Dose optimization for cancer treatments with considerations for late-onset toxicities.

Given that novel anticancer therapies have different toxicity profiles and mechanisms of action, it is important to reconsider the current approaches for dose selection. In an effort to move away from considering the maximum tolerated dose as the optimal dose, the Food and Drug Administration Project Optimus points to the need of incorporating long-term toxicity evaluation, given that many of these novel agents lead to late-onset or cumulative toxicities and there are no guidelines on how to handle them. Numerous methods have been proposed to handle late-onset toxicities in dose-finding clinical trials. A summary and comparison of these methods are provided. Moreover, using PI3K inhibitors as a case study, we show how late-onset toxicity can be integrated into the dose-optimization strategy using current available approaches. We illustrate a re-design of this trial to compare the approach to those that only consider early toxicity outcomes and disregard late-onset toxicities. We also provide proposals going forward for dose optimization in early development of novel anticancer agents with considerations for late-onset toxicities.

Incorporating patient-reported outcomes in dose-finding clinical trials with continuous patient enrollment.

Dose-finding clinical trials in oncology estimate the maximum tolerated dose (MTD), based on toxicity obtained from the clinician's perspective. While the collection of patient-reported outcomes (PROs) has been advocated to better inform treatment tolerability, there is a lack of guidance and methods on how to use PROs for dose assignments and recommendations. The PRO continual reassessment method (PRO-CRM) has been proposed to formally incorporate PROs into dose-finding trials. In this paper, we propose two extensions of the PRO-CRM, which allow continuous enrollment of patients and longer toxicity observation windows to capture late-onset or cumulative toxicities by using a weighted likelihood to include the partial toxicity follow-up information. The TITE-PRO-CRM uses both the PRO and the clinician's information during the trial for dose assignment decisions and at the end of the trial to estimate the MTD. The TITE-CRM + PRO uses clinician's information solely to inform dose assignments during the trial and incorporates PRO at the end of the trial for the estimation of the MTD. Simulation studies show that the TITE-PRO-CRM performs similarly to the PRO-CRM in terms of dose recommendation and assignments during the trial while almost halving trial duration in case of an accrual of two patients per observation window. The TITE-CRM + PRO slightly underperforms compared to the TITE-PRO-CRM, but similar performance can be attained by requiring larger sample sizes. We also show that the performance of the proposed methods is robust to higher accrual rates, different toxicity hazards, and correlated time-to-clinician toxicity and time-to-patient toxicity data.

Novel Approaches for Dynamic Visualization of Adverse Event Data in Oncology Clinical Trials: A Case Study Using Immunotherapy Trial S1400-I (SWOG).

PURPOSE: Clinical trial adverse event (AE) data are increasingly complex and high-dimensional, especially for trials evaluating novel targeted agents and immunotherapies. Standard approaches to summarize and analyze AEs remain generally tabular, failing to describe the nature of AEs. Novel dynamic and data visualization methods are needed to enable a more comprehensive assessment of the overall toxicity profile of treatments. METHODS: We developed methods for visualizing the numerous categorizations and types of AEs along with a dynamic approach to better reflect its highly dimensional nature without sacrificing the reporting of rare events. Circular plots displaying the proportion of maximal-grade AEs by system organ classes (SOCs) and butterfly plots displaying the proportion of AEs by severity for each AE term were developed to enable comparisons of AE patterns by treatment arm. These approaches were applied to a randomized phase III trial (S1400I; ClinicalTrials.gov identifier: NCT02785952) comparing nivolumab with nivolumab plus ipilimumab in patients with stage IV squamous non-small-cell lung cancer. RESULTS: Our visualizations revealed that patients randomly assigned to nivolumab and ipilimumab had higher rates of grade 3 or higher AEs compared with nivolumab alone for several SOCs, including musculoskeletal (5.6% v 0.8%), skin (5.6% v 0.8%), vascular (5.6% v 1.6%), and cardiac (4% v 1.6%) toxicities. They also suggested a pattern of higher prevalence of moderate GI and endocrine toxicities and showed that although the rates of cardiac and neurologic toxicities were similar, the types of events were discordant. CONCLUSION: The graphical approaches we proposed enable a more comprehensive and intuitive evaluation of toxicity types by treatment groups, which is not apparent in tabular and descriptive reporting methods.

Surv-CRM-12: A Bayesian phase I/II survival CRM for right-censored toxicity endpoints with competing disease progression.

The growing interest in new classes of anti-cancer agents, such as molecularly-targeted therapies and immunotherapies with modes of action different from those of cytotoxic chemotherapies, has changed the dose-finding paradigm. In this setting, the observation of late-onset toxicity endpoints may be precluded by treatment and trial discontinuation due to disease progression, defining a competing event to toxicity. Trial designs where dose-finding is modeled in the framework of a survival competing risks model appear particularly well-suited. We aim to provide a phase I/II dose-finding design that allows dose-limiting toxicity (DLT) outcomes to be delayed or unobserved due to competing progression within the possibly long observation window. The proposed design named the Survival-continual reassessment method-12, uses survival models for right-censored DLT and progression endpoints. In this competing risks framework, cause-specific hazards for DLT and progression-free of DLT were considered, with model parameters estimated using Bayesian inference. It aims to identify the optimal dose (OD), by minimizing the cumulative incidence of disease progression, given an acceptable toxicity threshold. In a simulation study, design operating characteristics were evaluated and compared to the TITE-BOIN-ET design and a nonparametric benchmark approach. The performance of the proposed method was consistent with the complexity of scenarios as assessed by the nonparametric benchmark. We found that the proposed design presents satisfying operating characteristics in selecting the OD and safety.

Sex Differences in Risk of Severe Adverse Events in Patients Receiving Immunotherapy, Targeted Therapy, or Chemotherapy in Cancer Clinical Trials.

PURPOSE: Women have more adverse events (AEs) from chemotherapy than men, but few studies have investigated sex differences in immune or targeted therapies. We examined AEs by sex across different treatment domains. METHODS: We analyzed treatment-related AEs by sex in SWOG phase II and III clinical trials conducted between 1980 and 2019, excluding sex-specific cancers. AE codes and grade were categorized using the Common Terminology Criteria for Adverse Events. Symptomatic AEs were defined as those aligned with the National Cancer Institute's Patient-Reported Outcome-Common Terminology Criteria for Adverse Events; laboratory-based or observable/measurable AEs were designated as objective (hematologic v nonhematologic). Multivariable logistic regression was used, adjusting for age, race, and disease prognosis. Thirteen symptomatic and 14 objective AE categories were examined. RESULTS: In total, N = 23,296 patients (women, 8,838 [37.9%]; men, 14,458 [62.1%]) from 202 trials experiencing 274,688 AEs were analyzed; 17,417 received chemotherapy, 2,319 received immunotherapy, and 3,560 received targeted therapy. Overall, 64.6% (n = 15,051) experienced one or more severe (grade ≥ 3) AEs. Women had a 34% increased risk of severe AEs compared with men (odds ratio [OR] = 1.34; 95% CI, 1.27 to 1.42; P < .001), including a 49% increased risk among those receiving immunotherapy (OR = 1.49; 95% CI, 1.24 to 1.78; P < .001). Women experienced an increased risk of severe symptomatic AEs among all treatments, especially immunotherapy (OR = 1.66; 95% CI, 1.37 to 2.01; P < .001). Women receiving chemotherapy or immunotherapy experienced increased severe hematologic AE. No statistically significant sex differences in risk of nonhematologic AEs were found. CONCLUSION: The greater severity of both symptomatic AEs and hematologic AEs in women across multiple treatment modalities indicates that broad-based sex differences exist. This could be due to differences in AE reported, pharmacogenomics of drug metabolism/disposition, total dose received, and/or adherence to therapy. Particularly large sex differences were observed for patients receiving immunotherapy, suggesting that studying AEs from these agents is a priority.

Misalignment between COVID-19 hotspots and clinical trial sites.

Hundreds of interventional clinical trials have been launched in the United States to identify effective treatment strategies for combating the coronavirus disease 2019 (COVID-19) pandemic. However, to date, only a small fraction of these trials have completed enrollment, delaying the scientific investigation of COVID-19 and its treatment options. This study presents novel metrics to examine the geographic alignment between COVID-19 hotspots and interventional clinical trial sites and evaluate trial access over time during the evolving pandemic. Using temporal COVID-19 case data from USAFacts.org and trial data from ClinicalTrials.gov, U.S. counties were categorized based on their numbers of cases and trials. Our analysis suggests that alignment and access have worsened as the pandemic shifted over time. We recommend strategies and metrics to evaluate the alignment between cases and trials. Future studies are warranted to investigate the impact of the misalignment of cases and clinical trial sites on clinical trial recruitment.

Would the Recommended Dose Have Been Different Using Novel Dose-Finding Designs? Comparing Dose-Finding Designs in Published Trials.

Simulation studies have shown that novel designs such as the continual reassessment method and the Bayesian optimal interval (BOIN) design outperform the 3 + 3 design by recommending the maximum tolerated dose (MTD) more often, using less patients, and allotting more patients to the MTD. However, it is not clear whether these novel designs would have yielded different results in the context of real-world dose-finding trials. This is a commonly mentioned reason for the continuous use of 3 + 3 designs for oncology trials, with investigators considering simulation studies not sufficiently convincing to warrant the additional design complexity of novel designs. METHODS: We randomly sampled 60 published dose-finding trials to obtain 22 that used the 3 + 3 design, identified an MTD, published toxicity data, and had more than two dose levels. We compared the published MTD with the estimated MTD using the continual reassessment method and BOIN using target toxicity rates of 25% and 30% and toxicity data from the trial. Moreover, we compared patient allocation and sample size assuming that these novel designs had been implemented. RESULTS: Model-based designs chose dose levels higher than the published MTD in about 40% of the trials, with estimated and observed toxicity rates closer to the target toxicity rates of 25% and 30%. They also assigned less patients to suboptimal doses and permitted faster dose escalation. CONCLUSION: This study using published dose-finding trials shows that novel designs would recommend different MTDs and confirms the advantages of these designs compared with the 3 + 3 design, which were demonstrated by simulation studies.

Seamless phase I/II design for novel anticancer agents with competing disease progression.

Molecularly targeted agents and immunotherapies have prolonged administration and complicated toxicity and efficacy profiles requiring longer toxicity observation windows and the inclusion of efficacy information to identify the optimal dose. Methods have been proposed to either jointly model toxicity and efficacy, or for prolonged observation windows. However, it is inappropriate to address these issues individually in the setting of dose-finding because longer toxicity windows increase the risk of patients experiencing disease progression and discontinuing the trial, with progression defining a competing event to toxicity, and progression-free survival being a commonly used efficacy endpoint. No method has been proposed to address this issue in a competing risk framework. We propose a seamless phase I/II design, namely the competing risks continual reassessment method (CR-CRM). Given an observation window, the objective is to recommend doses that minimize the progression probability, among a set of tolerable doses in terms of toxicity risk. In toxicity-centered stage of the design, doses are assigned based on toxicity alone, and in optimization stage of the design, doses are assigned integrating both toxicity and progression information. Design operating characteristics were examined in a simulation study compared with benchmark performances, including sensitivity to time-varying hazards and correlated events. The method performs well in selecting doses with acceptable toxicity risk and minimum progression risk across a wide range of scenarios.

Diffuse optical tomography breast imaging measurements are modifiable with pre-surgical targeted and endocrine therapies among women with early stage breast cancer.

PURPOSE: Diffuse optical tomography breast imaging system (DOTBIS) non-invasively measures tissue concentration of hemoglobin, which is a potential biomarker of short-term response to neoadjuvant chemotherapy. We evaluated whether DOTBIS-derived measurements are modifiable with targeted therapies, including AKT inhibition and endocrine therapy. METHODS: We conducted a proof of principle study in seven postmenopausal women with stage I-III breast cancer who were enrolled in pre-surgical studies of the AKT inhibitor MK-2206 (n = 4) or the aromatase inhibitors exemestane (n = 2) and letrozole (n = 1). We performed DOTBIS at baseline (before initiation of therapy) and post-therapy in the affected breast (tumor volume) and contralateral, unaffected breast, and measured tissue concentrations (in µM) of total hemoglobin (ctTHb), oxyhemoglobin (ctO2Hb), and deoxyhemoglobin (ctHHb), as well as water fraction (%). RESULTS: We found consistent decreases in DOTBIS-measured hemoglobin concentrations in tumor volume, with median percent changes for ctTHb, ctHHb, ctO2Hb, and water fraction for the entire cohort of - 27.1% (interquartile range [IQR] 37.5%), - 49.8% (IQR 29.3%), - 33.5% (IQR 47.4%), and - 3.6% (IQR 10.6%), respectively. In the contralateral breast, median percent changes for ctTHb, ctHHb, ctO2Hb, and water fraction were + 1.8% (IQR 26.7%), - 8.6% (IQR 29.3%), + 6.2% (IQR 29.5%), and + 1.9% (IQR 30.7%), respectively. CONCLUSION: We demonstrated that DOTBIS-derived measurements are modifiable with pre-surgical AKT inhibition and endocrine therapy, supporting further investigation of DOTBIS as a potential imaging assessment of response to neoadjuvant targeted therapies in early stage breast cancer.

A Phase I Study of the Combination of Pexidartinib and Sirolimus to Target Tumor-Associated Macrophages in Unresectable Sarcoma and Malignant Peripheral Nerve Sheath Tumors.

PURPOSE: To evaluate the safety and tolerability in phase I first-in-human combination therapy with pexidartinib, an inhibitor of colony-stimulating factor-1 receptor, and sirolimus, an mTOR inhibitor, to target tumor-associated macrophage (TAM) polarization in soft tissue sarcomas (STS). PATIENTS AND METHODS: This multicenter phase I study used the time-to-event continual reassessment method (TITE-CRM) to study the combination of sirolimus, doses ranging from 2 to 6 mg, with pexidartinib, doses ranging from 600 to 1,000 mg, both provided continuously on a 28-day cycle, in patients with advanced sarcoma. A total of 24 patients [8 malignant peripheral nerve sheath tumor, 3 tenosynovial giant cell tumor (TGCT), 5 leiomyosarcoma, and 8 with other sarcoma subtypes] were enrolled. The median age was 46 years, 56% were male, and 61% had >2 prior lines of therapy. RESULTS: The recommended phase II dose was 2 mg of sirolimus combined with 1,000 mg of pexidartinib daily. Of the 18 evaluable subjects, 5 experienced dose-limiting toxicities (2 elevated aspartate aminotransferase/alanine aminotransferase, 2 elevated sirolimus trough levels, and 1 grade 5 dehydration). Most common grade 2 or higher treatment-related adverse events included anemia, fatigue, neutropenia, and lymphopenia. Clinical benefit was observed in 12 of 18 (67%) evaluable subjects with 3 partial responses (all in TGCT) and 9 stable disease. Tissue staining indicated a decreased proportion of activated M2 macrophages within tumor samples with treatment. CONCLUSIONS: Pexidartinib can be safely administered with sirolimus. These findings support further investigation of this combination to determine clinical efficacy. Clinicaltrials.gov identifier NCT02584647.

Quantifying normal lung in pulmonary fibrosis: CT analysis and correlation with %DLCO.

BACKGROUND: Chest CT scans are routinely obtained to monitor disease progression in pulmonary fibrosis. However, radiologists do not employ a standardized system for quantitative description of the severity of the disease. Development and validation of a grading system offers potential for enhancing the information that radiologists provide clinicians. STUDY DESIGN AND METHODS: Our retrospective review analyzed 100 sequential patients with usual interstitial pneumonitis (UIP) on HRCT scans from 2018 and 2019. A radiologic scoring system evaluated the percent of normal lung on the basis of a 0-5 point scale per lobe (findings for the right middle lobe were included in the right upper lobe score), yielding an overall additive numerical score on a scale of 20 (completely normal lung) to 0 (no normal lung). Two radiologists quantified the percentage of normal lung by consensus agreement. Percent DLCO as well as demographic data were obtained from the medical record. Statistical analysis was performed using Spearman correlation to assess correlation between grade and percent DLCO. RESULTS: 96 patients met the inclusion criteria; average age was 71, 68% were male. Score on CT scan ranged from 18 to 4; average 10.9, SD 3.58. The single-breath diffusing capacity (percent DLCO) ranged from 88% to 17%; mean 44.5%, SD 14.3%. Spearman's correlation for CT score and percent DLCO was 0.622, P < 0.001. CONCLUSION: This scoring system quantifying the amount of normal lung on chest CT of patients with UIP correlated significantly with percent DLCO (P < 0.001) and appears to offer a promising quantitative measure to assess severity of disease.

Designing Dose-Finding Phase I Clinical Trials: Top 10 Questions That Should Be Discussed With Your Statistician.

In recent years, the landscape in clinical trial development has changed to involve many molecularly targeted agents, immunotherapies, or radiotherapy, as a single agent or in combination. Given their different mechanisms of action and lengths of administration, these agents have different toxicity profiles, which has resulted in numerous challenges when applying traditional designs such as the 3 + 3 design in dose-finding clinical trials. Novel methods have been proposed to address these design challenges such as combinations of therapies or late-onset toxicities. However, their design and implementation require close collaboration between clinicians and statisticians to ensure that the appropriate design is selected to address the aims of the study and that the design assumptions are pertinent to the study drug. The goal of this paper is to provide guidelines for appropriate questions that should be considered early in the design stage to facilitate the interactions between clinical and statistical teams and to improve the design of dose-finding clinical trials for novel anticancer agents.

A simulation study of approaches for handling disease progression in dose-finding clinical trials.

In traditional dose-finding studies, dose-limiting toxicity (DLT) is determined within a fixed time observation window where DLT is often defined as a binary outcome. In the setting of oncology dose-finding trials, often patients in advanced stage of diseases are enrolled. Therefore, disease progression may occur within the DLT observation window leading to treatment discontinuation and rendering the patient unevaluable for DLT assessment. As a result, additional patients have to be enrolled, increasing the sample size. We propose and compare several practical approaches for handling disease progression which occurs within the DLT observation window, while in the framework of the time-to-event continual reassessment method (TITE-CRM) which allows using partial observations. The approaches differ on the way they define an evaluable patient and in the way incomplete observations are included. The practical approaches, which we call strategies A, B and C, are illustrated and contrasted in the context of a single simulated trial, and compared via simulations under various scenarios of dose-progression relationship, in the setting of advanced soft-tissue sarcoma.

Towards clinical data-driven eligibility criteria optimization for interventional COVID-19 clinical trials.

OBJECTIVE: This research aims to evaluate the impact of eligibility criteria on recruitment and observable clinical outcomes of COVID-19 clinical trials using electronic health record (EHR) data. MATERIALS AND METHODS: On June 18, 2020, we identified frequently used eligibility criteria from all the interventional COVID-19 trials in ClinicalTrials.gov (n = 288), including age, pregnancy, oxygen saturation, alanine/aspartate aminotransferase, platelets, and estimated glomerular filtration rate. We applied the frequently used criteria to the EHR data of COVID-19 patients in Columbia University Irving Medical Center (CUIMC) (March 2020-June 2020) and evaluated their impact on patient accrual and the occurrence of a composite endpoint of mechanical ventilation, tracheostomy, and in-hospital death. RESULTS: There were 3251 patients diagnosed with COVID-19 from the CUIMC EHR included in the analysis. The median follow-up period was 10 days (interquartile range 4-28 days). The composite events occurred in 18.1% (n = 587) of the COVID-19 cohort during the follow-up. In a hypothetical trial with common eligibility criteria, 33.6% (690/2051) were eligible among patients with evaluable data and 22.2% (153/690) had the composite event. DISCUSSION: By adjusting the thresholds of common eligibility criteria based on the characteristics of COVID-19 patients, we could observe more composite events from fewer patients. CONCLUSIONS: This research demonstrated the potential of using the EHR data of COVID-19 patients to inform the selection of eligibility criteria and their thresholds, supporting data-driven optimization of participant selection towards improved statistical power of COVID-19 trials.

Dose-finding design and benchmark for a right censored endpoint.

Dose-finding trials aim to determine a safe dose to be tested in larger trials for efficacy. In oncology, designs generally assume conventional monotonic increasing dose-toxicity relationships, mostly with binary outcomes (e.g., dose-limiting toxicity or not), measured in the first cycle of therapy or for a fixed number of cycles. However, with new anti-cancer agents such as molecularly targeted therapies and immunotherapies, late-onset toxicities have become more frequent. Designs with prolonged observation windows and censored endpoints analyzed using survival models, appear particularly suited to these settings. Moreover, in this setting, the observation of the late-onset toxicity endpoint could be precluded by trial discontinuation due to death, progression, patient withdrawal, or physician discretion, defining a competing event to toxicity. We propose extensions of the Continual Reassessment Method (CRM) dose-finding design using survival working models for right-censored endpoints and for handling treatment discontinuation in the toxicity observation window, namely the Survival-CRM (Surv-CRM) and the informative survival-CRM (iSurv-CRM). We also developed a benchmark approach for its evaluation. In a simulation study, we compared the performance of the Surv-CRM and iSurv-CRM, to those of the Time-to-event (TITE)-CRM and the nonparametric benchmark. The performance of the proposed methods was consistent with the complexity of scenarios as assessed by the nonparametric benchmark. Without treatment discontinuations, the Surv-CRM provides proportions of correct dose selection close to those of the TITE-CRM with fewer observed toxicities and patients assigned to overtoxic dose levels. In the presence of treatment discontinuation, the iSurv-CRM outperforms the TITE-CRM in identifying the correct dose level.

A Phase I Trial of Intravesical Cabazitaxel, Gemcitabine and Cisplatin for the Treatment of Nonmuscle Invasive bacillus Calmette-Guérin Unresponsive or Recurrent/Relapsing Urothelial Carcinoma of the Bladder.

PURPOSE: For patients with bacillus Calmette-Guérin unresponsive or recurrent/relapsing nonmuscle invasive bladder cancer, multi-agent intravesical trials have been limited. In this study we investigate the safety of intravesical cabazitaxel, gemcitabine and cisplatin in the salvage setting. MATERIALS AND METHODS: This was a dose escalation, drug escalation trial for patients with bacillus Calmette-Guérin unresponsive or recurrent/relapsing nonmuscle invasive bladder cancer who declined or were ineligible for radical cystectomy. All patients underwent a 6-week induction regimen of sequentially administered cabazitaxel, gemcitabine and cisplatin. Complete response was defined as no cancer on post-induction transurethral bladder tumor resection and negative urine cytology, while partial response allowed for positive cytology. Responders continued with maintenance cabazitaxel and gemcitabine monthly for the first year and bimonthly for the second year. RESULTS: A total of 18 patients were enrolled. Mean age was 71 years, median followup was 27.8 months (range 16.3 to 46.9) and mean number of previous rounds of intravesical therapies before trial enrollment was 3.7. Nine patients (50%) had received intravesical chemotherapy after bacillus Calmette-Guérin and 7 (39%) were previously treated in a phase I clinical trial setting. At enrollment 6 (33%) subjects had T1 disease and 13 (72%) had carcinoma in situ. There were no dose limiting toxicities. Initial partial and complete response rates were 94% and 89%, respectively. At 1 year recurrence-free survival was 0.83 (range 0.57 to 0.94) and at 2 years estimated recurrence-free survival was 0.64 (0.32 to 0.84). CONCLUSIONS: In this high risk and highly pretreated cohort of bacillus Calmette-Guérin unresponsive or recurrent/relapsing nonmuscle invasive bladder cancer cases combination intravesical cabazitaxel, gemcitabine and cisplatin was a well tolerated and potentially effective regimen.

Phase 1 Study of Erlotinib and Metformin in Metastatic Triple-Negative Breast Cancer.

BACKGROUND: Epidermal growth factor receptor (EGFR) is frequently overexpressed in metastatic triple-negative breast cancer (mTNBC). One strategy for overcoming resistance to EGFR inhibition is concomitant inhibition of downstream signaling. The antidiabetic drug metformin inhibits both MAPK and PI3K/mTOR pathway signaling. We evaluated the combination of erlotinib and metformin in a phase 1 study of patients with mTNBC. PATIENTS AND METHODS: Patients with mTNBC who had received at least one prior line of therapy for metastatic disease were eligible. Erlotinib dose was fixed at 150 mg daily. Metformin dose escalation was planned according to a 3 + 3 design. Dose-limiting toxicities (DLT) were assessed during the first 5 weeks of therapy. The primary objective was to determine the maximum tolerated dose of metformin with fixed-dose erlotinib. Secondary endpoints were response rate, stable disease rate, and progression-free survival. RESULTS: Eight patients were enrolled. The median number of prior therapies for metastatic disease was 2.5 (range, 1-6). No DLT events were reported during the DLT assessment period. Most adverse events were grade 1/2. Grade 3 diarrhea despite maximum supportive care required dose reduction of metformin in one patient. Grade 3 rash led to study withdrawal in one patient. No grade 4 adverse events were reported. The best observed response was stable disease in 2 patients (25%). Median progression-free survival was 60 days (range, 36-61 days). CONCLUSION: Erlotinib and metformin were well tolerated in a population of pretreated mTNBC patients but did not demonstrate efficacy in this population.

Incorporating patient-reported outcomes in dose-finding clinical trials.

Oncology dose-finding clinical trials determine the maximum tolerated dose (MTD) based on toxicity outcomes captured by clinicians. With the availability of more rigorous instruments for measuring toxicity directly from patients, there is a growing interest to incorporate patient-reported outcomes (PRO) in clinical trials to inform patient tolerability. This is particularly important for dose-finding trials to ensure the identification of a well-tolerated dose. In this paper, we propose three extensions of the continual reassessment method (CRM), termed PRO-CRMs, that incorporate both clinician and patient outcomes. The first method is a marginal modeling approach whereby clinician and patient toxicity outcomes are modeled separately. The other two methods impose a constraint using a joint outcome defined based on both clinician and patient toxicities and model them either jointly or marginally. Simulation studies show that while all three PRO-CRMs select well-tolerated doses based on clinician's and patient's perspectives, the methods using a joint outcome perform better and have similar performance. We also show that the proposed PRO-CRMs are consistent under robust model assumptions.