Frequency, characterization, and prognostic analysis of PIK3CA gene mutations in Chinese esophageal squamous cell carcinoma

Human Pathology (2014) 45, 352–358

www.elsevier.com/locate/humpath

Original contribution

Frequency, characterization, and prognostic analysis of PIK3CA gene mutations in Chinese esophageal squamous cell carcinoma☆,☆☆ Jun Hou PhD a,1 , Dongxian Jiang PhD a,1 , Jingchuan Zhang PhD b , Paul R. Gavine PhD b , Songtao Xu PhD c , Yalan Liu PhD a , Chen Xu PhD a , Jie Huang PhD a , Yunshan Tan PhD a , Hao Wang PhD c , Yachao Lu PhD b , Li Zheng PhD b , Yingyong Hou PhD a,⁎, Lijie Tan PhD c,⁎ a

Department of Pathology, Zhongshan Hospital, Fudan University, Shanghai 200032, PR China Innovation Center China, AstraZeneca Global R&D, Zhangjiang Hi-Tech Park, Shanghai 201203, PR China c Department of Thorax, Zhongshan hospital, Fudan University, Shanghai 200032, PR China b

Received 21 July 2013; revised 15 September 2013; accepted 18 September 2013

Keywords: Esophageal squamous cell carcinoma; Mutation; PIK3CA gene

Summary PIK3CA gene mutations are found in numerous cancers but correlate differently with prognosis. Although the frequency of PIK3CA gene mutation in esophageal squamous cell carcinoma (ESCC) has been previously studied, a prognostic analysis has not been reported. Ninety-six surgically resected ESCC tissues were collected from Chinese patients and DNA was extracted. Gene mutations in PIK3CA (exons 9 and 20), EGFR (exons 18, 19, 20 and 21), KRAS (exons 2 and 3), and BRAF (exons 11 and 15) were screened using mutant-enriched liquid chip technology. PIK3CA gene mutations were identified in 12 of 96 ESCC cases (12.5%). No mutations were identified in EGFR, KRAS or BRAF genes in this study. Correlations between clinicopathological features and PIK3CA mutation status were analyzed and finally, patient survival information was used to determine the prognostic significance of PIK3CA mutation. Interestingly, the frequency of PIK3CA mutation was higher in female ESCC patients (31.3%, 5/16) than in males (8.8%, 7/80), and higher in patients with non–lymph node metastasis (19.6%, 10/51, P = .013) than in patients with lymph node metastasis (4.4%, 2/45, P = .025). Furthermore, patients with PIK3CA-mutated tumors showed a trend towards favorable overall survival (P = .085) but not disease-free survival (P = .238), suggesting that PIK3CA gene status may be a favorable predictive marker in ESCC patients. © 2014 Elsevier Inc. All rights reserved.

☆

Competing interests: The authors declare that they have no competing interests. Grant support: Identification and establishment of Chinese patient-derived esophageal carcinoma xenograft mouse model from Shanghai Science and Technology Project (No 11140902502). ⁎ Corresponding authors. Yingyong Hou is to be contacted at Department of Pathology, Zhongshan Hospital, Fudan University, Shanghai 200032, PR China, Lijie Tan, Department of Thorax, Zhongshan Hospital, Fudan University, Shanghai 200032, PR China. E-mail addresses: [email protected] (Y. Hou), [email protected] (L. Tan). 1 Jun Hou and Dongxian Jiang contributed equally to this project. ☆☆

0046-8177/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.humpath.2013.09.011

PIK3CA mutation in esophageal carcinoma

1. Introduction The phosphatidylinositol 3-kinase (PI3K) signaling pathway has long been suggested to play a pivotal role in cell growth and survival, with frequent pathway activation commonly documented in the genesis and progression of numerous human cancers [1]. PIK3CA is one of the most commonly mutated oncogenes in human cancer, and highly frequent oncogenic somatic mutations occur in the p110α PI3K catalytic subunit [1–6], especially in cancers of the colon [7], breast [8] and endometrium [9]. More than 75% of these mutations cluster in the helical (exon 9) and kinase domains (exon 20) of the PIK3CA gene. Mutations in the three mutation hotspots in PIK3CA (G1624A:E542K, G1633A: E545K, and A3140G:H1047R) have been shown to elevate lipid kinase activity, thereby leading to activation of downstream Akt signaling [6]. The successful development of prior targeted therapies (such as imatinib mesylate (anti-BCR/ABL and c-kit) and gefitinib and erlotinib [anti-EGFR]), which act through highly selective targeting of deregulated kinases, offer hope for the development of therapies targeted specifically against mutant PIK3CA [4,10–13]. It is well established that esophageal carcinoma is characterized by striking variations in geographical distribution and a high mortality rate [14]. Furthermore, esophageal squamous cell carcinoma (ESCC) represents the predominant histological subtype in Asian areas, while in contrast, the incidence of esophageal adenocarcinoma now exceeds that of ESCC in some western European countries [15]. Based on previous literature, the frequency of PIK3CA mutation appears to vary between esophageal cancer subtypes. Phillips et al identified PIK3CA mutations in 11.8% (4/35) of ESCC and 6% (3/50) of adenocarcinoma using single-strand conformational polymorphism and denaturing high-performance liquid chromatography [16]. There are however, conflicting reports regarding PIK3CA mutation rates in ESCC. Mori et al [17] identified mutations in 2/88 (2.2%) of ESCC using polymerase chain reaction (PCR)–direct sequencing. In contrast, Akagi et al [18] failed to identify PIK3CA mutations in ESCC (reviewed by Samuels and Waldman [5]). More recently, Shigaki et al [19] reported PIK3CA exon 9 and/or exon 20 mutations in 46/219 (21%) ESCC patients using a pyrosequencing approach. In order to determine the incidence and clinicopathological significance of PIK3CA gene mutation in Chinese ESCC patients, we performed an analysis of PIK3CA tumor mutation status and assessed correlations with relevant clinicopathological features in 96 ESCC patient samples. Furthermore, we reviewed patient survival data and explored the utility of PIK3CA mutation as a predictive marker of favorable patient survival.

353 tion at the Zhongshan Hospital (Department of Thorax Surgery, Zhongshan Hospital). Prior written informed consent was obtained from all patients, and the study protocol received ethics board approval from Zhongshan hospital, Fudan University. Patients received no prior therapy. Sections were reviewed by a pathologist to confirm the ESCC diagnosis. The following patient characteristics were collected: gender, age, tumor site (upper, middle, and lower region of esophagus), histological grade, coagulative necrosis, nerve and vascular infiltration, mitotic index (numbers recorded as ≤20 per 10 high power fields [HPF], 20-50/ 10HPF, or ≥50/10HPF), lymph node metastasis, and stage, as previously reported [20].

2.2. Gene mutation screening A single representative block, containing at least 70% neoplastic cells was selected for DNA extraction. Genomic DNA was extracted using the QIAamp Mini kit (Qiagen) according to the manufacturer's instructions. “Hotspot” EGFR, KRAS, BRAF and PIK3CA gene mutations were screened using a mutantenriched PCR method [21]. Screening was performed by SurExam Bio-Tech Co. Ltd. (Guangzhou Technology Innovation Base, Science City, Guangzhou, P.R. China), and 4 types of exon 9 and 2 types of exon 20 mutations were examined. For EGFR, exons 18, 19, 20, and 21 were screened. For KRAS, exons 2 and 3 were screened. For BRAF, exons 11 and 15 were screened, while for PIK3CA, exons 9 and 20 were analyzed.

2.3. Follow-up information Follow-up information for the 96 patients after surgery and treatment was provided by the referring clinicians, or else obtained directly from patients and their family members as standard procedure. The date of last follow up was April 22, 2013. Disease-free survival (DFS) and overall survival (OS) were measured from the time of surgery to the time of first recurrence (or most recent follow-up) or death.

2.4. Statistical analysis A χ2 test was used for univariate analysis; crosstabulations with qualitative variables were analyzed with Fisher exact test. Kaplan-Meier analysis was used to calculate DFS and OS. Log-rank test of survival analysis was used to compare DFS and OS as functions of variables and to identify significant differences. P b .05 were recorded as significant.

2. Materials and methods

3. Results

2.1. Patients and tissue samples

3.1. Characterization of ESCC patients

During March to October 2010, 96 primary ESCC samples were obtained intraoperatively during esophagectomy resec-

The clinicopathological features of the 96 ESCC patients are presented in Table 1. Patient age ranged from 37 to 81

354 Table 1

J. Hou et al. Correlation between PIK3CA gene mutation and ESCC patient clinicopathological features using univariate analysis (n = 96)

Clinicopathological features Sex Male Female Age at diagnosis (y) b60 ≥60 Smoking Yes (M/F) No (M/F) Alcohol consumption Yes (M/F) No (M/F) Tumor site Upper (M/F) Middle (M/F) Lower (M/F) Differentiation G1 G2 G3 Nerve invasion Yes No Vascular invasion Yes No Necrosis Yes No Lymph node metastases Yes No Mitoses (/10HPF) ≤20 20~50 ≥50 Stage I (A + B) II (A + B) III (A + B + C) PIK3CA mutation⁎ Yes No

n

PIK3CA mutation (%)

P

DFS .013⁎

80 16

7 (8.8) 5 (31.3)

35 61

3 (8.6) 9 (16.8)

32 (31:1) 64 (49:15)

4 (12.5) 8 (12.5)

27 (27:0) 69 (53:16)

4 (14.8) 8 (11.6)

1 (1:0) 33 (26:7) 62 (53:9)

0 8 (24.2) 4 (6.5)

0 61 35

0 8 (13.1) 4 (11.4)

34 62

2 (5.55) 10 (14.3)

19 77

2 (10.5) 10 (13.0)

58 38

6 (10.3) 6 (15.8)

44 52

2 (4.5) 10 (19.2)

29 37 30

3 (10.3) 4 (10.8) 5 (16.7)

4 53 39

0 10 (18.9) 2 (5.1)

12 83

– –

Prognosis

0.097

OS 0.266

.378

0.259

0.181

1.000

0.484

0.568

.668

0.738

0.439

.041⁎

0.742

0.794

.810

0.240

0.531

.147

0.028

0.013⁎

.771

0.123

0.952

.430

0.249

0.416

.030

b0.0001⁎

b0.0001⁎

.706

0.465

0.146

.107

b0.0001⁎

b0.0001⁎

0.238

0.085

– –

Abbreviations: DFS, disease-free survival; OS, overall survival. ⁎ P b .05.

years with a median age of 62 years. Of these, 35 patients were younger than 60 years and 61 were older than 60 years. The cohort comprised 80 males and 16 females; 32 (31 male and 1 female) patients had histories of smoking (P = .012) and 27 male patients of regular alcohol consumption (P = .006). Of the ESCC tumor samples, 1 (male patient) was located in the upper esophagus, 33 (26 males and 7 females) in the middle and 62 (53 males and 9

females) in the lower area; the female patients had a greater incidence of mid-esophageal tumors (43.8%) than male patients (32.5%), although no significant difference was identified. Based on cellular differentiation, 61 tumors were of grade II (G2) and 35 were grade III (G3). None were grade 1 (G1). Five tumors had invaded to the submucosa, 24 to the muscularis propria and 67 to the adventitia. Nerve infiltration

PIK3CA mutation in esophageal carcinoma presented in 34 cases, vascular infiltration in 19, coagulative necrosis in 58, and lymph node metastases were recorded in 44 cases. Four cases were stage IB, 29 stage IIA, 24 stage IIB, 19 stage IIIA, 17 stage IIIB, and 3 stage IIIC. The representative histological features are shown in Fig. 1.

3.2. Gene mutation analysis All 96 ESCC patient tumors were screened for gene mutations in EGFR, KRAS, BRAF and PIK3CA. PIK3CA gene mutations were identified in 12 (12.5%) of 96 primary tumors using mutant-enriched PCR methodology. PIK3CA exon 9 (activation loop) mutations were identified in 10 cases, 4 cases of E542K (GAA AAA) and 6 cases of E545K (GAG to AAG), while exon 20 (catalytic domain) mutations were identified in 2 cases: H1047R [(CAT to CGT; Table 2). No co-occurrence of exon 9 and exon 20 mutations were identified in any patient samples. Interestingly, none of the patient samples harbored gene mutations in EGFR, KRAS or BRAF.

3.3. Analysis of patient outcomes Among the 96 ESCC patients, only 1 patient was lost to follow-up. DFS and OS data were generated for the

355 Table 2 PIK3CA gene mutation characteristics in ESCC using mutant-enriched liquid chip technology Case Age (y) Gender Pathology Stage Exon Mutation 011 030 037 039 041 053 058 060 070 074 084 095

56 61 66 65 71 70 61 63 54 49 74 72

Female Female Male Male Female Male Female Male Male Male Female Male

ESCC ESCC ESCC ESCC ESCC ESCC ESCC ESCC ESCC ESCC ESCC ESCC

IIB IIA IIB IIA IIA IIIA IIB IIB IIA IIB IIB IIIB

9 9 20 9 9 9 9 9 20 9 9 9

E542K E545K H1047R E542K E542K E542K E545K E545K H1047R E545K E545K E545K

remaining 95 patients. Thirty-seven patients experienced tumor recurrences, including among others lymph node, liver metastasis and bone metastasis. DFS ranged from 1 to 37 months, with a mean value of 29.4 months. There were 32 deaths, and of these, 30 were ESCC-related. OS ranged from 1 to 37 months with a mean value of 31.9 months. The 3-year DFS and OS incidences for this cohort of ESCC patients were 60.8% and 67.0%, respectively.

Fig. 1 Representative histological images of ESCC patient (hematoxylin and eosin). A and B, Squamous cell carcinoma of esophagus with intramural invasion. B, Moderately differentiated carcinoma with necrosis. C, Poorly differentiated carcinoma with nerve infiltration. D, Vascular emboli.

356

J. Hou et al.

3.4. Relationship of PIK3CA mutation with clinicopathological parameters Within this study, we correlated the clinicopathological parameters of the ESCC patients with PIK3CA mutational status (Table 1). First, we found a significant difference in the PIK3CA mutation rate between female (31.3%, 5 out of 16) and male ESCC patients (8.8%, 7/80; P = .013). Secondly, we found a significant difference in the rate of PIK3CA mutation between ESCC patients without lymph node metastasis (19.2%, 10/52) and with lymph node metastasis (4.5%, 2/44; P = .030). Furthermore, PIK3CA mutation was frequently detected in ESCC tumors originating from the mid-region esophagus (P = .041).

3.5. Correlation of clinicopathological features and PIK3CA mutation with prognosis in Chinese ESCC Consistent with our previous experience, ESCC patients with lymph node metastasis and high TNM stage had unfavorable prognosis, both in terms of DFS and OS (all P b .0001). In addition, nerve invasion presented with unfavorable prognosis in our cohort (DFS P = .028, OS P = .013). All remaining clinicopathological features showed no correlation with either DFS or OS. All data are shown in Table 1. With regard to PIK3CA mutation, we identified marginal favorable prognostic value in OS of patients with tumors harboring PIK3CA mutation (P = .085; Table 1 and Fig. 2). No prognostic value was associated with PIK3CA mutation and DFS (P = .238, Fig. 3).

4. Discussion Somatic mutations within the PIK3CA gene were first identified in 2004 [2] in a variety of malignant human

Fig. 2 Kaplan–Meier representation of cumulative overall survival of ESCC patients with or without PIK3CA gene mutation.

Fig. 3 Kaplan–Meier representation of cumulative disease-free survival of ESCC patients with or without PIK3CA gene mutation.

tumors including colon [22], breast [23–25], endometrial [9], lung [23,24], and brain [26], with these mutations predominantly occurring in E542K or E545K and H1047R hotspots, found on exons 9 and 20 of the PIK3CA gene, respectively. Therefore, these 3 PIK3CA hotspot mutations were screened for within our cohort of ESCC patients.

4.1. Frequency of PIK3CA mutation in ESCC In this study, we investigated 96 Chinese ESCC patients to assess whether this phenomenon also exists in ESCC patients. Ten PIK3CA mutations in exon 9 and two PIK3CA mutations in exon 20 were identified, with no identified cases of mutational co-occurrence. The frequency of PIK3CA mutation in EC varies across studies and indeed, so do the gene sequencing methodologies used. Phillips et al identified PIK3CA mutations in 11.8% (4/35) of ESCC and 6% (3/50) of adenocarcinoma using a single-strand conformational polymorphism/denaturing high-performance liquid chromatography sequencing method [16]. Mori et al identified mutations in 2.2% (2/88) of ESCC using PCR–direct sequencing. Shigaki et al [19] reported the highest frequency of PIK3CA mutation (21%) in ESCC by employing a pyrosequencing approach. Herein, we report PIK3CA mutation in 12.5% of ESCC using a mutant-enriched liquid chip sequencing approach. Clearly, variety within the sequencing methodologies described above is likely to influence reporting of the “true” PIK3CA mutation rate in EC. Importantly however, in our experience (and that of others) the mutant-enriched liquid chip sequencing approach represents a superior method to that of PCR–direct sequencing [21]. We acknowledge this variety in PIK3CA mutation incidence and highlight the requirement for further studies to compare multiple detection methods to screen the same sample set.

PIK3CA mutation in esophageal carcinoma

4.2. PIK3CA mutation rates are significantly higher in female ESCC patients and mid-esophageal tumors Our data demonstrate that PIK3CA mutations occur with higher frequencies in female ESCC patients (31.3%), compared to male patients (8.8%), and higher frequencies in middle esophagus (24.2%), compared to lower esophagus (6.5%). Although female patients had a greater incidence of middle ESCC, the PIK3CA mutation rate sex difference takes on greater significance as female patients had tumors predominantly unassociated with Barrett's esophagitis, and generally lacked any histories of smoking and drinking. In 2 previous studies, PIK3CA mutations were more frequent in male patients than in female patients—9.8% (6/ 61) versus 4.3% (1/23) in ESCC and AC [16] and 15.8% (3/ 19) versus 6.7% (1/15) in ESCC only [16], although neither comparison was statistically significant. Based on Mori et al's [17] data, 2 samples with PIK3CA mutation were derived from male patients and two cell lines (KYSE501 and TE-5) with PIK3CA mutation originated from female patients. Of relevance to this discussion is the fact that esophageal cancer incidence is characterized by a striking variation in geographical distribution. Furthermore, ESCC represents the predominant histological subtype in Asian areas, while in contrast, esophageal adenocarcinoma is predominant in Western countries. Taken together, it's possible that geographical, histopathological factors, sequencing methodology and perhaps etiology may play important roles in the observed spectra and incidence of PIK3CA mutation and certainly warrant further investigation in the future.

4.3. PIK3CA mutation shows an inverse correlation with lymph node metastasis and a trend toward favorable survival in ESCC Across multiple cancer types, PIK3CA gene mutation is associated with differing trends in prognosis. Patients with PIK3CA gene mutation in ER+/HER2− breast cancer were associated with relatively low level mTORC1 signaling and a good prognosis following tamoxifen monotherapy [27]. Abe et al [28] established that PIK3CA gene mutation may serve as a biomarker for good prognosis in ovarian clear cell cancer. In contrast however, several researchers have determined that PIK3CA gene mutations are associated with unfavorable prognosis in colorectal cancer [30], although this correlation has not been established in other cohorts [29]. As discussed, in studies of ESCC, PIK3CA gene mutation has been detected in 2.2% to 21% of analyzed cases [14,15]. In a small number of cases (n = 80) [30], no significant differences in prognosis were observed between ESCC patient groups with or without PIK3CA mutation. However, in larger ESCC cohorts (n = 219), favorable prognosis has

357 been correlated with PIK3CA gene mutation [19]. In our cohort of 96 ESCC patients, PIK3CA mutations predominantly occurred in ESCC patients with non–lymph node metastases, indicating that PIK3CA mutations in ESCC are associated with better outcomes. Furthermore, when correlated with OS and DFS, a trend toward favorable prognosis was observed in OS. In summary, we report PIK3CA gene mutation in 12.5% of tumor samples from a cohort of 96 ESCC Chinese patients. A higher frequency of PIK3CA mutation was observed in female ESCC patients, in patients without lymph node metastasis, and in middle ESCC. Indeed, the sex differences in PIK3CA mutation rates are striking given the lack of association with Barrett's esophagitis and an absence of heavy smoking/drinking histories in the female patients. Furthermore, our study demonstrates a trend toward favorable prognosis in ESCC patients with tumors harboring PIK3CA gene mutation, suggesting that PIK3CA status may be a favorable predictive biomarker in ESCC patients. Longer follow-up in this study, or future studies with greater numbers of patients are, however, required to fully assess the prognostic value of PIK3CA gene mutation in ESCC patients.

References [1] Samuels Y, Wang Z, Bardelli A, et al. High frequency of mutations of the PIK3CA gene in human cancers. Science 2004;304:554. [2] Samuels Y, Velculescu VE. Oncogenic mutations of PIK3CA in human cancers. Cell Cycle 2004;3:1221-4. [3] Samuels Y, Ericson K. Oncogenic PI3K and its role in cancer. Curr Opin Oncol 2006;18:77-82. [4] Samuels Y, Diaz Jr LA, et al. Mutant PIK3CA promotes cell growth and invasion of human cancer cells. Cancer Cell 2005;7:561-73. [5] Samuels Y, Waldman T. Oncogenic mutations of PIK3CA in human cancers. Curr Top Microbiol Immunol 2010;347:21-41. [6] Zhao L, Vogt PK. Class I PI3K in oncogenic cellular transformation. Oncogene 2008;27:5486-96. [7] Liao X, Morikawa T, Lochhead P, et al. Prognostic role of PIK3CA mutation in colorectal cancer: cohort study and literature review. Clin Cancer Res 2012;18:2257-68. [8] Razis E, Bobos M, Kotoula V, et al. Evaluation of the association of PIK3CA mutations and PTEN loss with efficacy of trastuzumab therapy in metastatic breast cancer. Breast Cancer Res Treat 2011;128:447-56. [9] Konopka B, Janiec-Jankowska A, Kwiatkowska E, et al. PIK3CA mutations and amplification in endometrioid endometrial carcinomas: relation to other genetic defects and clinicopathologic status of the tumors. HUM PATHOL 2011;42:1710-9. [10] Wallin JJ, Edgar KA, Guan J, et al. Gdc-0980 is a novel class I PI3K/mTOR kinase inhibitor with robust activity in cancer models driven by the PI3K pathway. Mol Cancer Ther 2011;10:2426-36. [11] Stein RC, Waterfield MD. PI3-kinase inhibition: a target for drug development? Mol Med Today 2000;6:347-57. [12] Yuan J, Mehta PP, Yin MJ, et al. Pf-04691502, a potent and selective oral inhibitor of PI3K and mTOR kinases with antitumor activity. Mol Cancer Ther 2011;10:2189-99. [13] Brachmann SM, Hofmann I, Schnell C, et al. Specific apoptosis induction by the dual PI3K/mTOR inhibitor nvp-bez235 in HER2 amplified and PIK3CA mutant breast cancer cells. Proc Natl Acad Sci U S A 2009;106:22299-304.

358 [14] Montgomery E, Field JK, Boffetta P, Daigo Y, Shimizu M, Shimoda T. Squamous cell carcinoma of the oesophagus. In: Bosman FT, Carneiro F, Hruban RH, Theise ND, editors. WHO Classification of Tumours of the Digestive System. 4th ed. Lyon: IARC Press; 2010. p. 18-24. [15] Pennathur A, Gibson MK, Jobe BA, Luketich JD. Oesophageal carcinoma. Lancet 2013;381:400-12. [16] Phillips WA, Russell SE, Ciavarella ML, et al. Mutation analysis of PIK3CA and PIK3CB in esophageal cancer and Barrett's esophagus. Int J Cancer 2006;118:2644-6. [17] Mori R, Ishiguro H, Kimura M, et al. PIK3CA mutation status in Japanese esophageal squamous cell carcinoma. J Surg Res 2008;145: 320-6. [18] Akagi I, Miyashita M, Makino H, et al. Overexpression of PIK3CA is associated with lymph node metastasis in esophageal squamous cell carcinoma. Int J Oncol 2009;34:767-75. [19] Shigaki H, Baba Y, Watanabe M, et al. PIK3CA mutation is associated with a favorable prognosis among patients with curatively resected esophageal squamous cell carcinoma. Clin Cancer Res 2013;19:2451-9. [20] Kaifi JT, Gusani NJ, Jiang Y, et al. Multidisciplinary management of early and locally advanced esophageal cancer. J Clin Gastroenterol 2011;45:391-9. [21] Wu S, Zhu Z, He J, Luo X, Xu J, Ren-Heidenreich L. A novel mutantenriched liquid chip technology for the qualitative detection of somatic mutations in KRAS gene from both serum and tissue samples. Clin Chem Lab Med 2010;48:1103-6. [22] Benvenuti S, Frattini M, Arena S, et al. PIK3CA cancer mutations display gender and tissue specificity patterns. Hum Mutat 2008;29:284-8.

J. Hou et al. [23] Bachman KE, Argani P, Samuels Y, et al. The PIK3CA gene is mutated with high frequency in human breast cancers. Cancer Biol Ther 2004;3:772-5. [24] Lee JW, Soung YH, Kim SY, et al. PIK3CA gene is frequently mutated in breast carcinomas and hepatocellular carcinomas. Oncogene 2005;24:1477-80. [25] Saal LH, Holm K, Maurer M, et al. PIK3CA mutations correlate with hormone receptors, node metastasis, and Erbb2, and are mutually exclusive with PTEN loss in human breast carcinoma. Cancer Res 2005;65:2554-9. [26] Broderick DK, Di C, Parrett TJ, et al. Mutations of PIK3CA in anaplastic oligodendrogliomas, high-grade astrocytomas, and medulloblastomas. Cancer Res 2004;64:5048-50. [27] Loi S, Haibe-Kains B, Majjaj S, et al. PIK3CA mutations associated with gene signature of low mTORC1 signaling and better outcomes in estrogen receptor-positive breast cancer. Proc Natl Acad Sci U S A 2009;107:10208-13. [28] Abe A, Minaguchi T, Ochi H, et al. PIK3CA overexpression is a possible prognostic factor for favorable survival in ovarian clear cell carcinoma. HUM PATHOL 2013;44:199-207. [29] Eklof V, Wikberg ML, Edin S, et al. The prognostic role of KRAS, BRAF, PIK3CA and PTEN in colorectal cancer. Br J Cancer 2013;108: 2153-63. [30] Maeng CH, Lee J, van Hummelen P, et al. High-throughput genotyping in metastatic esophageal squamous cell carcinoma identifies phosphoinositide-3-kinase and BRAF mutations. PLoS One 2012;7:e41655.