A proposal for a new PhD level curriculum on quantitative methods for drug development.

This paper provides an overview of "Improving Design, Evaluation and Analysis of early drug development Studies" (IDEAS), a European Commission-funded network bringing together leading academic institutions and small- to large-sized pharmaceutical companies to train a cohort of graduate-level medical statisticians. The network is composed of a diverse mix of public and private sector partners spread across Europe, which will host 14 early-stage researchers for 36 months. IDEAS training activities are composed of a well-rounded mixture of specialist methodological components and generic transferable skills. Particular attention is paid to fostering collaborations between researchers and supervisors, which span academia and the private sector. Within this paper, we review existing medical statistics programmes (MSc and PhD) and highlight the training they provide on skills relevant to drug development. Motivated by this review and our experiences with the IDEAS project, we propose a concept for a joint, harmonised European PhD programme to train statisticians in quantitative methods for drug development.

Hyperprogression during anti-PD-1/PD-L1 therapy in patients with recurrent and/or metastatic head and neck squamous cell carcinoma.

BACKGROUND: Pembrolizumab and nivolumab are immune checkpoint inhibitors targeting PD-1 that have recently been approved in pretreated recurrent and/or metastatic head and neck squamous cell carcinoma (R/M HNSCC) patients. In the clinic, some patients seem not only not to benefit from anti-PD-L1/PD-1 agents but rather to experience an acceleration of tumor growth kinetics (TGK). PATIENTS AND METHODS: We retrospectively compared TGK on immunotherapy and TGK on last treatment in patients with R/M HNSCC treated with PD-1/PD-L1 inhibitors in four French centers. The TGK ratio (TGKR, ratio of the slope of tumor growth before treatment and the slope of tumor growth on treatment) was calculated. Hyperprogression was defined as a TGKR ≥ 2. RESULTS: From September 2012 to September 2015, 34 patients were identified. Patterns of recurrence included exclusive loco-regional recurrence in 14 patients, exclusive distant metastases in 11 patients, and both in 9 patients. No pseudo-progression was observed. Hyperprogression was observed in 10 patients (29%), including 9 patients with at least a locoregional recurrence, and only 1 patient with exclusively distant metastases. Hyperprogression significantly correlated with a regional recurrence (TGKR ≥ 2: 90% versus TGKR < 2: 37%, P = 0.008), but not with local or distant recurrence. Hyperprogression was associated with a shorter progression-free survival (PFS) according to RECIST (P = 0.003) and irRECIST (P = 0.02), but not with overall survival (P = 0.77). CONCLUSIONS: Hyperprogression was observed in 29% of patients with R/M HNSCC treated with anti-PD-L1/PD-1 agents and correlated with a shorter PFS. It occurred in 39% of patients with at least a locoregional recurrence and 9% of patients with exclusively distant metastases. No pseudo-progressions were reported. Mechanisms and causality of hyperprogression should further be assessed through prospective controlled studies.

Randomized phase II trial comparing molecularly targeted therapy based on tumor molecular profiling versus conventional therapy in patients with refractory cancer: cross-over analysis from the SHIVA trial.

Background: Several studies used the ratio of progression-free survival (PFS) on genotype-matched treatment to PFS on genotype-unmatched treatment to assess the efficacy of therapy guided by patients' tumor molecular profiling. We evaluated the PFS ratio from patients who cross-over in the SHIVA trial. Patients and methods: The primary end point of the SHIVA trial was to compare PFS on molecularly targeted agents (MTAs) based on tumor molecular profiling and treatment at physician's choice (TPC) in patients with any kind of cancer who had failed standard-of-care therapy. The experimental treatment included only marketed MTAs given outside their indications according to a pre-specified treatment algorithm. Patients were allowed to cross-over at disease progression in both arms. Response was evaluated according to RECIST 1.1 at randomization and at cross-over. We evaluated the ratio of PFS on MTA (PFSMTA) to PFS on TPC (PFSTPC) in patients who crossed-over. Results: Among 741 patients enrolled in the SHIVA trial, 197 were randomized, and 95 crossed-over, including 70 patients from the TPC to the MTA arm and 25 patients from the MTA to the TPC arm. Two patients crossed-over in the TPC arm without disease progression. The PFSMTA/PFSTPC ratio exceeded 1.3 in 37% of patients who crossed-over from the TPC to the MTA arm. The PFSMTA/PFSTPC ratio exceeded 1.3 in 61% of patients who crossed-over from the MTA arm to the TPC arm. Conclusions: The cross-over analysis of the SHIVA trial identified 37% of patients who crossed-over from TPC to MTA with a PFSMTA/PFSTPC ratio exceeding 1.3.

Patients' perceived tolerance of side effects in phase I cancer clinical trials: A qualitative study.

This qualitative study aimed to explore cancer patients' perceived tolerance of side effects in phase I drug trials. Patients with solid tumours receiving molecularly targeted agents with/without chemotherapy were eligible for inclusion. In-depth semi-structured interviews were carried out with 17 patients with a median [range] age of 63 [41-72] years. Treatment was discontinued in seven patients. Verbatim transcripts of the audio-taped interviews were analysed using a constructivist grounded theory approach. Four conceptual categories emerged from data analysis, labelled "suffering from side effects" comprising a range of symptoms, psychosocial or role disturbances; "striving to cope with side effects" reflecting psychological strategies for managing side effects; "hoping" reflecting expectations about treatment efficacy and relief from side effects; and "appraisal of care." Among patients remaining in the trial, treatment was currently perceived as fairly tolerable. For most respondents, whether still in a trial or not, treatment discontinuation could not be justified by the non-tolerance of treatment side effects. These results question the adequacy of patient-perceived tolerance reports to determine an optimal drug dose for phase II trials. Confronted with patients' hopes and inappropriate beliefs, communication is challenging in phase I trials and could benefit from facilitating psychosocial interventions.

Statistical controversies in clinical research: requiem for the 3 + 3 design for phase I trials.

BACKGROUND: More than 95% of published phase I trials have used the 3 + 3 design to identify the dose to be recommended for phase II trials. However, the statistical community agrees on the limitations of the 3 + 3 design compared with model-based approaches. Moreover, the mechanisms of action of targeted agents strongly challenge the hypothesis that the maximum tolerated dose constitutes the optimal dose, and more outcomes including clinical and biological activity increasingly need to be taken into account to identify the optimal dose. PATIENTS AND METHODS: We review key elements from clinical publications and from the statistical literature to show that the 3 + 3 design lacks the necessary flexibility to address the challenges of targeted agents. RESULTS: The design issues raised by expansion cohorts, new definitions of dose-limiting toxicity and trials of combinations are not easily addressed by the 3 + 3 design or its extensions. CONCLUSIONS: Alternative statistical proposals have been developed to make a better use of the complex data generated by phase I trials. Their applications require a close collaboration between all actors of early phase clinical trials.

Designs of drug-combination phase I trials in oncology: a systematic review of the literature.

BACKGROUND: Combining several anticancer agents can increase the overall antitumor action, but at the same time, it can also increase the overall observed toxicity. Adaptive dose-escalation designs for drug combinations have recently emerged as an attractive alternative to algorithm-based designs, and they seem more effective in combination recommendations. These methods are not used in practice currently. Our aim is to describe international scientific practices in the setting of phase I drug combinations in oncology. MATERIAL AND METHODS: A bibliometric study on phase I dose-finding combination trials was conducted using the Medline(®) PubMed database between 1 January, 2011, and 31 December 2013. Sorting by abstract, we selected all papers involving a minimum of two agents and then retained only those in which at least two agents were dose-escalated. RESULTS: Among the 847 references retrieved, 162 papers reported drug-combination phase I trials in which at least two agents were dose-escalated. In 88% of trials, a traditional or modified 3 + 3 dose-escalation design was used. All except one trial used a design developed for single-agent evaluation. Our study suggests that drug-combination phase I trials in oncology are very safe, as revealed by the calculated median dose-limiting toxicity rate of 6% at the recommended dose, which is far below the target rate in these trials (33%). We also examined requirements of phase I clinical trials in oncology with drug combinations and the potential advantages of novel approaches in early phases. CONCLUSION: Efforts to promote novel and innovative approaches among statisticians and clinicians appear valuable. Adaptive designs have an important role to play in early phase development.

Designs of preoperative biomarkers trials in oncology: a systematic review of the literature.

BACKGROUND: The identification of predictive and pharmacodynamics (PD) biomarkers of efficacy of anticancer-targeted therapies is not always straightforward. To address this problem, preoperative trials have been set up. The present study aimed at evaluating how these trials are designed. DESIGN: We retrieved all preoperative oncology trials, defined as preoperative trials having a PD end point. RESULTS: Only 56 trials met our selection criteria. Of these, 27 trials (48%) were randomized. Forty-nine trials (88%) evaluated at least a noncytotoxic agent. In 37 trials (66%), a single agent was administered. The most prevalent tumor type was breast cancer (59%). Median duration of accrual was 28 months (range: 9-98). In these trials, there was a mean of two patients included per month (range: 0-7). The date of surgery was fixed before study entry in 35 trials (62%), while surgery was set up after preoperative therapy in the remaining 21 trials (38%). In the former trials, median duration of preoperative therapy was 17 days (range: 1-112), whereas in the latter trials it ranged from 4 to 29 weeks. The primary end point was a PD end point in 26 of the 45 trials (58%) in which it was mentioned. One percent of patients could not undergo surgery as per protocol due to an adverse event. Statistically significant predictive and PD biomarkers were identified in 17 (30%) and 27 trials (48%), respectively. CONCLUSION: Preoperative biomarkers trials are infrequent but safe and feasible. These trials often permit the identification of predictive and PD biomarkers.

A new approach to integrate toxicity grade and repeated treatment cycles in the analysis and reporting of phase I dose-finding trials.

BACKGROUND: Safety assessment beyond the dose-limiting toxicity evaluation period provides relevant information to define the recommended phase II dose (RP2D) of a new treatment. We retrospectively analyzed three phase I trials to illustrate two indicators: per-cycle probability of graded toxicity and cumulative probability of severe toxicity over the treatment period. PATIENTS AND METHODS: Data were collected from two continual reassessment method (CRM) trials (T1: aviscumine in solid tumors with short time on treatment; T2: erlotinib + radiotherapy in brainstem gliomas with longer time on treatment) and one 3 + 3 design (T3: liposomal doxorubicin + cyclophosphamide combination in ovarian carcinoma). The probability of severe and moderate or severe toxicity per cycle was estimated at each dose level with mixed proportional odds model. The cumulative probability of severe toxicity was also estimated with the time-to-event CRM. RESULTS: Eighty-three patients were included in the three trials; 94, 96 and 72 treatment cycles were administered, in T1, T2 and T3, respectively. Moderate toxicities were at least twice as frequent as severe toxicities. An increased probability of toxicity over time was detected in T3 [P = 0.04; per-cycle probability of severe toxicity: 27% (cycle 1) to 59% (cycle 6) at the RP2D]. At the RP2D, 37% of patients experienced at least one severe toxicity over the first six cycles in T2, and 78% in T3. CONCLUSIONS: Dedicated methods can be used to analyze toxicities from all cycles of treatment. They do not delay accrual and should be integrated in the analysis and reporting of phase I dose-finding trials.

Randomised proof-of-concept phase II trial comparing targeted therapy based on tumour molecular profiling vs conventional therapy in patients with refractory cancer: results of the feasibility part of the SHIVA trial.

BACKGROUND: The SHIVA trial is a multicentric randomised proof-of-concept phase II trial comparing molecularly targeted therapy based on tumour molecular profiling vs conventional therapy in patients with any type of refractory cancer. RESULTS of the feasibility study on the first 100 enrolled patients are presented. METHODS: Adult patients with any type of metastatic cancer who failed standard therapy were eligible for the study. The molecular profile was performed on a mandatory biopsy, and included mutations and gene copy number alteration analyses using high-throughput technologies, as well as the determination of oestrogen, progesterone, and androgen receptors by immunohistochemistry (IHC). RESULTS: Biopsy was safely performed in 95 of the first 100 included patients. Median time between the biopsy and the therapeutic decision taken during a weekly molecular biology board was 26 days. Mutations, gene copy number alterations, and IHC analyses were successful in 63 (66%), 65 (68%), and 87 (92%) patients, respectively. A druggable molecular abnormality was present in 38 patients (40%). CONCLUSIONS: The establishment of a comprehensive tumour molecular profile was safe, feasible, and compatible with clinical practice in refractory cancer patients.

REBECA: a phase I study of bevacizumab and whole-brain radiation therapy for the treatment of brain metastasis from solid tumours.

BACKGROUND: Brain metastases (BMs) are associated with a poor prognosis. Standard treatment comprises whole-brain radiation therapy (WBRT). As neo-angiogenesis is crucial in BM growth, combining angiogenesis inhibitors such as bevacizumab with radiotherapy is of interest. We aimed to identify the optimal regimen of bevacizumab combined with WBRT for BM for phase II evaluation and provide preliminary efficacy data. PATIENTS AND METHODS: In this multicentre single-arm phase I study with a 3 + 3 dose-escalation design, patients with unresectable BM from solid tumours received three cycles of bevacizumab at escalating doses [5, 10 and 15 mg/kg every 2 weeks at dose levels (DL) 0, 1 and 2, respectively] and WBRT (30 Gy/15 fractions/3 weeks) administered from day 15. DL3 consisted of bevacizumab 15 mg/kg with WBRT from day 15 in 30 Gy/10 fractions/2 weeks. Safety was evaluated using NCI-CTCAE version 3. BM response (RECIST 1.1) was assessed by magnetic resonance imaging at 6 weeks and 3 months after WBRT. RESULTS: Nineteen patients were treated, of whom 13 had breast cancer. There were no DLTs. Grade 1-2 in-field and out-field toxicities occurred for five and nine patients across all DLs, respectively, including three and six patients (including one patient with both, so eight patients overall) of nine patients in DL3. One patient experienced BM progression during treatment (DL0). At the 3-month post-treatment assessment, 10 patients showed a BM response: one of three treated at DL0, one of three at DL1, two of three at DL2 and six of seven at DL3, including one complete response. BM progression occurred in five patients, resulting in two deaths. The remaining patient died from extracranial disease progression. CONCLUSION: Bevacizumab combined with WBRT appears to be a tolerable treatment of BM. DL3 warrants further efficacy evaluation based on the favourable safety/efficacy balance. ClinicalTrials.gov Identifier: NCT01332929.

Evaluation of the relevance of the growth modulation index (GMI) from the FFCD 0307 randomized phase III trial comparing the sequence of two chemotherapeutic regimens.

BACKGROUND: Precision medicine trials disrupted the paradigm of randomized controlled trials in large populations. Patient selection may be based on molecular alterations rather than on primary tumor location. In small patient populations, the growth modulation index (GMI) has been developed to evaluate treatment efficacy by using each patient as its own control. The FFCD 0307 randomized phase III trial compared two sequences of chemotherapy in advanced gastric cancer, which represents a unique opportunity to evaluate the relevance of the GMI. PATIENTS AND METHODS: In the FFCD 0307 trial, patients with advanced gastric cancer were randomized between two chemotherapy sequences [ECX followed by FOLFIRI at disease progression (arm A) versus FOLFIRI followed by ECX (arm B)]. GMI was defined as the ratio of the progression-free survival on second treatment (PFS2) to the time to progression on first treatment (TTP1). Sequence benefit was defined as a GMI exceeding 1.3 (GMI-high). GMI was correlated with overall survival (OS). OS1 and OS2 were measured from first randomization and second-line failure to death. RESULTS: Four hundred and sixteen patients were randomized (209 in arm A, 207 in arm B). One hundred and seventy-five patients (42%) received the two sequences and were assessable for GMI (97 in arm A, 79 in arm B). The median GMI was higher in arm A than in arm B (0.62 versus 0.47, P = 0.04). Patients with a high GMI had a longer OS1 (median 14.9 versus 11.5 months, NS). Median OS2 was doubled in the GMI-high group (3.4 versus 1.6 months, NS). CONCLUSION: GMI analyses suggest that ECX followed by FOLFIRI might represent a better therapeutic strategy than FOLFIRI followed by ECX. High GMI was associated with prolonged survival.

Tumour growth kinetics assessment: added value to RECIST in cancer patients treated with molecularly targeted agents.

BACKGROUND: Treatment effect is categorised into four classes by RECIST based on the evolution of the size of target lesions and the occurrence of new lesions, irrespective of tumour growth kinetics before treatment. This study aimed at evaluating the added value of tumour growth kinetics assessment to RECIST in patients treated with molecularly targeted agents (MTAs). METHODS: On-study imaging, along with pre-baseline imaging, of patients treated with MTA(s) in clinical trials at Institut Curie were centrally reviewed. The tumour growth ratio (TGr), defined as the ratio of the slope of tumour growth before treatment and the slope of tumour growth on treatment between the nadir and disease progression, was calculated for each patient. RESULTS: A total of 50 patients included in 18 trials were eligible for the study. Among the 44 patients who withdrew from the study because of disease progression according to the investigators' assessment, 18 patients (41%) had a TGr <0.9. Among these 18 patients, 5 had disease progression according to RECIST 1.1 based on our retrospective reassessment of on-study imaging and occurrence of no new lesion during study treatment. CONCLUSION: Our preliminary results suggest that a substantial proportion of patients treated with MTAs have discontinued treatment although being potentially benefitted from them.

Impact of the timing of tumor assessments on median progression-free survival in clinical trials in advanced cancer patients.

BACKGROUND: Survival-based surrogate endpoints such as progression-free survival (PFS) are commonly used in oncology clinical trials. The evaluation-time bias in the assessment of median disease progression in randomized trials has been suggested by several simulation studies, but never demonstrated in the clinic. We aimed to demonstrate the existence of potential evaluation-time bias by assessing the impact of the timing of tumor assessments on median PFS from control arms without any active treatment of randomized controlled trials involving advanced cancer patients. MATERIALS AND METHODS: A systematic literature search of English language publications from 1 January 2000 to 7 January 2021 was performed using MEDLINE (PubMed). Eligible trials for our meta-analysis included all randomized clinical trials evaluating anticancer drugs in adult patients with advanced cancers with a control arm without any anticancer drug consisting of best supportive care with or without a placebo. We performed a meta-regression analysis to analyze the correlation between the timing of the first tumor assessment and median PFS in patients randomized in the control arms without any active treatment. RESULTS: Of 3551 studies screened, 97 eligible trials were retrieved involving 36  747 patients, including 14  229 patients randomized into the control arms. A later first tumor assessment correlated with a prolonged median PFS (R2 = 0.44, P < 10-5). CONCLUSIONS: Our results confirm the existence of potential evaluation-time bias in clinical research that had been suggested by simulation studies. The timing of tumor assessments should be kept the same in precision medicine trials using the PFS ratio as an efficacy endpoint.

A comparison of model choices for the Continual Reassessment Method in phase I cancer trials.

Determination of the maximum tolerated dose (MTD) is the main objective of phase I trials. Trials are typically carried out with restricted sample sizes. Model-based approaches proposed to identify the MTD (including the Continual Reassessment Method or CRM) suppose a simple model for the dose-toxicity relation. At this early stage of clinical development, the true family of models is not known and several proposals have been done. Asymptotic convergence of the recommendation to the true MTD can be obtained with a one-parameter model even in case of model misspecification. Nevertheless, operating characteristics with finite sample sizes can be largely affected by the choice of the model. In this paper, we evaluate and compare several models in a simulation framework. This framework includes a large class of dose-toxicity relations against which to test the competing models, an 'optimal' method that provides efficient non-parametric estimates of the probability of dose limiting toxicity to serve as a benchmark and as a graphic representation. In particular we explore the use of a one-parameter versus a two-parameter model, we compare the power and the logistic models and finally we investigate the impact of dose recoding on the operating characteristics. Comparisons are carried out with both a likelihood approach and a Bayesian approach for model estimations. We show that average performances of a one-parameter model are superior and that the power model has good operating characteristics. Some models can speed up dose escalation and lead to more aggressive designs. We derive some behavior related to the choice of model and insist on the use of simulations under several scenarios before the initiation of each new trial in order to determine the best model to be used.

Prognosis of advanced hepatocellular carcinoma: comparison of three staging systems in two French clinical trials.

OBJECTIVE: The objective of this study was to assess the performance of three staging systems [Okuda, Cancer of the Liver Italian Program (CLIP) and Barcelona Clinic Liver Cancer group (BCLC)], for predicting survival in patients with hepatocellular carcinoma (HCC) and to explore how to improve prognostic classification among French patients with HCC whose main etiology is alcoholic cirrhosis. METHODS: We have pooled two randomized clinical trials in palliative condition from the Fédération Francophone de Cancerologie Digestive. They had included 416 and 122 patients. Performances of Okuda, CLIP and BCLC scores have been compared using Akaike information criterion, discriminatory ability (Harrell's C and the Royston's D statistics), monotonicity of gradients and predictive accuracy (Schemper statistics Vs). To explore how to improve classifications, univariate and multivariate Cox model analyses were carried out. RESULTS: The pooled database included 538 patients. The median survival was 5.3 months (95% confidence interval 4.6-6.2). For all statistics CLIP staging system had a better prognostic ability. Performances of all staging systems were rather disappointing. World Health Organization performance status (WHO PS) for CLIP or alpha-fetoprotein for BCLC allowed a significant improvement of prognostic information. CONCLUSION: Our results indicate that CLIP staging seems to be most adapted to palliative setting and that it could be better by associating WHO PS.

Phase II study of XR 5000 (DACA), an inhibitor of topoisomerase I and II, administered as a 120-h infusion in patients with non-small cell lung cancer.

XR5000 is a tricyclic carboxamide-based cytotoxic agent that binds to DNA by intercalation and stimulates DNA cleavage by inhibition of both topoisomerase I and II. The aim of this study was to evaluate the antitumoral activity and safety profile of XR5000 given as second-line chemotherapy in patients with advanced non-small cell lung cancer (NSCLC). Patients received XR5000 at the dose of 3010 mg/m(2) as a 120-h central venous infusion every 3 weeks. The 15 patients (median age 56 years, range 48-71 years) enrolled had an Eastern Cooperative Oncology Group (ECOG) performance status of 0 (3 patients), 1 (11 patients) or 2 (1 patient). A total of 32 cycles of XR5000 (median 2, range 1-6) were given to 14 patients. No objective response (assessed according to World Health Organization (WHO) criteria) was documented in the 12 evaluable patients by an external review panel; in 4 out of the 12 patients disease stabilisation was recorded. The following toxicities graded according to the Common Toxicity Criteria (CTC) version 2.0. were observed: one grade 3 and two grade 4 granulocytopenia, one grade 3 and one grade 4 thrombocytopenia, one grade 3 deep venous thrombosis, one grade 3 fatigue, and grade 3 undocumented epileptic seizures which led to death in 2 patients. With only 4 out of 12 patients reaching stable disease when using this dose and regimen, further evaluation of XR5000 in advanced NSCLC is not justified.

Multicentre phase II and pharmacokinetic study of RFS2000 (9-nitro-camptothecin) administered orally 5 days a week in patients with glioblastoma multiforme.

A phase II trial was instigated to investigate the antitumour activity, the safety and the pharmacokinetic parameters of RFS2000, a recently identified oral topoisomerase I inhibitor, given once daily (1.5 mg/m(2)/day) as first-line chemotherapy treatment for patients with advanced glioblastoma multiforme (GBM). Between 9 March and 15 September 2000, 17 patients were entered onto the trial. 15 patients were considered eligible. A total of 49 cycles (range 1-8) were administered. Grade 3-4 toxicity was observed in 5 patients. Neutropenia and thrombocytopenia were common toxicities. Pharmacokinetic analysis showed that 9-nitro camptothecin (9-NC) could be detected in the plasma and is progressively converted into 9-amino-camptothecin (9-AC). The response rate was poor, with 5 patients experiencing tumour stabilisation and 10 progressing. Thus, the results do not support the further evaluation of RFS2000 as a single agent in patients with recurrent GBM.