Diagnostic application of PIK3CA mutation analysis in Chinese esophageal cancer patients

The phosphatidylinositol 3-kinase (PI3K) signaling pathway has long been suggested to play a pivotal role in the growth and survival of the cell, and pathway activation is frequently found in the oncogenesis of human cancers [1]. High frequencies of somatic mutations conferring oncogenic potential have been found in the PIK3CA gene, which encodes the PI3K catalytic subunit p110α [2]–[7]. More than 30% of various solid tumor types were found to contain the mutations, and it was frequently mutated in cancers of the colon, breast, brain, liver, stomach and lung [3],[4],[8]. Mutations in the three mutation hotspots in PIK3CA (G1624A:E542K, G1633A:E545K, and A3140G:H1047R) have been shown to elevate the lipid kinase activity and thereby leading to the activation of the downstream Akt signaling pathway [7]. Targeted therapies such as Imatinib mesylate (anti-BCR/ABL and c-kit), Gefitinib and Erlotinib (anti-EGFR) that appear to impart a high degree of specificity for translocated/mutated oncogenesis give hope that therapies targeted specifically against mutant PIK3CA can be developed [5],[9]–[12]. A recent report extends previous studies suggesting that the acquisition of PIK3CA mutations may be an important molecular event in the etiology of esophageal cancer (EC) and the mutations are associated with their clinical outcome [13]–[16].

The frequency of PIK3CA mutation in EC varied from 0% to 21%, which could likely introduce some bias in the statistical analyses of their clinical significance [6],[15]–[18]. The difference might be due to a difference in the patient cohorts or the methods used to assess PIK3CA mutation. What’s more, somatic mutations of PIK3CA in EC samples are less than that observed in colorectal (18.8-32.0%) and breast (8.3-40%) tumors [2],[8],[19]. Therefore, standard methods with high sensitivity and speciality will be conductive to examine the predictive and prognostic significance of PIK3CA mutations in EC.

The first method used for mutation detection was direct Sanger sequencing of PCR products. However, it has been limited in its ability to identify low levels of mutation-bearing tumor cells [20],[21]. The development on new molecular techniques and the increased knowledge of the genes implicated in the therapy response has boosted the development of several new detection technology including pyrosequencing, polymerase chain reaction-restricted fragment length polymorphism (PCR-RELP), Scorpion amplification refractory mutation system (Scorpion-ARMS), length analysis of fluorescently labeled polymerse chain reaction (PCR), denaturing high-performance liquid chromatography (DHPLC), and mutant-enrich liquid chip (ME-Liquidchip). However, uncertainty exists regarding which detection method can be applied in a reproductive, sensitive, and simple manner in the routine diagnostic setting. In previous studies, researchers compared the detection rate of PIK3CA, EGFR, KRAS and BRAF mutation in colorectal or lung cancers by different methods [22]–[25]. To our knowledge, no comparative assessment of these novel technologies has been available in EC up until now.

The knowledge of the mutational status of many genes implicated in response to the therapy (KIT, EGFR, BRAF, KRAS, PI3K, etc.) [29]–[31] is a real fact in the clinical management of the patients. By combining the copious amount of sequencing data over the past year, we find that the PIK3CA gene is one of the two most commonly mutated genes identified in human cancers (the other being KRAS) [2]–[4],[6],[8]. In most tissue types, mutations predominantly cluster within the helical and kinase domains of the PIK3CA subunit. Exon 9 mutations are located in the helical domain of p110a and are considered to abrogate inhibitory intermolecular interaction between p85 and p110. In contrast, exon 20 mutations are located near the activation loop and are considered to produce constitutive kinase activity. Many researchers suggested that these genetic alterations may be kinase activating and oncogenic in different human malignant tumors, including colon, breast cancer, endometrial, lung, and brain tumors [32]–[37]. Recently, Song and Lin identified PIK3CA mutations in EC with whole-genome sequencing, whole-exome sequencing or array comparative genomic hybridization analysis, and suggested their functional relevance in EC [13],[14]. On the other hand, the use of some potent PI3K/mTOR inhibitors, even nonsteroidal anti-inflammatory drugs (NSAID) was thought associate with longer survival among patients with mutated-PIK3CA colorectal cancer, but not among patients with wild-type PIK3CA cancer [38]–[41]. Therefore, PIK3CA mutation screening would greatly improve therapeutic strategies for EC in the future.

According to the current study, the different methods used in examination provide different mutation results, such as KRAS and EGFR in the same samples [22],[42]. Several methods have been applied to detection of PIK3CA gene mutation in EC [15],[17],[18],[43], it is necessary to compare the assays.

Until recently, the gold standard and most widely available method for mutation detection was Sanger sequencing, which is the only approach that permits to detect all possible sequence variants present in the target sequence, but it has several drawbacks. In particular, it is time consuming and has low sensitivity (20–50%) for the detection of mutant cells in FFPE tissue. Moreover, subjective interpretation is difficult to avoid if signal/noise ratio is low. Pyrosequencing is a sequencing-by-synthesis approach based on sequential addition of dNTPs followed by release of a pyrophosphate molecule that differs from the chain termination dye of the Sanger method, and particularly efficient for investigating mutations in short sequences. So pyrosequencing assay for PIK3CA mutation detection is certainly useful, because most activating PIK3CA mutations cluster in the hotspots of exons 9 and 20. In some cohort the minimum of mutated alleles detected with pyrosequencing was 5%. The ARMS kit has many advantages, including a closed system to prevent contamination, fast turnaround time (which may be relevant in some clinical situations), user-friendly, software-assisted data interpretation that prevents subjective interpretation. ME-Liquidchip is a novel technology, which integrates the sensitive mutant enriched PCR and quantitative high-throughput liquidchip (suspension array) to detect DNA somatic mutations in EGFR, KRAS, BRAF, and PIK3CA genes from tissues or serum samples. It has been reported that ME-Liquidchip is capable of detecting as few as 20 copies of mutant alleles with a sensitivity limit of at least mutant: wild-type ratio of 0.1%.

Here, we report our study of PIK3CA mutations by Sanger sequencing, pyrosequencing, ME-Liquidchip and ARMS in a clinical setting. With 106 EC samples, a total of 12 PIK3CA mutations were identified. The prevalence and distribution of PIK3CA mutations found in our study by ME-Liquidchip (11.3%) were in agreement with previous data reported in esophageal cancer [16],[17],[43]. All mutations were independently reassessed by ARMS. ME-Liquidchip and ARMS detected five mutated cases were not found by Sanger sequencing and pyrosequencing and they detect all mutations that had been detected by the other two methods, suggesting a higher sensitivity.

As to Sanger sequencing and pyrosequencing, the detected rate of PIK3CA mutation was 6.6% and 5.7%, respectively. Sanger sequencing detected two tumors harboring E545K and H1047R mutation, but the same samples were considered wt according to pyrosequencing. On the contrary, a tumor harboring a E545K mutation in terms of pyrosequencing was not detected by Sanger sequencing. Because the three mutations were detected by ME-Liquidchip and ARMS, we are pretty sure that they are true false negative in Sanger sequencing or pyrosequencing. Interestingly, the mutation frequency detected by Sanger sequencing is slightly higher compared with that of pyrosequencing. In previous study, the latter was considered to be with the very high sensitivity [15],[24],[42]. Most probably, they reflect problems related to low tumor DNA enrichment in the sample or the predesigned kit.

On cost-effectiveness grounds, Sanger sequencing is the method with the better defined track record and is the least expensive, relying on instruments and cheap reagents that are widely available. Pyrosequencing, ME-Liquidchip and ARMS require expensive reagents and consumables along with purchase or lease of specific instruments required by the different kits. The cost-effectiveness ratio of molecular diagnostic tests depends on a number of factors that can vary considerably, depending on the characteristics of each molecular diagnostic laboratory or the demands of specific institutions. It is important to point out that we have performed our analysis in a consecutive series of 106 EC tumor samples, corresponding to the everyday practice of most molecular diagnostic laboratories.

It’s reported there was an different distribution of PIK3CA mutations between adenocarcinoma and SCC histological subtypes of EC [16],[43]. However, all our detected mutant samples were esophageal SCC. Striking variation in geographical distribution may be the reason. It’s known to us, esophageal SCC is the predominant histological type in Asian areas, in contrast, esophageal adenocarcinoma is the predominant histological in Western countries [44]. Our study is retrospective, the results still need to be further confirmed by more investigation.