同写意年度巨献!!7月25-26日,邀您相聚金鸡湖畔,与5000人一起为中国医药创新“同写意”!2025年5月20日,FDA的肿瘤药物咨询委员会(ODAC)对罗氏的Glofitamab(商品名Columvi)联合GemOx治疗复发或难治性弥漫性大B细胞淋巴瘤(DLBCL)的申请进行了评估。ODAC以8比1的投票结果反对将STARGLO试验结果应用于美国患者。下面来看下药物简介和听证会FDA立场(包括区域差异、人种选择(美国样本量少)、预后因素失衡、治疗实践差异等)。TONACEA01Glofitamab-gxbm简介Glofitamab-gxbm(商品名Columvi)是一种双特异性抗体药物,主要用于治疗复发或难治性弥漫性大B细胞淋巴瘤(DLBCL)和滤泡性淋巴瘤引起的大B细胞淋巴瘤(LBCL)的成年患者,特别是在接受两线或多线全身治疗后仍未缓解的患者。主要成分:格菲妥单抗(Glofitamab)是一种靶向 CD20 和 CD3 的双特异性抗体,通过与 T 细胞表面的 CD3 抗原和 B 细胞表面的 CD20 抗原结合,激活、扩增并重定向 T 细胞以杀伤肿瘤细胞。吉西他滨和奥沙利铂(GemOx)是传统的化疗药物组合,前者是一种抗代谢药物,后者是一种铂类化合物,两者协同作用可抑制癌细胞生长,并与格菲妥单抗发挥协同抗肿瘤效果。作用机制• 格菲妥单抗:作为 CD20xCD3 T 细胞接合剂,其 2:1 的独特结构包含两个 CD20 结合域和一个 CD3 结合域,能更有效地将 T 细胞拉近肿瘤细胞,增强免疫介导的细胞毒性,从而实现对恶性 B 细胞的选择性杀伤。• 吉西他滨和奥沙利铂:吉西他滨通过抑制 DNA 合成来阻止癌细胞增殖,奥沙利铂则通过形成 DNA 交联破坏其结构和功能,两者联用可产生协同作用,增强治疗效果。美国批准情况• 加速批准:2023 年 6 月 15 日,FDA加速批准glofitamab-gxbm(Columvi,基因泰克公司)用于治疗经二线或二线以上系统治疗后复发或难治的弥漫大B细胞淋巴瘤(DLBCL,NOS)或由滤泡淋巴瘤引起的大B细胞淋巴瘤(LBCL)(Food and Drug Administration granted accelerated approval to glofitamab-gxbm (Columvi, Genentech, Inc.) for relapsed or refractory diffuse large B-cell lymphoma, not otherwise specified (DLBCL, NOS) or large B-cell lymphoma (LBCL) arising from follicular lymphoma, after two or more lines of systemic therapy.)。• 补充生物制品许可申请:2024 年 12 月 9 日,罗氏宣布 FDA 已受理其关于 Glofit-GemOx 用于治疗复发或难治性 DLBCL 患者的补充生物制品许可申请(sBLA),这些患者需至少接受过一线既往治疗,且不适合进行自体干细胞移植(ASCT)。FDA 预计将在 2025 年 7 月 20 日之前就该申请作出决定。此次sBLA的提交主要基于STARGLO研究的结果。TONACEA02FDA监管近日会议进展与争议01会议背景2025年5月20日,美国食品药品监督管理局(FDA)召开了肿瘤药物顾问委员会(ODAC)会议,本次ODAC会议,是为了讨论STARGLO引起的关注。STARGLO(GO41944研究)被指定为确认性试验,以验证glofitamab-gxbm的预期临床疗效,并支持在早期治疗中与吉西他滨和奥沙利铂联用。本次会议的目的是就 STARGLO 研究结果的稳健性征求咨询委员会的意见,因为地区亚组之间多个终点的治疗效果不一致,以及 STARGLO 群体和总体试验结果是否适用于拟议的美国患者群体。02药物简介• 药物名称:glofitamab-gxbm• 申请人:Genentech, Inc• 适应症:用于治疗成人复发或难治性DLBCL患者,这些患者不适合进行自体干细胞移植(ASCT)。03研究背景• DLBCL概述:DLBCL是非霍奇金淋巴瘤(NHL)中最常见的亚型,占所有病例的30-40%。DLBCL是一种快速进展的恶性疾病,对患者的生命构成严重威胁。• 未满足的治疗需求:尽管有多种治疗方案,但仍有大量患者需要更有效的治疗选择,尤其是那些不适合ASCT或CAR-T细胞疗法的患者。04临床试验设计(图片来源于Oncologic Drugs Advisory Committee (ODAC) Meeting PDF)• STARGLO研究:一项全球多中心、随机、开放标签的Ⅲ期临床试验,旨在评估glofitamab与GemOx(吉西他滨和奥沙利铂)联合用药(Glofit-GemOx)与R-GemOx(利妥昔单抗+GemOx)相比,在不适合ASCT的R/R DLBCL患者中的疗效和安全性。• 研究设计:共纳入274名患者,按2:1随机分配至Glofit-GemOx组和R-GemOx组。主要终点为总生存期(OS),关键次要终点包括无进展生存期(PFS)、完全缓解率(CR)和CR持续时间(DOCR)。2021 年 2 月至 2023 年 3 月期间,全球 13 个国家的 62 个研究机构共招募了 274 名患者。该研究在所有国家同时启动,但各国研究机构的启动时间各不相同。美国是研究机构数量最多的国家(10 个研究机构),而各地区研究机构数量最多的是欧洲(49% 的研究机构,招募了 32% 的患者),其次是亚太地区(中国、台湾、韩国和澳大利亚)(35% 的研究机构,招募了 59% 的患者)和北美(16% 的研究机构,招募了 9% 的患者)。(图片来源于Oncologic Drugs Advisory Committee (ODAC) Meeting PDF)(图片来源于Oncologic Drugs Advisory Committee (ODAC) Meeting PDF)05研究结果• 总生存期(OS):Glofit-GemOx组的中位OS显著延长,为25.5个月(95% CI 18.3,NE),而对照组R-GemOx组为12.9个月(95% CI:7.9,18.5),风险比(HR)为0.62(95% CI: 0.43, 0.88)。• 无进展生存期(PFS):Glofit-GemOx组的中位PFS为13.8个月,而R-GemOx组为3.6个月,HR为0.40(95% CI: 0.28, 0.57)。• 完全缓解率(CR):Glofit-GemOx组的CR率为50.3%,而R-GemOx组为22.0%,差异显著(p<0.0001)。(图片来源于Oncologic Drugs Advisory Committee (ODAC) Meeting PDF)06安全性• 不良事件(AE):Glofit-GemOx组的AE发生率高于R-GemOx组,但大多数AE为低级别且可管理。• 细胞因子释放综合征(CRS):Glofit-GemOx组中的患者出现CRS,大多数为低级别,且在首次给药后逐渐减少。• 神经系统毒性:Glofit-GemOx组中的患者出现神经系统毒性,大多数为低级别。• 感染:Glofit-GemOx组中的患者出现感染,大多数为低级别。07FDA的立场以下解析FDA对人种和一些其他综合性立场,供参考,英文原文请参考后文参考链接2:1、FDA 对 STARGLO 结果是否适用于美国患者群体和美国医疗实践的立场:FDA 关于 STARGLO 结果对美国患者群体和美国医疗实践的适用性的立场: FDA 关于 “从肿瘤多区域临床开发计划中产生临床证据 ”的指导草案强调,FDA 在评估多区域肿瘤试验时的首要考虑因素是试验结果对美国预期使用人群和美国标准肿瘤医疗的适用性。这些研究产生的证据应来源于研究人群,以便根据美国患者的疾病或病情以及美国医疗实践标准来解释研究结果。在美国或与美国类似的地区注册了足够比例的人群的试验中,整体研究结果的普遍性问题较小。然而,在 STARGLO 试验中,由于美国患者入组人数有限(占意向治疗人群的 9%,其中 15 名患者接受了 Glofit-GemOx 治疗,10 名患者接受了 R-GemOx 治疗),且观察到不同地区亚组之间的治疗效果存在差异,因此将研究结果外推至美国患者人群和美国医疗实践的可能性较小。STARGLO 试验在患者入组分布上存在明显的不平衡,近一半的意向治疗(ITT)人群来自亚洲地区。具体来说,48%的 ITT 患者来自中国大陆、台湾和韩国,其中 29% 的患者来自中国大陆。方案规定至少有80名中国患者入组,“以确定glofitamab在GemOx基础上的疗效和安全性,为在中国的上市申请提供潜在支持”(方案v1)。该方案未对非亚洲国家提出此类要求,以确保患者分布更加均匀。在亚洲地区观察到的疗效结果显示,患者的存活率和疾病反应明显改善。相比之下,由澳大利亚、北美和欧洲患者组成的其余半数人群的疗效结果则明显不同。在亚洲地区入组患者身上观察到的疗效似乎对整个 STARGLO 试验结果产生了很大影响。值得注意的是,STARGLO 试验中这两个大的地区亚组之间存在如此显著的不一致性,令人担忧整体研究结果的稳健性。鉴于亚洲地区入组患者的结果可能对总体结果产生重大影响,FDA对这一人群的关键特征因素进行了全面评估。该评估旨在确定这些结果是否适用于美国患者群体,是否符合美国治疗复发/难治性弥漫大 B 细胞淋巴瘤 (R/R DLBCL) 的标准肿瘤治疗实践。STARGLO试验显示,基于内在和外在因素,在亚洲地区入组的患者群体与经过一线治疗后复发/难治性弥漫大B细胞淋巴瘤(R/R DLBCL)并被认为不符合自体干细胞移植条件的美国可比患者群体之间存在显著差异。英文版原文(上下滑动查看更多)The FDA’s Position on Applicability of STARGLO Results to a U.S. Patient Population and U.S. Medical Practice: The FDA's draft guidance on "Generating Clinical Evidence from Oncology Multiregional Clinical Development Programs" emphasizes that the primary consideration for the FDA when assessing multiregional oncology trials is the applicability of results to the intended use population in the United States and to U.S. standard oncological medical care. Evidence generated from these studies should be derived from study populations that allow for interpretation of results within the context of U.S. patients with the disease or condition and U.S. medical practice standards. In trials where an adequate proportion of the population is enrolled in the United States or regions similar to the United States, the generalizability of the overall study results is less concerning. However, in the case of the STARGLO trial, given the limited U.S. patient enrollment (9% of the Intent-to-Treat population, with 15 patients treated on the Glofit-GemOx arm and 10 patients on the R-GemOx arm) and the observed differential treatment effect between regional subgroups, the extrapolation of study results to the U.S. patient population and U.S. medical practice is less certain. The STARGLO trial presents a notable imbalance in its distribution of patient enrollment, with nearly half of the Intent-to-Treat (ITT) population recruited from the Asian Region. Specifically, 48% of the ITT population consisted of patients enrolled in China, Taiwan, and Korea with 29% of patients enrolling from China. As previously noted, the protocol specified that at least 80 patients from China be enrolled “in order to characterize the efficacy and safety profile of glofitamab in addition to GemOx to potentially support a marketing application in China” (protocol v1). The protocol did not include such requirements for the Non-Asian countries to ensure a more evenly distributed population. Efficacy outcomes observed in patients from the Asian Region demonstrated marked improvements in survival and disease response. In contrast, the remaining half of the population, comprising patients enrolled in Australia, North America, and Europe, showed markedly different efficacy outcomes. The efficacy measures observed in the patients enrolled from the Asian Region seemed to have exerted a strong influence on the overall STARGLO trial results. Notably, the presence of such significant inconsistencies between these two large regional subgroups in the STARGLO trial raises concerns about the robustness of the overall study results. Given the potential substantial impact of the results from patients enrolled in the Asian Region on the overall outcomes, the Agency conducted a thorough assessment of key factors characterizing this population. This evaluation aimed to determine whether these results could be applicable to the U.S. patient population and align with U.S. standard oncological care practices in the treatment of relapsed/refractory diffuse large B-cell lymphoma (R/R DLBCL). The STARGLO Trial revealed significant differences between the patient population enrolled in the Asian Region based on intrinsic and extrinsic factors to what would be expected in a comparable U.S. patient population with relapsed/refractory diffuse large B-cell lymphoma (R/R DLBCL) following one prior line of therapy and considered ineligible for autologous stem cell transplantation.2、STARGLO 试验的设计和实施对研究结果的解释有多方面的影响。• 中国亚群所需的最低患者人数:该试验被设计为一项多地区试验,在北美、欧洲、澳大利亚和亚洲进行招募。FDA 指出,试验总人数在各地区之间并不平衡,亚洲国家的入组患者人数较多,几乎占了总人数的一半。值得注意的是,试验方案规定了中国亚群的最低要求。具体来说,方案规定,“在中国大陆、香港和台湾经中国国家药品监督管理局(NMPA)认可的医疗机构”招募患者,以确保中国亚人群中的患者总数达到约 80 人。“提出这一要求的理由是为了确定除 GemOx 外,glofitamab 的疗效和安全性特征,以便为在中国的上市申请提供潜在支持"(方案 v1)。如果在方案的全球招募阶段未能招募到这一数量的患者,可在延长的中国招募阶段招募更多患者。由于最终从中国招募了 80 名患者,因此无需扩大招募范围。为确保平等代表性,方案中未对其他地区提出此类要求。• 更新分析的原因(2024 年 2 月 16 日的 CCOD):FDA 指出,随后进行的更新分析不仅是为了抵消 COVID-19 大流行的潜在影响。申请人最初向药监局提交的初步数据是基于 73% 的估计 OS 事件进行的中期分析。当时,观察到基于种族和地区的疗效结果不一致,FDA对研究结果的稳健性表示担忧。申请人认为,非亚洲国家的随访时间较短(中位数为 8.7 个月),而亚洲国家的随访时间较长(中位数为 14.6 个月),这可能是导致结果不一致的原因,并认为进一步的随访将显示出更一致的治疗效果。申请人建议增加一年的随访时间,以帮助解决观察到的不一致问题。• R-GemOx 治疗侵袭性淋巴瘤的时间表:每三周给药一次(Q3W)与每两周给药一次(Q2W)的对比研究对照组是每三周给药一次(Q3W)的 R-GemOx。虽然R-GemOx可以每三周给药一次,但由于R/R DBCL的侵袭性,大多数研究和许多机构的给药频率更高,为每两周一次。根据文献(Gnaoui,2007 年)(Mounier,2013 年)(Crump,2016 年)(Corazzelli,2015 年),临床实践指南推荐每两周使用一次 R-GemOx。• R-GemOx 在美国患者群体中的使用情况 R-GemOx 在美国并不常用,根据最近的 RWD 和使用报告,其使用率为 2-8% (Yamshon,2025 年)(申请人回复,BLA 761309-S-001,SD 202,10-25-2025 年收到)。在使用时,R-GemOx 通常用于因年龄或合并症而不适合强化治疗的患者(Cazelles,2021 年)(Gnaoui,2007 年)。据观察,R-Gem-Ox 在美国的使用情况令人质疑:R-Gem-Ox 是否适合作为美国患者的对照组。• 美国入组滞后正如申请人所指出的,亚洲地区的患者在研究初期就已入组,而欧洲和北美地区的患者则在研究中后期才入组。申请人认为,欧洲和美国患者入组较晚的部分原因是 COVID-19 大流行的影响。然而,申请人也指出,在美国有更多的研究机构拒绝参与 STARGLO。申请人在美国接触了 50 多个研究机构,其中 11 个研究机构已启动,10 个研究机构招募了患者(申请人提交,BLA 761309-S001,模块 2,临床概述,SD 194,2024 年 9 月 20 日收到)。虽然申请人未提及,但拒绝参与 STARGLO 的研究机构比例较高,这可能反映了研究者缺乏兴趣。• 2:1 随机化:一般来说,采用 1:1 的随机比例时随机化效率最高。虽然肿瘤学中使用 2:1 随机化,但在样本量较小的试验中,这种不平衡的比例会减少对照组的信息量,从而限制比较分析的进行。此外,2:1 随机化可能会带来偏差,因为试验组可能会被认为是更好的治疗方法,这在开放标签试验中会影响治疗组之间的依从性和行为。• Obinutuzumab预处理为降低CRS风险,在开始使用glofitamab之前,要进行Obinutuzumab预处理。Obinutuzumab是第二代CD20单克隆抗体,在多种非霍奇金淋巴瘤治疗中已被证明优于利妥昔单抗(Townsend,2023年)(Goede,2015年)。在 Glofit-GemOx 治疗组中使用奥比妥珠单抗后,由于奥比妥珠单抗与利妥昔单抗之间的差异,可能会产生不同的影响。英文版原文(上下滑动查看更多)The FDA’s Position There are several aspects of the STARGLO trial design and conduct that impact the interpretation of the study results. • Minimum Number of Patients Required for China Subpopulation: The trial was designed as a multiregional trial, enrolling in North America, Europe, Australia, and Asia. The FDA notes that the total trial population was not balanced by region and had an large representation of patients enrolled in Asian countries, which accounted for almost half of the population. It is important to note that the protocol stipulated a minimum requirement for a China subpopulation. Specifically, the protocol states that patients were to be enrolled “at sites in China (including the mainland, Hong Kong, and Taiwan regions)by the China’s National Medical Products Administration (NMPA) to ensure a total of up to approximately 80 patients in a China subpopulation.” The rationale for this requirement was “to characterize the efficacy and safety profile of glofitamab in addition to GemOx to potentially support a marketing application in China” (protocol v1).” If this number of patients was not obtained during the global enrollment phase of the protocol, additional patients could be enrolled in an extended China enrollment phase. Extended enrollment was not needed, as ultimately 80 patients were enrolled from China. No such requirements were included in the protocol for other regions to ensure equal representation. • Reasons for the Updated Analysis (CCOD of 16 February 2024): The FDA notes that a subsequent updated analysis was not only conducted to offset the potential impact of the COVID-19 pandemic. The Applicant initially presented topline data to the Agency based on the interim analysis conducted at 73% of the estimated OS events. At that time, the inconsistent efficacy results based on race and region were observed and the Agency conveyed their concern about the robustness of the study results. The Applicant posited that the shorter follow-up in Non-Asian countries (median 8.7 months) compared to the longer follow-up in the Asian countries (median 14.6 months) was the likely reason for the inconsistent results and posited that further follow-up would reveal a more consistent treatment effect.The additional year of follow-up was proposed by the Applicant to help address the inconsistencies observed.• Schedule of R-GemOx in Aggressive Lymphoma: Every Three Week Dosing (Q3W) versus Every Two Week Dosing (Q2W) The comparator arm is R-GemOx administered every three weeks (Q3W). While R-GemOx can be administered once every three weeks, most studies and many institutions administer the regimen more frequently, once every two weeks, due to the aggressive nature of R/R DBCL. The use of R-GemOx dosing every two weeks is recommended in clinical practice guidelines based on the literature (Gnaoui, 2007) (Mounier, 2013) (Crump, 2016) (Corazzelli, 2015)• Utilization of R-GemOx in a U.S. Patient Population R-GemOx is not a regimen commonly used in the U.S., with reported use of 2-8% per recent RWD and utilization reports (Yamshon, 2025) (Applicant response, BLA 761309-S-001, SD 202 received 10-25-2025). When used, R-GemOx is generally reserved for patients who are not suitable for intensive therapy due to age or comorbidities (Cazelles, 2021) (Gnaoui, 2007). The observed utilization of R-Gem-Ox in the U.S. calls into question the suitability of R-Gem-Ox as a control arm for the U.S. patient population.• Delayed Enrollment in the U.S. As noted by the Applicant, the Asian regions enrolled patients early in the study, whereas Europe and North America accrued through the mid- to late- phases of the study. The Applicant attributes the late enrollment of patients from Europe and the U.S. in part to the effects of the COVID-19 pandemic. However, the Applicant also notes that in the U.S. a higher number of sites declined participation in STARGLO. Over 50 sites were approached in the U.S. with 11 sites activated and 10 sites that enrolled patients (Applicant Submission, BLA 761309-S001, Module 2, Clinical Overview, SD 194 received 9-20-2024). Although not mentioned by the Applicant, the high rate of sites declining participation in STARGLO may reflect a lack of investigator interest.• 2:1 Randomization: In general, randomization is most efficient when utilizing a 1:1 randomization ratio. While 2:1 randomization is used in oncology, in trials with small sample sizes, this imbalanced ratio reduces the amount of information on the control arm, limiting comparative analyses. Additionally, a 2:1 randomization may introduce bias as the experimental arm may be considered a better treatment, which in an open-label trial can impact adherence and conduct between treatment arms.• Obinutuzumab Pre-treatment To mitigate the risk of CRS, obinutuzumab pre-treatment is administered prior to the initiation of glofitamab. Obinutuzumab is a 2nd-generation CD20 monoclonal antibody that has been shown to be superior to rituximab in several non-Hodgkin lymphoma settings (Townsend, 2023) (Goede, 2015). With the use of obinutuzumab in the Glofit-GemOx arm, there may be a differential impact given the differences between obinutuzumab and rituximab.3、FDA 对 STARGLO 患者群体的立场FDA 对 ITT 群体的组成在地区代表性以及对美国患者群体的普适性方面表示担忧。STARGLO 研究是一项多地区试验,招募了来自多个国家和地区的患者。不过,从这些地区的人口分布来看,亚洲国家(中国、台湾和韩国)的入组患者占很大比例。来自北美、欧洲和澳大利亚的患者合计占 ITT 总人数的 52%,而来自亚洲国家的患者则占总人数的 48%(FDA 表 1)。如前所述,方案规定至少要有80名中国患者入组,“以确定在GemOx基础上使用glofitamab的疗效和安全性,从而为在中国的上市申请提供潜在支持”(方案vl)。该方案并未对非亚洲国家提出此类要求,以确保研究对象的分布更加均匀。地区未作为分层因素,也未预先指定作为单独的动力分析。然而,申请人指定为 “北美”、“欧洲 ”和 “世界其他地区 ”的地区亚组被纳入了预先指定的亚组分析。FDA 不同意 STARGLO 群体反映了美国 R/R DLBCL 患者预期群体的临床特征。在美国注册的患者人数有限,而来自亚洲地区的患者人数众多。虽然大多数DLBCL病例确实发生在自称为白人的患者中,但美国DLBCL患者的总体人群并不单一。美国 SEER 2016-2020 年的数据估计,2024 年的 DLBCL 病例中,87% 的患者为白人,7.5% 为黑人,4.8% 为亚裔或太平洋岛民,0.5% 为美洲印第安人/阿拉斯加原住民(SEER 癌症统计;美国人口普查)。就种族而言,估计有 11% 的病例发生在西班牙裔患者身上,而大多数病例(89%)估计发生在非西班牙裔患者身上(SEER 癌症统计数据;美国人口普查)。就年龄中位数和性别分布而言,整个患者群体的人口统计学特征与美国患者群体相似。具体来说,美国患者的中位年龄为 67 岁,男性略占多数(56%)(SEER 癌症统计;美国人口普查)。不过,这些人群的其他患者特征似乎有所不同。目标患者人群是那些被认为不符合移植条件的患者。这些患者通常年龄较大,或因表现状况、合并症或对先前治疗反应不佳而不适合接受移植。如 FDA 表 2 所示,35% 的患者因拒绝而被认为不符合移植条件。正如在 “不符合移植条件的原因 ”一文中进一步讨论的所描述,拒绝移植的原因并不明确,这可能反映出患者本来可以接受强化化疗和自体干细胞移植。在地区分组分析中,这一差异以及其他差异似乎被放大了。疾病生物学方面的潜在差异也存在不确定性,特别是不同地区之间的起源细胞,可能会对患者的病程和治疗反应产生影响。这些考虑因素将在《起源细胞差异》中进一步讨论。英文版原文(上下滑动查看更多)The FDA's Position on STARGLO's Patient PopulationThe FDA has concerns with the composition of the ITT population in terms of regional representation as well as generalizability to a U.S. patient population.The STARGLO study is a multi regional t rial as it enrolled patients from multiple countr ies and regions. However, distr ibution of the population across those regions demonstrates an large representation of patients enrolled in Asian countries (China, Taiwan, and Korea). While the combined grouping of patients from North America, Europe, and Australia resulted in 52% of the ITT population, 48% of the total population was enrolled solely from Asian count ries (FDA Table 1). As noted, the protocol specified t hat at least 80 patients from China be enrolled "in order to characterize the efficacy and safety profi le of glofitamab in addition to GemOx to potentially support a marketing application in China" (protocol vl). The protocol did not include such requirements for the Non-Asian count ries to ensure a more evenly distr ibuted population. Region was not used as a stratification factor nor pre-specified as a separate powered analysis. However, regional subgroups, which the Applicant designated as "North America," "Europe," and "Rest of World," were included in the pre-specified subgroup analyses (Figure 8 Subgroup Analysis of OS by Key Baseline Risk Factors (STARGLO Updated Analysis, ITT Population).FDA does not agree that the STARGLO population mirrors the cl inical profile of the intended population of U.S. patients with R/R DLBCL. There was a limited number of patients enrolled in the U.S. and an large representation of the patients from Asian regions. While the majority of DLBCL cases do occur in patients identifying as White, the overall U.S. population with DLBCL is not homogenous. The US SEER data from 2016-2020 estimates the number of cases of DLBCL in 2024 to occur in 87% of patients identifying as White, 7.5% as Black, 4.8% as Asian or Pacific Islander, and 0.5% as American Indian/Alaska Native (SEER Cancer Statistics; US Census). In terms of ethnicity, 11% of cases were estimated to occur in Hispanic patients, while the majority of cases (89%) were estimated to occur in patients who are not Hispanic (SEER Cancer Statistics; US Census). The demographic profile of the overall patient population exhibits similarities to the U.S. patient population in terms of median age and gender distribution. Specifically, patients in the U.S. have a reported median age of 67 years and a slight male predominance (56%) (SEER Cancer Statistics; US Census). However, other patient characteristics appear to differ between these populations. The intended patient population are those who are considered ineligible for transplant. These patients are typically older or unfit due performance status, comorbidity, or insufficient response to prior treatment. As FDA Table 2 shows, 35% of the population was considered ineligible due to patient refusal. As discussed further in Reasons for Transplant Ineligibility, the reason for the refusal is not clear and could reflect a patient that would otherwise be a candidate for intensive chemotherapy and autologous stem cell transplantation. This difference, as well as others, appear magnified in the regional subgroup analyses. There is also uncertainty regarding potential differences in disease biology, specifically cell of origin between the regions that can play role in patient disease course and response to treatment. These considerations are further discussed in Cell of Origin Differences.FDA 对委员会要点考量的立场: 多地区试验有许多优点,例如,通过产生支持在多个地区使用一种药物的证据,可以快速累积和更有效地进行临床开发。多区域临床试验的另一个好处是,有可能发现预测区域治疗差异的因素。不过,这种地区差异通常会在试验实施前确定,并采取措施确保试验仍能产生统计学上可靠、临床上有意义的结果,且适用于预期使用人群。根据 ICH E17 指南,多地区临床试验应 “假设治疗效果适用于整个目标人群,特别是试验所包括的地区”(ICH E17)。该指南还规定,可能存在的内在和/或外在因素会对不同地区患者的治疗反应产生不同影响。因此,在规划多区域试验时应考虑这些因素。研究应确保有效评估治疗的一致性。ICH E17 指南指出,如果预期治疗效果存在重大差异,仍可进行多地区试验,但可能需要排除某些地区或某一地区内的特定亚组。STARGLO 试验表现出明显的地区效应,ITT 组中亚洲地区参与者的比例过高可能会放大这种效应。该研究的特点是美国的入组人数有限,而且在总生存期、无进展生存期和应答率/完全应答结果方面存在显著的地区差异。多种患者特异性因素被认为是造成这些地区差异的潜在原因,包括患者年龄、不符合移植条件的原因、接受过的既往疗法类型、原细胞特征以及因疾病进展而停药的比例。此外,试验实施和分析的一些方面可能进一步影响了亚洲和非亚洲地区之间观察到的不同结果,包括疾病评估的时间、暴露和治疗中止的差异以及潜在的偏倚。这些观察到的差异及其潜在的诱因引起了人们对试验结果的稳健性及其是否适用于预期的美国患者人群的极大担忧。因此,在解释和推广 STARGLO 试验结果时应仔细考虑这些发现。英文版原文(上下滑动查看更多)The FDA’s Position:Multiregional trials have numerous advantages, such as allowing rapid accrual and more efficient clinical development by generating evidence to support use of a drug in multiple regions. One additional benefit of multiregional clinical trials is the potential to identify factors that may predict regional treatment differences. However, typically this regional difference is identified prior to implementation of the trial and measures are taken to ensure that the trial will still produce statistically robust and clinically meaningful results that are applicable to the intended use population. Per the ICH E17 guidance, multiregional clinical trials should be “planned under the assumption that the treatment effect applies to the entire target population, particularly to the regions included in the trial.” (ICH E17). This guidance also stipulates that intrinsic and/or extrinsic factors could be present that may have an impact on patients’ response to therapies differently across regions. Therefore, these factors should be considered with planning multiregional trials. Studies should ensure an effective assessment of the consistency of a treatment. In cases where a major difference is expected in treatment effects, the ICH E17 guidance notes that multiregional trials can still be conducted but may require exclusion of some regions or a defined subgroup within a region.The STARGLO trial exhibited a notable regional effect, potentially amplified by the disproportionate representation of Asian regional participants in the ITT population. The study was characterized by limited enrollment from the United States and significant regional disparities in overall survival, progression-free survival, and response rate/complete response outcomes. Multiple patient-specific factors were identified as potential contributors to these regional differences, including patient age, reasons for transplant ineligibility, exposure to types of prior therapies, cell of origin characterization, and discontinuation rate due to progressive disease. Additionally, aspects of trial conduct and analyses may have further influenced the divergent results observed between Asian and Non-Asian regions, including the timing of disease assessments, differences in exposure and treatment discontinuation, and potential bias. These observed differences and their potential contributing factors raise substantial concerns regarding both the robustness of the trial results and their applicability to the intended United States patient population. Consequently, these findings warrant careful consideration in the interpretation and generalizability of the STARGLO trial results.08关键问题FDA对STARGLO研究结果的稳健性和适用性表示担忧,尤其是在区域差异显著的情况下。FDA强调,尽管研究结果在亚洲地区显示出积极的疗效,但在非亚洲地区,包括美国,这些结果的适用性尚需进一步验证。FDA认为,尽管Glofit-GemOx在亚洲地区显示出显著的疗效,但由于区域差异和样本量不平衡,这些结果可能不适用于美国患者群体。区域差异:研究结果显示,亚洲地区的患者在OS、PFS、CR率等方面表现出显著的治疗效果,而非亚洲地区的患者则未观察到类似的效果。这种区域差异引发了对研究结果稳健性和适用性的担忧。美国患者适用性:由于美国患者在研究中的参与度较低(仅占9%),FDA对将这些结果推广至美国患者群体持谨慎态度。— 总结 — FDA对STARGLO研究结果的稳健性和适用性表示担忧,主要问题包括以下几点:研究结果稳健性问题:• 区域差异显著:STARGLO试验在全球范围内开展,但结果显示亚洲患者亚组的治疗效果显著优于非亚洲患者亚组。在亚洲地区,Glofitamab联合GemOx治疗组的死亡风险降低了61%,而非亚洲地区,治疗组的死亡风险反而比对照组高出6%。这种区域间的疗效差异表明,研究结果可能受到特定地区患者特征的影响,导致整体疗效数据的稳健性受到质疑。• 美国患者样本量小:试验中美国患者的入组比例仅为9%,共25名患者。样本量过小使得对美国患者群体的疗效评估缺乏足够的统计学支持,难以准确判断该治疗方案在美国患者中的真实效果。• 预后因素失衡:在北美患者中,Glofitamab组与利妥昔单抗组相比,原发性难治性疾病患者比例更高(80.0% vs 40.0%),国际预后指数(IPI)评分更高(66.6% vs 50.0%),疾病分期更晚(80.0% vs 66.7%)。这种预后因素的失衡使得在小样本亚组中随机化平衡预后因素的效果大打折扣,进一步增加了对研究结果稳健性的担忧。研究结果适用性问题:• 患者特征差异:亚洲患者亚组与非亚洲患者亚组在年龄、种族、疾病负担等方面存在显著差异。例如,亚洲患者的中位年龄为62岁,而非亚洲患者为71岁;非亚洲患者中大多数为白人,黑人患者仅占2%。这些差异可能影响治疗效果的外推性,使得基于亚洲患者数据得出的结论难以直接适用于美国患者。• 治疗实践差异:不同地区的医疗实践和治疗标准也存在差异。例如,美国的新抗淋巴瘤治疗标准与其他地区不同,这可能影响对治疗效果的评估。此外,亚洲地区患者拒绝移植的比例远高于非亚洲地区(65% vs 7%),这也限制了对患者适应症的全面分析。申请人应注意的问题:• 加强美国患者入组:在未来的临床试验中,申请人应确保美国患者的入组比例足够高,以提供更具代表性的数据,从而更好地评估治疗方案在美国患者中的疗效和安全性。• 关注区域差异:在设计和分析多区域临床试验时,申请人需要更加关注不同地区的患者特征和治疗效果的差异,并采取措施减少这些差异对研究结果的影响。• 平衡预后因素:在试验设计中,应更加注重随机化过程的质量,确保各治疗组之间的预后因素平衡,以提高研究结果的可靠性和稳健性。• 考虑医疗实践差异:申请人应充分考虑不同地区的医疗实践和治疗标准差异,确保试验结果能够反映真实世界中的临床实践情况。药物上市不易,以史为鉴。参考来源:1.https://www.gene.com/media/press-releases/15061/2025-05-20/genentech-provides-update-on-fda-advisor2.chrome extension://efaidnbmnnnibpcajpcglclefindmkaj/https://www.fda.gov/media/186525/download3. https://www.onclive.com/view/fda-odac-votes-against-the-applicability-of-starglo-data-for-glofitamab-plus-chemo-for-u-s-patients-with-r-r-dlbcl