Targeted PI3K/AKT/mTOR therapy for metastatic carcinomas of the cervix: A phase I clinical experience.

BACKGROUND
Activated PI3K/AKT/mTOR pathway frequently occurs in metastatic or recurrent cervical carcinomas. However, the clinical benefits of matched therapy, a therapeutic approach targeting a specific mutational abnormality, have not yet been established.


METHODS
We analyzed the outcomes of patients with metastatic or recurrent cervical carcinomas who had a test for PIK3CA mutation and/or PTEN loss/mutation, and received ≥1 phase I therapeutic regimen between January 2006 and June 2013.


RESULTS
Patients with adenocarcinoma had fewer PIK3CA mutations (14%), and survived longer (median, 14.2 months) than those with squamous cell carcinoma (48% and 7.2 months; p = 0.016, and 0.001, respectively). Matched therapy targeting the activated PI3K/AKT/mTOR pathway led to a favorable rate of SD ≥ 6 months/CR/PR (53%) and significantly longer progression-free survival (median, 6.0 months) than non-matched therapy (11% and 1.5 months; p = 0.08 and 0.026; respectively). In patients with squamous cell carcinoma of the cervix, the presence of PIK3CA mutations was associated with a significantly longer overall survival (median, 9.4 months) than the absence of PIK3CA mutations (median, 4.2 months; p = 0.019).


CONCLUSIONS
Matched therapy targeting the activated PI3K/AKT/mTOR pathway provided meaningful clinical benefits. Thus, further evaluation of PI3K/AKT/mTOR pathway targeted therapy is warranted, especially in metastatic or recurrent squamous cell carcinoma.


INTRODUCTION
Cancer of the uterine cervix is one of the most common gynecologic malignancies and causes of death worldwide [1]. In the United States, more than 12,000 women were diagnosed with cervical cancer in 2013, and more than 4,000 patients died from this disease [2]. Squamous cell carcinoma and adenocarcinoma account for about 95% of all cervical cancers [3]. Most patients who are diagnosed with early-stage disease have a high rate of long-term survival after they have undergone curative surgical resection and/or chemoradiotherapy [4,5]. Patients with recurrent or metastatic cervical cancers who are not amenable to radical local excision or curative chemoradiotherapy, have a poor prognosis [6,7]. Palliative systemic therapies such as platinum-based doublets plus an anti-angiogenic agent as the standard of care for firstline treatment yield modest survival gains of 3.7 months [8][9][10]. Subsequent conventional chemotherapeutic regimens result in increased toxicity with limited clinical www.impactjournals.com/oncotarget benefit [11]. The overall poor prognosis of these patients warrants the urgency for the development of novel therapeutic regimens [12,13].
There are three classes of phosphoinositide 3-kinase (PI3K) isoforms [14]. Class I PI3Ks are heterodimeric lipid kinases composed of the p110 catalytic subunit and the p85 regulatory subunit. Three major p110 isoforms have been described: p110α (PIK3CA), p110β (PIK3CB), and p110δ (PIK3CD) [15,16]. PI3K and PTEN counterpoise each other on conversion between phosphatidyl-inositol-4-5-bisphosphate (PIP2) and phosphatidyl-inositol-3, 4, 5-trisphosphate (PIP3) [17,18]. Upon activation through either PIK3CA mutation or PTEN loss/mutation, PI3K phosphorylates PIP2 to PIP3, which facilitates recruitment and activation of AKT to initiate a cascade of downstream signaling events including the mTOR complex, a major downstream pathway [16]. Therefore, a regimen including a PI3K inhibitor and/or an mTOR inhibitor can be utilized to target PIK3CA mutation and/or PTEN loss/mutationmediated activated PI3K/AKT/mTOR pathway [19]. This strategy is defined as matched therapy: a therapeutic regimen including an agent (either as a single agent or as a part of a combination regimen) that is known to directly inhibit a specific mutation, and/or to inhibit its key downstream pathways, such as treatment with a PI3K inhibitor, an mTOR inhibitor, or these agent-based regimens for a cancer patient carrying a PIK3CA mutation and/or PTEN loss/mutation [20].
Our previous study and others have documented that the activated PI3K/AKT/mTOR pathway is frequently identified in patients with metastatic or recurrent cervical cancers [21][22][23][24][25]. We hypothesized that cervix cancer patients with aberrations in this pathway would achieve clinical benefit (defined as objective response and prolonged stable disease) when treated with PI3K/AKT/ mTOR pathway targeted agents. In this article, we further our previous study by limiting patients who only have either squamous cell carcinoma or adenocarcinoma of the cervix in order to reach definitive conclusions, which will be applied to facilitate future drug development by choosing appropriate therapeutic regimens for appropriate cancer patients. Herein we document a clear observation of antitumor responses with acceptable toxicity from PI3K/AKT/mTOR matched therapy, representing an important new therapeutic strategy for a patient population possessing limited treatment options.

Patient selection
Consecutive patients with metastatic or recurrent cervical carcinoma (either squamous cell carcinoma or adenocarcinoma) who were referred to the Department of Investigational Cancer Therapeutics (A Phase I Clinical Trials Program) at MD Anderson between January 1, 2006, and June 30, 2013, had a test for PIK3CA mutation and/or PTEN loss/mutation in a Clinical Laboratory Improvement Amendments-certified molecular diagnostic laboratory, and received treatment in at least one phase I clinical trial, were evaluated. Follow-up was defined as the time of the initial office visit after the patient was referred to the phase I clinic until death or February 28, 2014. This study was conducted in accordance with MD Anderson's Institutional Review Board guidelines.

Data collection
Two individuals independently reviewed patients' electronic medical records at MD Anderson and crosschecked the collected data. Clinical information that was collected included race, prior treatment history (e.g., surgery, radiation therapy, and chemotherapy), date of birth, Eastern Cooperative Oncology Group performance status at the initial phase I clinic visit, mutation profiling of the tumor specimen including PIK3CA mutation and/or PTEN loss/mutation status, phase I clinical trial therapies, and clinical outcomes: severe adverse event (SAEs), progression-free survival (PFS), overall survival (OS), and objective responses including complete remission (CR), partial response (PR), and stable disease for 6 months or longer (SD ≥ 6months).
SAEs were defined as toxic effects that were grade 3 or higher according to the National Cancer Institute Common Terminology Criteria for Adverse Events v3.0 or v4.0 (http://ctep.cancer.gov/reporting/ctc.html) [26]. Clinical objective responses were evaluated according to the Response Evaluation Criteria in Solid Tumors version 1.0 or 1.1 per individual study protocols [27,28]. PFS was defined as the interval from the date of initial treatment to the first objective documentation of disease progression, the time of death if it occurred, or the last date of contact on February 28, 2014, at which time the patients' data were censored. OS was estimated from the date of the initial phase I clinical trial therapy to death or the last date of contact on February 28, 2014, at which time the patients' data were censored.
Enrollment of an eligible patient into a specific phase I trial depended on the availability of the phase I trial at the time of presentation and the preference of the treating physician according to good clinical practice. If a phase I trial was unsuccessful, another available phase I trial was introduced as long as a patient was eligible and willing to participate.

Statistical analyses
Categorical data were described using contingency tables. Continuously scaled measures were summarized with descriptive statistical measures (i.e., the median with the range), whereas PFS and OS rates were estimated using the Kaplan-Meier method. Patients who were still alive at the time of data analysis were censored at that time. Fisher's exact test was used to assess the associations between categorical variables and mutation status. Statistical inferences were based on two-sided tests at a significance level of p < 0.05. Statistical analyses were carried out using SPSS Statistics version 22 (IBM Inc., Armonk, NY).

Study population
Fifty-five consecutive patients with metastatic or recurrent adenocarcinoma (n = 24) or squamous cell carcinoma (n = 31) who had a test for PIK3CA mutation and/or PTEN loss/mutation and underwent treatment in at least one phase I clinical trial in the Phase I Clinical Trials Program were included in this study. Patient characteristics and molecular profiling are listed in Table 1. All patients had received at least one prior systemic chemotherapeutic regimen for metastatic or recurrent disease (median 2 regimens, range 1 to 4) and underwent molecular marker studies in a Clinical Laboratory Improvement Amendments-certified molecular diagnostic laboratory.

Severe adverse events
All patients were included for analyses of SAEs; a total of 450 cycles of therapy were administered according to 38 phase I clinical trials [21]. During their first phase I trial therapy, approximately 20% of patients (n = 11/55) experienced 19 episodes of SAEs: neutropenia and thrombocytopenia (n = 5 each, 9%); hyponatremia (n = 2, 4%); and hypokalemia, hypomagnesaemia, hypophosphatemia, mucositis, fatigue, bowel perforation, and bacterial infection (n = 1 each, 2%). During the second phase I trial therapy, only 1 (9%) of 11 patients had an SAE (fatigue). Subsequently, no patient experienced an SAE before all 4 patients progressed within the initial 8 weeks of their phase I trial therapy. In patients who received PI3K/AKT/mTOR pathway targeted therapeutic regimens, no apparent difference was observed between patients with PI3KCA mutation and/or PTEN loss/ mutation and those without.
Of 11 patients who received their second phase I clinical trial therapy, 27% achieved SD ≥ 6 months/CR/ PR (PR = 1 and SD ≥ 6 months = 2), and the median PFS in this group was 4.2 months (95% CI, 3.0-5.4). Of 4 patients who received their third phase I clinical trial therapy, no patient achieved SD ≥ 6 months/CR/PR.

DISCUSSION
In this study, we identified that patients with previously treated, locally advanced or metastatic cervical cancer harboring mutations in the PI3K/AKT/mTOR www.impactjournals.com/oncotarget We conclude from these observations that targeted therapy among these patients with pathway potentiation is a viable strategy for further development.

Figure 2: Kaplan-Meier plots for survivals. In
Several observations are of note. In general, patients with metastatic or recurrent squamous cell carcinoma of the cervix were significantly younger, and had a higher prevalence of PIK3CA mutations. However, they displayed lower antitumor activity to currently available phase I clinical trials at MD Anderson associated with significantly shorter OS than patients with metastatic or recurrent adenocarcinoma of the cervix. Second, matched therapy targeting the activated PI3K/AKT/mTOR pathway in patients with metastatic or recurrent squamous cell carcinoma of the cervix led to a favorably higher rate  were associated with a median OS of 6 months (n=3; 95% CI, 2-10), similar to PTEN intact (in blue) with 7 months (n = 21; 95% CI, 3.1-10.9; p = 0.762). In Figure 5B for patients with adenocarcinoma, PTEN aberrations (in green) were associated with a median OS of 11.3 months (n = 4; 95% CI, 0-22.6), similar to PTEN intact (in blue) with 15.7 months (n = 16; 95% CI, 7.1-24.3; p = 0.791).
of SD ≥ 6 months/CR/PR as well as significantly longer PFS and OS than non-matched therapy. However, in patients with metastatic or recurrent adenocarcinoma of the cervix, there was no significant difference in OS associated with matched therapy in spite of a higher rate of SD ≥ 6 months/CR/PR and significantly longer PFS, which might be owing to at least three potential factors: lower prevalence of the activated PI3K/AKT/ mTOR pathway, intrinsic sensitivity to novel phase I trial therapy available at MD Anderson phase I service, and the presence of PIK3CA mutation and/or PTEN loss/ mutation that cannot stratify for aggressiveness of disease. Third, PIK3CA mutation, but not PTEN loss/mutation, was associated with significantly longer OS, indicating the differential effects of these genetic aberrations on sensitivity to matched therapy, as well as matched patients and the preferential choices of the treating physicians in the Phase I Clinical Trials Program, MD Anderson for assigning patients to matched therapy.
Of 136 patients with PIK3CA-mutation and/or PTEN-loss/mutation advanced solid tumors seen in our phase I clinic, 25% achieved SD ≥ 6 months/CR/PR (95%CI, 0.18-0.33) after receiving matched therapy targeting the activated PI3K/AKT/mTOR pathway, and a median PFS of 2.5 months (95% CI, 1.8-3.2) [31]. These outcomes were significantly lower than those reported in the current study of patients with PIK3CAmutation and/or PTEN-loss/mutation metastatic or recurrent carcinoma receiving matched therapy. Several confounding factors might have contributed to these differences. Almost half of the patients with metastatic or recurrent cervical carcinoma, especially those with squamous cell carcinoma, presented with PIK3CA mutation and/or PTEN loss/mutation, and this proportion was significantly higher in patients with advanced solid tumors (48% versus 22%, p = 0.002), suggesting that the activated PI3K/AKT/mTOR pathway is a driving mechanism for the survival of cervical carcinoma cells. Another factor is that concurrent mutations or specific molecular profiles might be more important than a single gene mutation. PIK3CA mutation was more prevalent in patients with KRAS mutation than in patients with wildtype KRAS [31], whereas in this report, a low frequency of KRAS mutation was found in metastatic or recurrent cervical carcinomas, and in squamous cell carcinomas specifically, no KRAS mutation was identified in any of the 26 tested patients (Table 1). Since PIK3CA mutation and/ or PTEN loss/mutation with simultaneous KRAS mutation were associated with significantly lower antitumor activity and shorter PFS than PIK3CA mutation and/or PTEN loss/mutation without simultaneous KRAS mutation, two or more coexisting mutations constitute different classes of mutation profiles that predict responses to various biologically targeting agents and/or their combinations and render matched therapy more complicated than expected. When developing efficacious regimens to target the activated PI3K/AKT/mTOR pathway, metastatic or recurrent squamous cell carcinomas of the cervix might be an appropriate clinical model to be tested in early-phase clinical trials because of their high prevalence of PIK3CA mutation and/or PTEN loss/mutation as previously reported [15,[32][33][34][35] and their lack of simultaneous KRAS mutation, in agreement with the hypothesis that coexisting KRAS mutation become resistant to regimens targeting the activated PI3K/AKT/mTOR pathway.
When considering the clinical relevance of our findings, several limitations should be kept in mind. First, matched therapy includes not only phase I clinical trials of a single agent targeting the activated PI3K/ AKT/mTOR pathway, but also phase I clinical trials of an agent targeting this pathway in combination with other biologically targeted agents or conventional chemotherapeutic agents. The importance of using an agent to target the activated PI3K/AKT/mTOR pathway might be overestimated since tumor control might be induced by simultaneous inhibition of other key targets and/or processes by the combination regimens. Second, patients with poor clinical outcomes may have been selectively excluded from being referred to a phase I trial because of rapid tumor progression, poor performance status, insufficient organ function, severe comorbidity, economic issues, and inaccessibility of treatment. Third, molecular studies of mutation profiles were usually conducted on archival tumor specimens whenever available, regardless of whether the specimens had been treated in a phase I clinical trial. Finally, we had a limited sample size available for subgroup analyses, which confounded the ability to validate statistical significance in category assessment. Therefore, conclusions from this retrospective study should be considered preliminary evidence to generate hypotheses, which require further validation in larger prospective studies.
In conclusion, our results showed that almost half of the patients with metastatic or recurrent squamous cell carcinoma of the cervix had PIK3CA mutation and/or PTEN loss/mutation without coexisting KRAS mutation, providing an appropriate patient population to the test efficacy of a regimen including a single agent targeting the PI3K/AKT/mTOR pathway or a combination regimen with either another biologically targeted agent and/or conventional chemotherapeutic agent. Matched therapy targeting a specific mutation has provided apparent clinical benefits for cancer patients. Complicated mutation profiles might be more important and useful than a single gene mutation in predicting antitumor activity and clinical efficacy to a specific regimen [36][37][38][39][40]. Therefore, future development and evaluation of novel regimens targeting the activated PI3K/AKT/mTOR pathway for the purpose of translating high antitumor activity to prolonged survival benefit is warranted in larger prospective clinical trials for the treatment cervical carcinoma, especially squamous cell carcinoma.