies in East Asian population have shown that the mutation rate of the EGFR gene in LBCSs of NSCLC detected by RT-PCR and ARMS-PCR was 30.1%-50.5%(11–12). In our study, the EGFR mutation rates detected by NGS in pleural effusion LBCSs were found to be 62.6% (139/222), which are similar to previous literature reported by NGS detection results using cytological specimens in East Asian population (55.3–68.0%) (13–14). And surpassing the EGFR gene mutation rates by RT-PCR or AMRS-PCR.This may be due to the fact the selected cases were confirmed by immunohistochemistry as lung adenocarcinoma, excluding cases of adenosquamous carcinoma or squamous cell carcinoma. Additionally, the homogeneous selection of cases in this study, all being stage III to IV cases either with pleural effusion or with metastatic lymph nodes, may contribute to the higher mutation rates observed. Furthermore, the use of NGS methods in this study may have increased the sensitivity of gene mutation detection.
In the analysis of 13 paired pleural effusion LBCSs and MLNBSs, 76.9% showed complete consistency in molecular typing results (Kappa = 0.669, P < 0.001). The consistentcy rate of EGFR reached 84.6% (Kappa = 0.740, P < 0.001), indicating a high consistency with MLNBSs. The inconsistent cases all exhibited EGFR and KRAS mutations in pleural effusion specimens, but not in lymph node biopsy specimens, suggesting that pleural effusion specimens may be more sensitive than metastatic lymph node specimens in detecting EGFR and KRAS genes, although the specific mechanism requires further study. It has also been found in literature and previous studies that the expression rate of PD-L1 in cytology specimens is higher than that in primary tumor histological specimens, and some literature approximately 25.0% in pleural effusion, underscoring the role of pleural effusion as an optimal candidate for molecular typing (15–18). The specific mechanism, however, necessitates further investigation.
The incidence of EGFR mutations and EGFR T790M mutations in patients who received EGFR-TKI treatment was found to be significantly higher compared to those without prior EGFR-TKI treatment. Moreover, the frequency of other gene point mutations (KRAS, BRAF) apart from the EGFR gene also exhibited a relatively elevated rate of mutation rates compared to those before treatment, possibly due to the development of EGFR TKI resistance mutations in certain patients. Several previous studies have demonstrated that the majority of patients undergoing EGFR-TKI treatment will eventually develop drug resistance, with approximately 50%-60% of these cases being attributed to either the presence of the EGFR T790M mutation or multiple point mutations in EGFR, such as E709G, G719S, S768I, and L861Q [19]. According to the study conducted by Villadolid J et al. NSCLC patients harboring a KRAS gene mutation exhibited an augmented likelihood of drug recurrence and metastasis, with the presence of this mutation being one of the pivotal mechanisms underlying primary drug resistance in EGFR-TKI naive patients (19–20). Furthermore, our investigation also revealed a significantly higher incidence rate of KRAS mutation in the treated group compared to the untreated group (6.15% > 3.50%), suggesting that KRAS mutation may contribute to resistance to TKI drugs.
NSCLC patients who undergo molecular typing using primary and metastatic lesion puncture histological specimens or surgical resection specimens to detect EGFR mutations and receive EGFR-TKI treatment can achieve an average PFS of 8.5 to 17.2 months (21–25). In our study, the median PFS after EGFR-TKI treatment with EGFR mutations was 12.3 months (95% CI: 10.8–13.9). The PFS of patients with stage III and IV lung adenocarcinoma was consistent with the previous findings (PFS: 10.9–11.2 months) (26). Among the two patients who had an EGFR mutation detected in LBCSs but not in MLNBSs, one of them achieved a significantly longer PFS of 14.0 months after EGFR-TKI treatment compared to the average PFS of 12.3 months, indicating an effective treatment response. Overall, molecular typing results obtained from LBCSs of pleural effusion are reliable and contribute to prolonged survival time following targeted therapy. In addition, the average PFS of EGFR-TKI treated patients with wild type EGFR detected in LBCSs was significantly lower compared to that of patients with EGFR mutation (4.1 vs 12.3, χ²=32.154, P < 0.001), further confirming the reliability of molecular typing results based on pleural effusion LBCSs. Cytology samples are simple to obtain, less invasive, particularly beneficial for late-stage patients with pleural effusion who do not require secondary biopsies. For patients, this approach offers convenience and economic benefits while reducing pain associated with specimen collection. Furthermore, the high sensitivity of pleural effusion LBCSs in gene mutation detection facilitates accurate selection of clinical molecular typing.
The assessment of tumor cell content is a crucial component in routine molecular testing. The recommended threshold for tumor cell content in routine molecular detection is typically set at 20% (27–28), which poses a challenge for analyzing samples with tumor cell content below 20%. It has been reported that EGFR mutations can be detected in cytology samples with tumor cell content below 20% (29). Therefore, we included samples with tumor cell content less than 20% and more than 10% in NGS detection, and the differences in gene mutation detection rates among samples with different tumor cell contents after EGFR-TKI treatment and their relevance to clinical targeted therapy were discussed. Although previous studies by Gu and Forest F et al. (30–31) have reported similar frequencies of EGFR mutation in cytological specimens with tumor cell content > 5% and < 20% compared to those with tumor cell content ≥ 20%, our study found a significantly lower EGFR mutation rates in the G1 and G2 groups compared to the G3 group. However, due to the limited number of cases, there was no statistical difference between the G1 and G2 groups (P = 0.160). Additionally, Forest F observed a lower KRAS mutation rate in samples with tumor cell content > 5% and < 20% compared to those with tumor cell content > 20% (11.1% vs 28.2%). In our study, the KRAS mutation rate in samples with tumor cell content < 20% was slightly lower than that in samples with tumor cell content > 20%, but no statistically significant difference was found (3.8% vs 4.1%, P > 0.05). The cytopathologist's experience in assessing tumor cell content is important. According to Li et al. (32), the frequency of EGFR T790M mutation in biopsy samples with tumor cell content ≥ 20% was significantly higher compared to samples with tumor cell content < 20%. However, there is limited research on the differences in gene mutation detection rates in LBC samples with different tumor cell content after EGFR-TKI treatment. In this study, 49 cases of pleural effusion specimens after EGFR-TKI treatment were categorized into G1 group (2 cases, 10%~15%), G2 group (2 cases, 15%~20%), and G3 group (45 cases, ≥ 20%) based on tumor cell content to investigate the difference in gene mutation detection rates. Our outcome revealed that the EGFR mutation rate was notably higher in G3 compared to G1 and G2 (73.3% vs 50% vs 50%), and only G3 exhibited the presence of EGFR T790M mutation while it was absent in both G1 and G2 groups. In our preliminary study, we conducted genetic testing on 82 liquid-based cytological specimens after EGFR-TKI treatment, including pleural effusion, fine needle aspiration (FNA), and brush biopsy (FOB); similarly, EGFR T790M mutation was not detected in three cytological specimens with tumor cell content < 20%.This is a noteworthy issue, and of course, the small number of specimens warrants further research by increasing the sample size. Combining the above results, for untreated NSCLC patients, when the tumor cell content is less than 20% during molecular typing, NGS testing on other kind samples for gene variations can still be attempted to increase the chances of receiving EGFR-TKI treatment. However, for patients who have already undergone EGFR-TKI therapy, it is advisable to select specimens with a tumor cell content exceeding 20% for molecular typing to obtain more dependable results of molecular detection.