Ruth Zaslansky, Phlipp Baumbach, Winfried Meissner and the PAIN OUT collaborative
Dept. of Anaesthesiologyand Intensive Care, University Hospital Jena, Jena, Germany.

As an international venture, the PAIN OUT perioperative pain registry, offers a unique opportunity for evaluating patients’ pain responses to surgery across different countries. We compared patient reported outcomes (PROs) from a cohort of American patients after orthopedic surgery with patients’ reports from 13 countries (‘International’).1 American patients rated some PROs higher than the ‘International’ patients, specifically ‘worst pain since surgery’. However, for other PROs differences were not so apparent. This was puzzling. Were the differences in ‘worst pain’ assessments related to patient’s culture? Related to differences in care? Is this a methods question? ‘Worst pain’ is one of the most commonly used PROs in pain studies. Since this analysis, evaluations of other patient cohorts indicate that the country from which a patient comes, contributes very little to the variance of single PROs.

‘Culture’ is a difficult concept to define and to operationalize. Organizations such as the European Commission, United Nations and different European NGOs recommend that a person’s culture/ethnicity/race is included in studies.2 Effects of culture/ethnicity/race on responses of pain should not be ignored as there may be need to adapt assessment tools and/or treatments when they are applied in countries or to people who speak languages other than where they were first developed. Clinicians, researchers and regulatory authorities, should be aware about these ‘cultural’ components so they are taken into account when these stakeholders carry out their respective tasks.3,4
During the talk, we will describe and discuss findings from PAIN OUT and from the literature with the aim of proposing a framework for studies in pain.

References
1. Zaslansky R, Meissner W, Chapman CR. Pain after orthopaedic surgery: differences in patient reported outcomes in the United States vs internationally. An observational study from the PAIN OUT dataset. Br J Anaesth. 2018;120(4):790-797. doi:10.1016/j.bja.2017.11.109
2. European Commission, Directorate-General for Justice and Consumers, Farkas, L., Analysis and comparative review of equality data collection practices in the European Union : data collection in the field of ethnicity, Publications Office, 2020, https://data.europa.eu/doi/10.2838/447194
3. Sharma, S, Ferreira-Valente, A, de C. Williams, A.C., Abbott,JH , Pais-Ribeiro, J , Jensen, M.P , Group Differences Between Countries and Between Languages in Pain-Related Beliefs, Coping, and Catastrophizing in Chronic Pain: A Systematic Review, Pain Medicine, pnz373, https://doi.org/10.1093/pm/pnz373
4. Feng, Y., Herdman, M., van Nooten, F. et al. An exploration of differences between Japan and two European countries in the self-reporting and valuation of pain and discomfort on the EQ-5D. Qual Life Res 26, 2067–2078 (2017). https://doi.org/10.1007/s11136-017-1541-5

Chronic neuropathic and chronic pelvic pain syndromes consist of mixed groups of patients presenting with a variety of different symptoms and sensory signs. A simple but very meaningful approach for assessing the symptoms of patients is the use of Patient Reported Outcome Measures (PROMs), i.e. health outcomes directly reported by the patient. So far, clinical trials on treatment efficacy mainly rely on the patient’s ratings of spontaneous pain intensity as the primary outcome parameter. It is very likely, however, that other outcome measures might be additionally or even more important for the patients’ wellbeing, quality of life and cognitive and functional ability than spontaneous pain intensity sensations. In addition, trials might use different outcome domains and outcome measures making comparison of their results, for example in Cochrane analysis, very difficult. Within this presentation, recent results from systematic reviews aiming to provide evidence for the current use of outcome measures in clinical trials related to neuropathic and pelvic pain will be presented. Briefly, the results show that within studies of both pain aetiologies, several outcome domains in different combinations were assessed and several different measurement instruments were used for each outcome domain. This high heterogeneity between studies in both neuropathic and pelvic pain syndromes indicate why comparison of treatment response is rather difficult. Thus, official recommendations of a core set of PROMs that should be used as a minimum outcome set in clinical trials to assess and predict treatment effect is desperately needed.

Winfried Meissner on behalf of the PAIN OUT collaborative
Dept. of Anaesthesiology and Intensive Care, University Hospital Jena, Jena, Germany

Registry data offer both opportunities and risks when used in clinical pain studies. On the one hand, they have the potential to depict everyday clinical practice, are often based on large case numbers and can therefore also be used to analyze rare events (e.g. complications). On the other hand, registry data are often collected under less well-controlled conditions, data quality and completeness can pose challenges. Especially in the case of patient-reported outcomes (PROs), we know that without standardized survey conditions, the validity of PROs is often insufficient – for example, pain-associated PROs from routine clinical documentation are often of little use. When analyzing register data, it is also important to consider the purpose for which the register is used. For example, data used for reimbursement or quality assurance may be subject to bias. Furthermore, PROs from registries should not always be regarded as dependent variables of an intervention, so it must be carefully examined whether a causal chain between interventions and PROs can be assumed.

Data analysis from large registers often requires advanced analysis methods and extensive adjustment to correct any co-variables. At the same time, however, they offer analysis options that go beyond conventional RCTs, e.g. by applying AI methods to large amounts of data. Quasi-experimental analyzes are also possible, for example by using the variance in clinical practice for analyzing associations with PROs.

Placebo responses in clinical trials have been reported to grow over the last decades, and for pain are among the highest reported. This impedes successful trials, as it reduces the potential space for new drugs to show efficacy over placebo. Understanding, and ideally managing, the factors contributing to analgesic placebo responses in clinical trials is therefor crucial for brining new pain medications forward. Here, multiple possible factors for increased placebo analgesia response will be discussed, using mixed methods of theoretical modelling, systematic literature reviews and meta-analyses, data mining approaches in large cross-company datasets from international consortia like TransCelerate, IMI Europain, and IMI PainCare, as well as individual patient phenotyping and genotyping data from single clinical trials. Both aspects of trial design, like collation of outcome measures and trial inclusion criteria, as well as individual patient aspects like demographic factors and potential genetic risk factors will be discussed.

The term nociplastic pain was adapted by the International Association for the Study of Pain (IASP) in 2017, as a third mechanistic descriptor in addition to nociceptive and neuropathic pain. Nociplastic pain is defined as “pain that arises from altered nociception” that cannot be explained by nociceptive or neuropathic pain mechanisms. The terminology harmonizes with the International Classification of Diseases (ICD-11) where nociplastic pain conditions are classified as “chronic primary pain”, highlighting the conception of nociplastic pain as a disease in it’s own right, to contrast with “secondary chronic pain”, e.g. nociceptive and neuropathic pain conditions, where pain is a symptom. However, as patients often suffer from a combination of several pain types, clinically useful criteria are needed to facilitate the determination of pain mechanisms. To that end, the IASP terminology task force developed clinical criteria and a grading system for nociplastic pain affecting the musculoskeletal system. The principles of the classification algorithm will be presented, including hypersensitivity phenomena and co-morbidities. The usefulness of the criteria will be demonstrated by carefully guiding the audience through the algorithm using relevant patient examples, with special emphasis on patients simultaneously presenting with several pain types that need to be distinguished in order to provide correct treatment and facilitate relevant follow-ups. The strengths and pitfalls regarding the concept of nociplastic pain will be discussed from a clinical and research perspective.

Machine learning methods and artificial intelligence are gaining increasing popularity in all fields of research and practice, including medicine. White box models, like linear/logistic regression and decision trees, possess great interpretability however generally lack high-end precision. Black box models, like neural networks/deep learning yield great predictability, however it is unclear and nontransparent how results were obtained. So far, in medicine white box models are still used to a much greater extent, due to the enormous importance of explainability in medical settings. The new research field of XAI – Explainable Artificial Intelligence – is dedicated to make results of powerful and precise black box models more explainable, so that these models can be used in a wider range of applications and can also be applied in highly regulated environments like the health care or finance sector.

The difference and importance of both, global and local explainability, is demonstrated specifically for the field of medicine. Different methods like LIME, ICE, PDP, VI and SHAP will be presented to make results of black box machine learning methods more explainable. The usage and interpretation of the presented tools will be illustrated at a specific pain trial setting.

The growing appreciation of the importance of the gut microbiome in health and disease is revolutionizing many fields of clinical medicine. In recent years, evidence is mounting that the gut microbiome also plays a role in chronic pain. In fibromyalgia, the composition of the gut microbiome of affected women shows significant alterations as compared to healthy controls, as certain bacterial taxa are found at either increased- or decreased abundance. Furthermore, differences in the metabolic activity of gut bacteria is associated with altered circulating metabolic end-products in the blood of women with fibromyalgia including short-chain fatty acids and secondary bile acids. To investigate the possible causal role of the gut microbiome in nociplastic pain, a mouse model was used: Microbiome transplantation from women with fibromyalgia, but not from healthy controls, to germ-free mice led to pain hypersensitivity. Taken together, gut microbiome composition is compositionally associated with fibromyalgia, its function is altered in women with fibromyalgia and finally, it seems to play a causal role in the pathogenesis of the syndrome. These recent observations pave the way for the development of much needed new diagnostic and therapeutic modalities for fibromyalgia and possibly other forms of chronic pain, which will be discussed.