自21世纪初以来,医药产品中的活性药物成分(APIs)进入环境并对生态系统造成不利影响的可能性受到了越来越多的监管审查。活性药物化合物被设计成在人体中引起特定的生物效应——通常是低剂量的——但如果未使用或排出的原料药进入城市废水流,并最终进入自然水体,也有可能对野生动物造成意想不到的不利影响。尽管关于环境中的药物的研究在21世纪激增,但专门针对这一主题的全球法规数量仍然有限。
人们早已认识到,制药废水会将原料药和其他加工化学品作为点源排放到受纳水体中。因此,地方辖区历来对这些排放进行单独监管,以降低对附近生态系统的风险。本章将讨论单独的环境法规,这些法规是针对制药产品的使用可能导致更广泛的社会排放和更广泛的环境风险而制定的。
这里将详细介绍美国的环境评估 (environmental assessment, EA) 和欧盟的环境风险评估 (environmental risk assessment, ERA) 法规要求,这些要求与产品的使用有关,是产品注册档案的强制性内容,已在全球范围内采用。这些法规的重点是评估小分子活性成分及其主要代谢物或生物制品,但不包括辅料。
针对原料药的监管 EA/ERAs 主要通过比较两个浓度值来评估风险:首先是对原料药在水、土壤或沉积物中因消费者使用而非生产而产生的浓度的估计,其次是对水、土壤或沉积物中可能对生活在这些栖息地的生物安全的最高浓度的估计。
然后对这两个浓度进行比较,如果原料药的环境浓度高于预期的安全浓度,则不能排除环境风险。针对特殊风险的监管性 EA/ERAs 还包括额外的检查,包括在食物链中累积导致捕食者二次中毒、破坏废水处理过程以及本章稍后详述的其他风险。
进行 EA/ERAs 的典型策略是利用高估风险的简化假设(simplifying assumptions)进行初步或筛选级评估。如果使用这些简化假设预测不出风险,那么监管机构就可以放心,可能的环境风险很低,评估也就结束了。
如果不能排除风险,则进行额外的研究,提供推翻简化假设的数据,从而对风险做出更切合实际的估计。这一过程根据需要反复进行,以得出风险结论,同时避免浪费资源进行不必要的测试。目前全球有两种主要的原料药环境影响测试战略,一种在美国,另一种在欧盟,这两种战略都采用这种分层方法进行数据收集和评估。
1
全球监管格局
制定规范性的、针对药品的环境风险评估 (environmental risk assessment, ERA) 指导仅限于欧盟和美国。在这些司法管辖区之外,适用于人类制药产品 ERA 的监管框架要么是定性的,要么不是专门针对制药的,要么根本不存在。加拿大、中国和澳大利亚等国缺乏针对药品的监管框架,但通过适用于其他产品类别的一般化学评估计划来评估药品的环境风险。
许多国家(包括南美和非洲地区)目前都不要求对人类药物进行环境风险评估。随着对环境中药物存在的监管力度加大,预计将制定新的监管框架,并加强现有的要求。
- 加拿大: 加拿大卫生部目前通过一般的新物质计划对药品进行评估,但正在制定修正案,以便根据《食品和药品法》建立一个专门针对药物活性成分的单独监管框架。
- 欧盟: 2018 年,欧盟对ERA 指南进行了修订,尽管目前只是草案形式,但预计将取代撰写本报告时仍在使用的 2006 年版本。
- 中国: 2022 年 5 月,中国国务院发布了一项行动计划,其中包括加强新化学品环境风险管理框架的目标。
- 日本:在日本,药品在一般化学品计划内进行评估,直到 2016 年发布了药品专用指南。由于该药品专用指南不是规定性的,因此预计提交的材料将遵循美国或欧盟的指南。
在没有具体药品指南的国家,当要求进行 EA/ERA 时,一般认为可以采用欧盟使用的框 架,因为它是全球最完善、最规范、最全面的药品法规。
在美国和欧盟,如果申请人假定药物的所有残留物都具有与原料药相同的环境归宿和影响特征,则可以忽略原料药的代谢物。这被认为是一种保护性方法,因为预计代谢物的毒性不会超过母体原料药。如果申请人有数据表明,主要代谢物具有与 EA/ERA 结果相关的显著不同特性,他们可以通过对代谢物进行全面的 EA/ERA 来加以说明。但这种做法很少采用,因为它需要大量资源,需要对代谢物分子进行全套的环境归宿和效应测试。
01
美国的小分子药物环境评估
在美国,药品审批申请必须提交给食品与药物管理局 (FDA) 的药品评估与研究中心 (CDER)。提交的申请必须符合分类排除的要求,表明该行动(即药品审批)不会对人类环境质量产生重大影响,或者为拟议行动提供完整的环境评估。Figure 10-1 提供了一个流程图,用于确定一项行动是否符合分类排除的条件或是否需要完整的环境影响评估。至少属于以下一种情况的行动通常符合分类排除的条件:
1.研究性新药申请 (IND)
2.将导致生产环境中自然存在的活性分子的变化,但需注意该变化不会显著改变其在环境中的浓度或分布;
3.不会增加活性分子使用量的行动(例如,与以前批准的相同适应症相比,剂量减少或持续时间缩短)
4.将增加活性分子的使用,但导致水中的水生环境引入浓度(EICaquatic)低于 1 μg/L(ppb)的行动。
对于上述列表中的第4项,FDA指导文件中规定的计算方法将EICaquatic触发浓度与上市批准后5年内原料药的最大年销售量联系起来。这个计算假设API在全美统一使用,并在排放到环境之前在废水中稀释。计算的结果是,在美国,任何销售数量低于每年约45,000公斤的原料药都符合分类排除的条件,并且不需要EA。每年的准确计算值会有所不同,视乎最近污水处理厂呈报的污水量而定。根据FDA指南的这一部分,大多数原料药符合分类排除。
在确定类别排除之前,必须考虑可能表明药物可能对人类环境造成严重危害的特殊情况。如果发现特殊情况,则必须完成完整的EA以获得上市批准。与特殊情况有关的行动的例子包括,药物的活性成分已知是有毒的,或在预期浓度下可能对暴露环境产生持久影响,药物可预见地对受保护物种或受保护物种的关键栖息地造成伤害,药物的活性成分来自自然物种并威胁到该物种的继续存在(即过度捕捞)。
对于上述清单中的第 4 项,FDA指导文件中规定的计算方法是,将 EICaquatic 触发浓度与原料药在获准上市后 5 年内的预期最大年销售量联系起来。该计算方法假定该原料药在全美范围内统一使用,并在排放到环境中之前在废水中被稀释。计算的结果是,任何原料药在美国的年销售量低于约 45,000 千克时,均符合分类豁免的条件,无需进行环境影响评估。具体的计算值每年都不同,这取决于最近报告的每年进入美国公有处理厂(POTWs;废水处理厂)的废水量。根据 FDA 准则的这一部分,大多数原料药都符合分类排除的条件。
在做出分类排除的决定之前,必须考虑可能表明药物可能对人类环境造成严重危害的特殊情况。如果特殊情况得到确认,则必须完成全面的环境评估,拟议的药物才能获得上市批准。与特殊情况相关的行动包括:活性分子已知有毒或在预期浓度下可能对暴露环境产生持久影响的药物;可预见会对受保护物种或受保护物种的重要栖息地造成伤害的药物;活性分子来自天然物种并威胁到该物种继续生存的药物(即过度采伐)。
如果拟批准的药物不符合分类排除的标准,则必须提交完整的环境影响评估。EA 的基本要求是提供足够的信息,以确定拟议的行动是否会对人类环境产生重大影响。与下文所述的欧盟环境影响评估程序相比,FDA的环境影响评估程序规定较少。如果申请人认为需要进行 EA,则应联系 CDER,以确定在进行 EA 之前应遵循的适当测试和评估途径。FDA 要求的一般方法如下。
计算药物导致的活性分子在相关环境区域(即水、土壤和空气)中的预测暴露浓度。然后,针对每个环境区划,将预期暴露浓度与可能对暴露生物产生重大不利影响的浓度进行比较。
作为评估的一部分,要对占初始剂量 10%以上的活性分子和代谢物或降解物的环境归宿(environmental fate)进行评估。报告环境归宿评估的相关物理和化学特性,包括水溶性、解离常数、辛醇/水分配系数(logKOW)、蒸气压或亨利定律常数。如果活性分子的特性表明其有在环境中大量吸附(附着)固体的趋势,则还应包括有机碳-水分配系数(logKOC)的特征。根据环境归宿评估结果,确定活性分子的相关环境区划。
针对每个相关环境区划计算 EIC(environmental introduction concentration)。除稀释外,计算中还可考虑其他因素,包括新陈代谢、废物处理过程中的清除以及环境损耗(如水解、光解、生物降解)。水处理后,附着在生物固体上的物质可能会通过生物固体土地应用进入陆地环境。
对于这些 API,EICterrestrial 的计算基于分配特性、耗竭机制的影响以及生物固体的土地应用率。医药产品通常不需要大气环境 EIC,因为活性分子通常不会释放到空气中。在相关情况下,例如医用气体,可以计算大气环境的 EIC。除 EIC 外,还要计算每个相关环境区划的预期环境浓度 (EEC)。EEC 考虑了影响活性分子最初进入环境后的环境浓度的因素,包括额外的消耗、稀释、吸附和生物累积。
为确定相关的效应浓度,可通过申办方协助的测试、同行评审的文献来源或两者收集实验得出的毒性数据。计算机建模的结果也可用于辅助。FDA 指南建议采用分层方法进行毒性测试(图 10-2)。这包括微生物抑制测试,所有药物审批都要求进行微生物抑制测试,以评估产品的处置是否会对废物处理过程产生不利影响。
如果预期活性分子会在环境中迅速完全耗尽,则无需进行进一步的环境影响测试。如果一种物质的水解半衰期少于 24 小时,有氧生物降解半衰期少于 8 小时,或土壤生物降解半衰期少于 5 天,则该物质被认为会在环境中迅速消耗。
• 如果需要进行额外的毒性测试,将分三级进行。
• 一级测试描述了至少用一种合适的测试生物进行的急性毒性测试。
• 二级测试也是急性毒性测试,但使用的是水生生物、陆生生物或两者的最低基本组。
• 三级测试是基于第 1 级、第 2 级或第 1 级和第 2 级测试结果的慢性毒性测试。
如果通过接触和影响评估,确定了潜在的重大环境影响,则可在环境评估提交文件的"缓解措施"一节中解决这些问题。缓解措施的例子包括使用标签说明(如“不要冲水”)、废物收集要求(如医院收集)以及限制可向个人开出的剂量。环境评估提交文件的“建议行动的替代方案”部分包括讨论“不采取行动”的替代方案,以及可能对有关环境构成最小风险的替代方案。
如果建议的行动没有确定对环境的影响,本节也会说明该结果。提交后,FDA将评估EA文件以及公众意见,以评估所提供信息的准确性。根据这一评估,FDA决定是否有必要采取缓解措施,以及是否发布“环境影响声明”(EIS)或“无重大影响发现”(FONSI)。
如果FDA确定该行动可能会显著影响人类环境质量,则会发布环境影响报告书。环评报告将讨论已确定的重大影响,以及避免或尽量减少这些影响的备选方案。环境影响报告书是一份公开文件,包括环境影响报告书或环境影响报告书的摘要,说明拟议的行动为何不会对人类环境造成重大影响,并说明环境影响报告书是不必要的。
FDA 要求报告实验得出的物理和化学特性,包括水溶性、解离常数、KOW 和蒸汽压或亨利定律常数。可能与评估相关的其他研究包括土壤和沉积物吸附研究、鱼类生物富集研究,以及评估耗竭机制的研究,包括代谢、水解、有氧生物降解、土壤生物降解和光解。
建议在开始任何毒性测试之前咨询 CDER。一般来说,所有药物审批都要求进行微生物呼吸抑制试验(microbial respiration inhibition test)。进一步的毒性研究要求针对相关的活性分子。
典型的水生毒性基础研究包括鱼类、水生无脊椎动物和藻类的急性试验。典型的陆生毒性研究基础套件包括植物早期生长试验以及蚯蚓和土壤微生物毒性试验。
• 试验方法和生物体应符合 FDA《环境评估技术手册》、EPA(40 CFR 797)或经济合作与发展组织(OECD)确定的标准。所有研究都应符合 FDA 的良好生产规范或良好实验室规范 (GLP) 规定。
• FDA 于 2016 年发布了与具有潜在雌激素、雄激素或甲状腺激素通路活性的药物有关的补充指南。对具有此类活性的活性分子的药物的批准提供了一个例子,说明通常可能符合分类排除的条件,但由于特殊情况,需要进行全面的环境评估。在这种情况下,具有雌激素、雄激素或甲状腺活性的活性分子在浓度低于 1 μg/L 时会对暴露的水生生物的生殖和发育造成影响,而这一浓度值通常被用作确定是否符合分类排除条件的阈值。此外,无论1 级和2 级毒性测试的结果如何,都应对此类原料药进行3 级毒性研究。
• 还应注意的是,根据FDA的指导,生产和处置信息无需作为环境评估的一部分提交。我们假定,药品的生产和处置符合其他法规,可防止对人类环境造成重大不利影响。但是,如果出现使这一假设失效的特殊情况,则应在环境影响评估中对情况、缓解措施和可能的替代方案进行讨论。
02
欧盟的小分子药物环境风险评估
在欧盟,除个别情况外,新的上市许可申请必须进行环境风险评估。需要进行初步筛选,以确定是否需要进行全面的 ERA环境风险评估(Environmental Risk Assessment)。例如,对于由肽、蛋白质或类似化学实体组成的原料药,申请人可以提交不需要进行研究的理由。此外,对于已在该欧盟成员国销售的原料药,只有在新用途可能导致环境暴露增加的情况下,才需要进行ERA(例如 因为新产品可能会吸引更多的人在该成员国使用该原料药)。Figure 10-3 显示了根据 2006 年 EMA 指南6 执行 ERA 所需的决策过程。
在需要进行 ERA 时,环境风险结论并不能成为欧盟拒绝批准上市的理由。相反,后续研究或减少暴露的措施,如要求贴标签,以阻止消费者将未使用的药物倒入废水(水槽或马桶)中。
一旦启动,ERA 程序最多包括三个阶段:
• 第一阶段
• 第二阶段 A 级,以及
• 第二阶段 B 级。
Table 10-1 列出了每个阶段所需的研究以及这些研究在 ERA 报告中的主要使用方式。API 在Phase I只需进行一项环境研究(即确定辛醇-水分配系数)。如果Phase I评估得出的结论是没有预期风险,则ERA 终止。如果不能排除风险,则需要在Phase II A 级中进行额外的研究和评估步骤。大多数原料药都需要进行Phase II A 级评估,ERA 结果中有六个决策点,可触发Phase II B 级中最多六个额外的评估途径。相比之下,Phase II B 级流程有多个独立的评估途径,只需完成Phase II A 级 ERA 结果触发的特定途径。
由于采用分阶段 ERA 方法(需要初步研究结果来确定是否需要进行后续研究),因此通常无法提前预测 ERA 的全套研究(或时间表和成本)。预测化学品环境归宿和影响的计算机模型(如免费提供的 EPIWeb 套装)可用于估算可能的评估路径,因为此类软件可预测某些所需研究的结果。
所有原料药都必须在Phase I进行持久性、生物蓄积性和毒性 (PBT) 综合属性评估。EMA 准则要求进行筛选性 PBT 评估,以确定是否可以排除这种担忧。在此,辛醇-水分配系数是一种替代测量方法,用于衡量原料药从水(预计使用后部分原料药会在水中产生)进入水生生物(如鱼类)脂肪储存库的倾向,进而导致捕食者(如鸟类)的暴露和潜在毒性。如果原料药具有高度疏水性(见Table 10-1 ),则不能排除其属于 PBT 或极持久性极生物累积性 (vPvB) 物质的可能性,必须根据REACH 指南进行全面评估。
PBT评估和II期A级评估的关键研究是确定中性分子的辛醇-水分配系数(logP或logow);logD表示可电离分子)。在药物开发过程中,通常在生物学相关的pH值为7.4时检测原料药的logP,但由于研究设计与推荐方法的差异,这些结果通常不被ERA接受。欧盟天然水的pH值范围相对较宽,因此EMA要求可电离APIs的评估范围为pH 5至pH 9原料药通常在该pH范围内具有解离常数值,因此环境暴露是中性和电离形式的混合物。原料药在脂肪中积累的倾向(因此在食物网中比水生生物更高的捕食者中)取决于存在哪种形式。因此,可电离APIs需要辛醇-水分布常数(logD),该常数表示电离形式和中性形式的相对疏水性,以及每种形式在相关pH范围内的丰度。
在进行 PBT 评估的同时,还针对对鱼类等生物的直接水生影响进行了Phase I ERA,其中有三种可能的途径,其中前两种较为罕见:
1.原料药的最大日剂量很低,预计环境暴露量不会上升到令人担忧的水平,因此无需进行额外研究即可完成 ERA;
2.该原料药是一种内分泌活性物质,即使暴露浓度很低,预计也会造成环境风险,因此需要专门的 ERA 流程,需要与监管机构协商,制定定制的测试和评估计划;或
3.需要进行一整套研究(见Table 10-1 ),以确定原料药的潜在环境归宿和影响,从而得出六个不同的风险决策点。
Phase I评估不允许申请人考虑产品的销售量或预期销售量。而是采用了一种计算方法,假定 1%的人口以每日最大剂量使用原料药,药物被冲入污水处理厂,在被稀释后进入自然水域。对于大多数原料药来说,1% 的市场渗透率是一个严重的高估,因此申请人可以选择使用第三方流行病学数据,表明所显示的疾病流行率低于总人口的 1%,以此将该假设修改为更切合实际的估计值。如果地表水中的预测环境浓度(PECsurfacewater) 小于 0.01 μg/L,则 ERA 可在Phase I结束。
该计算方法(使用默认的 1%市场渗透率)意味着,任何最大日剂量小于 2 毫克的原料药都不需要在Phase I后进行评估。但这一限制不适用于 "内分泌活性 "原料药,因为预计对环境产生不利影响的浓度低于其他类别的化合物。对于具有内分泌活性的原料药,必须咨询 EMA 以确定适当的 ERA 要求。
申请人委托或进行的新研究应采用 GLP 方法进行,但对于现有的原料药,如果研究质量足够高,可使用公共领域的研究来履行检测义务。不接受来自硅学模型的数据。
Phase II A 级评估结束时的六个决策要点旨在排除对环境的六个不同方面的保护,具体如下。
1.对水生生物的直接毒性
2.地下水源(地下水)
3.微生物
4.捕食物种的生物累积和二次中毒
5.对陆生(土栖)生物的毒性
6.对沉积物栖息生物的毒性
如果不能排除一个或多个保护目标的风险,则需要在Phase II B 级对这些目标进行额外的测试和评估步骤。Phase II B 级评估需要进行额外的测试,重点是更好地了解风险的可能性和风险的程度。很少有原料药会触发所有的Phase II B 级评估途径,许多 ERA 在Phase II A 级评估之后就终止了,结论是预计不会有风险。Phase II B 级评估途径中有几项要求 ERA 遵循欧盟一般化学品评估计划 REACH(化学品注册、评估、许可和限制)中的规范性技术指南。
如果在Phase II B 级 ERA 结束时,不能排除风险,则 ERA 必须提出申请人将采取的缓解措 施,以降低已确定的风险在自然界中显现的可能性。常见的缓解策略包括贴标签,警告消费者不要通过废水或家庭垃圾处理未使用的医药产品,以及实施药品返还计划,防止未使用的医药产品释放到环境中。
风险结论不会影响对原料药的批准,但在需要进行环境影响评估时,遗漏环境影响评估可能会危及授权。因此,如果对原料药进行的某些研究采用的方案与建议的方案不同,或者出现了其他偏离指南的情况,那么谨慎的做法是尽可能在最高的确定性水平上完成 ERA,并执行审查员要求的任何后续研究,作为批准后的要求,以解决任何遗留问题。
03
美国生物制品的环境评估
美国生物制品审批的环境评估程序与传统药物审批的环境评估程序类似,但有一些具体的不同之处。首先,生物制品的环境评估要提交给 FDA 内的另一个中心,即生物制品评估与研究中心 (CBER)。FDA 于 2015 年发布了 1998 年指南的补充文件,明确了基因疗法、有载体疫苗和相关重组病毒或微生物产品(GTVV)的特定 EA 要求。
Figure 10-4 是确定生物批准是否符合分类排除条件的流程图。如果拟议的行动是一项 IND,如果该物质是天然存在的,或者如果该行动不会增加该物质的使用,则生物批准通常符合条件。如果 GTVV 包含来自同属不同物种的蛋白质编码序列,如果 GTVV 与野生型物质之间的区别仅限于减弱点突变或缺失,如果 GTVV 在制造步骤中被杀死或灭活,或者如果 GTVV 由转基因人体细胞组成,则该 GTVV 被认为是天然存在的。
与 EICaquatic 临界值相关的常规药物审批分类排除不适用于生物制品。如果生物制品与可能引起争议的环境影响、基本未知的环境影响、对受保护物种的不利影响或违反联邦、州或地方法律有关,这些因素就构成了特殊情况,生物制品的审批就不符合分类排除的条件。
生物制品的环境影响评估与传统药物产品的环境影响评估在评估潜在环境问题的方法上有所不同。生物制品的评估不是基于分级毒性测试的结果,而是基于相关物质及其可能进入的环境的特征。GTVV 的表型属性会被考虑在内,包括其分布、宿主范围和滋养性,以及是否具有毒性、致病性或已知对动物、植物或微生物具有毒性。
还讨论了 GTVV 的环境限制,如其生长或繁殖所需的基质,以及其环境中是否可能存在这些基质,其对控制方法(即抗生素、抗病毒药物或杀菌剂)的敏感性,以及可能促进或限制其在环境条件下生存的其他特征。
如果 GTVV 的任何特征使其能够超越自然生物,则要讨论这种选择性优势及其影响,以及接触 GTVV 对非目标种群的任何其他潜在影响。重要的考虑因素包括可能特别容易受到感染的脆弱种群,以及可能影响环境影响的可能性或严重性的潜在遗传不稳定性。根据对严重性和可能性的综合评估,每种已确定的潜在环境影响都被划分为 "高"、"中 "或 "低至可忽略不计 "风险。
与药品审批的环境影响评估报告一样,生物制品审批的环境影响评估报告也包括对拟议行动的缓解措施和替代方案的讨论。生物制品的缓解措施包括在生物封闭措施下处理相关物质、尽量减少与脆弱人群或物种的接触以及尽量减少气溶胶的形成。FDA 将根据其对 EA 中所提供信息的评估结果,如前所述发布 EIS((Environmental Impact Statement) 或 FONS(Finding of No Significant Impact)。
可进行研究,以支持将潜在环境影响分为 "高风险"、"中度风险 "或 "低至可忽略风险"。可用于为这些风险分类提供信息的实验数据包括:进入环境的 GTVV 量的特征、其在环境中的降解率、其被暴露生物体吸收的频率、导致暴露生物体感染的水平、其遗传稳定性以及与自然生物体相比的相对适应性。
04
欧盟生物制品的环境风险评估
EMA 要求符合其转基因生物 (genetically modified organism, GMO) 定义的生物药品必须进行 ERA(Environmental Risk Assessment)根据该定义,转基因生物含有转基因,或源自不同类型生物的遗传物质,并以人工方式引入。生物制品的ERA程序旨在解决这些转基因可能在自然界中以意想不到的有害方式传播和表达所带来的风险。
通常所说的转基因生物的定义包括从转基因生物中提取的物质(如蛋白质、碳水化合物)。如果这些材料不能复制或转移遗传物质,则不属于本条例的重点。
与小分子的ERA程序不同的是,转基因生物的ERA程序是由申请人的专家风险评估员与监管审查员协商后根据具体情况确定的。ERA必须涉及准则中大致列出的六个方面:
1.确定可能造成不良影响的特征;
2.评估每种潜在不利影响的后果及其严重程度;
3.评估每种潜在不利影响发生的可能性;
4.估计第 1 项中确定的转基因生物的每种特性所带来的风险;
5.选择应对风险的管理策略;
6.确定总体风险。
与小分子化合物的定量风险评估程序不同,这里采用的是一个等级系统,用 "高"、"中"、"低 "和 "可忽略不计 "来估算潜在后果及其程度、可能性和不良影响风险。定性和定量数据都可用于支持评估。必要时,风险缓解措施可包括设计容器以尽量减少生物体逃逸,说明意外溢出时如何灭活转基因生物,以及处置说明。
EMA 的转基因生物ERA 准则第 4 节介绍了监管程序和时间表。值得注意的是,由于转基因生物ERA程序的临时性质,该指南建议 "潜在申请人在提交申请前六个月至一年要求与EMA举行提交前会议(pre-submission meeting)",以商定测试和评估方法。
案例研究:欧盟的萘普生ERA(Case Study: Naproxen ERA in the EU)
为了帮助理解小分子药物的 ERA 流程(这是全球最普遍采用和接受的框架),我们将以萘普生对环境的影响作为案例进行研究。
萘普生是一种非甾体抗炎药(NSAID),用于缓解疼痛和炎症。由于这是一种非处方原料药,公共领域有许多关于其环境归宿和影响的研究,这些研究是针对活性分子的钠盐和酸进行的。两者都会导致在自然水体中接触到相同的活性分子,因此必须对研究报告中的所有浓度进行校正,以反映等效浓度(盐或酸)。在本示例中,浓度被校正为等同于钠盐。
2007 年发布的一份萘普生环境风险评估报告为执行本例 ERA 提供了丰富的数据来源,其中包括水溶性、解离常数、logD、熔化温度和生物降解性等数据。
PECsurfacewater 计算需要输入最大日剂量(1100 毫克),计算结果为 5.5 微克/升,超过了 0.01 微克/升的阈值,表明需要进行Phase II A 级 ERA。萘普生的最高对数值(1.06)28 低于 PBT 评估的触发值(4.5)。因此,无需进行 PBT 评估。
虽然有多项研究支持对萘普生的环境归宿进行评估,但在本例中,为说明ERA 过程如何起作用而选择的关键研究是那些包含足够细节以评估研究设计质量和数据可靠性并得出对环境有保护作用的值的研究(当多项研究得出不同结果时)。
测得的最大 logKoc 为 3.00, 在长达 30 天(少于标准 OECD 308 试验的 180 天)的沉积物转化试验中,50% 的消散时间长达 6.6 天,从水到沉积物的转移可忽略不计。这种非标准测试可用于 ERA,但容易受到审查人员的质疑,他们可能会要求进行标准沉积物转化研究,作为批准后的要求。
如果标准测试未能通过Phase II A 级评估,那么批准后要求还包括需要根据Phase II B 级准则进行全面的沉积物评估。
相关可靠的水生毒性测试包括四项微生物研究、四项藻类研究、四项无脊椎动物研究、五项鱼类研究和一项蝌蚪研究。测试中最敏感的物种是日本稻鱼、萘普生钠当量的无不良影响浓度为 55 μg/L。该值是预测水中无影响浓度或 PNEC 的基础。毒性值除以 10 的评估系数,以考虑不确定性,从而确定水的 PNEC 为 5.5 μg/L。
在同一项研究中,水蚤 Moina macrocopa 七天繁殖未观察到不良影响的萘普生钠当量浓度为 362 μg/L,31 从而得出地下水的 PNEC 为 36.2 μg/L。微生物的预测无效应浓度是根据对活性污泥微生物联合体的微生物抑制的最低不利影响水平 2.52 毫克/升32 和评估系数 10 得出的,微生物的预测无效应浓度为 252 微克/升。
Phase II A 级保护的六个目标评估如下:
1.地表水:PECsurfacewater 等于 PNECwater,因此不能排除风险,需要对水进行第 II 阶段 B 级评估。
2.地下水:PECsurfacewater × 25% 小于 PNECgroundwater,因此可以排除该终点的风险。
3.微生物:PECsurfacewater 小于 PNECmicroorganism,因此可以排除该终点的风险。
4.生物累积:logD(1.06)小于阈值(3.0),因此可以排除该终点的风险。
5.陆地(土壤):logKoc(3.00)小于阈值(4.0),因此可以排除该终点的风险。
6.沉积物:在沉积物转化测试中,没有出现从水到沉积物的明显转移,因此可以排除该终点的风险。
在Phase II B 级中,必须对地表水途径进行评估,这就需要使用 SimpleTreat 模型来完善 PEC 值,使其更符合实际情况,并考虑到废水处理厂对萘普生的去除情况。根据 REACH 法规,欧盟使用 SimpleTreat 模型进行安全评估。该模型中包含了用于欧盟监管目的的建议默认值。
在一项为期 28 天的固有生物降解性试验中,经过 13 天的滞后期,萘普生的需氧量达到了理论最大需氧量的 78%,研究结束时达到了 98%,这表明萘普生具有固有的生物降解性。因此,可在 SimpleTreat 模型中使用 0.1/hr 的废水处理厂降解率常数,以考虑 PECsurfacewater 值中的降解情况。根据模型预测,进入废水处理厂的萘普生有 54% 被去除,从而减轻了暴露程度,并将 PECsurfacewater 值降至 2.55 μg/L。现在,该值小于 PNECwater,排除了地表水终点的风险。在此 ERA 中,无需对萘普生进行风险缓解。
以上是从公开文献中选取的萘普生ERA研究的简化解释。当然,这也有局限性,因为在评估现有研究的可靠性和相关性以选择最合适的研究作为决策依据时会出现复杂性。大多数公开发表的研究都是在大学进行的,没有采用 GLP 标准,也没有像商业实验室的研究那样提供详尽的文献资料。
因此,风险评估人员对研究的适当性得出不同的结论,从而得出相互矛盾的风险结论,是很常见的。在实践中,这可能要求申请人在审查过程中以合理的科学依据为其研究选择进行辩护,但仍有可能无法说服他们认为其原料药不会对环境造成风险。在这种情况下,监管审查人员可能会要求进行一项或多项补救研究,作为批准后的要求,以提高 ERA 中数据的质量。
— 结论—
近几十年来,对原料药潜在环境影响的关注和对制定强有力的药物批准法规的兴趣有所增加。目前,美国和欧盟为评估潜力提供了最全面、定量的药物特定监管框架使用原料药带来的环境风险。通过比较使用的原料药在水、土壤或沉积物中可能出现的浓度与在这些栖息地中生活的生物可能安全的水、土壤或沉积物中的最高浓度,主要确定了EA/ERA(美国的环境评估 (environmental assessment, EA) 和欧盟的环境风险评估 (environmentalrisk assessment, ERA))中的小分子的潜在风险。
然而,EA/ERA所需的信息是否被认为是完整的取决于所使用的监管框架和药物的特征。对于生物制品,只有具有“高”、“中”和“低”风险结果的定性EA/ERA是可能的,并且需要与美国或欧盟的监管机构协商确定自定义测试和评估计划。此外,EA/ERA要求也在不断变化,因为世界各地的司法管辖区在不断发展和为进一步完善药物批准的环境安全和风险考虑方法。
英文版原文
(上下滑动查看更多)
Chapter 10 Regulatory Environmental Risk Assessment of Human Pharmaceuticals
The potential for active pharmaceutical ingredients (APIs) from medicinal products to enter the environment and lead to adverse effects to ecosystems has received increased regulatory scrutiny since the early 2000s. Active drug compounds are designed to elicit specific biological effects in humans – often at low doses – but also have the potential to cause unintended adverse effects to wildlife if unused or excreted APIs enter municipal wastewater streams, and ultimately natural waters.1 Although research addressing pharmaceuticals in the environment ballooned in the 2000s there are still only a limited number of global regulations aimed specifically at this topic.
It has long been recognized that pharmaceutical manufacturing effluents can release APIs among other process chemicals into receiving waters as point sources. As such, local jurisdictions have traditionally regulated these emissions on an individual basis to mitigate risks to nearby ecosystems. This chapter addresses separate environmental regulations that were developed to address the potential for the use of pharmaceutical products to result in more widespread societal emissions and broader environmental risks. Regulatory requirements for environmental assessment (EA) in the US and environmental risk assessment (ERA) in the EU related to product use, which are mandated as an element of product registration dossiers and have been adopted globally, are detailed here. These regulations are focused on evaluating small molecule active ingredients and their major metabolites or biologics but not excipients, under the assumption that the biologically active portions of these products are most likely to lead to risk.
Regulatory EA/ERAs for APIs primarily evaluate risk by comparing two concentration values: first an estimate of the concentration of the API expected to occur in water, soil, or sediment due to consumer use and not manufacture, and second an estimate of the highest concentration in water, soil, or sediment that is likely to be safe for organisms living in those habitats. These two concentrations are then compared and if the environmental concentration of the API is higher than the expected safe concentration, environmental risk cannot be ruled out. There are additional checks included in regulatory EA/ERAs for special risks as well, including accumulation in the food chain leading to secondary poisoning in predators, disruption to wastewater treatment processes, and others detailed later in this chapter.
A typical strategy for performing EA/ERAs is to conduct an initial, or screening-level, assessment using simplifying assumptions that overestimate risk. If there is no risk predicted using these simplifying assumptions, then regulators are reassured that the likely environmental risk is low, and the assessment ends.
If risk cannot be ruled out, additional studies are conducted to supply data that overrides the simplifying assumptions, allowing a more realistic estimate of risk. This process is done iteratively as needed to make a conclusion on risk while avoiding wasting resources to conduct unnecessary tests. There are two major global strategies for testing the environmental impact of APIs, one in the US and one in the EU, both of which use this tiered approach to data collection and assessment.
一、Global Regulatory Landscape
The development of prescriptive, pharmaceutical-specific ERA guidance has been limited to the EU and the US. Outside of those jurisdictions, regulatory frameworks applicable to ERA of human pharmaceutical products are either qualitative, not specific to pharmaceuticals, or nonexistent. Examples of countries that lack pharmaceutical-specific regulatory frameworks, but which assess environmental risks of pharmaceuticals through a general chemical assessment program applicable to other product categories include Canada, China, and Australia.2,3 Environmental risk assessments for human pharmaceuticals are not currently required in many countries, including regions of South America and Africa.
As regulatory attention on the presence of pharmaceuticals in the environment intensifies, new regulatory frameworks are expected to be developed, and existing requirements are expected to be strengthened.
• Canada: While Health Canada currently assesses pharmaceuticals through the general New Substances Program, amendments are in development that would create a separate regulatory framework specific to pharmaceutical active ingredients under the Food and Drug Act.4
• EU: In 2018, the EU ERA guidelines were revised, although currently in draft format, they are expected to supersede the 2006 version still in use at the time of writing.
• China: In May 2022, China State Council released an Action Plan that includes a goal of enhancing the framework governing environmental risk management of new chemicals.
• Japan: In Japan, pharmaceuticals were evaluated within a general chemical program until 2016, when pharmaceutical-specific guidance was published.3,8 Because this pharmaceutical-specific guidance is not prescriptive, submissions areexpected to follow either US or EU guidance.
In countries where specific pharmaceutical guidance is absent, when an EA/ERA is requested, it is generally considered acceptable to follow the framework used in the EU, as it is the most well-developed, prescriptive, and comprehensive pharmaceutical EA/ERA regulation in force globally.
In both the US and EU, metabolites of the API can be ignored if the applicant assumes that all residues from the drug have the same environmental fate and effects characteristics as the API.
This is considered to be a protective approach, as it is expected that no metabolites would be more toxicologically potent than the parent API. If the applicant has data to show that a major metabolite has significantly different characteristics relevant to the EA/ERA outcome, they can account for this by performing a full EA/ERA for the metabolite. This is rarely done however, because it is resource intensive, requiring a full suite of environmental fate and effects testing for the metabolite molecule.
01
Environmental Assessment of Small Molecules in the US
In the US, drug approval applications are submitted to the Food and Drug Administration (FDA) Center for Drug Evaluation and Research (CDER).9 Submissions must either claim a categorical exclusion, indicating the action (i.e., the approval of the drug) will not significantly affect the quality of the human environment, or provide a full EA for the proposed action. Figure 10-1 provides a process flow diagram for determining whether an action qualifies for a categorical exclusion or requires a full EA. An action that falls into at least one of the following categories typically qualifies for a categorical exclusion:
1.An investigational new drug application (IND)
2.Actions that will result in the production of an active moiety that naturally occurs in the environment, with the caveat that the action does not significantly alter its concentration or distribution in the environment
3.Actions that will not increase the use of the active moiety (e.g., lower dosage or shorter duration than previously approvedfor the same indication)
4.Actions that will increase the use of the active moiety but result in an aquatic environmental introduction concentration (EICaquatic) below 1 μg/L (ppb) in water.
For item 4 in the above list, the calculation prescribed in the FDA’s guidance document9 relates the EICaquatic trigger concentration to the maximum annual sales volume of the API expected in the 5 years after marketing approval. This calculation assumes that the API is used uniformly throughout the US and is diluted in wastewater before emission to the environment. The result of the calculation is that any API sold in quantities below a rate of about 45,000 kg per year in the US qualifies for a categorical exclusion, and no EA is required. The exact value of the calculation varies each year depending on the most recently reported volume of wastewater entering publicly owned treatment works (POTWs; wastewater treatment plants) in the US annually. Most APIs qualify for categorical exclusion under this section of the FDA’s guideline.9
Before a determination of categorial exclusion is reached, extraordinary circumstances that may indicate the potential for a drug to cause serious harm to the human environment must be considered. If extraordinary circumstances are identified, a full EA must be completed for the proposed drug to obtain marketing approval. Examples of actions associated with extraordinary circumstances include drugs with active moieties that are known to be toxic or are likely to have a lasting impact on the exposed environment at expected concentrations, drugs that foreseeably cause harm to protected species or the critical habitats of protected species, and drugs with active moieties that are derived from natural species and threaten the continued existence of the species (i.e., overharvesting).
If the proposed approval of a drug does not meet the criteria for a categorical exclusion, a complete EA must be submitted. The basic requirement for an EA is that sufficient information is provided to determine whether the proposed action will significantly impact the human environment. The FDA EA process is less prescriptive than the EU ERA process described below. If the applicant believes that an EA is required, they should contact CDER to determine the appropriate testing and assessment pathway to follow before undertaking the EA. The general approaches required by FDA are described below.
Predicted exposure concentrations of the active moiety in relevant environmental compartments (i.e., water, soil, and air) resulting from the drug are calculated. Then, for each environmental compartment, the expected exposure concentration is compared with the concentration that is likely to have a significant adverse impact on exposed organisms.
As part of this evaluation, the environmental fate of the active moiety and metabolites or degradants that account for more than 10% of the initial dose, is evaluated. Relevant physical and chemical characteristics that inform the environmental fate assessment are reported, including water solubility, dissociation constant(s), octanol/water partition coefficient (logKOW), and vapor pressure or Henry’s Law constant. If the properties of the active moiety indicate a tendency for substantial sorption (adherence) to solids in the environment, characterization of the organic carbon-water partition coefficient (logKOC) should also be included. Based on the environmental fate assessment, the relevant environmental compartments for the active moiety are identified.
EICs are calculated for each of the relevant environmental compartments. Besides dilution, additional factors including metabolism, removal during waste treatment, and environmental depletion (e.g., hydrolysis, photolysis, biodegradation) can also be factored into this calculation. Following water treatment, substances that adhere to biosolids may be introduced to the terrestrial environment through biosolid land application. For these APIs, EICterrestrial is calculated based on partitioning characteristics, the impact of depletion mechanisms, and the rate of biosolid land application. An EIC for the atmospheric environment is not usually required for pharmaceutical products because the active moiety is not typically released into the air. When relevant, such as for medical gases, an EICatmospheric can be calculated. In addition to an EIC, an expected environmental concentration (EEC) is calculated for each relevant environmental compartment. The EEC accounts for factors that influence the environmental concentration of the active moiety after its initial introduction to the environment, including additional depletion, dilution, sorption, and bioaccumulation.
To determine relevant effect concentrations, experimentally derived toxicity data are collected through sponsor-facilitated testing, peer-reviewed literature sources, or both. The results of computer modelling can also be used in a supportive capacity.
FDA guidance9 suggests using a tiered approach to toxicity testing (Figure 10-2). This includes microbial inhibition testing, which is required for all drug approvals to assess whether disposal of the product will adversely impact waste treatment processes.
In the circumstance that a rapid and complete depletion of the active moiety in the environment is expected, no further environmental effects testing is required. A substance is considered to rapidly deplete in the environment if it is associated with a hydrolysis half-life of less than 24 hours, an aerobic biodegradation half-life of less than 8 hours, or a soil biodegradation half-life of less than 5 days.
• If additional toxicity testing is required, it will proceed through three tiers.
• Tier 1 testing describes acute toxicity testing performed with at least one suitable test organism.
• Tier 2 testing also describes acute toxicity testing, but it is performed with the minimum base set of aquatic organisms, terrestrial organisms, or both.
• Tier 3 testing is chronic toxicity testing based on the results of Tier 1, Tier 2, or both Tier 1 and Tier 2 testing.
If, through the exposure and effects evaluation, potential significant environmental impacts are identified, these may be addressed in the ‘Mitigation Measures’ section of the EA submission document. Examples of mitigation measures include the use of label directions (e.g., ‘do not flush’), waste collection requirements (e.g., hospital collection), and limitations on the number of doses that can be prescribed to an individual. The ‘Alternatives to the Proposed Action’ section of the EA submission document includes a discussion of the ‘no action’ alternative as well as alternatives that may pose the least risk to the environment of concern. If no environmental impact has been identified for the proposed action, that finding is also stated in this section.
Following submission, the FDA evaluates the EA document along with public input to assess the accuracy of the information provided. Based on this evaluation, the FDA determines whether mitigation measures are necessary and whether to issue an ‘Environmental Impact Statement’ (EIS) or a ‘Finding of No Significant Impact’ (FONSI). An EIS is issued if the FDA determines that the action may significantly affect the quality of the human environment. The EIS would provide a discussion of the significant impacts identified and the alternatives available to avoid or minimize these impacts. A FONSI is a public document that either includes the EA or a summary of the EA, justifying why the proposed action will not significantly impact the human environment, and stating that an EIS is not necessary.
The FDA requires reporting of experimentally derived physical and chemical characteristics including water solubility, dissociation constant(s), KOW, and vapor pressure or Henry’s Law constant. Additional studies that may be relevant to the assessment include soil and sediment sorption studies, fish bioconcentration studies, and studies evaluating depletion mechanisms including metabolism, hydrolysis, aerobic biodegradation, soil biodegradation, and photolysis.
Consultation with CDER is recommended before any toxicity testing is initiated. Generally, the microbial respiration inhibition test is required for all drug approvals. Further toxicity study requirements are specific to the active moiety of interest.
The typical aquatic toxicity study base set includes acute tests with fish, aquatic invertebrates, and algae. The typical terrestrial toxicity study base set includes the plant early growth test, and toxicity tests with earthworms and soil microbes.
Test methods and organisms should meet the standards identified by the FDA Environmental Assessment Technical Handbook, the EPA (40 CFR 797), or the Organization for Economic Cooperation and Development (OECD). All studies should be compliant with either FDA’s good manufacturing practice or good laboratory practice (GLP) regulations.
• Supplemental guidance related to drugs that have potential estrogenic, androgenic, or thyroid hormone pathway activity was published by the FDA in 2016.10 Approvals for drugs with active moieties with such activity provide an example of actions that may typically qualify for a categorical exclusion, but due to extraordinary circumstances a full EA is required. In this case, active moieties with estrogenic, androgenic, or thyroid activity have been shown to cause reproductive and developmental effects in exposed aquatic organisms at concentrations less than 1 μg/L, the value typically used as the threshold to determine qualification for categorical exclusion. In addition, Tier 3 toxicity studies should be conducted for such APIs irrespective of the results of Tier 1 and Tier 2 toxicity testing.
• It should also be noted that in accordance with the FDA guidance manufacturing and disposal information is not required to be submitted as part of the EA. It is assumed that pharmaceutical products are produced and disposed of in compliance with other regulations that will prevent a significant adverse impact on the human environment. If there are extraordinary circumstances that invalidate this assumption, however, a discussion of the circumstances, mitigating measures, and possible alternatives, should be included in the EA.
02
Environmental Risk Assessment of Small Molecules in the EU
In the EU, an ERA is required for new marketing authorization applications with some exceptions. An initial screening step is needed to identify whether a full ERA is required. For instance, an applicant can submit justification that there is no need for studies for APIs consisting of peptides, proteins, or similar chemical entities. Also, for APIs already marketed in that EU member state, an ERA is only required if the new use is likely to result in increased environmental exposure (e.g. because the new product is likely to induce a larger population to use the API in that member state). Figure 10-3 shows the decision process required for performing an ERA according to the 2006 EMA guideline.6 The types of applications and concomitant need for an ERA is described in Section 2 of the EMA guideline.
When an ERA is required, a conclusion of environmental risk is not grounds for denying a marketing authorization in the EU. Rather, follow-up studies or exposure mitigation measures, such as labeling requirements to discourage consumers from disposing unused medication into wastewater (sink or toilet), could be required.
Once initiated the ERA process consists of a maximum of three phases:
• Phase I,
• Phase II Tier A, and
• Phase II Tier B.
Table 10-1 lists the studies required during each of these phases and the primary ways in which they are used in the ERA report. Only one environmental study is required for APIs in Phase I (i.e., determination of the octanol-water partition coefficient). If the Phase I assessment reaches a conclusion of no expected risk, the ERA is terminated. If risk cannot be ruled out, additional studies and assessment steps are required in Phase II Tier A. The Phase II Tier A assessment is required for most APIs and there are six decision points in the ERA outcome that can trigger up to six additional assessment pathways in Phase II Tier B. When a Phase II Tier A ERA is required, all of the studies recommended must be conducted and assessment pathways evaluated. In contrast, the Phase II Tier B process has multiple independent assessment pathways, and only the specific pathways triggered by the Phase II Tier A ERA outcome need to be completed.
Because of the phased ERA approach – with initial study results required to determine whether subsequent studies will be needed – it is usually not possible to predict the full suite of studies (or timeline and cost) for the ERA in advance. Computer models for predicting the environmental fate and effects of chemicals such as the freely available EPIWeb suit can be used to estimate a likely assessment path, because such software predicts the outcome of some of the required studies.
All APIs must undergo an assessment for the combination of persistence, bioaccumulation, and toxicity (PBT) attributes in Phase I. The PBT assessment proceeds in parallel to the rest of the ERA steps. The EMA guideline6 requires a screening PBT assessment to determine if this concern can be ruled out. Here, the octanol-water partition coefficient is a surrogate measure for the propensity of the API to move from water (where some of the API is expected to occur after use) into the fat stores of aquatic organisms (e.g., fish), which then leads to exposure and potential toxicity to predators (e.g., birds). If the API is highly hydrophobic (see Table 10-1), the possibility that it is a PBT or a very persistent very bioaccumulative (vPvB) substance cannot be ruled out, and full assessment according to REACH guidance24 is necessary.
A critical study for the PBT assessment and the Phase II Tier A assessment is a determination of the octanol-water partition coefficient (logP or logKow for neutral molecules; logD for ionizable molecules). It is common for logP of APIs to be tested during drug development at the biologically relevant pH of 7.4, but these results are usually not accepted in an ERA because of differences in the study design from the recommended method.
Natural waters in the EU have a relatively broad pH range, leading the EMA to require evaluation from pH 5 to pH 9 for ionizable APIs.25 It is common for APIs to have a dissociation constant value in this pH range, such that environmental exposure is to a blend of the neutral and ionized forms. The propensity of the API to accumulate in fat (and hence in predators that are higher up in food webs than aquatic organisms) depends on which form is present. As a result, the octanol-water distribution constant (logD), which accounts for the relative hydrophobicity of the ionized and neutral forms, and the abundance of each form over the relevant pH range, is required for ionizable APIs.
In parallel with the PBT assessment, there is a Phase I ERA for direct aquatic effects to organisms such as fish, in which there are three potential paths, the first two of which are rare:
1.The maximum daily dose of the API is so low that environmental exposure is not expected to rise to a level of concern, so the ERA can be concluded with no additional studies required;
2.The API is an endocrine active substance and is expected to result in environmental risk even at very low exposure concentrations, so a specialized ERA process is necessary requiring consultation with regulatory authorities to devise a custom testing and assessment plan; or
3.A suite of studies (see Table 10-1) is required to characterize the potential environmental fate and effects of the API leading to six separate decision points on risk.
The Phase I assessment does not allow the applicant to consider the amount of product sold or expected to be sold. Rather, a calculation is used which assumes that 1% of the population uses the API at the maximum daily dose and that the drug is washed into sewage treatment plants where it is diluted before entering natural waters. The market penetration of 1% is a gross overestimate for most APIs, so the applicant has the option to use third-party epidemiological data showing that the indicated disease prevalence is less than 1% of the total population as a way to modify that assumption to a more realistic estimate. If the resulting predicted environmental concentration in surface water (PECsurfacewater) is less than 0.01 μg/L, the ERA can be concluded at the Phase I stage. The calculation (using the default 1% market penetration) means that any API with a maximum daily dose less than 2 mg does not require assessment after the Phase I stage. This limit does not apply for APIs that are ‘endocrine active’, however, because the expectation is that adverse environmental effects can occur at concentrations less than for other classes of compounds. For endocrine active APIs, EMA must be consulted to determine appropriate ERA requirements.
New studies commissioned or conducted by the applicant should be conducted using GLP, but for existing APIs, studies from the public domain can be used to fulfil testing obligations if the study quality is adequate. Data from in silico models are not accepted.
The six decision points at the conclusion of a Phase II Tier A assessment are meant to rule out concern for six separate aspects of the environment targeted for protection, as listed below.
1. Direct toxicity to aquatic organisms
2. Underground water sources (groundwater)
3. Microorganisms
4. Bioaccumulation and secondary poisoning in predator species
5. Toxicity to terrestrial (soil-dwelling) organisms
6. Toxicity to sediment-dwelling organisms
If risk cannot be ruled out for one or more of these targets of protection, additional testing and assessment steps are required in Phase II Tier B for those targets. Phase II Tier B assessments require additional testing focused on better understanding the likelihood of risk and the magnitude of that risk. It is rare for an API to trigger all of the Phase II Tier B assessment pathways, and many ERAs are terminated after Phase II Tier A with a conclusion that risk is not expected. Several of the Phase II Tier B assessment pathways require the ERA to follow prescriptive technical guidance from the EU’s general chemical assessment program, REACH – Registration, Evaluation, Authorisation and Restriction of Chemicals.
If, at the end of the Phase II Tier B ERA, risk cannot be ruled out, the ERA must propose mitigation measures that the applicant will undertake to reduce the likelihood that the risk identified would be manifested in nature. Common strategies for mitigation include labeling to warn consumers not to dispose of unused medicinal products via wastewater or household waste, and to implement a drug return program that would prevent release of unused medicinal products into the environment. A finding of risk will not influence approval of the API, but omission of an ERA when one is required can endanger the authorization. As a result, if some studies available for an API were conducted using protocols different than those recommended, or if another deviation from the guideline occurred, it is prudent to complete the ERA at the highest level of certainty possible and perform any follow-up studies requested by reviewers as a post-approval requirement to settle any remaining questions.
03
Environmental Assessment of Biologics in the US
The EA process for biologic approvals in the US is similar to that already described for conventional drug approvals with some specific differences. Firstly, EAs for biologics are submitted to a different center within the FDA, the Center for Biologics Evaluation and Research (CBER). A supplement to the FDA’s 1998 guidance was published in 2015 to clarify EA requirements specific to gene therapies, vectored vaccines, and related recombinant viral or microbial products (GTVVs).
A flow diagram representing the process to determine if a biologic approval qualifies for a categorical exclusion is shown in Figure 10-4. Biologic approvals generally qualify if the proposed action is an IND, if the substance is naturally occurring, or if the action will not increase the use of the substance. A GTVV is considered naturally occurring if it includes a protein-coding sequence from a different species within the same genus, if the differentiation between the GTVV and a wild-type substance is limited to attenuating point mutations or deletions, if the GTVV has been killed or inactivated during a manufacturing step, or if the GTVV consists of genetically modified human cells. The conventional drug approval categorical exclusion related to the EICaquatic threshold does not apply to biologics. If the biologic product is associated with environmental effects that are likely to be controversial, environmental effects that are largely unknown, adverse effects on protected species, or violations of federal, state, or local law, these factors constitute extraordinary circumstances, and the biologic approval would not qualify for a categorical exclusion.
EAs for biologic products differ from EAs for conventional drug products in their approach to evaluating potential environmental issues. Rather than an assessment based on the results of tiered toxicity testing, biologics are assessed based on characteristics of both the substance of interest and the environment into which its likely to be introduced. Phenotypic attributes of a GTVV are considered, including its distribution, host range, and tropism and whether it is virulent, pathogenic, or known to be toxic to animals, plants, or microbes. The environmental limitations of a GTVV are discussed, such as which substrates are necessary for its growth or reproduction and if those substrates are likely to be present in its environment, its susceptibility to control methods (i.e., antibiotics, antivirals, or biocides), and other characteristics that might promote or limit its survival under environmental conditions. If the GTVV has any traits that would allow it to out-compete natural organisms, this selective advantage and its implications are discussed as well as any other potential impacts of exposure on non-target populations. Important considerations include vulnerable populations that may be particularly susceptible to infection and potential genetic instability that may influence the likelihood or severity of environmental impacts. Based on the combined assessment of severity and likelihood, each identified potential environmental effect is categorized as posing ‘high,’ ‘moderate,’ or ‘low to negligible’ risk.
As is the case with EA submissions for drug approvals, EAs for biologic approvals include a discussion of mitigation measures and alternatives to the proposed action. Examples of mitigation measures for biologic products include handling substances of interest under biocontainment measures, minimizing contact with vulnerable populations or species and minimizing aerosol formation. The FDA will issue an EIS or a FONSI as previously described, depending on the outcome of their assessment of the information provided in the EA.
Studies may be performed to support the categorization of potential environment effects as ‘high risk’, ‘moderate risk’, or ‘low to negligible risk.’Experimental data that can be used to inform these risk categorizations include characterization of the amount of the GTVV that enters the environment, its degradation rate in the environment, the frequency at which it is taken up by exposed organisms, the level that results in infection in exposed organisms, its genetic stability and its relative fitness compared with naturally occurring organisms.
04
Environmental Risk Assessment of Biologics in the EU
The EMA requires ERAs for biologic medicinal products that meet their definition of a genetically modified organism (GMO).27 According to that definition, a GMO contains a transgene, or genetic material originating from a different type of organism, introduced using artificial means. The ERA process for biologics is meant to address risks due to the potential for these transgenes to spread and be expressed in nature in unexpected and harmful ways.
The definition of GMO in common parlance includes material (e.g., proteins, carbohydrates) derived from GMOs. These materials are not the focus of this regulation if the material is not capable of replication or of transferring genetic material.
Unlike the ERA process for small molecules where there is a well-defined list of tests and evaluation procedures, the GMO ERA procedure is determined on a case-by-case basis in consultation between the applicant’s expert risk assessor and the regulatory reviewer(s). The ERA must address six areas, outlined broadly in the guideline:
1. Identify characteristics that may cause adverse effects;
2. Evaluate the consequences for each potential adverse effect, and its magnitude;
3. Evaluate the likelihood of occurrence for each potential adverse effect;
4. Estimate the risk posed by each characteristic of the GMO identified in item 1;
5. Select management strategies to address risks;
6. Determine the overall risk.
In contrast with the quantitative risk assessment procedure for small molecules, here a ranking system using the qualities ‘high,’ ‘moderate,’ ‘low,’ and ‘negligible’ are used for estimates of the potential consequences and their magnitudes, likelihoods, and risks of adverse effects. Both qualitative and quantitative data can be used to support the assessment. Risk mitigation, when needed, could include designing containers to minimize organism escape, instructions for inactivation of GMOs upon accidental spillage, and directions for disposal.
The regulatory procedure and timeline are described in Section 4 of the EMA’s GMO ERA guideline. Notably, because of the ad hoc nature of the GMO ERA process, it is recommended in the guideline that “prospective applicants request pre-submission meetings with the EMA six months to one year in advance of submission of the application” to agree on a testing and assessment approach.
Case Study: Naproxen ERA in the EU
To help understand the ERA process for small molecules – which is the framework most commonly undertaken and accepted worldwide – the environmental impact of naproxen will be used as a case study.
Naproxen is a non-steroidal anti-inflammatory drug (NSAID) indicated for the relief of pain and inflammation. As this is an over-the-counter API, there are many studies on its environmental fate and effects available in the public domain which have been conducted with both the sodium salt and acid of the active moiety. Both result in exposure to the same active moiety in natural waters, so all concentrations reported in the studies must be corrected to reflect the equivalent concentrations (either salt or acid). For this example, concentrations were corrected to be equivalent to the sodium salt.
An environmental risk assessment for naproxen published in 200728 is a rich source of data for performing this example ERA, including data on aqueous solubility, dissociation constant, logD, melting temperature, and biodegradability.
The PECsurfacewater calculation requires the maximum daily dose (1100 mg) as an input and results in a value of 5.5 μg/L, which exceeds the threshold value of 0.01 μg/L , indicating the need for a Phase II Tier A ERA.
The highest logD value for naproxen (1.06)28 is less than the trigger value for PBT assessment (4.5). Hence no PBT assessment is necessary.
While multiple studies support an assessment of the environmental fate of naproxen, key studies selected for this example of how the ERA process works are those that include adequate detail to assess the quality of the study design and reliability of the data and result in values protective for the environment (when multiple studies have different outcomes).29-32
Naproxen is not readily biodegradable because it does not reach the ‘pass’ criteria in an OECD 301 test.29 The maximum measured logKoc was 3.00, and in a sediment transformation test conducted for up to 30 days (less than the 180-day duration of a standard OECD 308 test), the 50% dissipation time was up to 6.6 days, with negligible shifting from water to sediment. This non-standard test can be used in the ERA, but would be vulnerable to challenge by reviewers, who could require a standard sediment transformation study as a post-approval requirement.
If the standard test fails the Phase II Tier A assessment, then the post-approval requirement would also include a need to conduct a full sediment assessment according to Phase II Tier B guidelines.
Relevant and reliable aquatic toxicity testing available included four microorganism studies, four algae studies, four invertebrate studies, five fish studies, and one study with tadpoles. The most sensitive species among those tested was the Japanese rice fish,
Oryzias latipes, which had a no observed adverse effects concentration of 55 μg/L naproxen sodium equivalents.31 This value forms the basis of the predicted no effect concentration or PNEC in water. The toxicity value is divided by an assessment factor of 10 to account for uncertainty to determine a PNECwater of 5.5 μg/L. The PNECgroundwater was derived from invertebrates in the same study, where the water flea Moina macrocopa seven-day reproduction no observed adverse effects concentration was 362 μg/L naproxen sodium equivalents31 yielding a PNECgroundwater of 36.2 μg/L. The PNECmicroorganism was derived from the lowest adverse effects level for microbial inhibition to an activated sludge microbial consortium of 2.52 mg/L32 with an assessment factor of 10 to obtain a PNECmicroorganism of 252 μg/L.
The six Phase II Tier A protection targets are evaluated as follows:
1. Surface water: the PECsurfacewater is equal to the PNECwater, so risk cannot be ruled out and Phase II Tier B assessment for water is required.
2. Groundwater: the PECsurfacewater × 25% is less than the PNECgroundwater, so risk can be ruled out for this endpoint.
3. Microorganisms: the PECsurfacewater is less than the PNECmicroorganism so risk can be ruled out for this endpoint.
4. Bioaccumulation: logD (1.06) is less than the threshold value (3.0) so risk can be ruled out for this endpoint.
5. Terrestrial (soil): logKoc (3.00) is less than the threshold value (4.0), so risk can be ruled out for this endpoint.
6. Sediment: No significant shifting from water to sediment occurred in a sediment transformation test, so risk can be ruled out for this endpoint.
In Phase II Tier B, the surface water pathway must be evaluated, which requires the use of the SimpleTreat model to refine the PEC value to be more realistic, accounting for removal of naproxen in the wastewater treatment plant. The SimpleTreat model is used in the EU under REACH for safety assessments.
Recommended default values for EU regulatory purposes are included in the model. In a 28-day inherent biodegradability test, after a lag time of 13 days, 78% of the theoretical maximum oxygen demand was reached, and 98% was reached by the end of the study showing that naproxen is inherently biodegradable.28 Hence a degradation rate constant in the wastewater treatment plant of 0.1/hr could be used in the SimpleTreat model to account for degradation in the PECsurfacewater value. The model predicts that 54% of the naproxen entering the wastewater treatment plant is removed, mitigating exposure and reducing the PECsurfacewater to 2.55 μg/L. This value is now less than the PNECwater, ruling out risk for the surface water endpoint. No risk mitigation would be necessary for naproxen in this ERA.
This is a simplified explanation of an ERA for naproxen using studies selected from the open literature. This is of course a limitation as complexity arises when evaluating the reliability and relevance of available studies to select the ones most appropriate as the basis for decision making. Most published studies are conducted at universities and do not use GLP or provide as thorough documentation as studies from commercial laboratories. As a result, it is common for risk assessors to arrive at different conclusions on the appropriateness of studies and hence reach conflicting conclusions about risk. In practice, this can require the applicant to defend their study selections with sound science in the review process, and potentially still not prevail in their opinion that their API poses no environmental risk. In this case, one or more remedial studies could be required by regulatory reviewers as a post-approval requirement to improve the quality of the data in the ERA.
二、Conclusion
Concern over the potential environmental impact of APIs and interest in developing robust regulations for pharmaceutical approvals have increased in recent decades. Currently, the US and the EU provide the most comprehensive, quantitative pharmaceutical-specific regulatory frameworks for assessing the potential environmental risks posed by API use. Potential risks are primarily determined for small molecules in EAs/ERAs by comparing the concentration of the API expected to occur in water, soil, or sediment due to consumer use with the highest concentration in water, soil, or sediment that is likely to be safe for organisms living in those habitats. However, the information required for an EA/ERA to be considered complete varies significantly based on the regulatory framework used and the characteristics of the pharmaceutical of interest. For biologics, only qualitative EA/ERAs with outcomes such as ‘high,’ ‘medium,’ and ‘low’ risk are possible, and require custom testing and assessment plans determined in consultation with regulatory authorities in the US or EU. Furthermore, EA/ERA requirements are in flux, as jurisdictions worldwide continue to develop and further refine approaches to consider environmental safety and risk for pharmaceutical approvals.
参考文献:(上下滑动查看更多)
All references verified 9 March 2023.
1. Kidd KA, et al. Collapse of a fish population after exposure to a synthetic estrogen. PNAS. 2007;104(21):8897-8901.
2. Chemical Inspection and Regulation Service. Guidance documents for new chemical substance notification in China (China REACH)– non official. Current as of 2010. https://www.cirs-reach.com/China_ Chemical_Regulation/Guidance_New_Chemical_Registration_China.pdf
3. Jose J, et al. Comparison of the regulatory outline of ecopharmacovigilance of pharmaceuticals in Europe, USA, Japan and Australia. Sci Total Environ. 2020;709:134815
4. Government of Canada. Forward Regulatory Plan 2022-2024: Regulations amending the Food and Drug Regulations(Environmental Risk Assessment and Management of Ingredients in Drugs). Current as of 30 November 2022. https://www.canada.ca/en/
health-canada/corporate/about-health-canada/legislation-guidelines/acts-regulations/forward-regulatory-plan/plan/environmental-riskassessment-
active-ingredients-drugs.html
5. European Medicines Agency. Guideline on the environmental risk assessment of medicinal products for human use. EMEA/CHMP/SWP/4447/00 Rev. 1. Draft. 2018. Current as of 15 November 2018. https://www.ema.europa.eu/en/news/revised-guideline-assess-riskhuman-medicines-environment
6. European Medicines Agency. Guideline on the environmental risk assessment of medicinal products for human use – first version. EMEA/CHMP/SWP/4447/00 corr 2. 2006. Current as of 13 January 2015. https://www.ema.europa.eu/en/environmental-risk-assessmentmedicinal-products-human-use-scientific-guideline
7. Luo W, LaMotte R. China sets out a regulatory roadmap to control new chemicals. Beveridge & Diamond. 31 May 2022. https://www. bdlaw.com/publications/china-sets-out-a-regulatory-roadmap-tocontrol-new-chemicals/
8. Kaminishi M, de Roode DF. Pharmaceuticals in the environment: The necessity of harmonizing EU guidelines in developing a Japanese ERA Scheme on pharmaceuticals. Current as of 26 May 2021. https://www.criver.com/eureka/pharmaceuticals-environment
9. Food and Drug Administration. Environmental assessment of human drug and biologics applications [guidance]. Current as of 22 February 2018. https://www.fda.gov/regulatory-information/search-fdaguidance documents/environmental-assessment-human-drug-andbiologics-applications
10. Food and Drug Administration. Environmental assessment: Questions and answers regarding drugs with estrogenic, androgenic, or thyroid activity [guidance]. Current as of 24 August 2018. https://www.fda.gov/regulatory-information/search-fda-guidance-documents/environmental-assessment-questions-and-answers-regarding-drugsestrogenic-
androgenic-or-thyroid
11. Organization for Economic Cooperation and Development.Test no.107: Partition coefficient (n-octanol/water): shake flask method. In:OECD guidelines for the testing of chemicals, section 1. Current as of 27 July 1995. https://doi.org/10.1787/9789264069626-en
12. Organization for Economic Cooperation and Development. Test no. 123: Partition coefficient (1-octanol/water): slow-stirring method. In:OECD guidelines for the testing of chemicals, section 1. Current as of 30 June 2022.https://doi.org/10.1787/9789264015845-en
13. Organization for Economic Cooperation and Development. Test no.106: Adsorption -- desorption using a batch equilibrium method. In:OECD guidelines for the testing of chemicals, section 1. Current as of 21 January 2000. https://doi.org/10.1787/9789264069602-en.
14. United States Environmental Protection Agency. Fate, transport and transformation test guidelines OPPTS [Office of Prevention, Pesticides and Toxic Substances] 835.1110 activated sludge sorption isotherm. Current as of April 1996. https://nepis.epa.gov/Exe/ ZyPURL.cgi?Dockey=P100G6ZO.txt
15. Organization for Economic Cooperation and Development. Test no. 121: Estimation of the adsorption coefficient (Koc ) on soil and on sewage sludge using high performance liquid chromatography (HPLC). In: OECD guidelines for the testing of chemicals, section 1. Current as of 22 January 2021. https://doi.org/10.1787/9789264069909-en
16. Organization for Economic Cooperation and Development. Test no. 301: Ready biodegradability. In: OECD guidelines for the testing of chemicals, section 3. Current as of 17 July 1992. https://doi. org/10.1787/9789264070349-en
17. Organization for Economic Cooperation and Development. Test no. 310: Ready biodegradability - CO2 in sealed vessels (headspace test). In: OECD guidelines for the testing of chemicals, section 3. Current as of 26 September 2014. https://doi.org/10.1787/9789264224506-en
18. Organization for Economic Cooperation and Development. Test no. 308: Aerobic and anaerobic transformation in aquatic sediment systems. In: OECD guidelines for the testing of chemicals, section 3. Current as of 24 April 2022. https://doi.org/10.1787/9789264070523-en
19. Organization for Economic Cooperation and Development. Test no. 201: freshwater alga and cyanobacteria, growth inhibition test. In: OECD guidelines for the testing of chemicals, section 2. Current as of 28 July 2011. https://doi.org/10.1787/9789264069923-en
20. Organization for Economic Cooperation and Development. Test no. 211: Daphnia magna reproduction test. In: OECD guidelines for the testing of chemicals, section 2. Current as of 2 October 2012. https://doi.org/10.1787/9789264185203-en
21. Organization for Economic Cooperation and Development. Test no. 210: Fish, early-life stage toxicity test. In: OECD guidelines for the testing of chemicals, section 2. Current as of 26 July 2013. https://doi.org/10.1787/9789264203785-en
22. Organization for Economic Cooperation and Development. Test no. 209: Activated sludge, respiration inhibition test (carbon and ammonium oxidation). In: OECD guidelines for the testing of chemicals, section 2. Current as of 22 July 2010. https://doi.
org/10.1787/9789264070080-en
23. United States Environmental Protection Agency. Estimation Program Interface Suite™ v 4.11. Current as of 25 August 2022. https://www.epa.gov/tsca-screening-tools/download-epi-suitetm-estimationprogram-interface-v411
24. European Chemicals Agency. Guidance on information requirements and chemical safety assessment. Part C: PBT/vPvB assessment.Current as of 3 July 2017. http://dx.doi.org/10.2823/139408
25. European Medicines Agency. Questions and answers on Guideline on the environmental risk assessment of medicinal products for human use EMA/CHMP/SWP/44609/2010 Rev. 1. Current as of 26 May 2016. https://www.ema.europa.eu/en/documents/scientific-guideline/questions-answers-guideline-environmental-risk-assessmentmedicinal-products-human-use-revision-1_en.pdf
26. Food and Drug Administration. Determining the need for and content of environmental assessments for gene therapies, vectored vaccines, and related recombinant viral or microbial products. Current as of 16 May 2019. https://www.fda.gov/regulatory-information/search-fdaguidance-documents/determining-need-and-content-environmentalassessments-gene-therapies-vectored-vaccines-and-related
27. European Medicines Agency. Guideline on environmental risk assessments for medicinal products consisting of, or containing, genetically modified organisms (GMOs). EMEA/CHMP/BWP/473191/2006 – Corr. 2006. Current as of 11 December 2006. https://www.ema.europa.eu/en/environmental-risk-assessmentsmedicinal-products-containing-consisting-genetically-modified
28. Straub JO, Stewart KM. Deterministic and probablilistic acute-based environmental risk assessment for naproxen for Western Europe.Environ Tox Chem. 2007;26(4):795-806.
29. Lahti M, Oikari A. Microbial transformation of pharmaceuticals naproxen, bisprolol, and diclofenac in aerobic and anaerobic environments. Arch Environ Contam Toxicol. 2011;61:202-210.
30. Carballa M, et al. Determination of the solid-water distribution coefficient (Kd) for pharmaceuticals, estrogens, and musk fragrances in digested sludge. Wat Res. 2008;42(1-2):287-295.
31. Kwak K, et al. Chronic toxicity and endocrine disruption of naproxen in freshwater waterfleas and fish, and steroidogenic alteration using H295R cell assay. Chemosphere. 2018;204:156-162.
32. W Wang S, et al. Effects of pharmaceutically active compounds on a mixed microbial community originating from a municipal wastewater treatment plant. Environ Sci Technol. 2008;42(4):1091-1095.
声明:原英文原文版权归美国法规事务协会(RAPS)所有,本译文供参考,如有任何建议,请联系我们。
同写意媒体矩阵,欢迎关注↓↓↓