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.

Randomised Phase 1 clinical trials in oncology.

The aims of Phase 1 trials in oncology have broadened considerably from simply demonstrating that the agent/regimen of interest is well tolerated in a relatively heterogeneous patient population to addressing multiple objectives under the heading of early-phase trials and, if possible, obtaining reliable evidence regarding clinical activity to lead to drug approvals via the Accelerated Approval approach or Breakthrough Therapy designation in cases where the tumours are rare, prognosis is poor or where there might be an unmet therapeutic need. Constructing a Phase 1 design that can address multiple objectives within the context of a single trial is not simple. Randomisation can play an important role, but carrying out such randomisation according to the principles of equipoise is a significant challenge in the Phase 1 setting. If the emerging data are not sufficient to definitively address the aims early on, then a proper design can reduce biases, enhance interpretability, and maximise information so that the Phase 1 data can be more compelling. This article outlines objectives and design considerations that need to be adhered to in order to respect ethical and scientific principles required for research in human subjects in early phase clinical trials.

Rolling continual reassessment method with overdose control: An efficient and safe dose escalation design.

In phase 1 dose escalation studies, dose limiting toxicities (DLTs) are defined as adverse events of concern occurring during a predefined time window after first dosing of patients. Standard dose escalation designs, such as the continual reassessment method (CRM), only utilize this binary DLT information. Thus, late-onset DLTs are usually not accounted for when CRM guiding the dose escalation and finally defining the maximum tolerated dose (MTD) of the drug, which brings safety concerns for patients. Previously, several extensions of CRMs, such as the time-to-event CRM (TITE-CRM), fractional CRM (fCRM) and the data augmented CRM (DA-CRM), have been proposed to handle this issue without prolonging trial duration. However, among the model-based designs, none of the designs have explicitly controlled the risk of overdosing as in the escalation with overdose control (EWOC) design. Here we propose a novel dose escalation with overdose control design using a two-parameter logistic regression model for the probability of DLT depending on the dose and a piecewise exponential model for the time to DLT distribution, which we call rolling-CRM design. A comprehensive simulation study has been conducted to compare the performance of the rolling-CRM design with other dose escalation designs. Of note, the trial duration is significantly shorter compared to traditional CRM designs. The proposed design also retains overdose control characteristics, but might require a larger sample size compared to traditional CRM designs.

Assessment of various continual reassessment method models for dose-escalation phase 1 oncology clinical trials: using real clinical data and simulation studies.

BACKGROUND: The continual reassessment method (CRM) identifies the maximum tolerated dose (MTD) more efficiently and identifies the true MTD more frequently compared to standard methods such as the 3 + 3 method. An initial estimate of the dose-toxicity relationship (prior skeleton) is required, and there is limited guidance on how to select this. Previously, we compared the CRM with six different skeletons to the 3 + 3 method by conducting post-hoc analysis on a phase 1 oncology study (AZD3514), each CRM model reduced the number of patients allocated to suboptimal and toxic doses. This manuscript extends this work by assessing the ability of the 3 + 3 method and the CRM with different skeletons in determining the true MTD of various "true" dose-toxicity relationships. METHODS: One thousand studies were simulated for each "true" dose toxicity relationship considered, four were based on clinical trial data (AZD3514, AZD1208, AZD1480, AZD4877), and four were theoretical. The 3 + 3 method and 2-stage extended CRM with six skeletons were applied to identify the MTD, where the true MTD was considered as the largest dose where the probability of experiencing a dose limiting toxicity (DLT) is ≤33%. RESULTS: For every true dose-toxicity relationship, the CRM selected the MTD that matched the true MTD in a higher proportion of studies compared to the 3 + 3 method. The CRM overestimated the MTD in a higher proportion of simulations compared to the 3 + 3 method. The proportion of studies where the correct MTD was selected varied considerably between skeletons. For some true dose-toxicity relationships, some skeletons identified the true MTD in a higher proportion of scenarios compared to the skeleton that matched the true dose-toxicity relationship. CONCLUSION: Through simulation, the CRM generally outperformed the 3 + 3 method for the clinical and theoretical true dose-toxicity relationships. It was observed that accurate estimates of the true skeleton do not always outperform a generic skeleton, therefore the application of wide confidence intervals may enable a generic skeleton to be used. Further work is needed to determine the optimum skeleton.

Optimizing the Therapeutic Window of Targeted Drugs in Oncology: Potency-Guided First-in-Human Studies.

Many targeted therapies are administered at or near the maximum tolerated dose (MTD). With the advent of precision medicine, a larger therapeutic window is expected. Therefore, dose optimization will require a new approach to early clinical trial design. We analyzed publicly available data for 21 therapies targeting six kinases, and four poly (ADP-ribose) polymerase inhibitors, focusing on potency and exposure to gain insight into dose selection. The free average steady-state concentration (Css ) at the approved dose was compared to the in vitro cell potency (half-maximal inhibitory concentration (IC50 )). Average steady-state area under the plasma concentration-time curve, the fraction unbound drug in plasma, and the cell potency were taken from the US drug labels, US and European regulatory reviews, and peer-reviewed journal articles. The Css was remarkably similar to the IC50 . The median Css /IC50 value was 1.2, and 76% of the values were within 3-fold of unity. However, three drugs (encorafenib, erlotinib, and ribociclib) had a Css /IC50 value > 25. Seven other therapies targeting the same 3 kinases had much lower Css /IC50 values ranging from 0.5 to 4. These data suggest that these kinase inhibitors have a large therapeutic window that is not fully exploited; lower doses may be similarly efficacious with improved tolerability. We propose a revised first-in-human trial design in which dose cohort expansion is initiated at doses less than the MTD when there is evidence of clinical activity and Css exceeds a potency threshold. This potency-guided approach is expected to maximize the therapeutic window thereby improving patient outcomes.

Early 3+3 Trial Dose-Escalation Phase I Clinical Trial Design and Suitability for Immune Checkpoint Inhibitors.

PURPOSE: Despite the expansion of immune checkpoint inhibitor (ICI) indications, the relationship between ICI dose and toxicity or response is not well established. To understand this correlation, we performed a meta-analysis of ICI trials that used dose escalation. EXPERIMENTAL DESIGN: We searched PubMed and abstracts presented at (inter)national meetings for trials using FDA-approved ICIs. The reported rates of grade 3-5 adverse events (G3-5 AE), immune-related adverse events (irAE), and response were correlated with doses within each ICI using marginal exact generalized linear models. RESULTS: A total of 74 trials (7,469 patients) published between January 2010 and January 2017 were included. For ipilimumab, the incidence of G3-5 AEs was 34% with a significant 27% reduced risk in lower doses (P = 0.002). However, no relationship was observed between dose and irAEs or response. For nivolumab, the incidence of G3-5 AEs was 20.1% which was lower in non-small cell lung cancer (NSCLC) compared with renal cell carcinoma (RCC) or melanoma (P ≤ 0.05) with no dose-toxicity relationship. In melanoma and NSCLC, a dose-response association was observed, which was not observed in RCC. For pembrolizumab, the incidence of G3-5 AEs was 13.3%, which was lower in melanoma compared with NSCLC (P = 0.03) with no dose-toxicity relationship. In melanoma, lower dose levels correlated with decreased odds of response (P = 0.01), a relationship that was not observed in NSCLC. CONCLUSIONS: Our analysis shows a lack of consistent dose-toxicity or dose-response correlation with ICIs. Therefore, dose escalation is not an appropriate design to conduct ICI studies. Here we present an innovative trial design for immune-modulating agents.

Phase 1 Trial of MLN0128 (Sapanisertib) and CB-839 HCl (Telaglenastat) in Patients With Advanced NSCLC (NCI 10327): Rationale and Study Design.

INTRODUCTION: There are currently no approved targeted therapies for lung squamous-cell carcinoma (LSCC) and KRAS-mutant lung adenocarcinoma (LUAD). About 30% of LSCC and 25% of KRAS-mutant LUAD exhibit hyperactive NRF2 pathway activation through mutations in NFE2L2 (the gene encoding NRF2) or its negative regulator, KEAP1. Preclinical data demonstrate that these tumors are uniquely sensitive to dual inhibition of glycolysis and glutaminolysis via mammalian target of rapamycin (mTOR) and glutaminase inhibitors. This phase 1 study was designed to assess safety and preliminary activity of the mTOR inhibitor MLN0128 (sapanisertib) in combination with the glutaminase inhibitor CB-839 HCl. METHODS: Phase 1 dose finding will use the queue-based variation of the 3 + 3 dose escalation scheme with the primary endpoint of identifying the recommended expansion dose. To confirm the acceptable tolerability of the recommended expansion dose, patients will subsequently enroll onto 1 of 4 expansion cohorts (n = 14 per cohort): (1) LSCC harboring NFE2L2 or (2) KEAP1 mutations, or (3) LUAD harboring KRAS/(KEAP1 or NFE2L2) coalterations, or (4) LSCC wild type for NFE2L2 and KEAP1. The primary endpoint of the dose expansion is to determine the preliminary efficacy of MLN0128/CB-839 combination therapy. CONCLUSION: This phase 1 study will determine the recommended expansion dose and preliminary efficacy of MLN0128 and CB-839 in advanced non-small-cell lung cancer with a focus on subsets of LSCC and KRAS-mutant LUAD harboring NFE2L2 or KEAP1 mutations.

Screening for Phase 1 Clinical Trials: An Opportunity to Evaluate Psychological States and Reeducate Patients on Prognosis?

Disease progression or recurrence after a period of remission can be a challenging event for individuals seeking cancer treatment. Those referred for possible phase 1 trial enrollment are often motivated to participate in these studies with hope for a cure despite approximately 5% response rates in this setting. Addressing such commonly held misunderstandings during the initial evaluation for phase 1 trial eligibility could provide a valuable opportunity to improve physician communication by identifying signs of distress or psychiatric conditions, addressing underlying psychological biases, and encouraging adaptive coping strategies.

A phase I dose-finding design with incorporation of historical information and adaptive shrinking boundaries.

Although many novel phase I designs have been developed in recent years, few studies have discussed how to incorporate external information into dose-finding designs. In this paper, we first propose a new method for developing a phase I design, Bayesian optimal interval design (BOIN)[Liu S et al. (2015), Yuan Y et al. (2016)], for formally incorporating historical information. An algorithm to automatically generate parameters for prior set-up is introduced. Second, we propose a method to relax the fixed boundaries of the BOIN design to be adaptive, such that the accumulative information can be used more appropriately. This modified design is called adaptive BOIN (aBOIN). Simulation studies to examine performances of the aBOIN design in small and large sample sizes revealed comparable performances for the aBOIN and original BOIN designs for small sample sizes. However, aBOIN outperformed BOIN in moderate sample sizes. Simulation results also showed that when historical trials are conducted in settings similar to those for the current trial, their performance can be significantly improved. This approach can be applied directly to pediatric cancer trials, since all phase I trials in children are followed by similar efficient adult trials in the current drug development paradigm. However, when information is weak, operating characteristics are compromised.

A comparison of phase I dose-finding designs in clinical trials with monotonicity assumption violation.

BACKGROUND/AIMS: In oncology, new combined treatments make it difficult to order dose levels according to monotonically increasing toxicity. New flexible dose-finding designs that take into account uncertainty in dose levels ordering were compared with classical designs through simulations in the setting of the monotonicity assumption violation. We give recommendations for the choice of dose-finding design. METHODS: Motivated by a clinical trial for patients with high-risk neuroblastoma, we considered designs that require a monotonicity assumption, the Bayesian Continual Reassessment Method, the modified Toxicity Probability Interval, the Bayesian Optimal Interval design, and designs that relax monotonicity assumption, the Bayesian Partial Ordering Continual Reassessment Method and the No Monotonicity Assumption design. We considered 15 scenarios including monotonic and non-monotonic dose-toxicity relationships among six dose levels. RESULTS: The No Monotonicity Assumption and Partial Ordering Continual Reassessment Method designs were robust to the violation of the monotonicity assumption. Under non-monotonic scenarios, the No Monotonicity Assumption design selected the correct dose level more often than alternative methods on average. Under the majority of monotonic scenarios, the Partial Ordering Continual Reassessment Method selected the correct dose level more often than the No Monotonicity Assumption design. Other designs were impacted by the violation of the monotonicity assumption with a proportion of correct selections below 20% in most scenarios. Under monotonic scenarios, the highest proportions of correct selections were achieved using the Continual Reassessment Method and the Bayesian Optimal Interval design (between 52.8% and 73.1%). The costs of relaxing the monotonicity assumption by the No Monotonicity Assumption design and Partial Ordering Continual Reassessment Method were decreases in the proportions of correct selections under monotonic scenarios ranging from 5.3% to 20.7% and from 1.4% to 16.1%, respectively, compared with the best performing design and were higher proportions of patients allocated to toxic dose levels during the trial. CONCLUSIONS: Innovative oncology treatments may no longer follow monotonic dose levels ordering which makes standard phase I methods fail. In such a setting, appropriate designs, as the No Monotonicity Assumption or Partial Ordering Continual Reassessment Method designs, should be used to safely determine recommended for phase II dose.

Pivotal Considerations for Optimal Deployment of Healthy Volunteers in Oncology Drug Development.

Oncology drug development is among the most challenging of any therapeutic area, with first-in-human trials expected to deliver information on both safety and activity. Until recently, therapeutic approaches in oncology focused on cytotoxic chemotherapy agents, ruling out even the possibility of enrolling normal healthy volunteers (NHVs) in clinical trials due to safety considerations. The emergence of noncytotoxic modalities, including molecularly targeted agents with more favorable safety profiles, however, has led to increasing numbers of clinical pharmacology studies of these agents being conducted in NHVs. Beyond rapid enrollment and cost savings, there are other advantages of conducting specific types of studies in NHVs with the goal of more appropriate dosing decisions in certain subsets of the intended patient populations, allowing for enrollment of such patients in therapeutic trials from which they might otherwise have been excluded. Nevertheless, the decision must be carefully weighed against potential disadvantages, and although the considerations surrounding conduct of clinical trials using NHVs are generally well-defined in most other therapeutic areas, they are less well-defined in oncology.

Bayesian cancer clinical trial designs with subgroup-specific decisions.

Two illustrative applications are presented of Bayesian clinical trial designs that make adaptive subgroup-specific decisions based on elicited utilities of patient outcomes to quantify risk-benefit trade-offs. The first design is for a randomized trial to evaluate effects of nutritional prehabilitation on post-operative morbidity in esophageal cancer patients undergoing surgery. The second design is for a dose-finding trial of natural killer cells to treat advanced hematologic malignancies, with five time-to-event outcomes. Each design is based on a Bayesian hierarchical model that borrows strength between subgroups. Computer simulation is used to evaluate each design's properties, including comparison to a simpler design ignoring treatment-subgroup interactions. The simulations show that accounting prospectively for treatment-subgroup interactions yields designs with very desirable properties, is greatly superior to a simplified comparator design that ignores subgroups if treatment-subgroup interactions actually exist, and each design is robust to deviations from the assumed underlying model.

Adaptive dose-finding based on safety and feasibility in early-phase clinical trials of adoptive cell immunotherapy.

BACKGROUND/AIMS: Dose feasibility is a challenge that may arise in the development of adoptive T cell therapies for cancer. In early-phase clinical trials, dose is quantified either by a fixed or per unit body weight number of cells infused. It may not be feasible, however, to administer a patient's assigned dose due to an insufficient number of cells harvested or functional heterogeneity of the product. The study objective becomes to identify the maximum tolerated dose with high feasibility of being administered. This article describes a new dose-finding method that adaptively accounts for safety and feasibility endpoints in guiding dose allocation. METHODS: We propose an adaptive dose-finding method that integrates accumulating feasibility and safety data to select doses for participant cohorts in early-phase trials examining adoptive cell immunotherapy. We sequentially model the probability of dose-limiting toxicity and the probability of feasibility using independent beta-binomial models. The probability model for toxicity borrows information across all dose levels using isotonic regression, allowing participants infused at a lower dose than his or her planned dose to contribute safety data to the dose-finding algorithm. We applied the proposed methodology in a single simulated trial and evaluated its operating characteristics through extensive simulation studies. RESULTS: In simulations conducted for a phase I study of adoptive immunotherapy for newly diagnosed glioblastoma, the proposed method demonstrates the ability to identify accurately the feasible maximum tolerated doses and to treat participants at and around these doses. Over 10 hypothesized scenarios studied, the percentage of correctly selecting the true feasible and maximum tolerated dose ranged from 50% to 90% with sample sizes averaging between 21 and 24 participants. A comparison to the only known existing method accounting for safety and feasibility yields competitive performance. CONCLUSION: We have developed a new practical adaptive dose-finding method to assess feasibility in early-phase adoptive cell therapy trials. A design that incorporates feasibility, as a function of the quantity and quality of the product manufactured, in addition to safety will have an impact on the recommended phase II doses in studies that evaluate patient outcomes.

Concomitant intraperitoneal and systemic chemotherapy for extensive peritoneal metastases of colorectal origin: protocol of the multicentre, open-label, phase I, dose-escalation INTERACT trial.

INTRODUCTION: Cytoreductive surgery and hyperthermic intraperitoneal chemotherapy (CRS-HIPEC) has become standard of care for patients with peritoneal metastases of colorectal origin with a low/moderate abdominal disease load. In case of a peritoneal cancer index (PCI) score >20, CRS-HIPEC is not considered to be beneficial. Patients with a PCI >20 are currently offered palliative systemic chemotherapy. Previous studies have shown that systemic chemotherapy is less effective against peritoneal metastases than it is against haematogenous spread of colorectal cancer. It is suggested that patients with peritoneal metastases may benefit from the addition of intraperitoneal chemotherapy to systemic chemotherapy. Aim of this study is to establish the maximum tolerated dose of intraperitoneal irinotecan, added to standard of care systemic therapy for colorectal cancer. Secondary endpoints are to determine the safety and feasibility of this treatment and to establish the pharmacokinetic profile of intraperitoneally administered irinotecan. METHODS AND ANALYSIS: This phase I, '3+3' dose-escalation, study is performed in two Dutch tertiary referral centres. The study population consists of adult patients with extensive peritoneal metastases of colorectal origin who have a good performance status and no extra-abdominal metastases. According to standard work-up for CRS-HIPEC, patients will undergo a diagnostic laparoscopy to score the PCI. In case of a PCI >20, a peritoneal access port will be placed in the abdomen of the patient. Through this port we will administer intraperitoneal irinotecan, in combination with standard systemic treatment consisting of 5-fluorouracil/leucovorin with oxaliplatin and the targeted agent bevacizumab. Therapy consists of a maximum of 12 cycles 2-weekly. ETHICS AND DISSEMINATION: This study protocol is approved by a research medical ethics committee (Rotterdam, Netherlands) and the Dutch Competent Authority (CCMO, The Hague, Netherlands). The results of this trial will be submitted for publication in a peer-reviewed scientific journal. TRAIL REGISTRATION NUMBER: NL6988 and NL2018-000479-33; Pre-results.

Autologous bone marrow expanded mesenchymal stem cells in patellar tendinopathy: protocol for a phase I/II, single-centre, randomized with active control PRP, double-blinded clinical trial.

INTRODUCTION: Patellar tendon overuse injuries are common in athletes. Imaging may show a change in tissue structure with tendon thickening and disruption of the intratendinous substance. We wish to test the hypothesis that both autologous bone marrow expanded mesenchymal stem cells and autologous leukocyte-poor platelet-rich plasma (LP-PRP) implanted into the area of the disrupted tendinopathic patellar tendon will restore function, but tendon regeneration tissue will only be observed in the subjects treated with autologous bone marrow expanded mesenchymal stem cells. METHODS AND ANALYSIS: This is a single-centre, pilot phase I/II, double-blinded clinical trial with randomisation with active control. Twenty patients with a diagnosis of patellar tendinopathy with imaging changes (tendon thickening and disruption of the intratendinous substance at the proximal portion of the patellar tendon) will be randomised in a 1:1 ratio to receive a local injection of either bone-marrow autologous mesenchymal stem cells (MSC), isolated and cultured under GMP at The Institute of Biology and Molecular Genetics (IBGM) (Spain) or P-PRP. The study will have two aims: first, to ascertain whether a clinically relevant improvement after 3, 6 and 12 months according to the visual analogue scale (VAS), Victorian Institute of Sport Assessment for patellar tendons (VISA-P) and dynamometry scales (DYN) will be achieved; and second, to ascertain whether the proposed intervention will restore tendon structure as determined by ultrasonography (US), Doppler ultrasonography (DUS), and innovative MRI and ultrasound techniques: Magnetic Resonance T2 FAT SAT (UTE, Ultrashort Echo TE) sequence and Ultrasound Tissue Characterization (UTC). Patients who are randomised to the P-PRP treatment group but do not achieve a satisfactory primary endpoint after 6 months will be offered treatment with MSC. TRIAL REGISTRATION: NCT03454737.

A utility-based Bayesian optimal interval (U-BOIN) phase I/II design to identify the optimal biological dose for targeted and immune therapies.

In the era of targeted therapy and immunotherapy, the objective of dose finding is often to identify the optimal biological dose (OBD), rather than the maximum tolerated dose. We develop a utility-based Bayesian optimal interval (U-BOIN) phase I/II design to find the OBD. We jointly model toxicity and efficacy using a multinomial-Dirichlet model, and employ a utility function to measure dose risk-benefit trade-off. The U-BOIN design consists of two seamless stages. In stage I, the Bayesian optimal interval design is used to quickly explore the dose space and collect preliminary toxicity and efficacy data. In stage II, we continuously update the posterior estimate of the utility for each dose after each cohort, using accumulating efficacy and toxicity from both stages I and II, and then use the posterior estimate to direct the dose assignment and selection. Compared to existing phase I/II designs, one prominent advantage of the U-BOIN design is its simplicity for implementation. Once the trial is designed, it can be easily applied using predetermined decision tables, without complex model fitting and estimation. Our simulation study shows that, despite its simplicity, the U-BOIN design is robust and has high accuracy to identify the OBD. We extend the design to accommodate delayed efficacy by leveraging the short-term endpoint (eg, immune activity or other biological activity of targeted agents), and using it to predict the delayed efficacy outcome to facilitate real-time decision making. A user-friendly software to implement the U-BOIN is freely available at www.trialdesign.org.

Sequential or combined designs for Phase I/II clinical trials? A simulation study.

BACKGROUND: Phase I and Phase II clinical trials aim at identifying a dose that is safe and active. Both phases are increasingly combined. For Phase I/II trials, two main types of designs are debated: a dose-escalation stage to select the maximum tolerated dose, followed by an expansion cohort to investigate its activity (dose-escalation followed by an expansion cohort), or a joint modelling to identify the best trade-off between toxicity and activity (efficacy-toxicity). We explore this question in the context of a paediatric Phase I/II platform trial. METHODS: In series of simulations, we assessed the operating characteristics of dose-escalation followed by an expansion cohort (DE-EC) designs without and with reassessment of the maximum tolerated dose during the expansion cohort (DE-ECext) and of the efficacy-toxicity (EffTox) design. We investigated the probability to identify an active and tolerable agent, that is, the percentage of correct decision, for various dose-toxicity activity scenarios. RESULTS: For a large therapeutic index, the percentage of correct decision reached 96.0% for efficacy-toxicity versus 76.1% for dose-escalation followed by an expansion cohort versus 79.6% for DE-ECext. Conversely, when all doses were deemed not active, the percentage of correct decision was 47% versus 55.9% versus 69.2%, respectively, for efficacy-toxicity, dose-escalation followed by an expansion cohort and DE-ECext. Finally, in the case of a narrow therapeutic index, the percentage of correct decision was 48.0% versus 64.3% versus 67.2%, respectively, efficacy-toxicity, dose-escalation followed by an expansion cohort and DE-ECext. CONCLUSION: As narrow indexes are common in oncology, according to the present results, the sequential dose-escalation followed by an expansion cohort is recommended. The importance to re-estimate the maximum tolerated dose during the expansion cohort is confirmed. However, despite their theoretical advantages, Phase I/II designs are challenged by the variations in populations between the Phase I and the Phase II parts and by the lagtime in the evaluation of toxicity and activity.

Phase I trials as valid therapeutic options for patients with cancer.

For many years, oncology phase I trials have been referred to as 'toxicity trials' and have been believed to have low clinical utility other than that of establishing the adverse event profile of novel therapeutic agents. The traditional distinction of clinical trials into three phases has been challenged in the past few years by the introduction of targeted therapies and immunotherapies into the routine management of patients with cancer. This transformation has especially affected early phase trials, leading to the current situation in which response rates are increasingly reported from phase I trials. In this Perspectives, we highlight key elements of phase I trials and discuss how each one of them contributes to a new paradigm whereby preliminary measurements of the clinical benefit from a novel treatment can be obtained in current phase I trials, which can therefore be considered to have a therapeutic intent.