Environmental risk assessment for human medicines

An environmental risk assessment for a human medicine in the European Union studies the medicine's potential effects on the environment and sets the precautions they call for. Whether the product is a genetically modified organism decides which guideline governs the assessment and when it is reviewed: a modified product is assessed nationally before a clinical trial and again centrally at marketing authorisation. Every product's assessment is filed in Module 1.6 of the marketing-authorisation dossier and shapes the storage and disposal wording of its labelling, and a medical writer can support its preparation.

Obligation, scope and legal basis

An environmental risk assessment (ERA) is the process by which the potential effects of a medicine on the environment are studied and the right precautions are taken. ^{1, 2} In the European Union it is a legal obligation: an ERA must accompany every new marketing-authorisation application, whatever the type of medicine. ^{1, 3, 2} Its aim is to identify the environmental risks and the ways to mitigate them at the user level. ⁴

The requirement applies to essentially every medicine, with some exceptions, and not only in the centralised procedure but in the other European procedures too. ¹ Radiopharmaceuticals carry additional requirements. ¹ Without an ERA a medicine is not approved. ^{5, 6} Appropriate details of the assessment are included in the European Public Assessment Report of an approved medicine, so that this information is available to the public. ²

Human medicines may enter the environment during their manufacture, use and disposal. ² Under Directive 2001/83/EC, the scope covers the risks arising from the use, storage and disposal of the medicine; the risks arising from its synthesis or manufacture lie outside the assessment. ^{7, 3} Those risks are expected to come into scope through an ongoing European Commission initiative. ⁷ The assessment is based on the use of the product and on the physico-chemical, ecotoxicological and fate properties, such as degradation and persistence, of its active substance. ² Its legal basis is Article 8(3) of Directive 2001/83/EC together with Directive 2001/18/EC. ³

Medicines reach the environment by various routes. ⁶ A highly toxic product, such as a chemotherapy agent or an advanced therapy used in a very specific setting, may be given to the patient only once or twice, in hospital, by staff and at very low concentration. ⁸ Medicines used daily, by contrast, end up in the environment: metformin, the most commonly used oral therapy for diabetes in Europe, is eliminated not only through household waste but also by being excreted unmetabolised, so it too has an environmental footprint. ^{8, 9}

Minimising the environmental impact of pharmaceuticals is part of Good Clinical Practice and is stated in the 11th principle of the Declaration of Helsinki. ³

GMO status and the applicable guideline

Medicines can be classified in many ways, but for an ERA an important classification is whether the product is a genetically modified organism (GMO). ¹⁰

Advanced therapy medicinal products (ATMPs), a wave of potentially curative treatments emerging as technology advances, fall into three categories: gene therapy products, cell therapy products and tissue-engineered products. ⁹ They typically contain genetic material that has been altered. ¹¹ An organism counts as a GMO when genetic material has been introduced into it in a way that does not occur naturally by mating or recombination. ¹¹ Not all ATMPs are GMOs, and the same product type can be a GMO in one product and not in another. ¹¹

The COVID-19 vaccines show where the line falls: the mRNA vaccines deliver only a piece of genetic material and are not GMOs, while the adenoviral-vector vaccine is based on a modified organism, is a GMO and therefore required an ERA. ¹² Altered genetic material does not necessarily make a product a GMO; a product is one when an organism such as a viral or bacterial vector is modified and introduced. ¹² So even within one product type or one disease, whether a product is a GMO is decided by looking at the product itself and how it is made. ^{12, 13}

Two separate guidelines apply: one for medicinal products for human use that are not genetically modified, and a distinct one for genetically modified organisms. ⁷ The guideline for products that are not genetically modified is EMEA/CHMP/SWP/4447/00, in its 2006 version, Corr2, ^{3, 14} and in its revision 1, a draft dated 15 November 2018. ^{15, 16, 14} The guideline for products consisting of or containing genetically modified organisms is EMEA/CHMP/BWP/473191/2006 – Corr. ^{15, 16} Gene therapy products have their own guideline on the scientific requirements for the assessment, EMEA/CHMP/GTWP/125491/2006. ¹⁵

Non-GMO assessment

Phases

The non-GMO assessment is a stepwise procedure, similar to that for veterinary medicinal products. ³ It is iterative. ¹⁷ Its stages are counted differently from one account to another. ^{3, 7, 15}

One account describes a two-phase procedure, in which the results at the end of Phase I decide whether Phase II is required. ³

Another counts essentially three steps. ⁷ Phase I estimates the predicted environmental exposure; if a certain threshold is met, the assessment enters a Phase IIA fate-and-effect analysis, and if that step identifies a risk, a Phase IIB refinement and extended assessment follows. ¹⁷ The studies proceed from microorganisms, daphnia and fish in water to water sediments and finally to soil, for terrestrial effects. ¹⁷

A third counts three phases. ¹⁵ Phase 1 evaluates the product's potential for bioaccumulation and persistence in the environment, its so-called environmental exposure. ¹⁵ Significant anticipated exposure, or a specific risk associated with the active compound (the "however clause"), takes the evaluation to Phase 2, which determines what happens to the product when it enters the environment, its "fate", and its potential effect on the organisms living in the ecosystem. ¹⁵ Identified risks take it to Phase 3, which refines and extends the assessment. ¹⁵ A negative Phase 3 result is not cause for rejecting the authorisation of a human medicine. ¹⁵

Phase I: environmental exposure

The active pharmaceutical ingredient is usually the parent compound. ³ The assessment rests on a "total residue approach", with two assumptions: that the body does not metabolise the active ingredient and excretes it as the parent compound, and that metabolites have similar or lower toxicity than the parent. ³ Phase I exposure is estimated from the active ingredient alone, not from its route of administration, pharmaceutical form, metabolism or excretion. ¹⁶

Phase I studies may include:

• risk assessments of the possibility that an organism in the environment is exposed to the active ingredient and that ecotoxicity occurs ¹⁶
• persistent, bioaccumulative and toxic assessments, which evaluate how far the ingredient degrades in the environment, accumulates in organisms and is toxic, and address the intrinsic properties that make long-term environmental risks unpredictable ¹⁶
• complete literature reviews. ¹⁶

Drug substances with an octanol–water partition coefficient (log Kow) above 4.5 are screened stepwise for persistence, bioaccumulation and toxicity, and at each phase, if certain thresholds or risks are met, testing continues to the next step. ¹⁸ The first calculation is the predicted environmental concentration (PEC). ¹⁸ Its formula looks intimidating, but the applicant supplies only one value, the estimated maximum daily dose consumed per inhabitant; every other value is a default set in the guideline, which also shows how to calculate the fraction of market penetration. ¹⁸ The defined daily dose values it uses are a rather fixed estimate. ¹⁹

If the PEC in surface water falls below 0.01 µg/L and no other environmental concerns are apparent, the product is assumed not to present a high risk for the environment, testing stops there and the ERA is complete; at or above that value, the drug substance enters Phase II. ^{19, 16} Where a substance is already known to affect the reproduction of vertebrates or lower animals at concentrations below that threshold, all the remaining steps of the guideline apply as if the PEC had exceeded 0.01 µg/L. ¹⁹ The cut-off does not apply to active ingredients, such as endocrine-disrupting chemicals, that disrupt reproduction in vertebrates. ¹⁶ Certain substances, such as endocrine-disrupting chemicals and antiparasitics, undergo Phase II whatever their Phase I outcome. ³

Phase II: fate and effects

Phase II consists of two tiers, A and B. ²⁰ Its studies of the active ingredient may cover physico-chemical properties, environmental fate, ecotoxicological effects and mechanism of action, and they address environmental risk for soil, ground and surface water, the functioning of sewage treatment plants, sediment and secondary poisoning of predators. ²⁰ In Tier A, predicted no-effect concentrations are calculated for surface water, ground water and microorganisms; if the ratio is below one, the ingredient is considered safe and no more testing is required, and if it is above one (above 0.1 for microorganisms), Tier B fate-and-effects testing is required. ²⁰

Phase II studies are experimental work performed by toxicology teams, not by the medical writer. ²¹ They should ideally follow predefined standards such as OECD or ISO protocols, although other options can also be used, and all of them must be GLP compliant, with all the data taken into account cumulatively. ²¹ Tier A covers adsorption, biodegradability, aquatic sediment and an algae growth-inhibition test, then daphnia, fish and sludge. ²¹ Tier B moves into terrestrial fate and effects in soil, including growth and reproduction. ^{21, 22} Each study follows one fixed protocol. ^{21, 22} If animal studies are conducted, they should implement the replacement, reduction and refinement principles of animal welfare in accordance with Directive 2010/63/EU, be GLP compliant and follow test guidelines issued by the OECD, the European Commission or comparable guidelines. ²⁰

Justification in place of studies

Natural substances, and substances with a log Kow at or below 4.5, can be spared the studies, although the guideline does not call this a waiver. ²² The applicant still files the Module 1.6 document but includes a justification of why the studies are not necessary, which for the second group rests on the log Kow value. ²² The justification is made at the time of the marketing-authorisation application. ²² An ERA can consist solely of a justification for not submitting studies. ³

Revision 1 of the guideline

The European Medicines Agency (EMA) published revision 1 of the guideline for products that are not genetically modified on 30 November 2018 for a six-month public consultation, with comments invited by 30 June 2019. ² The revision is based on a 2014 concept paper and on the work of an expert group led by the Safety Working Party of the EMA's human medicines committee, builds on the twelve years of experience since the original guideline was published, and aims to facilitate the work of both applicants and regulators in the interest of environmental protection. ²

The revision introduces a decision tree clarifying when assessment studies are required, and more detailed technical guidance for applicants to increase the consistency of assessments. ² One of its most notable changes is the introduction of the term endocrine active substances, to include all compounds that affect development or reproduction. ² It gives guidance on estimating the exposure of predators to pharmaceuticals through the food chain, called secondary poisoning, as well as directly through the environment. ² It also proposes to limit one laboratory test method, the OECD 308 environmental fate test, to certain categories of substance, which will reduce the testing burden on applicants. ² In the interest of animal welfare, the guideline encourages applicants to share data generated for the assessment, implementing the replacement, reduction and refinement principles in accordance with Directive 2010/63/EU, to avoid unnecessary repetition of studies. ²

GMO assessment

The GMO assessment is somewhat similar to the non-GMO one but is a more detailed iteration. ¹⁷ It was created because advanced therapies differ significantly from small or chemically synthesised molecules, so a more focused assessment was needed. ¹³ Its objective is to identify and assess the potential harmful effects of the GMO on humans, animals, plants, microorganisms and the environment at large. ^{13, 23}

Two directives regulate GMOs and together are referred to as the GMO framework. ^{24, 25} The contained-use directive is written around containment measures and covers products handled in a controlled environment such as a laboratory or hospital. ²⁵ The deliberate-release directive covers products with no specific containment measures, which are introduced directly into the environment, as when transgenic plants are sown in a field. ²⁵ Both were designed for transgenic plants and animals rather than for medicinal products, so applying them to advanced therapies is not a perfect fit. ^{25, 26}

Principles and information needed

The GMO assessment follows these principles:

• Compare the modified organism with its unmodified wild-type parent; this is the most important principle. ²³
• Base the assessment on scientific knowledge, the literature and studies conducted with the GMO itself. ²³
• Treat every product case by case, because even similar products can differ in important respects. ²³ An ERA prepared for one product cannot simply be reused for another without careful checking. ²⁷
• Re-evaluate whenever new information becomes available, from clinical trials or other studies, and check whether the risks have changed. ²³

The evaluation needs the characteristics of both the modified and the parental organism; how the GMO was made, meaning which genes were inserted, deleted or altered; how it is released into the environment, which for a medicine means how it is administered to patients; what interactions it might have with the environment; and the monitoring and control strategy together with emergency response plans. ²⁸

Steps of the analysis

The annex of the deliberate-release directive sets out the analysis in six main steps:

1. Identify each potential hazard that may cause adverse effects. ²⁹
2. Identify the potential consequence of each. ²⁹
3. Evaluate how likely it is to occur. ²⁹
4. Estimate the risk each adverse effect poses, and repeat this for every identified hazard. ²⁹
5. Add the layer of risk mitigation or risk management. ^{29, 30}
6. From all of these, determine the overall risk of the GMO. ³⁰

Clinical-trial stage

A GMO ERA is assessed twice: at the clinical trial authorisation application, by the member state, and at marketing authorisation, centrally by the EMA. ³⁰ Because all ATMPs are mandatory for the centralised procedure, the GMO assessment at marketing authorisation forms part of the general centralised assessment. ³¹ A GMO product needs an ERA before a trial can start, whereas a non-GMO product needs one at marketing authorisation even if nothing was done at trial stage. ⁵

The Clinical Trials Regulation harmonises the assessment of clinical trials across the EU through a portal, CTIS, that holds the trial data. ³¹ An application is reviewed by the competent authority for its core documents and by the ethics committee for its patient-related documents, such as patient information sheets, informed consent, data protection rules and other administrative forms. ³² The GMO assessment sits in neither part: responsibility for it stays with the member states, which adds a third review layer, by the environmental body responsible for ERA in each state, for products that consist of or contain GMOs. ^{32, 24} Running trials nationally was already cumbersome for developers, and although the Clinical Trials Regulation may make the procedure somewhat more standardised and streamlined, the separate assessment by a different competent authority adds hurdles to trial authorisation. ²⁴

The member states also implemented the two directives very differently, producing different documentation, review times, approval frameworks and procedures. ²⁶ Countries differ in whether they treat a given product under contained use, under deliberate release, or under either; in whether the GMO assessment may be submitted in parallel with the clinical trial application or must be completed before it; and in whether a dedicated biosafety agency handles GMO aspects or, as in Germany, the competent authority that takes the trial application also handles them. ^{33, 34}

Good practice documents and pre-written ERAs

To harmonise and streamline the process across the EU, the European Commission introduced a common application form and specific ERAs; these harmonised documents, called good practice documents, aim to increase transparency and harmonise the GMO assessment across countries. ³⁵ The specific ERAs were developed for particular categories of medicinal product, those on which more information is already available, and take the form of a generic pre-written ERA document that already contains the six-step assessment for the writer to refer to. ^{35, 36}

Qualifying is not automatic. ³⁶ Even where a product appears to fall into a covered category, the applicant has to check the conditions before relying on the pre-written document, for instance that there is no risk of formation of replication-competent viruses. ³⁶ The common application form carries a dedicated ERA section into which either the good practice document or the pre-written ERA is placed; a product that qualifies needs no risk assessment of its own, and where no good practice document is in place a complete ERA has to be prepared. ^{36, 37}

Where no specific ERA document can be used, consultation with an experienced biosafety officer or regulatory expert is highly recommended. ³⁸ Some member states require the GMO documentation in the local language. ³⁸ Some also require additional information, so the requirements on the website of each country to which the trial application will be submitted should be checked. ²⁷

Module 1.6 of the dossier

In the CTD, Module 1.6 holds all ERA information, whether the product is a GMO or not: Module 1.6.2 for GMOs and Module 1.6.1 for non-GMOs, or a waiver, depending on the product. ^{27, 39, 3} Guidelines exist for both, along with question-and-answer documents. ^{27, 39} Before the application is filed, scientific advice is always available, and its questions can cover the ERA itself, the precautionary and safety measures, and the use and disposal of the medicine. ³⁹

Module 1.6.2 for GMOs

Module 1.6.2 can build on the GMO assessment prepared for the trial application. ⁴⁰ Its key elements are a copy of the competent authorities' written consent from that stage, a common application form similar to the one already submitted, and a complete ERA, either the pre-written product-specific one, referred to rather than reproduced, or one prepared in full. ⁴⁰ The results of any investigations or studies performed with the GMO that could help the assessment can be attached, and further elements are copies of the proposed product information, the monitoring plan, and the Summary Notification Information Format, a very short version of the ERA that will be publicly available. ⁴¹ Module 1.6.2 has to stand alone, because during assessment it goes to the individual member states and each must be able to make its own assessment from it. ⁴¹

Module 1.6.1 for non-GMOs

For Module 1.6.1, a non-GMO product has nothing from the trial stage to build on, and there is a guidance document but no template, so a good medical writer can help create a suitable one. ⁴² Its key elements are the predicted environmental concentration estimate, the possible risks to the environment, the precautionary and safety measures to be taken in line with them, and a proposal for labelling. ⁴²

Structure of the ERA document

The ERA of a human medicine has a well-defined structure. ⁴³ The introduction identifies the active ingredient clearly: company name or code, IUPAC name, CAS number, empirical and structural formulae, SMILES code and molecular weight. ⁴³ If relevant, a rationale for the absence of environmental studies follows; otherwise the Phase I and/or Phase II studies are summarised as text and tables, as required, with the full study reports and references listed in the annex. ⁴³ The document must end with the author's dated signature, information on the author's education, training and professional experience, and a statement of the author's relationship with the applicant. ⁴³

Labelling and risk mitigation

On the basis of the assessment, the amount of drug released into the environment can be minimised through disposal under very strict conditions, through risk-minimisation activities, and through labelling that makes correct storage and disposal possible. ^{17, 8} The assessment has implications for the wording on storage and disposal in the Summary of Product Characteristics and the patient information leaflet. ⁶ The disposal procedure varies from country to country, even for a drug approved centrally in Europe. ⁴⁴

Where an assessment cannot exclude environmental risk, the applicant proposes risk mitigation strategies to minimise release of the medicine into the environment. ⁴⁵ The key strategy is clear disposal instructions, such as returning used patches, delivery devices and unused medicines to pharmacies or recycling centres with designated collection boxes, and other strategies present information on potential environmental risk, proper storage and use in the labelling and package leaflet. ⁴⁵ For aquatic toxicology and fate studies, sharing the analytical verification data on the active ingredient on the applicant's website or in a general database is "encouraged", so that those in water management can monitor substances of concern, though the quantity and quality of data sharing are debatable. ⁴⁵

Evaluation questions and updates after authorisation

During evaluation, the EMA may also raise comments or questions on the assessment. ⁴⁴ A minor comment is that the wrong guideline was followed, which is usually not a serious problem but will be pointed out. ⁴⁴ A more substantial one is that the EMA disagrees with the Phase I evaluation outcome and perhaps asks for environmental studies, or judges the bioaccumulation-potential assessment incorrect and asks for additional studies to support a claim of low environmental risk. ⁴⁴

After the initial marketing authorisation, which is valid for five years, the only legal requirement is to update the ERA for type II variations. ^{44, 46} A typical type II variation is a new indication, for instance a drug used as an oral tablet that will now also be used as a dermal patch in another disease, a change that may increase environmental exposure. ⁴⁶

Shortcomings and proposed remedies

The regulation has shortcomings in four areas:

• Harmonisation. Module 1.6 is a nation-specific chapter, so ERA requirements are not necessarily harmonised across the EU as other parts of a submission are. Despite improvements, discrepancies with other environmental assessment guidelines remain: the guideline is not harmonised with the Classification, Labelling and Packaging Regulation, and the predicted no-effect concentration and Environment Quality Standard approaches differ. ⁴⁷
• Scope. The assessment puts the onus for user-created risks to the environment on manufacturers, so that they evaluate benefit–risk and offer mitigation, but it does not look at manufacturing processes and their release of active ingredient and other chemicals; including an assessment of risk during manufacture would increase the document's relevance. ⁴⁷
• Coverage. The guidelines do not require testing of antibiotics' impact on diatoms, critical microscopic planktonic algae; because diatoms generate about 20% of the earth's oxygen each year, they should be part of environmental assessments as well. ⁴⁷
• Transparency. The public cannot access the complete assessments or the official reports based on them, since the law requires only the publication of public assessment reports, which need not contain information from the assessments; authorisation holders may also refuse disclosure by claiming commercially confidential information. ⁴⁷

It has been argued that under environmental information law, releasing pharmaceuticals into the environment is an "emission into the environment", which gives a legal basis for full public disclosure of assessments and their official reports. ⁴⁷ Sharing data on active ingredients through publicly accessible databases is proposed as a resource-saving solution, a precondition for detecting emerging environmental trends and risks early and for avoiding resource waste from unnecessary repetition of animal studies and loss of knowledge; the Swedish Pharmaceuticals and Environment database is an example of such an effort. ⁴⁷ Coordinated efforts by governments, regulators and pharmaceutical companies to promote and facilitate data sharing from assessments are critical for planetary health. ⁴

The medical writer's contribution

A medical or scientific writer working for a pharmaceutical company can write the assessment in collaboration with the scientists and toxicologists involved, who can oversee and review it, because writers have a strong understanding of the science and experience in turning documents into a structured, well-written report. ^{43, 4} Such teamwork can produce a well-rounded document for submission to the EMA. ⁴³

Medical writers can support and prepare the applications for a GMO using the right source documents or in consultation with the people who work with the product, such as those who hold the manufacturing information and, at the clinical site, those who know the control measures. ³⁸ Beyond the ERA section itself, the common application form has administrative information, information about the GMO and information about control measures that a writer can help prepare. ³⁷

Across the development path, a medical writer can:

• get involved in common application forms, including the national-level ERAs for GMOs ⁴⁶
• develop specific ERA templates with a proposal for how to present the data ⁴⁶
• know the guidelines and, at the trial-application stage, gather information from different sources, acting as a mediator as with clinical teams ⁴⁸
• help the team build a coherent story and ensure the claims in the conclusion are supported by the summarised data ⁴⁸
• keep in view the effect of Module 1.6 on the Summary of Product Characteristics, and the reverse ⁴⁸
• take part in timeline planning and management. ⁴⁸

Overall responsibility for Module 1.6 does not rest with the writer. ⁴⁸

Medical communicators can also communicate the findings of assessments to the public in plain language. ^{43, 4} One view holds that everyone in the profession, including those working in science communication, should feel empowered to educate their communities at least on the correct disposal of medicines. ^{48, 49} Ideally, anyone unsure should check the package leaflet and, if it is no longer to hand, return the medicine to the pharmacy. ⁴⁹