Keywords

The WHO Blue Books

Diagnosis and classification of individual cancers underpin treatment and care of cancer patients, as well as research into cancer epidemiology, prevention, diagnosis, and treatment. Traditionally, cancer classification has been based on consensus of histopathological opinion, with very limited consideration of more recent aspects as advances in molecular pathology or the use of evidence-based practices to inform decisions. However, new technologies are transforming the field rapidly, and it has become increasingly clear that the traditional approach to cancer classification is insufficient. The understanding of cancer at a molecular level, the advances in prognosis and prediction, as well as the fast development of methods to assess related evidence has now reached a point where this information must be included in the decision-making process of the classification and diagnoses of cancer. Other fields such as digital pathology and image analysis are also producing new insights, challenging many diagnostic criteria and showing an urgent need to integrate these facets of diagnosis into cancer classification internationally using an appropriate evidence-based approach. Also, the fast development of related scientific areas as omics or biotechnology has added pressure to the process of assessing the evidence in a timely manner, contributing extensively to the information overload in the field. More than one million papers pour into the PubMed database each year (about two papers per minute) (Landhuis 2016), challenging experts who are already overwhelmed by publishing and other time-eaters.

The International Agency for Research on Cancer (IARC) has been responsible for the World Health Organization (WHO) Classification of Tumours since its third edition and is in charge of updating the classification on a regular basis, the fifth edition of this classification currently being under production. Published as the WHO Blue Books and the WHO Blue Books online (Fig. 1) (WHO Classification of Tumours 2019), this is an essential resource for pathologists and cancer researchers worldwide. They provide the standards against which tumors are classified and are key in cancer diagnosis, while also supporting cancer research, treatment, and prognosis. Each book summarizes the characteristics of one tumor type, including diagnostic criteria, pathological features, and associated molecular alterations. One edition covers all organ sites in 14 volumes, describing and illustrating in a strictly disease-oriented manner each single tumor type to provide the international standards for diagnosis and cancer research.

Fig. 1
figure 1

The World Health Organization (WHO) Classification of Tumours (fifth edition), published as the WHO Blue Books and the WHO Blue Books online

Full size image

The fifth edition of the WHO Classification of Tumours (WCT) contains on average 300 tumor types per book and provides a definition for each type followed by a description of relevant features with the same structure for all. Aspects such as etiology, pathogenesis, epidemiology, clinical features, macroscopic appearances, histology, cytology, molecular pathology, essential and desirable diagnostic features, staging, prognostic factors, and predictive biomarkers are summarized, and all require a review of the scientific literature to assess available evidence. Sometimes conflictive decisions need to be taken, and it is most relevant to inform these decisions with the best available evidence. To retrieve this evidence and to evaluate and summarize it while minimizing the risk of introducing bias, evidence-based practice methods have to be followed, and for this reason, increasingly systematic review methods have been introduced in the processes informing the WCT. But performing systematic reviews for all is not feasible. If each section describing one aspect related to one single tumor type were to require at least one formal review, that would mean conducting more than 3000 reviews for a book, which would not be practicable and highly inefficient. The current approach therefore relies largely on literature searches for published articles performed by the same subject experts that contribute to the books, being based on their individual perceptions of need for the assessment of the content of these books. Yet, these decisions affect the classification and hence the diagnosis and management of cancer patients worldwide. So, to minimize the risk of including biased information, an editorial board composed of standing members and content experts (mainly practicing pathologists) takes the final decisions agreeing in structured meetings on definitions and core criteria for each tumor systematically for every new edition.

Adapting the Current Procedure to Information Overload

Overreliance on expert consensus at editorial board meetings may lead to problems, and IARC undertakes continuous efforts to understand how this may affect the WCT (Table 1). In a recent analytical exercise (Uttley et al. 2020), the WCT group together with international experts identified key elements of the process that need to be taken into account before drafting recommendations for improvement of the evidence base of the classification. They acknowledged the need to review evidence in order to assess all relevant issues in a WHO Blue Books but also realized that different expertise and inevitable time constraints of participating experts would result in highly variable results with a potential for studies relevant to decisions to be missed. Not all issues need to be assessed by systematic reviews (Collins and Fauser 2005; Ioannidis 2016), a background search often being sufficient to assess statements not relevant to classification issues. These were also the findings of an informal exploration of the initial WCT fifth edition books showing each book to contain over 100 subjects described as “unknown” and more than 200 labelled as “clinically not relevant,” often without any additional description of the section or references (WHO Classification of Tumours 2019). Also, sections such as etiology and pathogenesis are frequently described as “unknown,” yielding uncertainty whether an exhaustive search of the literature had been performed to affirm the notion. This does not necessarily indicate that systematic reviews should be routinely performed, considering the impracticality to produce the required numbers in time. Many statements are more related to background information provided for the classification of the tumor type and less so to decisions relevant to classification. Insisting on systematic approaches would probably not improve the evidence base of the classification per se, while relying on simple background searches for this type of assessment would give authors the chance to invest more efforts in pressing issues: topics that need to be assessed by systematic reviews due to their potential for controversy.

Table 1 Problem arising from consensus-based approach and proposed solutions by an evidence-based approach
Full size table

In the previously mentioned analytical exercise (Uttley et al. 2020), authors discussed also that relying on consensus of the editorial board without a structured and controlled process of evidence synthesis may include unintended bias from content experts and influence decisions. Certain studies could be given more value, overlooked, or misunderstood depending on the clinical or research background of the authors (Young 2009). Authors also pointed out that the representativeness of the expert panel was key to avoid skewed decisions that may not be fully informed by the best and most relevant evidence (Doust and Del Mar 2004; McKee et al. 1991) and that language skills and personal and cultural differences in expert panels may lead to discussions being dominated by individuals (Doust and Del Mar 2004; McKee et al. 1991; Kea and Sun 2015). And lastly, concerns about previous errors in referenced evidence being carried forward during the subsequent updates were discussed.

Controversies Pertaining to Evidence-Based Approaches in Pathology

The WCT has an important influence of incorporating noncommercial knowledge into the diagnosis for patient management, as well as repercussions in clinical diagnosis and management of cancer patients, which is why much effort is invested in providing the most accurate synthesis of the scientific evidence to inform all decisions relevant to the international standards which underpin the diagnosis of individual tumor types. To mitigate the previously described problems, a progressive addition of evidence-based practices to the editorial process of the WCT has been planned, as well as the promotion of evidence-based pathology and the performance of systematic reviews in the field to facilitate the rapid incorporation of best available evidence. Systematic reviews are widely regarded as the cornerstone of evidence-based medicine, hence the best available evidence to inform decisions, because they include comprehensive literature searching, transparency in methods, and rigorous evidence appraisal (Higgins 2008; Mulrow 1994). However, even if performing systematic reviews to assess certain aspects of a tumor type could improve the reliability of decisions taken by the editorial board, important limitations exist to the regular application of such methods. Training in methods and expertise is required to perform well a systematic review (Gotzsche and Ioannidis 2012; Piehl et al. 2003; Uttley and Montgomery 2017), they are laborious, time-consuming, and difficult to interpret, and these difficulties increase with evidence of low grade. Traditional systematic review methods are often rigid, and best practice guidelines such as Cochrane (Higgins 2008) and PRISMA (Moher et al. 2009) yield numbers closely aligned with meta-analytic reviews for medical interventions and methods that are not necessarily appropriate for research in pathology. Also, critique of the traditional hierarchy of evidence-based medicine with systematic reviews at the top has been registered among clinicians and pathologists for oversimplifying complex clinical diagnostical issues (Williams and Garner 2002; Parker 2005; Murad et al. 2016). Indeed, they are subject to influence if not conducted as objectively as any other research, and managing the influence and expertise of team members in a systematic review remains critical to producing reliable and externally valid conclusions (Uttley and Montgomery 2017).

How to Adapt Evidence Synthesis to the Needs of Pathologists?

Although clearly essential for any evidence synthesis in a controversial question, systematic review methods would need to be adapted to the specific needs of the WCT and its editorial process. Timely assessment of the evidence is one of the major problems for WCT, and this is something that traditional, comprehensive systematic reviews struggle to provide. Modern, alternative review methods including mapping, rapid, or scoping reviews could improve the quality of scientific reviews used to inform decisions in the WCT (Dobbins 2017; Khangura et al. 2012; Polisena et al. 2015; Moher et al. 2015), even though they do increase the risk of introducing bias and errors during the review process and may not always provide the expected results (Dobbins 2017; Khangura et al. 2012; Haby et al. 2016). Therefore, full systematic reviews of the available evidence should be—for the moment—limited to the assessment of research questions that directly affect the classification of tumor or may have major consequences in cancer diagnosis. And a focus on how to solve these challenges for the WCT editorial process is needed.

Additional problems detected during the editorial process are the low number of high-quality evidence synthesis published in the field of pathology and therefore the difficulties for the WCT authors to retrieve ready-to-use high-level evidence to prove their statements. Scientific production in the field of pathology is frequently at the lowest levels of the evidence pyramid, with publications as case reports, case series, or narrative reviews being common. This is well known and has been explained with motives as being performed by clinical “unfunded” faculty members, the predominantly observational nature of a pathologist’s work, or that anatomic pathology is subject to “interobserver variation” and has therefore an inherently higher proneness to error. Also, other quality issues are common as meta-analyses in diagnostic pathology showing less compliance with reporting recommendations such as PRISMA (Liu et al. 2017) or an underutilization of this method (Kinzler and Zhang 2015). Nowhere is this more apparent than in the field of molecular pathology, and it is there where most controversial topics arise with research groups suggesting different tumor classifications based on proposed molecular signatures not taking in consideration what type of evidence needs to be generated to prove such a relationship (Baxter 2003; González-Reymúndez and Vázquez 2020; Mamatjan et al. 2017). The issue of using validated methods for WCT is becoming increasingly important due to these rapid advances with more information becoming evident than from histopathological analysis alone as in former years. Systematic review methods may need to be adapted to assess this information, as some steps of a validated systematic review process may not meet current need, but evidence synthesis and evaluations are needed to assess how to incorporate single molecular advances into the WCT. Also, to be able to perform adequate systematic review studies in the field, standardized reporting recommendations such as the Standards for Reporting Diagnostic Accuracy Studies (STARD) (Cohen et al. 2016), Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) (Moons et al. 2015), and other standardized reporting guidance (www.equator.net) have to be followed by original research allowing for pooling and comparison of their results.

Plan of Action

Herein lies the challenge for the WHO Classification of Tumours: to develop a strategy that helps to move future editions acting at multiple levels toward a more evidence-based approach, where experts take decisions well informed by accurate synthesis of the best available evidence. Steps are already being taken to incorporate systematic and adapted review methods into the editorial process without compromising the much-needed expert consensus. The importance of systematic reviews and expert interpretation does not need to be viewed as mutually exclusive. One should not be seen as superior to another, as each represent necessary concepts of analysis and synthesis (Mickenautsch 2010) and evidence-based medicine has always stated that external evidence can inform, but never replace expertise (Sackett 1997). There is also an opportunity to ensure that research and evidence synthesis in the field of pathology and therefore for the WCT are fit for purpose without enforcing rigid guidelines, just addressing identified challenges in a coordinated manner. At the WCT, we propose an approach that is adapted to the needs of the WHO Blue Books which starts by requesting contributing authors to employ a series of “nonnegotiables” while performing the literature reviews that feed into key decisions for the classification. Table 2 shows the rationale and risk underlying each proposed “nonnegotiable.” It ends with the provision of training in systematic review methods to a broad spectrum of pathologists, oncologists, and researchers contributing with their work to the body of evidence consulted to classify tumors. Experts, who serve as authors and editors, require basic training on epidemiological essential knowledge to be able to critically appraise their findings and also methodological support in systematic review and literature searching methods that can be provided by the WCT secretariat. Other important challenges for the WCT such as the application of the differing evidence levels and considerations within the field of pathology are addressed by projects initiated by the WCT program, projects that search for consensus in the field to promote a comparable assessment of studies. Furthermore, a register is planned where topics in need for assessment are listed and prioritized based on WCT requirements and scientific considerations.

Table 2 Methodological nonnegotiables for systematic reviews for the purposes of tumor classification
Full size table

The IC3R Initiative

IARC and the WCT program have set ourselves the challenge to continuously revise our procedures to keep incorporating evidence-based practices into the current expert-led approach and follow continuous improvement principles for our editorial process. The privileged position of IARC in the field allowed us to initiate a representative international collaboration that joins actions to identify potential challenges and solutions for tumor classification and for cancer research in general: the International Collaboration for Cancer Classification and Research or in short, IC3R (https://ic3r.iarc.who.int/). This collaboration aims to provide a unique forum for coordinating evidence generation, for synthesis and evaluation, and for tumor classification and cancer research, as well as related issues as standard setting or developing best practice recommendations (Cree et al. 2021). IC3R facilitates the communication between the WCT editorial board and the multiple scientific communities producing the body of evidence used to inform decisions in the classification (Fig. 2). Member institutions include universities, research centers, and other interested parties that assign representatives to discuss and coordinate international efforts for the provision of concrete deliverables needed to set standards and to underpin the WHO Classification of Tumours and cancer research in general (Fig. 3).

Fig. 2
figure 2

Framework of the International Collaboration for Cancer Classification and Research (IC3R)

Full size image
Fig. 3
figure 3

Expected impact of the International Collaboration for Cancer Classification and Research (IC3R)

Full size image

The aims of the collaboration include—among others—the harmonization of cancer-related research; the production of evidence evaluations for clinical settings, cancer research, and epidemiology; the identification of research and information gaps; and the promotion of evidence-based pathology and evidence-based practice in related fields. All objectives align with evidence-based medicine principles that have already been successfully addressed in other specialized medical fields. For this reason, a project was proposed with the aim to reproduce successful examples of the promotion of evidence-based medicine in a medical field as the Cochrane Collaboration (https://www.cochrane.org/) or the Joanna Briggs Institute (https://jbi.global/).

The evidence-based pathology project of IC3R (Cree et al. 2021) is the first project to be initiated under the framework of the collaboration and addresses the challenges of producing an evidence-based tumor classification, including deliverables such as adapted evidence levels for the field, new systematic review methods, training in evidence-based pathology, systematic review tools, and the development of a network of reviewers and supervisors to further develop and spread evidence-based pathology. Such an evidence-based pathology movement, with sufficient resources and adequate coordination, could respond with evidence-based assessments to the predicted large number of decisions that will have to be addressed for tumor classification. And IC3R is ideally positioned to efficiently promote and coordinate such an endeavor, working with other partners in pathology and drawing on the expertise of the more widespread evidence-based movements. The project is planned as a multiple partner joint action, where partners such as the University of Ulm (Germany), the Cochrane Centre (Netherlands), and the Exeter Test Group of the University of Exeter (United Kingdom) contribute to different work packages with their distinct expertise. At the beginning, the focus will be on the design of new methodologies and the development of a training program for experts, to allow the construction of an extended network of evidence-based pathology hubs in subsequent phases.

Evi-Pat

The Evidence-Based Pathology Training Initiative (Evi-Pat) started working on the development of a new training program for this purpose in the summer of 2020 and developed a proposal for a new online training in evidence-based practice and systematic review methods especially developed for pathologists, oncologists, and experts in the field of cancer. This training focuses on adapting methods to the type of evidence used and produced in cancer-related practice and in teaching systematic review methods to experts in the field. Overcoming reluctance to the application of evidence-based medicine (EBM) principles and promoting the development of an evidence-based pathology movement among early career professionals in the field are further objectives of Evi-Pat. The collaborators are jointly developing a research project to design and produce this educational tool adapted to the times of COVID-19 restrictions with virtual and in-person elements and to compare its performance to an online-only experience in terms of knowledge transfer and user satisfaction. Also, the added value of such an adapted training program for pathology and oncology is meant to be evaluated, as well as the delivery of the module to pathologists and experts from the field of oncology.

This module will be the first deliverable of the work package on training from the evidence-based pathology project and will support effective teaching, research, and networking on evidence-based pathology and oncology for professionals in a manner that is resourceful and—hopefully—applicable in every place in the world with an Internet connection. Outcomes of the full research project will be used to develop a full evidence-based pathology training program over time and help to promote principles of evidence-based medicine in the field of pathology and oncology to better inform decisions of the WCT. Examples such as the Cochrane Collaboration (https://www.cochrane.org/) and the Campbell Collaboration (https://www.campbellcollaboration.org/) have already shown that a collaborative effort can improve the evidence base of decision-making in a whole field, and drawing on the expertise of the more widespread evidence-based movements is one of the impacts of promising predictors of this project.

Conclusion

We believe that an evidence-based approach to informing key decisions that feed into tumor classification will allow the WCT editorial board to mitigate the potential inclusion of biased decisions into the classification and also benefit authors by providing structured, transparent, and reliable methods for the synthesis of available evidence for each tumor type, as the same time as training to apply these methodologies. Our hope over time is that this approach will increase the rigor of the decisions feeding into WCT, by addressing critical questions and identifying research gaps, as well as reaching recommendations for research to inform future editions. Such an approach will maintain the reliability of tumor classification and help to provide solutions to challenges like the exponential rise in number of scientific publications and the need to manage new types of information such as evidence from genetics or big data.