Introduction
Lung cancer is now the most commonly diagnosed cancer and the leading cause of cancer death worldwide [
1]. In USA, 412,230 cases had lung cancer history and the new cases estimated in 2012 were 226,160. Most of lung cancers (56%) are diagnosed at an advanced stage as the typically asymptomatic in early stage. Lung cancer is classified into primarily two subgroups: small-cell lung cancer (SCLC) and non-small-cell lung cancer (NSCLC), and the later accounts for approximately 85% of all lung cancers. Although the notable progress has been made in lung therapy, this disease is still associated with a poor prognosis and has few effective treatment options. The overall 5-year survival rate for NSCLC is 17.1% [
2].
The chemotherapy efficacy is varied from different individual, even in patients with similar clinical and pathologic features, the outcome varies: some complete released, some are stable or even progression. So some authors consider NSCLC as a heterogeneous disease [
3]. The genetic factor may contribute greater to this diverse efficacy, the SNPs or gene expression products (mRNA and protein) could be special biomarkers. As the development of biochemistry, some notable biomarkers such as EGFR, VEGF, were identified to predict lung cancer treatment outcome, breast cancer susceptibility gene 1 (
BRCA1) has emerged as one of the most appealing genetic markers among them.
BRCA1 located in chromosome 17q21, and was identified as a breast and ovarian cancer susceptibility gene.
BRCA1 germline mutations have been correlated to the increasing risk of developing breast and ovarian cancer [
4,
5]. Recent studies have shown that the protein encoded by this gene is a nuclear phosphoprotein and has multiple roles not only in DNA damage repair but also in cell cycle checkpoint or cell death machinery [
6,
7]. A greater sensitivity to cisplatin with decreased
BRCA1 mRNA expression and a greater resistance to the paclitaxel with increased
BRCA1 mRNA expression was observed in breast cancer cell lines [
7,
8]. Also in tumour cells isolated from malignant effusions of NSCLC and gastric cancer patients, the same effect was observed [
9].
Followed by
in vitro studies, clinical studies explored this relationship. Taron
et al.[
10] firstly examined the potential role of
BRCA1 mRNA expression in predicting differential chemotherapy sensitivity in NSCLC, and found the patients with high
BRCA1 had poor outcome while those with low had better outcome. Followed by Taron, a series of studies evaluated the relationship between
BRCA1 level and chemotherapy outcome. The chemotherapy regimens were mainly focused on platinum-based and toxal-based treatment. However, the results were inconclusive due to the limited sample size and the limited statistics power. Current study provided a comprehensive assessment on the association between
BRCA1 level and the platinum- and toxal-based chemotherapy in NSCLC using meta-analysis.
Discussion
Although the relationship between BRCA1 expression and chemotherapy outcomes of NSCLC has been investigated by previous studies, the results were inconsistent and some were even conflicting. So a systematic review and meta-analysis based on the published literature was necessary to give further insights on this conflicting issue. Our meta-analysis showed that for platinum-based chemotherapy, low/negative BRCA 1 expression were associated with not only better ORR, but also longer OS and EFS, but for toxal-based chemotherapy, high/positive BRCA1 was associated with better ORR.
Platinum agents can bind to DNA and form complexes thus inducing intra- and inter-strand DNA, as well as DNA-protein cross-links and results in cell growth inhibition and apoptosis. As one of ant-tubulin agents, taxol inhibits cell division by enhancing formation and stabilization of microtubules and disrupts the mitotic spindle assembly, and a surveillance mechanism known as the spindle checkpoint at the metaphase-anaphase transition have been activated. It leads to the abrogation of the spindle checkpoint, results in unequal segregation of chromosomes and aneuploidy [
38].
Like excision repair cross complementing 1 (
ERCC1), xeroderma pigmentosum group D and G (
XPD,
XPG),
BRCA1 belongs to nucleotide excision repair (NER) system [
39], which has been reported to be the mayor repair system that reduced platinum-induced DNA damage.
BRCA1 involves in homologous recombination, nonhomologous end joining, mismatch repair and other effects though its interaction with other DNA repair gene such as
ATM,
ATR,
RAD51,
RAD50,
MRE11,
NBS1.
BRCA2 and so on [
7]. The reason that high/positive
BRCA1 could predict the good response to taxol is still not clear, 3 mechanisms had been proposed in explained this issue: (1) trigger cell cycle arrest in G2/M phase, (2) enhance apoptosis through a pathway involving H-Ras, MEKK4, JNK, and activation of caspases 8 and 9, (3) participate in spindle assembly checkpoint signaling [
6,
40].
BRCA1 gene showed an interesting outcome in NSCLC chemotherapy. Several cell studies and our meta-analysis based on clinical trials demonstrated low/negative
BRCA1 expression could benefit from platinum-based chemotherapy; in contrast, the high level of BRCA1 expression was in favor of toxal contained agents. This may confuse us, how could we determine chemotherapy choice properly? Rosell customized treated 84 patients based on their
BRCA1 expression: low, cisplatin plus gemcitabine (GP); intermediate, cisplatin plus docetaxel (DC); high, docetaxel alone. The median survival (MS) and 2-year survival of low
BRCA1 patients received GP regime was 11 month and 41.2%, which seem to be favorable with the traditional randomized trial treated with GP or pemetrexed plus cisplatin. The MS of high
BRCA1 patients received single-agent docetaxel was 11 month and had no detrimental effect when compared with a large phase III trial in patients treated with DC [
41]. If this hypothesis is validated, the NSCLC patients with high
BRCA1 should receive taxol based and non-platinum-contained adjuvant chemotherapy, which would be more economic, efficacy and less toxic effect for patients. However, more multi-center prospective clinical trials should be conducted to confirm this hypothesis.
Since
BRCA1 mRNA and protein level was associated with treatment efficacy, why other biomarkers such as SNPs in this gene could be a choice? But in another hand, it seems that gene expression level provides direct evidence and SNPs provide indirect evidence as it is usually gene product especially protein rather than gene itself play an import role in biochemical activity. Although SNPs are important gene variant that affect the protein expression, but many factors involve in protein synthesis. We found that studies evaluated the SNPs in
BRCA1 gene and the clinical outcome was limited. Su [
42] found that
BRCA1 S1613G was associated with platinum-based chemotherapy efficacy in objective response rate. In a large trail consisted of 300 NSCLC patients at stages III and IV, AACC haplotype but not single S1613G in
BRCA1 was associated with poor overall survival (hazard ratio = 2.097; 95%CI, 1.339 to 3.284) treated with platinum combination chemotherapy [
43]. However, whether SNPs or haplotypes were associated with clinical outcome was still under debate since lacking of enough evidence.
To interpret the results of meta-analysis, several important acknowledgments should be addressed. First, did the
BRCA1 assessment methodology consistently? As we know, IHC detects gene expression at protein level, while RT-PCR assays at mRNA level. From mRNA to protein, many factors such as transcription, post-transcriptional regulation, translation and post-translation may affect this process. Besides, RT-PCR uses the bulk tumor/tissue to extract RNA, while IHC can distinguish cell type and can read protein level only in cancer cell when compared with normal epithelial cell. Even in studies using IHC or RT-RCR assessment methodology, their cutoff value was inconsistently. Although in subgroup analysis based on
BRCA1 detecting methods in platinum-based treatment, both IHC and RT-PCR showed the significant association between
BRCA1 level and ORR, the potential heterogeneity may exist. Also, what’s the proper cutoff that could predict the chemotherapy efficacy to a great extent? We are looking forward the future researches explore this relationship. Second, is the platinum-based chemotherapy the pure platinum and the toxal-based chemotherapy the pure toxal?
BRCA1 gene shows the different mechanism and efficacy in platinum and toxal regimens. As cell experiments suggest that low/negative
BRCA1 benefit from platinum whereas high/negative BRCA1 benefit more from anti-tubulin regimen such as paclitaxel and docetaxel. But in practice, single agent in chemotherapy is impossible as the limited efficacy. Platinum is usually combined with anti-tubulin agents, for example, toxal and platinum (TP), docetaxel and carboplatin (DC). In our meta-analysis, we sorted the studies into platinum-based studies means that every patient received platinum agents (cisplatin, carboplatin or oxaliplatin), the toxal-based chemotherapy means that every patient received toxal contained agents (toxal, taxane or docetaxel). Although our meta-analysis showed that patients with low/negative
BRCA1 have better objective response rate and longer OS and EFS, and patients with high/positive
BRCA1 have better ORR, the confounding factors from chemotherapy agents may exist in studies. Third, is
BRCA1 an important predict or prognosis factor to the clinical outcome? Many factors may contribute to the ORR, OS as well as EFS, for example, age, smoking status, pathological type, tumor stage, the drug dosage and treatment cycle, also the genetic as well as gene-environment interaction also involve in disease progression, there were not enough baseline characters that ensure us to conduct stratified analysis. Four, were all relevant studies included in the analysis? This is impossible and difficult to assess. Although we searched and collected relevant studies until the December 10, 2012 to our best effort, there were still some studies with negative results did not presented the sufficient data and some were failed to get published. In our meta-analysis, only 3 Caucasian studies including 197 patients evaluated the ORR in platinum-based treatment. In toxal-based chemotherapy studies, only 4 studies consisted of 376 patients evaluated this association. The small sample size may mislead us and draw a wrong conclusion. Besides, except for one multi-center study [
31], our included samples were mainly distributed in some countries in East-Asian (Chinese and Japanese) and European (Spanish, Greece). So few studies could we found in other countries such us USA, Canada, UK, German, France and so on. Also, the African population was limited. This disequilibrium of population distribution may also affect our results.
Competing interest
The authors declare that they have no conflict of interest.
Authors’ contributions
YYL and XL conceived and designed the study, YYL and XYL participated in selecting study, extracting data, performing the statistical analysis and drafting the manuscript. XL has been involved in revising the manuscript critically for important intellectual content. All authors read and approved the final manuscript.