Exploratory composite endpoint demonstrates benefit of trilaciclib across multiple clinically meaningful components of myeloprotection in patients with small cell lung cancer

Abstract Chemotherapy‐induced myelosuppression is an acute, dose‐limiting toxicity of chemotherapy regimens used in the treatment of extensive‐stage small cell lung cancer (ES‐SCLC). Trilaciclib protects haematopoietic stem and progenitor cells from chemotherapy‐induced damage (myeloprotection). To assess the totality of the myeloprotective benefits of trilaciclib, including analysis of several clinically relevant but low‐frequency events, an exploratory composite endpoint comprising five major adverse haematological events (MAHE) was prospectively defined: all‐cause hospitalisations, all‐cause chemotherapy dose reductions, febrile neutropenia (FN), prolonged severe neutropenia (SN) and red blood cell (RBC) transfusions on/after Week 5. MAHE and its individual components were assessed in three randomised, double‐blind, placebo‐controlled Phase 2 trials in patients receiving a platinum/etoposide or topotecan‐containing chemotherapy regimen for ES‐SCLC and in data pooled from the three trials. A total of 242 patients were randomised across the three trials (trilaciclib, n = 123; placebo, n = 119). In the individual trials and the pooled analysis, administering trilaciclib prior to chemotherapy resulted in a statistically significant reduction in the cumulative incidence of MAHE compared to placebo. In the pooled analysis, the cumulative incidences of all‐cause chemotherapy dose reductions, FN, prolonged SN and RBC transfusions on/after Week 5 were significantly reduced with trilaciclib vs placebo; however, no significant difference was observed in rates of all‐cause hospitalisations. Additionally, compared to placebo, trilaciclib significantly extended the amount of time patients remained free of MAHE. These data support the myeloprotective benefits of trilaciclib and its ability to improve the overall safety profile of myelosuppressive chemotherapy regimens used to treat patients with ES‐SCLC.

K E Y W O R D S chemotherapy, myelopreservation, myeloprotection, myelosuppression, small cell lung cancer, trilaciclib

What's new?
Chemotherapy-induced myelosuppression is common in patients with small cell lung cancer and both the ensuing chemotherapy dose reductions/delays and the supportive care interventions are associated with risks. Trilaciclib has been found to decrease the incidence of chemotherapyinduced myelosuppression in three randomised, placebo-controlled clinical trials of patients with extensive-stage small cell lung cancer. Here, to further assess the myeloprotective effects of trilaciclib, an exploratory composite endpoint of major adverse haematological events was developed. Compared with placebo, trilaciclib reduced the cumulative incidence of major adverse haematological events and increased time to first event, indicating that trilaciclib improves the overall safety of myelosuppressive chemotherapy regimens.

| INTRODUCTION
Small cell lung cancer (SCLC) is an aggressive disease, with approximately 60% to 70% of patients having extensive-stage SCLC (ES-SCLC) at diagnosis. 1 Although platinum-based chemotherapy regimens have been the cornerstone of treatment for patients with ES-SCLC, 1,2 the addition of immunotherapy to chemotherapy has led to significant improvements in clinical outcomes and represents a new standard of care for patients with otherwise limited therapeutic options. [3][4][5] Most chemotherapeutic agents exert their effects by targeting highly proliferative cells; therefore, in addition to tumour cells, high-turnover normal tissues, such as hair follicles, mucosa and bone marrow, are also damaged. Chemotherapy-induced damage of haematopoietic stem and progenitor cells (HSPCs) in the bone marrow leads to myelosuppression, a common and sometimes life-threatening complication that most commonly manifests as neutropenia, anaemia and/or thrombocytopenia. 6 Indeed, haematological adverse events are among the most frequently reported toxicities associated with chemotherapy-based SCLC treatment regimens. 7 Chemotherapyinduced myelosuppression (CIM) can increase the risk of serious infections and bleeding and negatively impact patients' quality of life. [8][9][10][11] CIM is usually managed with chemotherapy dose reductions and/or delays, which may reduce the dose intensity of chemotherapy and potentially impair its therapeutic efficacy. [12][13][14] Moreover, supportive care interventions for CIM are generally used after the onset of symptoms and only address the deficiency of a single blood cell lineage. Such interventions also carry their own set of risks and limitations. For example, the use of erythropoiesis-stimulating agents (ESAs) to manage chemotherapy-induced anaemia is only effective in approximately 60% of patients and is associated with an increased risk of thromboembolic events. 15,16 Red blood cell (RBC) and platelet transfusions, used to treat anaemia and thrombocytopenia, respectively, are burdensome to patients and carry risks of transfusion reactions, thromboses, alloimmunisation and immunosuppression, and infection. 9,[17][18][19] Although primary prophylaxis with granulocyte colony-stimulating factors (G-CSFs) can reduce the risk of febrile neutropenia (FN) and improve overall survival, use of G-CSFs can lead to adverse effects such as bone pain. [20][21][22] In addition to these challenges, the use of supportive care interventions and the need for hospitalisations owing to CIM and its consequences can incur substantial healthcare resource use and financial costs, resulting in a considerable economic burden on patients, caregivers and healthcare systems. 10,23,24 A treatment that can proactively protect against CIM, thereby reducing the need for supportive care and hospitalisations, would therefore be particularly valuable.
Trilaciclib transiently arrests CDK4/6-dependent HSPCs in the G1 phase of the cell cycle during chemotherapy exposure, thereby protecting bone marrow function from chemotherapy-induced damage (myeloprotection or myelopreservation). 25,26 The myeloprotective effects of trilaciclib differ from the myelosuppressive effects of oral CDK4/6 inhibitors approved for the treatment of hormone receptor-positive breast cancer.
The chronic administration of oral CDK4/6 inhibitors results in myelosuppression owing to the sustained blockade of HSPC proliferation in the bone marrow 27 ; by contrast, IV infusions of trilaciclib are completed within 4 hours prior to chemotherapy on each day chemotherapy is administered, allowing for more control over HSPC cycle arrest. As a result, HSPCs are protected from damage during exposure to chemotherapy, but are able to resume proliferation afterwards. 26,28 Trilaciclib has been evaluated in three independent, global, multicentre, randomised, double-blind, placebo-controlled Phase 2 trials in which patients with ES-SCLC received trilaciclib or placebo prior to chemotherapy. [29][30][31] In each individual trial, and in a pooled analysis of data from all three trials, administering trilaciclib prior to chemotherapy led to a reduction in CIM across multiple lineages, along with a reduction in the use of supportive care measures, and improvements in quality of life. [29][30][31][32] Notably, the observed myeloprotective effects of trilaciclib were observed in the absence of any detrimental effects on antitumour efficacy. [29][30][31][32] Because trilaciclib affects multiple haematological and associated outcomes, a predefined, exploratory composite endpoint was used to assess the totality of benefit with trilaciclib across several clinically meaningful components of myeloprotection. Here, we describe the development of this composite endpoint of major adverse haematological events (MAHE) and present results from each of the three SCLC trials, along with results from the pooled analysis.

| Study design and participants
The three randomised Phase 2 trials were conducted at multiple sites across the United States and Europe. Full details of the study designs have been reported previously. [29][30][31] In study G1T28-05 (NCT03041311), patients were ineligible if they presented with symptomatic brain metastases requiring immediate treatment, and, for studies G1T28-05 and G1T28-02, if they had received prior systemic therapy for SCLC. For study G1T28-03, patients were excluded if they had a history of topotecan treatment, and they must have had disease progression during or after first-or second-line chemotherapy and been eligible to receive topotecan.
In each study, primary prophylaxis with G-CSF and use of ESAs was prohibited in cycle 1, although therapeutic G-CSF was allowed; after cycle 1, supportive care, including G-CSF and ESAs, was allowed as needed. RBC and platelet transfusions were allowed per investigator discretion throughout the entire treatment period.
Each study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines of the International Council for Harmonisation, and the protocols and all study-related materials were approved by the institutional review board or independent ethics committee of each participating site. All patients provided written informed consent.

| MAHE composite endpoint
The MAHE composite endpoint is comprised of five individual components that are clinically relevant but low-frequency consequences of CIM (Table 1). These include all-cause hospitalisations, all-cause che-

| Statistical analysis
The aim of the pooled analysis was to understand the effects of trilaciclib with greater statistical precision, particularly for low-frequency endpoints such as those related to RBCs and for specific patient populations with limited numbers in the individual trials. The pooled analysis set included all randomised patients in study G1T28-05, all randomised patients in the Phase 2 part of study G1T28-02 and patients receiving trilaciclib or placebo plus topotecan 1.5 mg/m 2 in the Phase 2 part of study G1T28-03. Breslow-Day testing was used to assess the statistical validity of data integration. 32 The treatment effect of trilaciclib compared to placebo was assessed using a negative binomial regression model, adjusting for duration of treatment in weeks or number of cycles. To adjust for potential variability between patients, the models included com- the pooled analysis and G1T28-05) and sensitivity to first-line treatment (sensitive vs resistant in G1T28-03) as these stratification factors have previously been shown in SCLC trials to predict treatment outcomes. [33][34][35] Study (G1T28-05, G1T28-02 and G1T28-03 in the pooled analysis) was also included as a fixed effect in order to adjust for potential variability between studies.

| Cumulative incidence of MAHE
In each of the three individual trials, aRRs of <1 indicated that, compared to placebo, administration of trilaciclib prior to chemotherapy resulted in a statistically significant reduction in the cumulative incidence of MAHE (Table 2). Cumulative incidence of MAHE was also clinically and statistically significantly lower for trilaciclib than for placebo in the pooled analysis (Table 2). Across all three individual trials the cumulative incidence of MAHE was higher in the placebo arm than in the trilaciclib arm by Week 3 and remained higher throughout the treatment period (up to Week 18 in study G1T28-05, Week 21 in G1T28-02 and Week 36 in G1T28-03; Figure 1).

| Cumulative incidence of individual MAHE components
The cumulative incidences of the separate MAHE components in the individual studies and in the pooled analysis are shown in Table 4.    Calculated as the total number of events divided by the total duration in weeks, or the total number of cycles with an event divided by the total number of cycles. b Calculated using the negative binomial method, adjusting for duration of treatment in weeks or number of cycles. c Stratification factors of ECOG PS (0 or 1 vs 2) and presence of brain metastases (yes vs no) included as fixed effects. d Stratification factor of ECOG PS (0 or 1 vs 2) included as a fixed effect. e Stratification factors of ECOG PS (0 or 1 vs 2) and sensitivity to first-line treatment included as fixed effects.
or the pooled analysis. This could be because this endpoint included all hospitalisations, including those due to reasons other than myelosuppression and is therefore not sensitive enough to register changes resulting from the administration of trilaciclib. Indeed, a separate ad hoc analysis of pooled safety data from these three SCLC trials showed that significantly fewer patients receiving trilaciclib were hospitalised specifically owing to CIM or sepsis compared to those receiving placebo, in line with the myeloprotective mechanism of action of trilaciclib. 32 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 Placebo 119 114 92 58 57 53 50 48 47 44 44 43 42 30 24 19 17 14 13 13 13 11 5 4 4 4 4 4 3 3 3 2 2 1 1 1 1 0 0 0 0  In addition to the impact of myelosuppression on patients' health and quality of life, the economic impact of haematological AEs and their management is also an important consideration for healthcare systems. It is notable, therefore, that a recent budget impact assessment for trilaciclib in decreasing the incidence of CIM in patients with ES-SCLC found that the incremental cost of trilaciclib was projected to be largely offset by a reduction in the costs of managing AEs related to myelosuppression. Consequently, the net financial impact was estimated to be a budgetary cost saving. 48 Other clinical contexts may also benefit from the assessment of treatment-induced myelosuppression. Thoracic radiotherapy is often administered as consolidation therapy to patients with ES-SCLC who respond to systemic chemotherapy. 49,50 In addition, the combination of immunotherapy plus radiotherapy is now being investigated in patients with ES-SCLC, on the basis of improvements shown in the non-SCLC setting. 51 Like cytotoxic chemotherapy, radiotherapy damages rapidly proliferating bone marrow cells, which can lead to haematological toxicity. 52 Indeed, in one retrospective analysis of patients with ES-SCLC, there was significantly more leukopenia in patients receiving chemotherapy and radiotherapy than in those receiving chemotherapy alone (53% vs 34%; P = .033). 53 Although the clinical effects of trilaciclib have not been investigated in patients receiving radiation therapy, it is feasible that the myeloprotective effects of trilaciclib when administered prior to chemotherapy may subsequently benefit patients treated in this setting by helping to protect HSPCs from damage before consolidation radiotherapy. The potential effects of trilaciclib in the context of radiotherapy therefore warrant further investigation.
Finally, although there is a theoretical concern that trilaciclib may antagonise the intended antitumor efficacy of chemotherapy in CDK4/6-dependent tumours, trilaciclib had no impact on the efficacy of chemotherapy in the SCLC studies, or in preclinical models of CDK4/6-dependent and -independent tumours. 25,[29][30][31][32]54 In each of the three randomised trials included in this analysis, progression-free survival and overall survival (OS) were similar between the trilaciclib and placebo arms. [29][30][31] Moreover, in study G1T28-05, median OS with trilaciclib prior to etoposide/carboplatin/atezolizumab (12.0 months) was similar to that observed with etoposide/carboplatin/atezolizumab in the pivotal IMPower133 study (12.3 months). 3,29 Strengths of this analysis include the predefined nature of the MAHE endpoint, and the consistent benefits observed with trilaciclib vs placebo prior to chemotherapy across multiple studies. The analysis was limited by the relatively small number of patients enrolled in these studies, as reflected by the fact that differences between trilaciclib and placebo were not consistently observed in individual studies for some of the MAHE components. Nonetheless, pooling of the datasets, which was supported statistically, allowed the effect of trilaciclib on these endpoints to be assessed with greater statistical accuracy. An additional limitation is that the impact of trilaciclib in combination with other commonly used second-or later-line chemotherapy options for SCLC (aside from topotecan) was not evaluated.
Overall, however, this analysis strengthens the conclusion that trilaciclib is well tolerated and acts as a myeloprotection agent, reducing CIM and its consequences across multiple haematopoietic cell lineages.
In conclusion, the robust improvements in the exploratory MAHE composite endpoint across the three studies and the pooled analysis further support the myeloprotective benefits of trilaciclib and its ability to improve the safety of chemotherapy regimens used to treat patients with ES-SCLC. Using the MAHE endpoint to assess clinical and health economic outcomes may help to ensure optimal patient care.
has participated in advisory boards for AstraZeneca, Boehringer Ingelheim and MSD, and has received support for attending meetings and/or travel from AstraZeneca and MSD. Dr Jie Xiao was an employee and shareowner of G1 Therapeutics, Inc. at the time of study and manuscript preparation. Drs Janet K. Horton and Rajesh K. Malik are employees and shareowners of G1 Therapeutics, Inc. Dr Janakiraman Subramanian has consulted for AstraZeneca, Boehringer Ingelheim, Eli Lilly, G1 Therapeutics, Jazz Pharma, Novartis, Pfizer and Takeda, has participated as a speaker for AstraZeneca, Boehringer Ingelheim and Eli Lilly and has provided writing support for AstraZeneca, Boehringer Ingelheim and Novartis.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.