Target-Based Therapeutic Matching in Early-Phase Clinical Trials in Patients with Advanced Colorectal Cancer and PIK3CA Mutations

The PIK3CA gene encodes the 110α subunit of phosphoinositide 3-kinase (PI3K) and is commonly mutated in a myriad of human cancers (1). PIK3CA mutations activate the PI3K/AKT/mammalian target of rapamycin (mTOR) pathway, which leads to carcinogenesis and tumor progression (2–4). Preclinical and early clinical data suggest that PIK3CA mutations can render tumors sensitive to PI3K/AKT/mTOR pathway inhibition, whereas simultaneous KRAS mutations can drive therapeutic resistance (3, 5–9). Many of the latest advances in cancer medicine have occurred when tumor-specific molecular abnormalities were matched with appropriately selected targeted therapies (10–12). Examples in solid tumors include treatment with KIT inhibitors in gastrointestinal stromal tumors with KIT mutations (13), EGF receptor (EGFR) inhibitors in non–small cell lung cancer harboring EGFR mutations (14), and BRAF inhibitors in melanoma with BRAF mutations (15, 16). It is plausible that matching patients with colorectal cancer harboring PIK3CA mutations with therapies targeting the PI3K/AKT/mTOR pathway may lead to improved therapeutic benefit, as has been suggested in breast and gynecologic cancers (7, 8). PIK3CA mutations occur in approximately 17% of colorectal cancers; however, there are limited data on the outcomes of matched targeting of the PI3K/AKT/mTOR pathway in these patients (17–20). We investigated patients with colorectal cancer referred to the Clinical Center for Targeted Therapy at MD Anderson Cancer Center (MD Anderson) for the presence of PIK3CA mutations and analyzed their treatment outcomes.

Patients with advanced colorectal cancer refractory to standard therapies referred for early clinical trials with targeted therapeutic agents to the Clinical Center for Targeted Therapy at MD Anderson were eligible for analysis provided they had adequate tissue available for mutation analysis. The registration of patients in the database, pathology assessment, and mutation analysis were performed at MD Anderson. All treatments and analyses were performed in accordance with MD Anderson IRB guidelines.

PIK3CA and KRAS mutations were investigated in archival formalin-fixed, paraffin-embedded tissue blocks or material from fine needle aspiration biopsy obtained from diagnostic and/or therapeutic procedures. All histologies were centrally reviewed at MD Anderson. PIK3CA and KRAS mutation testing was done at the Clinical Laboratory Improvement Amendment (CLIA)–certified Molecular Diagnostic Laboratory within the Division of Pathology and Laboratory Medicine at MD Anderson. DNA was extracted from microdissected, paraffin-embedded tumor sections and further studied using a PCR-based DNA sequencing method for PIK3CA mutations in codons c532 to c554 of exon 9 (helical domain) and c1011 to c1062 of exon 20 (kinase domain), which included the mutation hotspot region of the PIK3CA protooncogene by Sanger sequencing after amplification of 276- and 198-bp amplicons, respectively, using primers designed by the MD Anderson Molecular Diagnostic Laboratory. After January 2011, the assay used was mass spectrometric detection (Sequenom MassARRAY) to screen for the mutational hot spots in exon 1 (Q60K, R88Q, E110K, and K111N), exon 4 (N345K), exon 6 (S405S), exon 7 (E418K, C420R, and E453K), exon 9 [P539R, E542 (base 1 and 2), E545 (all 3 bases), and Q546 (base 1 and 2)], exon 18 (F909L), and exon 20 [Y1021 (base 1 and 2), T1025 (base 1), M1043I, M1043V, A1046V, H1047Y, H1047R, H1047L, and G1049R]. The mutations identified during the initial screening were confirmed by Sanger sequencing assay. The lower limit of detection is approximately 10%. In addition, whenever possible, mutation analyses for KRAS, NRAS codons 12 and 13, and 61 mutations of exons 2–3 and BRAF mutations in exon 15 were carried out using PCR-based DNA sequencing mutation, as previously described (21).

Consecutive patients with underlying PIK3CA mutations were offered, whenever possible, a clinical trial, which included an inhibitor of the PI3K/AKT/mTOR pathway. Treatment continued until disease progression or the occurrence of unacceptable toxicity. Treatment was carried out according to the requisites in the treatment protocols selected. Assessments, including history, physical examination, and laboratory evaluations, were performed as specified in each protocol, typically before the initiation of therapy, weekly during the first cycle, and then, at a minimum, at the beginning of each new treatment cycle. Efficacy was assessed from computed tomography (CT) scans and/or MRI at baseline before treatment initiation and then every 2 cycles (6–8 weeks). All radiographs were read in the Department of Radiology at MD Anderson and reviewed in the Department of Investigational Cancer Therapeutics tumor measurement clinic. Responses were categorized per Response Evaluation Criteria in Solid Tumors (RECIST) 1.0 (22). In brief, complete response (CR) was defined as the disappearance of all measurable and nonmeasurable disease; partial response (PR) was defined as at least a 30% decrease in the sum of the longest diameter of measurable target lesions; progressive disease was defined as at least a 20% increase in the sum of the longest diameter of measurable target lesions, or unequivocal progression of a nontarget lesion, or the appearance of a new lesion; and stable disease was defined as neither sufficient shrinkage to qualify for PR nor sufficient increase to qualify for progressive disease.

Two-way contingency tables were used to summarize the relationship between two categorical variables. A Fisher exact test was used to assess the association among categorical variables and PIK3CA mutation status. Progression-free survival (PFS), estimated by the Kaplan–Meier method, was defined as the time interval from the start of therapy to the first observation of disease progression or death, whichever occurred first. Patients alive and without disease progression were censored at the last follow-up date. All statistical analyses were carried out using SPSS 19 computer software (SPSS).

A total of 194 patients with advanced colorectal cancer were screened for the presence of PIK3CA mutations, starting January 2009. Their median age was 58 years (range, 25 to 81 years); 140 (72%) were White, 28 (14%) African American, 19 (10%) Hispanic, and 7 (4%) Asian. Detailed patient characteristics are listed in Table 1.

PIK3CA mutations were detected in 31 (16%) of the 194 patients, with 22 mutations in the kinase domain in exon 9, and 8 mutations in the helical domain in exon 20. The most frequent point mutation was E545K (1633G>A) in 11 (35%) patients, followed by E542K (1624G>A) in 8 patients (26%), and H1047L (3140A>T) in 3 (10%) patients (Table 2). No association between PIK3CA mutation and age or ethnicity or tumor site was identified (Table 1).

Of the 194 patients, 189 patients were tested for KRAS mutations and 90 (47%) had a mutation. Among patients with KRAS mutations, the most prevalent ones were G12D (35G>A) in 30% (27/90), G12V (35G>T) in 20% (18/90), G13D (38G>A) in 11% (10/90), and G12A (35G>C) in 11% (10/90) of patients. Patients with PIK3CA mutations were more likely to have simultaneous KRAS mutations compared with patients with wild-type (WT) PIK3CA (71%, 22/31 vs. 43%, 68/158; P = 0.006). KRAS mutations compared with WT KRAS were associated with PIK3CA mutations in exon 9 (18%, 16/90 vs. 6%, 6/99; P = 0.01), but not in exon 20 (6%. 5/90 vs. 3%, 3/99; P = 0.48; Table 3).

Of the 194 patients, 167 patients were tested for BRAF mutations and 11 (7%) had mutated BRAF (10 patients had a V600E mutation and 1 patient had a D594G mutation). All patients with a BRAF mutation were negative for KRAS mutations. Of these 11 patients with BRAF mutations, 2 (18%) had coexistent PIK3CA mutations. There was no difference in the incidence of BRAF mutations among those with or without PIK3CA mutations (7%, 2/27 vs. 6%, 9/140; P = 0.69).

Of the 31 patients with PIK3CA mutations, 17 (55%) were treated in clinical trials including a PI3K/AKT/mTOR pathway inhibitor. These patients had been refractory to a median of four prior therapies (range, 2–7). Most patients (11/17, 65%) received mTORC1 inhibitor (rapalog)-based therapy; 5 of 17 (29%) received PI3K inhibitor-based therapy, and 1 (6%) received an AKT inhibitor-based therapy (Supplementary Table S1). Most patients (13/17, 76%) received PI3K/AKT/mTOR inhibitors at 100% of the maximum tolerated dose/recommended phase II dose/FDA (U.S. Food and Drug Administration)-approved dose (range, 30%–100%). Of the 11 patients treated with mTORC1 inhibitors, 10 (91%) received the maximum tolerated dose/recommended phase II dose/FDA-approved dose. In contrast, only 1 (20%) of 5 patients treated with PI3K inhibitors was treated with the maximum tolerated dose (Supplementary Table S1). Of the 17 patients, none achieved PRs or CRs and only 1 (6%; 95% CI, 0.01–0.27) patient had stable disease for more than 6 months (stable disease ≥6; Fig. 1). This was similar to the stable disease ≥6/PR/CR rate of 16% (11/67; 95% CI, 0.09–0.27) in patients with colorectal cancer without PIK3CA mutations treated on the same protocols targeting the PI3K/AKT/mTOR pathway (P = 0.44).

We also reviewed 14 of 31 patients (45%) with PIK3CA mutations who were not treated with PI3K/AKT/mTOR inhibitors. Of these 14 patients, 5 were not treated because of ineligibility or patient/doctor preference and 9 received other experimental therapies, often because PIK3CA status was not available at the time of decision making. None (0%; 95% CI, 0.00–0.29) of these 9 patients attained a PR or stable disease ≥6, which was similar to stable disease ≥6/PR/CR rate of 6% (1/17; 95% CI, 0.01–0.27) in the remaining 17 patients treated with PI3K/AKT/mTOR inhibitors (P = 1.00).

The median PFS for these 17 patients with a PIK3CA mutation on target-matched therapy with PI3K/AKT/mTOR inhibitors was 1.9 months (95% CI; 1.5–2.3). There was no difference in median PFS in 11 patients with KRAS mutations versus 6 patients with WT KRAS (1.8 months; 95%CI, 1.4–2.2 vs. 1.9 months; 95%CI, 1.1–2.7; P = 0.59). The same 17 patients had a median PFS of 3.1 months (95% CI; 0.4–5.8) on their last FDA-approved therapy before referral to the Clinical Center for Targeted Therapy (P = 0.1; Fig. 2). Patients (n = 67) without PIK3CA mutations treated with the same therapies with PI3K/AKT/mTOR inhibitors had a similar median PFS (2.3 months; 95%CI, 2.1–2.6) as did patients with PIK3CA mutations treated with PI3K/AKT/mTOR inhibitors (1.9 months; 95% CI, 1.5–2.3; P = 0.44; Fig. 2).

In addition, we reviewed PFS in 9 patients with colorectal cancer and PIK3CA mutations who were treated with experimental therapies other than PI3K/AKT/mTOR inhibitors. The median PFS for those 9 patients was 1.9 months (95%CI, 1.0–2.7), which was similar to the median PFS of 1.9 months (95%CI 1.5–2.3) in 17 patients with PIK3CA mutations treated with matched therapy (P = 0.77; Fig. 2).

We identified PIK3CA mutations in 16% of the 194 patients with colorectal cancer referred to the Clinical Center for Targeted Therapy. The most common types of PIK3CA mutations associated with colorectal cancer in our patients were E545K and E542K mutations in exon 9. These results are similar to those in published studies (18, 19). Earlier studies suggested that H1047R mutations may be more sensitive to PI3K/AKT/mTOR pathway inhibition, but in our study, only 2 patients (6%) had this particular mutation and that can potentially account for their lack of therapeutic response (8).

We found a strong association between KRAS mutations and PIK3CA mutations in our patients that supports previously published data (18, 19, 23–25). We found that KRAS mutations were more frequently associated with exon 9 than exon 20 PIK3CA mutations. A similar association was noted by De Roock and colleagues in a study of 743 patients with metastatic colorectal cancer, but a larger study of 1,170 patients with predominantly localized colorectal cancer did not find a similar association (18, 19).

The paradigm for matching targeted therapy based on specific molecular defects identified in the cancer cell was established by the remarkable success of imatinib in chronic myeloid leukemia (26). This model has been successfully replicated in solid tumors such as BRAF V600E mutated melanoma and non–small cell lung cancer with ALK rearrangement (15, 16, 27). In the context of targeting PIK3CA-mutated cancers with PI3K/AKT/mTOR inhibitors, we have observed encouraging responses in heavily pretreated patients with breast and gynecologic malignancies and PIK3CA mutations treated with PI3K/AKT/mTOR pathway inhibitors (9, 28, 29). In the current study, 17 heavily pretreated patients with colorectal cancer and PIK3CA mutations treated with PI3K/AKT/mTOR inhibitors did not achieve better outcomes compared with patients with colorectal cancer without PIK3CA mutations treated with identical therapies (stable disease ≥6/PR/CR rate 6% vs.16%; P = 0.44) and their outcomes were strikingly similar response rates of 4% to 11% previously reported from early-phase clinical trials in an unselected population, including a 1.4% response rate and 9% rate of stable disease at 6 months for unselected patients with colorectal cancer from our center (30–33). In addition, the median PFS in patients with PIK3CA mutations treated with PI3K/AKT/mTOR inhibitors demonstrated a trend toward inferiority to the median PFS achieved by the same patients on their last FDA-approved therapy (1.9 months vs. 3.1 months; P = 0.1). Our observations are in line with the experience from a major European cancer center reporting lack of therapeutic benefit with therapeutic matching (including PIK3CA) in colorectal cancer (20). Reasons why these patients do not derive similar benefits from PI3K/AKT/mTOR-targeted therapies as patients do with PIK3CA mutations and breast or gynecologic cancers remains unknown. Preclinical models demonstrated that simultaneous mutations in the mitogen-activated protein kinase (MAPK) pathway (RAS, RAF, or MEK) can negate the effect of mTOR inhibitors in patients with PIK3CA mutations (5, 6). Therefore, it is plausible that this strong association between PIK3CA and KRAS mutations in colorectal cancer can contribute to therapeutic resistance as, in our series, 71% of patients with colorectal cancer and PIK3CA mutations had simultaneous KRAS mutations. In contrast, in patients with breast and gynecologic malignancies with PIK3CA mutations, only 23% had coexisting KRAS mutations, which were often in different exons than in colorectal cancer (9). Furthermore, Shimizu and colleagues analyzed phase I therapies targeting PI3K and MAPK pathways in diverse advanced cancers and demonstrated superior outcomes with combined PI3K and MAPK inhibition in patients with cancers having alterations in both pathways (20, 34).

The majority of patients in our study were treated with rapalogs and it is possible that the outcomes with these inhibitors may be inferior due to feedback activation of AKT, as has been previously described in vitro and in patients (35, 36). Other biomarkers besides activating mutations in PIK3CA may be required to select for inhibitors of the PI3K pathway, including loss of the tumor suppressor PTEN or gene amplification of members the insulin growth factor family. Ongoing clinical efforts in PIK3CA-mutant colorectal cancer are using AKT or PI3K inhibitors and excluding concurrent KRAS mutations in an effort to isolate a responsive population.

Our study has several important limitations. First, there were only a small number of patients with PIK3CA mutations who were treated with PI3K/AKT/mTOR inhibitors; the majority of patients (65%) received mTORC1-targeting agents. Second, the analysis was retrospective. Third, patients received drugs that impacted different parts of the pathway (mTORC1 inhibitors, 11 patients; PI3K inhibitors, 5 patients; AKT inhibitor, 1 patient), although the latter might also imply that conclusions are not confined to one type of drug. Fourth, doses were variable. For instance, only 76% of patients received a maximum tolerated dose/recommended phase II dose/FDA-approved dose of a PI3K/AKT/mTOR inhibitor. Of interest, 91% of patients treated with mTORC1 inhibitors received a maximum tolerated dose/recommended phase II dose/FDA-approved dose compared with only 20% of patients treated with a PI3K inhibitor, which precludes firm conclusions about the latter class of drugs (32, 37). Fifth, none of the patients in our analysis received combined PI3K and MAPK-targeted therapy.

In conclusion, our data, although preliminary, suggest that targeted matched therapy with PI3K/AKT/mTOR pathway inhibitors (in particular mTORC1 inhibitors) in patients with metastatic colorectal cancer harboring PIK3CA mutations may be associated with minimal activity. The lack of benefit can conceivably be explained, in part, by the high prevalence of coexisting KRAS mutations driving therapeutic resistance.

F. Janku received commercial research grants from Novartis, Roche, Biocartis, Transgenomic, and Trovagene and is a consultant/advisory board member of Trovagene. S. Kopetz is a consultant/advisory board member of Roche, Sanofi, Amgen, Bristol-Myer Squibb, and Bayer. R. Kurzrock has research support from GlaxoSmithKline, Novartis, Merck, and Bayer. No potential conflicts of interest were disclosed by the other authors.

Conception and design: F. Janku, A.M. Tsimberidou, R. Kurzrock

Development of methodology: P. Ganesan, F. Janku, A.M. Tsimberidou, R. Luthra

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): P. Ganesan, F. Janku, A. Naing, D.S. Hong, A.M. Tsimberidou, G.S. Falchook, J.J. Wheler, S.A. Piha-Paul, S. Fu, R. Luthra, E.S. Kopetz

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): P. Ganesan, F. Janku, A. Naing, S. Fu, J.J. Lee, R. Luthra

Writing, review, and/or revision of the manuscript: P. Ganesan, F. Janku, A. Naing, D.S. Hong, A.M. Tsimberidou, G.S. Falchook, S.A. Piha-Paul, S. Fu, J.J. Lee, M.J. Overman, R.A. Wolff, R. Kurzrock

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): F. Janku, A.M. Tsimberidou, V.M. Stepanek

Study supervision: F. Janku, V.M. Stepanek

The authors thank Joann Aaron for scientific review and editing of the article.

This work was supported in part by NIH grant U01 CA062461 (to R. Kurzrock). Molecular testing was supported in part by the Sheikh Khalifa Bin Zayed Al Nahyan Institute for Personalized Cancer Therapy.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.