Jacob Garcia

Request for regulatory guidance for cancer cachexia intervention trials

Rome was not built in a day. Likewise, if one considers the evolution of systemic anti-cancer treatment, it took decades to go from acceptance of any therapy at all to single agents achieving isolated tumour responses (without prolongation of survival) to the current use of combination regimens as adjuvant therapy to surgery. Such incremental progress has led to improved quality of life and eventually survival for patients with some types of cancer. Cachexia and skeletal muscle wasting in cancer are significant clinical problems of high medical need for a large proportion of cancer patients and associated with very poor quality of life and very high mortality.1,2 An effective treatment of a complex multifactorial syndrome such as cachexia will likely evolve from a series of steps of discovery and new interventions before a comprehensive multimodal strategy can be identified that improves patient’s quality and quantity of life.3 There are reasons to be optimistic about the possibility that in the future, cachexia may be treated effectively. A number of drugs have already been developed that target key underlying mechanisms, namely, reduced food intake and altered metabolism and regulation of muscle mass, with the latter being split into pro-anabolic and anti-catabolic approaches. However, there is also some reason to be concerned because of the wide variability in current trial design, including different inclusion criteria, endpoints, analysis plans and the definition of best concomitant supportive care. Taken to the extreme, such differences in general approach have resulted in divergent opinions on what to consider a meaningful clinical endpoint by the European Medicines Agency (EMA) versus the US Food and Drug Administration (FDA). A result has been that in the clinical development programmes of some drugs, different endpoint assessments for American versus European regulatory authorities within the same trials using the same source data have been adopted. An example is the case of the POWER 1 and 2 trials testing the selective androgen receptor modulator enobosarm in patients with cancer suffering from muscle wasting.4–6 There has been a considerable influence from regulatory authorities on trial design. In the last 10 years, some regulatory authorities have consistently suggested that the design of randomized controlled trials testing treatments for cachexia should be aimed at demonstrating appropriate risk versus benefit, where benefit is defined as concomitant improvement in skeletal muscle mass (or lean mass) and relevant/meaningful physical function or improved survival. Whilst this is an admirable goal, from recent phase III trial results, this appears to be possibly unachievable with single modality interventions. Equally, it is not defined to whom the change is supposed to be ‘meaningful’: patient, caregiver, doctor, nurse or healthcare provider? The recent phase III trials of enobosarm used a co-primary end-point of lean body mass and stair climb power.4–6 Based on the FDA agreed co-primary responder analysis the trials failed to reach significance, principally because of lack of benefit in terms of the functional end point. In the responder analysis demanded by the FDA, an increase in performance of at least 10% for stair-climb power test was required for a patient to be considered to have benefited clinically [paper submitted]. Preventing a decline in performance was not considered in these analyses. In the analysis suggested by the EMA (which generally aims to assess clinically meaningful change regardless of direction), the same data were analysed using continuous data, and one of the two POWER trials may be considered successful, as both tests for change in lean mass and for stair-climb power showed significant changes over time. The two trials also had to be different, because of different background chemotherapy demanded in the inclusion criteria (i.e. taxane and non-taxane based). It is not clear, whether these are two trials in two orphan-type cancer indications (with different results), or are instead two trials in one indication with inconsistent results. It all depends on your approach to drug development (and maybe also on the regulatory body you talk to), but certainly it is a confusing situation by any standard. Similarly, the phase III (ROMANA) trials of the ghrelin receptor agonist anamorelin have shown significant benefit in terms of lean and fat body mass, but not for hand grip strength.7 These findings are not completely unexpected since whilst in healthy young individuals there is a strong positive correlation between muscle mass and muscle strength/power per se and between changes thereof,8,9 in older, sick individuals, the magnitude of strength generated by a certain unit of mass tends to be lower. These findings suggest that preservation/augmentation of muscle mass does not necessarily always translate into clinical benefits in non-muscular aspects of the cachexia syndrome as other factors may remain unchanged from a unimodal approach (e.g. targeting muscle anabolism). If other aspects of the cachexia syndrome remain unchanged (like systemic inflammation and catabolism or physical inactivity and undernutrition), can unimodal approaches lead to an increase in physical activity and/or preservation of independence? It appears that a more comprehensive approach to the cachexia syndrome is warranted if the reference outcomes of improved physical activity/preservation of independence are to be pursued. Still, preservation of function (and not its improvement) may also be a laudable aim of treatment development in cachexia, and regulatory guidance should permit for that. However, unlike areas such as hypertension where a given change in blood pressure is accepted as a surrogate for clinical events, it has to be recognized that in cachexia, the ‘relevance’ or ‘meaningfulness’ of a change in surrogates such as hand grip strength, stair-climb power, leg extension strength or timed sit-up-and-go is not known. Perhaps direct measures of patients’ daily physical activity would be better? In the related field of COPD-associated body wasting, exercise rehabilitation is well established with extensive guideline recommendations that are evidence-based.10 These guidelines have been developed over time and are multimodal in focus and are explicitly aimed at improving physical functioning and physical activity levels, nutritional status and quality of life. For patients with heart failure, chronic kidney disease, stroke or ageing-related frailty, such multimodal approaches are frequently considered,11–13 but evidence is so far weak compared with what has been achieved in COPD. Novel therapeutic agents under development for cachexia mostly focus on specific aspects of the syndrome (e.g. muscle anabolism, inflammation or appetite stimulation).14 Surely, phase III registration trials should assess safety in general, but efficacy specifically in relation to the target of the drug based on its mechanism of action. It may not be right to discard an intervention as ineffective because it does not yet affect a functional outcome, if, in fact – when inserted into a multimodal intervention that reflects the multifaceted aspect of the cachexia syndrome – the drug shows extended benefits that touch on issues such as health-related quality of life, patient-reported symptoms and tolerance of anti-cancer therapy. In the context of a complex disease process and a desire for multimodal therapies, regulatory advice on co-therapy with nutrition and exercise is also needed. Suggestions as to how best to include in this context supportive care in clinical trials 15 may also be helpful. We understand that this may include additional clinical trials for food products and supportive care approaches and surely this is acceptable, if the rules of the ‘game’ are clear for the good of our patients. Regulators need to be engaged in encouraging the testing of these modalities and their systematic inclusion in trial designs. In heart failure, such activities have already been initiated and aim to shift the development and authorization of medicines from the molecule paradigm to their evaluation in the context of the whole healthcare regimen.16 If a trial of a new agent incorporates these elements and is successful, it cannot lie with the pharmaceutical company to ensure that such adjuncts are available in precisely the same format everywhere in the world. Rather, the approved drug label may need to recommend such adjuncts for optimal effect. Clearly, this is not an easy field for new developments, but the medical need is great and the commercial returns for those who make it may be big. Once drugs are approved, the longer process of incorporating new agents into best clinical practice can begin. It should be clear to pharmaceutical companies, academic trialists and regulators that they may need to be more realistic about what can be achieved in a single step. Maybe also the adaptive licensing approach proposed by EMA 13 can help in this process of developing regulatory pathways. A willingness to consider current data with an open mind and a ‘Notice on Regulatory Guidance’ on cachexia trial design for cancer and beyond that cuts across continents would be a major step forwards to maintain drug development momentum, if there is to be genuine progress at this exciting juncture in the development of cachexia therapy. We want to help make that a reality whichever way we can.

Comparative effectiveness of granulocyte colony-stimulating factors to prevent febrile neutropenia and related complications in cancer patients in clinical practice: A systematic review.

Introduction Febrile neutropenia (FN) is a serious side-effect of myelosuppressive chemotherapy. Several clinical trials and observational studies have evaluated the effects of prophylactic granulocyte colony-stimulating factors (G-CSFs) on risk of FN and related complications; however, no systematic reviews have focused on effectiveness in routine clinical practice. Here, we perform a systematic review assessing the comparative effectiveness of prophylaxis with a long-acting G-CSF (pegfilgrastim) versus short-acting G-CSFs (filgrastim, lenograstim, and filgrastim biosimilars) in cancer patients in real-world clinical settings. Methods A systematic review was performed based on a pre-specified protocol and was consistent with the Cochrane Collaboration Handbook (2009) and the Centre for Reviews and Dissemination’s Guidance for Undertaking Reviews in Health Care (2011). MEDLINE, Embase, BIOSIS, Cumulative Index to Nursing and Allied Health Literature, and Cochrane Library databases were searched for articles published from January 2002 to June 2014. Congress databases (MASCC/ASCO/ESMO) and Google Scholar were searched for abstracts published from January 2012 to August 2014. Filgrastim (NEUPOGEN®), lenograstim and nivestim (a filgrastim biosimilar) were the only short-acting G-CSFs and pegfilgrastim (Neulasta®) was the only long-acting G-CSF described in eligible studies. Outcomes of interest were FN, FN-related hospitalisation and other FN-related complications (death, chemotherapy dose delays and reductions, antimicrobial treatment, severe neutropenia and costs and resource use). Results Of 1259 unique records identified, 18 real-world observational studies met predefined inclusion criteria; 15 were retrospective studies, and 3 were prospective studies. Multiple tumour types, chemotherapy regimens and geographical regions were included. Seven studies provided statistical comparisons of the risk of FN; risk of FN among patients receiving prophylaxis with pegfilgrastim versus short-acting G-CSF was significantly lower in three studies, numerically lower in three studies, and numerically higher in one study. Six studies provided statistical comparisons of the risk of FN-related hospitalisation; risk of FN-related hospitalisation among patients receiving prophylaxis with pegfilgrastim versus short-acting G-CSF was significantly lower in all six studies, though some variation was seen in subanalyses. Data for other outcomes were sparse with available results being generally consistent with the results seen for risk of FN and FN-related hospitalisation. Conclusions Based on the findings from this review of real-world comparative effectiveness studies, risks of FN and FN-related complications were generally lower for prophylaxis with pegfilgrastim versus prophylaxis with short-acting G-CSFs.

Burden of Chemotherapy-Induced Febrile Neutropenia Hospitalizations in US Clinical Practice, by Use and Patterns of Prophylaxis with Colony-Stimulating Factor

IntroductionEvidence suggests that many cancer chemotherapy patients who are candidates for colony-stimulating factor (CSF) prophylaxis do not receive it or receive it inconsistent with guidelines, and that such patients have a higher risk of febrile neutropenia hospitalization (FNH). Little is known about the number and consequences of FNH by use/patterns of CSF prophylaxis in US clinical practice.MethodsA retrospective cohort design and private healthcare claims data were employed. Study population comprised adults who received a chemotherapy course with a high-risk regimen, or an intermediate-risk regimen (if ≥1 FN risk factor present), for non-metastatic breast cancer or non-Hodgkin’s lymphoma (NHL); each chemotherapy cycle within the course and each FNH episode within the cycles were identified. Consequences included mortality, inpatient days, and costs (US

Reply to Barroso-Sousa R et al.

2013) during FNH. Use (yes/no) and patterns (agent, administration day/duration) of CSF prophylaxis were evaluated within cycles in which FNH episodes occurred.ResultsAmong all FNH episodes (n=6,355; 109 episodes per 1,000 patients), 41.3% (95% CI: 40.1-42.5) occurred among patients who did not receive CSF prophylaxis in that cycle, and 8.8% (8.1-9.5) occurred among those who received CSF prophylaxis on the same day as chemotherapy. Among FNH episodes occurring in patients who received daily CSF agents (2% of CSF use), 56.1% (44.1-68.0) received prophylaxis <7 days during the cycle. Results for FNH consequences were comparable.ConclusionsIn this retrospective evaluation, one-half of FNH episodes, outcomes, and costs among cancer chemotherapy patients who were candidates for CSF prophylaxis occurred in those who either did not receive it or received it inconsistent with guidelines.

Chemotherapy Dose Intensity and Overall Survival Among Patients With Advanced Breast or Ovarian Cancer

G-CSF prophylaxis is routinely recommended for every cycle of dose-dense chemotherapy regimens such as ddAC-T [1e3]. In a small, retrospective, single-institution study, Barroso-Sousa et al. described G-CSF use during the T portion of ddAC-T: 87% of patients (n 1⁄4 135) received G-CSF during 1 cycle of T, 13% (n 1⁄4 21) did not. In contrast with previous studies showing G-CSF benefit during T [4,5], Barroso-Sousa reported no significant differences in dose delays or treatment completion. However, as the authors discussed, this study was limited by a small sample size, which precluded the ability to accurately compare outcomes. In addition, appropriate G-CSF use could not be adequately evaluated, as G-CSF use was at the physicians discretion and patient risk factors were not thoroughly assessed or controlled for. Furthermore, while high relative dose intensity (RDI) is associated with improved clinical outcomes in patients with early-stage breast cancer [6], the effect of G-CSF use on chemotherapy dose reductions and RDI were not described by Barroso-Sousa et al. We appreciate the hypothesis-generating data in this study. Given the limitations of this retrospective study, we look forward to the results of the prospective trial currently underway (NCT02698891) to describe baseline myelotoxicity associated

Abstract P6-10-01: Risk of chemotherapy-induced febrile neutropenia (FN) in patients (pts) with non-metastatic breast cancer (BC) and documented risk factors for FN

Background The effects of chemotherapy dose intensity on patient outcomes in advanced cancer are not well understood. We studied the association between chemotherapy relative dose intensity (RDI) and overall survival (OS) among patients with advanced breast or ovarian cancer. Patients and Methods This retrospective cohort study included adults with advanced breast or ovarian cancer who received first‐line myelosuppressive chemotherapy (January 2007 to December 2010) in US Oncology Network community practices. Dose delays ≥ 7 days, dose reductions ≥ 15%, and RDI relative to standard regimens were described. OS was measured by the Kaplan‐Meier method and Cox proportional hazards models. Results Among 874 patients with advanced breast cancer, 33.2% experienced dose delays ≥ 7 days, 48.7% experienced dose reductions ≥ 15%, and 38.9% had RDI < 85%. In the multivariable Cox proportional hazards model, Eastern Cooperative Oncology Group performance status 1/2 versus 0 (hazard ratio [HR] = 1.45; 95% confidence interval [CI], 1.15‐1.82) and triple‐negative status (HR = 3.14; 95% CI, 1.15‐8.62) were significantly associated with mortality. Among 170 patients with advanced ovarian cancer, 43.5% experienced dose delays ≥ 7 days, 48.2% experienced dose reductions ≥ 15%, and 46.5% had RDI < 85%. In the multivariable Cox proportional hazards model, dose reductions ≥ 15% (HR = 1.94; 95% CI, 1.09‐3.46) and other tumor histology (vs. nonserous adenocarcinoma; HR = 3.55; 95% CI, 1.38‐9.09) were significantly associated with mortality. Conclusion Dose delays, dose reductions, and reduced RDI were common. In advanced breast cancer, health status and triple‐negative disease were significantly associated with mortality. In advanced ovarian cancer, dose reductions and tumor histology were significantly associated with mortality. These results can help identify potential risk factors and characterize the effect of chemotherapy dose modification strategies on mortality. Micro‐Abstract A retrospective cohort study of patients with advanced breast (n = 874) and ovarian cancer (n = 170) was conducted to evaluate the association between relative dose intensity of first‐line intravenous myelosuppressive chemotherapy and overall survival. Factors associated with increased mortality were identified for both cancer types. Knowledge of potential risk factors may help inform the effect of dose modification strategies on mortality.

Dose Delays, Dose Reductions, and Relative Dose Intensity in Patients With Cancer Who Received Adjuvant or Neoadjuvant Chemotherapy in Community Oncology Practices

Background: Clinical practice guidelines recommend primary prophylactic colony-stimulating factor (CSF) for pts with cancer receiving myelosuppressive chemotherapy when their risk of FN is high (≥20%). Evaluating FN risk in pts who receive regimens that are not documented as high risk in guidelines can be challenging; these pts may be at a high risk of developing FN based on a combination of regimen and pt risk factors. This retrospective study estimated FN risk among subgroups of pts with non-metastatic BC receiving 1 of 3 commonly used, non-high risk chemotherapy regimens. Methods: Pt-level data from 2 US healthcare claims databases comprising medical and outpatient pharmacy claims from commercial and Medicare supplemental plans were pooled. Eligible pts were ≥18 years old and had initiated a course of TC, TCH, or non-dose-dense AC/AC-T for non-metastatic BC between July 1, 2003 and June 30, 2012. Occurrence of FN during any cycle of the first chemotherapy course was identified using diagnosis codes for neutropenia, fever, or infection. Risk factors, documented in guidelines and the published literature, were evaluated during the 12 months before chemotherapy initiation. The percentage of pts, FN risk, and relative risk (RR, compared to pts with no risk factors) for subgroups of pts with specific risk factors are presented. Results: 50,893 pts were included in the analysis. FN risks by chemotherapy regimen and subgroup are shown in the table. A high proportion of pts (74%-78%) had ≥1 risk factor for FN, and these pts had a higher FN risk than pts with no risk factors. 55.4%, 56.7%, and 42.2% of all pts who received TC, TCH, or AC/AC-T, respectively, received CSF prophylaxis in cycle 1. Conclusions: FN risk assessments are needed for pts who are receiving non-high risk regimens. Citation Format: Derek Weycker, Xiaoyan Li, Richard Barron, Hongsheng Wu, Phuong Khanh Morrow, Hairong Xu, Maureen Reiner, Jacob Garcia, Shivani Mhatre, Gary Lyman. Risk of chemotherapy-induced febrile neutropenia (FN) in patients (pts) with non-metastatic breast cancer (BC) and documented risk factors for FN [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P6-10-01.