The role of patient-reported outcome measures in trials of artificial intelligence health technologies: a systematic evaluation of ClinicalTrials.gov records (1997-2022)

The extent to which patient-reported outcome measures (PROMs) are used in clinical trials for artificial intelligence (AI) technologies is unknown. In this systematic evaluation, we aim to establish how PROMs are being used to assess AI health technologies. We searched ClinicalTrials.gov for interventional trials registered from inception to Sept 20, 2022, and included trials that tested an AI health technology. We excluded observational studies, patient registries, and expanded access reports. We extracted data regarding the form, function, and intended use population of the AI health technology, in addition to the PROMs used and whether PROMs were incorporated as an input or output in the AI model. The search identified 2958 trials, of which 627 were included in the analysis. 152 (24%) of the included trials used one or more PROM, visual analogue scale, patient-reported experience measure, or usability measure as a trial endpoint. The type of AI health technologies used by these trials included AI-enabled smart devices, clinical decision support systems, and chatbots. The number of clinical trials of AI health technologies registered on ClinicalTrials.gov and the proportion of trials that used PROMs increased from registry inception to 2022. The most common clinical areas AI health technologies were designed for were digestive system health for non-PROM trials and musculoskeletal health (followed by mental and behavioural health) for PROM trials, with PROMs commonly used in clinical areas for which assessment of health-related quality of life and symptom burden is particularly important. Additionally, AI-enabled smart devices were the most common applications tested in trials that used at least one PROM. 24 trials tested AI models that captured PROM data as an input for the AI model. PROM use in clinical trials of AI health technologies falls behind PROM use in all clinical trials. Trial records having inadequate detail regarding the PROMs used or the type of AI health technology tested was a limitation of this systematic evaluation and might have contributed to inaccuracies in the data synthesised. Overall, the use of PROMs in the function and assessment of AI health technologies is not only possible, but is a powerful way of showing that, even in the most technologically advanced health-care systems, patients' perspectives remain central.

Embedding patient-reported outcomes at the heart of artificial intelligence health-care technologies

Integration of patient-reported outcome measures (PROMs) in artificial intelligence (AI) studies is a critical part of the humanisation of AI for health. It allows AI technologies to incorporate patients' own views of their symptoms and predict outcomes, reflecting a more holistic picture of health and wellbeing and ultimately helping patients and clinicians to make the best health-care decisions together. By positioning patient-reported outcomes (PROs) as a model input or output we propose a framework to embed PROMs within the function and evaluation of AI health care. However, the integration of PROs in AI systems presents several challenges. These challenges include (1) fragmentation of PRO data collection; (2) validation of AI systems trained and validated against clinician performance, rather than outcome data; (3) scarcity of large-scale PRO datasets; (4) inadequate selection of PROMs for the target population and inadequate infrastructure for collecting PROs; and (5) clinicians might not recognise the value of PROs and therefore not prioritise their adoption; and (6) studies involving PRO or AI frequently present suboptimal design. Notwithstanding these challenges, we propose considerations for the inclusion of PROs in AI health-care technologies to avoid promoting survival at the expense of wellbeing.

AB0836 Micronutrients deficiencies in Enteropathic Spondyloarthritis: the interplay with metabolism and HLA-B27 in disease phenotype

Background

Micronutrients play immunomodulatory roles through interactions with innate and adaptive immunity influencing the pro/anti-inflammatory balance. Serum levels depend on multiple factors such as gender, nutrition, and gut microbiome. Micronutrient deficiencies (MNDs) are associated with a pro-inflammatory status and co-morbidities in patients with chronic inflammatory diseases. No studies focused on potential interplay between MNDs and disease phenotype in Enteropathic Spondyloarthritis (ESpA) in which the combination of SpA with inflammatory bowel diseases (IBD) might dramatically affect micronutrients status.

Objectives

We analysed the occurrence of anemia (Hb ≤12 g/dl) and deficiencies of ferritin (Fe, ≤15 ng/dL), vitamin D [25(OH)D, ≤20 ng/ml], vitamin B12 (VB12, ≤200 pg/ml), and folic acid (FA, ≤ 4ng/mL) in ESpA patients. The interplay of MNDs with gender, metabolic parameters, HLA-B27 susceptibility, type of SpA and IBD, disease activity, and treatments was also explored.

Methods

Selection criteria of this cross-sectional descriptive study consisted of having a diagnosis of ESpA, regardless of its type, in an age of ESpA onset ≥18 and ≤80 years, among patients who were admitted to a combined Gastro-Intestinal and RHEumatologic “GI–Rhe” clinic (Tor Vergata University Hospital, Rome, Italy). Exclusion criteria were represented by active IBD, pregnancy or lactation, kidney and/or liver failure, alcohol abuse, neoplasia, ongoing supplementations. SpA disease activity was assessed by ASDAS-CRP and functional status by HAQ-S. All the enrolled patients underwent blood chemistry analysis to determine parameters including CRP, uric acid (SUA), Fe, 25(OH)D, VB12, FA, and HLA-B27 typing.

Results

We included 164 patients comprising 109 females and 55 age-matched males. A diagnosis of Crohn’s Disease (CD) occurred in 70% of patients while Ulcerative Colitis (UC) represented a third of the cohort. Peripheral (per) and axial (ax) SpA were equally distributed (50%vs50%). Patients with ax-SpA displayed a greater prevalence of CD than UC (p=0.02) while UC was prevalent in per-SpA (p=0.02). B27+ was revealed in 19% of our cohort: B27+ patients had a higher prevalence of ax-SpA (p=0.016) and a more severe disease activity (p=0.02) than B27-. Moreover, B27 positivity and uveitis were prevalent in ax-SpA compared to per-SpA (p=0.009 and p=0.01, respectively). According to univariate analysis, males showed higher SUA (P=0.004) and BMI (p=0.03) than females. Conversely, females showed a higher prevalence of anemia than males (p=0.002). A third of ESpA cohort showed FA (31.6%) and 25(OH)D (27.8%) deficiency while VB12 defect was less frequent (18.2%) and was registered almost entirely in B27- ESpA (p=0.02). CD-ESpA showed a lower mean VB12 (p=0.04) and a higher prevalence of ocular/skin co-morbidities (p=0.02) and ax-SpA (p=0.04) than UC-ESpA. Accordingly, CD-ESpA were on bDMARDs more than UC-ESpA (p=0.04).

Conclusion

Our findings document that FA and 25(OH)D deficiencies represent the main MNDs among ESpA patients while VB12 seems to be deficient mostly in patients with CD and almost exclusively in B27- patients. Otherwise, B27+ in ESpA results to be associated mainly with disease phenotype and treatments. In ESpA, the gender of patients appears to impact principally on dysmetabolism highlighting the role for nutritional interventions particularly in males. The interplay of MNDs with B27 and dysmetabolism in ESpA deserves further investigations also taking into account CD/UC localization and behavior.

References

[1]Park YE, et al. Incidence and risk factors of micronutrient deficiency in patients with IBD and intestinal Behçet’s disease: folate, vitamin B12, 25-OH-vitamin D, and ferritin. BMC Gastroenterol. 2021;21(1):32.doi: 10.1186/s12876-021-01609-8.

[2]Conigliaro P, et al. Impact of a multidisciplinary approach in enteropathic spondyloarthritis patients. Autoimmun Rev. 2016;15(2):184-90.doi: 10.1016/j.autrev.2015.11.002.

Disclosure of Interests

None declared

A rare cause of acute gastrointestinal hemorrhage: ileal lipoma Case report

Primary small bowel tumors account for 1-6% of all Gastrointestinal (G.I.) tract malignancies. Among these pedunculated lipomas are unusual. We report a case of a 66-year-old male with a history of G.I. hemorrhage and acute anemia, negative upper and lower endoscopies and a pedunculated lipoma in distal ileum, revealed by CT enterography. The patient was successfully treated by open surgery. Patients with G.I. hemorrhage and negative upper and lower endoscopies need an accurate evaluation of small bowel. Multislice CT enterography or Magnetic Resonance enteroclysis/ enterography represent the fastest and more accurate tools to obtain an exhaustive evaluation of small bowel. In case of small bowel tumors this diagnostic procedures can show site and stage and can even suggest histological type of such neoplasms, with a significant impact in the surgical planning, avoiding time consuming surgical exploration. In this patient multislice TC enterography allowed a correct diagnosis of benign lipoma due to its radiological density and absence of infiltration of the intestinal wall and surrounding tissues.

KEY WORDS

CT enterography, Gastrointestinal hemorrhage, Lipoma, Small bowel.

A new enrichment model for high sensitivity detection and downstream analyses of circulating tumor cells in breast cancer patients

Abstract #4162

The detection of circulating tumor cells (CTCs) in breast cancer patients have the potential to improve prognostication and the monitoring of response to treatment. Most CTC enrichment technologies are based on binding to anti-EpCAM antibodies. The sensitivity of such assays is limited by tumors that express no or undetectable levels of EpCAM. Improvements in CTC detection coupled with the development of systems to interrogate CTCs for therapeutic target expression could lead to novel applications for patient monitoring, clinical diagnosis and treatment. In this study, we describe a sensitive and reproducible enrichment method for CTCs. We defined cells as circulating tumor cells with three criteria: Positive for cytokeratin (CK+) and DAPI (nuclear) (DAPI+) and negative staining for CD45 (CD45-). We have previously reported that this system has a higher sensitivity for circulating tumor cell detection and provides a better platform for CTC downstream analyses compare to the methods currently available in the market. Herein, we describe the use of this platform for the evaluation of breast cancer biomarkers in CTCs. Blood samples from patients with metastatic breast cancer were used for CellSearch™ assay (Veridex , LLC ) and our CTC assay (A1000 CTC enrichment and detection kit, Genetix). We performed the CTC enrichment assay using the combination of anti-CK and anti-EpCAM antibodies. CTCs were identified with brightfield and fluorescence labeled anti-CK, anti-CD45 and DAPI (nuclear stain) images. The Ariol® system (Applied Imaging Corporation) was employed for automated cell image capture and analysis of CTCs on glass slides. CTCs enriched on the glass slides were used for CTC downstream analysis. Our CTC enrichment model is designed to have the capability to enrich all the three types of CTCs including CK+ & EpCAM+, CK+ & EpCAM-/low and CK-/low & EpCAM+ cells. Compared to the enrichment methods using anti-EpCAM or anti-cytokeratin antibody alone, a higher CTC detection rate and a larger dynamic CTC detected range were obtained with our new enrichment model. Interestingly there were clear CTC number differences with enrichment methods in the higher CTC count patient samples which indicate that the different enrichment methods may enrich different types of CTCs from patient blood samples.
 Results of DNA and RNA FISH analyses on enriched CTCs indicate that the CTCs on glass slides can be used for its downstream analyses directly or indirectly. Our method may have better performance on enrichment of heterogeneous CTCs and provide a better platform for CTCs profiling for biomarker evaluations and CTC downstream analyses.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 4162.

Effect of MT1 melatonin receptor deletion on melatonin-mediated phase shift of circadian rhythms in the C57BL/6 mouse

In the mouse suprachiasmatic nucleus (SCN), melatonin activates MT1 and MT2 G-protein coupled receptors, which are involved primarily in inhibition of neuronal firing and phase shift of circadian rhythms. This study investigated the ability of melatonin to phase shift circadian rhythms in wild type (WT) and MT1 melatonin receptor knockout (KO) C57BL/6 mice. In WT mice, melatonin (90 microg/mouse, s.c.) administered at circadian time 10 (CT10; CT12 onset of activity) significantly phase advanced the onset of the circadian activity rhythm (0.60 +/- 0.09 hr, n = 41) when compared with vehicle treated controls (-0.02 +/- 0.07 hr, n = 28) (P < 0.001). In contrast, C57 MT1KO mice treated with melatonin did not phase shift circadian activity rhythms (-0.10 +/- 0.12 hr, n = 42) when compared with vehicle treated mice (-0.12 +/- 0.07 hr, n = 43). Similarly, in the C57 MT1KO mouse melatonin did not accelerate re-entrainment to a new dark onset after an abrupt advance of the dark cycle. In contrast, melatonin (3 and 10 pm) significantly phase advanced circadian rhythm of neuronal firing in SCN brain slices independent of genotype with an identical maximal shift at 10 pm (C57 WT: 3.61 +/- 0.38 hr, n = 3; C57 MT(1)KO: 3.45 +/- 0.11 hr, n = 4). Taken together, these results suggest that melatonin-mediated phase advances of circadian rhythms of neuronal firing in the SCN in vitro may involve activation of the MT2 receptor while in vivo activation of the MT1 and possibly the MT2 receptor may be necessary for the expression of melatonin-mediated phase shifts of overt circadian activity rhythms.

Primary and secondary structure of 26S ribosomal RNA of Oenothera mitochondria

The primary structure of 26S ribosomal RNA from mitochondria of the dicotyledoneous plant Oenothera berteriana is inferred from the sequence of a cloned rDNA restriction fragment. A tentative secondary structure model valid for Oenothera and for the major part of maize mitochondrial 26S rRNA has been constructed in analogy to the refined german model for E. coli L-rRNA (Maly and Brimacombe 1983). The derived structure generally matches the eubacterial model providing further support to the E. coli consensus structure. Some structural features however show eukaryotic characteristics. Possible interactions between L-rRNA, 5S rRNA and initiator-tRNA are discussed.