O-220 An annotation-free embryo scoring system (iDAScore®) based on deep learning shows high performance for pregnancy prediction after single-vitrified blastocyst transfer

Study question

How is the performance of an automated embryo scoring system for pregnancy prediction after single-vitrified blastocyst transfer (SVBT) compared to other, annotation-dependent blastocyst grading systems?

Summary answer

Automatic embryo ranking by iDAScore shows a higher or equal performance, with regards to pregnancy prediction after SVBT, compared to manual, annotation-dependent grading systems.

What is known already

Blastocyst viability can be assessed by blastocyst morphology grades and/or morphokinetic parameters. However, morphological and morphokinetic embryo assessment is prone to both inter- and intra-observer variation. Recently, embryo ranking models have been developed based on artificial intelligence (AI) and deep learning. Such models rank embryos according to their potential for pregnancy only based on images and do not require any user-dependent annotation. So far, no study has independently assessed the performance of AI models compared to other embryo scoring models, including traditional morphological grading.

Study design, size, duration

A total of 3,014 SVBT cycles were retrospectively analysed. Embryos were stratified according to SART age groups. The quality and scoring of embryos were assessed by iDAScore v1.0 (iDAS, Vitrolife, Sweden), KIDScoreTM D5 v3 (KS; Vitrolife), and Gardner criteria. The performance of the pregnancy prediction for each embryo scoring model was compared using the area under curve (AUC) of the receiver operating characteristic curve for each maternal age group.

Participants/materials, setting, methods

Embryos were cultured in the EmbryoScope+ and EmbryoScopeFlex (Vitrolife). iDAS was automatically calculated using the iDAScore model running on the EmbryoViewer (Vitrolife). KS was calculated in EmbryoViewer after annotation of the required parameters. ICM and TE were annotated according to the Gardner criteria. The degree of expansion in all blastocysts was Grade 4 due to our freezing policy. Furthermore, Gardner’s scores were stratified into four grades (Excellent: AA, Good: AB BA, Fair: BB, Poor: others).

Main results and the role of chance

The AUCs of the < 35 years age group (n = 389) for pregnancy prediction were 0.72 for iDAS, 0.66 for KS and 0.64 for Gardner criteria. The AUC of iDAS was significantly higher (P < 0.05) compared to the other two models. For the 35–37 years age group (n = 514) the AUCs were 0.68, 0.68, and 0.65 for iDAS, KS and Gardner, respectively, and were not significantly different. The AUCs of the 38–40 years age group (n = 796) were 0.67 for iDAS, 0.65 for KS and 0.64 for Gardner criteria and where was not significantly different. The AUCs of the 41–42 years age group (n = 636) were 0.66, 0.66, and 0.63 for iDAS, KS and Gardner, respectively, and there was no significant difference among the pregnancy prediction models. For the > 42 years age group (n = 389) AUCs were 0.76 for iDAS, 0.75 for KS and 0.75 for Gardner criteria and not significantly different. Thus, for all age groups, iDAS was either highest or equal to the highest AUC, although a significant difference was only observed for the youngest age group.

Limitations, reasons for caution

In this study, SVBT was performed after minimal stimulation and natural cycle in vitro fertilisation (IVF). Therefore, we had only few cycles with elective blastocyst transfer. However, there was also no bias in selecting the embryos for SVBT.

Wider implications of the findings

Our results showed that objective embryo assessment by a completely automatic and annotation-free model, iDAScore, does perform as good or even better than more traditional embryo assessment or an annotation-dependent ranking tool. iDAS could be an optimal pregnancy prediction model after SVBT, especially in young and advanced age patients.

Trial registration number

not applicable

Clinical benefits of precision medicine in treating solid cancers: European Society of Medical Oncology-Magnitude of Clinical Benefit Scale score-based analysis

Background

Precision and matched cancer medicine has the potential to complement the existing biomarker approaches in cancer treatment. However, despite their promising potential, certain negative results have highlighted their limitations in molecular biology-driven treatment strategies. This study aimed to evaluate the clinical benefits of precision therapies.

Materials and methods

Three reviewers independently identified and assessed precision and matched cancer treatment studies published between January 2015 and December 2020. Clinical benefits of the treatments included in our cohort were assessed using two established frameworks; the European Society of Medical Oncology-Magnitude of Clinical Benefit Scale version 1.1 (ESMO-MCBS) and the American Society of Clinical Oncology Value Framework.

Results

Of the 290 eligible studies, 130 were for lung cancer, 51 for solid tumors, 24 for melanoma, and 24 for breast cancer. The common targets were: epidermal growth factor receptor (N = 66), serine/threonine-protein kinase B-Raf (N = 40), anaplastic lymphoma kinase (ALK) (N = 34), breast cancer protein (N = 26), phosphatidylinositol-3 kinase/protein kinase B/phosphatase and tensin homolog (PI3K/AKT/PTEN) pathway (N = 19), receptor tyrosine-protein kinase erbB-2 (HER2) (N = 19), mitogen-activated protein kinase (RAS/RAF/MAPK) pathway (N = 18), programmed death-ligand 1 (N = 12), fibroblast growth factor receptor (N = 8), and others (N = 43). The ESMO-MCBS scales ranged from 0 to 4. Based on the clinical benefit values, tumor mutational burden/mismatch repair-deficient/microsatellite instability-high for immunotherapy, anaplastic lymphoma kinase, and neurotrophic receptor tyrosine kinase therapeutic targets were considered high, whereas RAS/RAF/MAPK and PI3K/AKT/PTEN were considered low. Additionally, we found a significant difference between each average score (P < 0.001).

Conclusions

This study showed that precision and matched cancer therapies require further improvement. This is consistent with the views of the tumor board and of clinicians that precision strategies need to be revised to improve their therapeutic effects.

•

Of the 290 eligible studies, 130 were for lung cancer, 51 for solid tumors, 24 for melanoma, and 24 for breast cancer.

•

Precision strategies need to be revised to improve their therapeutic effects.

•

Therapeutic targets need to be narrowed down to improve the efficacy of precision medicine in the clinical setting.

Prognostic impacts of prehospital age shock index in patients with acute myocardial infarction

Introduction

Early identification of high-risk patients is the cornerstone of managing patients with acute myocardial infarction (AMI). Age Shock index (ASI; age multiplied by the ratio of heart rate/systolic blood pressure) has been reported to be similar to Global Registry of Acute Coronary Events (GRACE) risk score for predicting mortality in patients with AMI. However, prognostic impacts of prehospital ASI (pre-ASI) in patients with AMI remain unknown.

Methods

We analyzed of 2578 AMI patients who underwent emergency primary percutaneous coronary intervention (PCI) from January 2013 to March 2018, using data from Mie ACS Registry, a prospective and multicenter registry in Japan. Pre-ASI was recorded by emergency medical services at the first contact with the patient before admission, and in-hospital ASI (in-ASI) was recorded prior to PCI at admission. The primary end point was defined as all-cause death.

Results

Median follow-up duration was 753 days (497–838 days). All-cause death was observed in 230 (8.9%) patients. The ROC-AUC (Receiver operating characteristic-area under the curve) of pre-ASI for all- cause death was 0.76 (p&lt;0.001), which was similar to that of in-ASI (0.78, p&lt;0.001) (p=0.25 for pre-ASI versus in-ASI). The cut-off value for pre-ASI and in-ASI was for the prediction of all-cause death was both 45 with a sensitivity of 0.66 and a specificity of 0.78, with a sensitivity of 0.68 and a specificity of 0.76 respectively. According to the Kaplan-Meier survival analysis by combination of pre-ASI≥45 and in-ASI≥45, the patients with pre-ASI≥45 and in-ASI≥45 showed significantly higher all-cause mortality compared to the patients with pre-ASI≥45 and in-ASI&lt;45, the patients with pre-ASI&lt;45 and in-ASI≥45, and the patients with pre-ASI&lt;45 and in-ASI&lt;45 (p&lt;0.001) (Figure). The addition of pre-ASI≥45 to in-ASI≥45 (global chi-squared score: 205) resulted in a significantly increased global chi-squared score, suggesting the incremental prognostic value of pre-ASI (267; p&lt;0.001). Multivariate cox proportional hazard regression analysis for all-cause mortality demonstrated pre-ASI≥45 was a significant independent predictor (HR: 4.86; 95% CI: 3.36 to 7.02, p&lt;0.001). It was strongest predictor compared to left ventricular ejection fraction&lt;40% (HR: 2.45; 95% CI 1.67 to 3.58, p&lt;0.001), hemodialysis (HR: 3.45; 95% CI 1.66 to 7.17, p=0.001), door to balloon time&gt;90 minutes (HR: 1.66; 95% CI 1.18 to 2.34, p=0.004).

Conclusions

High pre-ASI predict increase mortality and assessment of both high pre-ASI and high in-ASI enhance risk stratification in patients with AMI. Early recognizing high pre-ASI may help us make better strategies and improve prognosis for high-risk AMI patients.

Figure 1

Funding Acknowledgement

Type of funding source: None

Premature acute coronary syndrome patients do not have a better prognosis for their age than mature ACS patients by propensity score match analysis

Background

Acute coronary syndrome (ACS) is the most important cardiovascular (CV) disease with a prevalence that increases with age. There is no data which compared the prognosis with premature ACS and mature ACS using propensity score matched analysis

Purpose

The purpose of this study was to compare the prognosis of premature ACS patients and mature ACS patients using propensity score matched analysis.

Methods

We analyzed of 4249 ACS patients (69.1±12.6, male 77%) including 773 premature ACS patients (50.1±6.8, male 78%) and 3476 mature ACS (73.3±9.3, male 77%) from January 2013 to December 2018, using data from Mie ACS Registry, a prospective and multicenter registry in Japan.

Premature onset of ACS was defined as younger than 65 years old in male and 55 years old in female.

Primary end point was as major adverse cardiac event (MACE) including cardiovascular death, non-fetal myocardial infarction, heart failure requiring admission and unstable angina.

Results

During median follow duration of 742 days ranging from409 to 828 days, 502 MACE were occurred. Premature ACS patients were younger and showed higher body mass index compared to mature ACS patients (50.1±6.8 vs 73.3±9.3 y.o., 25.5 vs 23.0, P&lt;0.001, respectively). However, premature ACS patients were more likely to be associated with ST elevation myocardial infarction, dyslipidemia, family history of coronary artery disease (CAD) and lower Killip classification compared to mature ACS patients (P&lt;0.01, respectively). Common CAD risk factors such as hypertension, diabetes mellitus and past history of CAD were less associated with premature ACS patients compared to mature ACS patients (P&lt;0.01, respectively). Unadjusted Kaplan-Meier survival curves demonstrated the favorable prognosis in premature ACS patients compared to mature ACS patients with hazard ratio of 0.57 (95% CI 0.45–0.71, P&lt;0.001, see Figure 1A). We compared a 1:1 propensity score-matched cohort of 1208 patients with or without premature onset of ACS adjusting the several factors mentioned above (n=604, respectively). Age could not be introduced as a factor of propensity score match when comparing premature and mature ACS patients. After propensity score-match, premature ACS patients is about 18 years younger than mature ACS patients (50.7±6.5 vs 68.5±8.2 y.o., P&lt;0.001). The average age of premature ACS was younger than that of mature ACS, but MACE by Kaplan-Meier survival analysis for premature ACS patients was equivalent to mature ACS patients (P=0.77, see Figure 1B).

Conclusion

Premature ACS patients are required very careful management because they might have factors with unfavorable prognosis, such as lifestyle habit and genetics, that may be beyond age.

Figure 1

Funding Acknowledgement

Type of funding source: None

Right coronary artery as a culprit artery for better prognosis in patients with acute myocardial infarction (AMI) with or without shock

Background

Although patients with acute myocardial infarction (AMI) complicated by cardiogenic shock, morbidity and mortality remain high even with early revascularization and modern intensive care.

Culprit artery and prognosis were associated in patients with acute myocardial infarction.

Purpose

Evaluation of short- and long-term prognosis of AMI with cardiogenic shock by right coronary artery (RCA) and left coronary artery (LCR)

Method

We investigated 3400 AMI patients (age 68.8±12.7 y.o.) were enrolled from Mie ACS registry. They were divided into 4 groups according to the culprit artery and presence or absence of cardiogenic shock: RCA without shock n=1114, RCA with shock n=74, LCA without shock n=2028, LCA with shock n=184. Primary endpoint was defined as all-cause mortality.

Results

During the median follow-up periods with 743 days, 12.6% of the patients experienced all-cause death. RCA and LAC with shock groups demonstrated significantly higher in-hospital mortality compared to groups without shock (p&lt;0.001, Figure 1A). Interestingly, after discharge, LCA with shock group showed significant higher all-cause mortality compared with other 3 groups. Surprisingly, RCA with shock group showed similar favorable prognosis to that of without shock groups (Figure 1B). Multivariate analyses for after discharge mortality showed that LCA with shock group was strongest independent poor prognostic factor with hazard ratio of 2.3 (95% CI 1.4–3.7), but RCA with shock group was not.

Conclusion

Association of cardiogenic shock is the hazardous risk factor for cases with AMI, especially LCA infarction. Surprisingly, RCA AMI cases with shock showed favorable prognosis as well as AMI without shock.

Kaplan-Meier survival curves

Funding Acknowledgement

Type of funding source: None

Prognostic impact of right ventricular overload in patients with acute coronary syndrome

Background

Right ventricular (RV) overload is associated with adverse outcome in patients with chronic heart failure. However, its prognostic value in acute coronary syndrome (ACS) patients remains unknown.

Purpose

The purpose of this study was to investigate the prevalence and prognostic impact of right ventricular overload in ACS patients.

Methods

We studied 2797 ACS patients from Mie ACS registry, a prospective and multicenter registry in Japan. They were divided into 4 subgroups according to the severity of RV overload and the extent of Left Ventricle Ejection Fraction (LVEF) assessed by echocardiography before hospital discharge. High RV overload was defined as trans-tricuspid pressure gradient (TRPG) ≥40mmHg and preserved LVEF was defined as ≥50%. The primary outcome was defined as 2-year all-cause mortality. Median follow up duration was 730 days (1–2215 days).

Results

High RV overload was detected in 76 patients (2.7%). In basic patients characteristics, high RV overload patients were significantly older and higher killip classification than low RV overload patients (P&lt;0.01, respectively). Laboratory data in high RV overload patients showed lower hemoglobin level and higher serum creatinine level than those in low RV overload patients (P&lt;0.01, respectively).

Echocardiographic findings in high RV overload represented lower LVEF, higher rate of moderate or severe mitral regurgitation and left atrial enlargement than those in low RV overload patients (P&lt;0.01, respectively).

During the follow-up periods (median 730 days), 260 (9.3%) patients experienced all-cause death. Multivariate cox hazard regression analysis for all-cause mortality demonstrated that high RV overload was an independent poor prognostic factor in the entire study population. Among patients with preserved LVEF, high RV overload resulted in an increased risk of all-cause mortality compared to low RV overload (P&lt;0.0001).

Conclusion

In ACS patients, high RV overload strongly contributes to worsening of prognosis regardless of the extent of LVEF.

Kaplan-Meier survival curve

Funding Acknowledgement

Type of funding source: None

A novel surgical training simulator for transanal total mesorectal excision

BACKGROUND

The transanal total mesorectal excision (TaTME) approach for rectal cancer has been gaining popularity. Although TaTME requires specific training, the opportunity to obtain this training is limited. We developed the first dry simulator that includes some important structural landmarks to provide training that includes all the procedural steps of TaTME.

METHODS

The model was structured based on a computed tomography scan data. The simulator mimics the rectum and surrounding key anatomical structures. All material components were made of polyvinyl alcohol and stained with various colors to easily identify these anatomical structures while undergoing the training.

RESULTS

Our simulator is compact in its size and can be easily packed, stacked, and stored. The materials mimic the real one and the model allows us to perform the entire step-by-step TaTME procedures with genuine clinical devices from the attachment of operative platform to the anastomosis. We have held the course 24 times to date, with more than 200 surgeons participating not only from domestic hospitals but also from other countries. The result of the satisfaction survey from 65 participants, which was rated from 0 to 10 points, was an average score of 8.6 (± 1.4) points.

CONCLUSIONS

We present the world's first TaTME simulator and believe that it will play an important role for the safe spread of TaTME surgery.

Long-term follow-up of the randomized trial of mesorectal excision with or without lateral lymph node dissection in rectal cancer (JCOG0212)

BACKGROUND

Japan Clinical Oncology Group (JCOG) 0212 (ClinicalTrials.gov NCT00190541) was a non-inferiority phase III trial of patients with clinical stage II-III rectal cancer without lateral pelvic lymph node enlargement. The trial compared mesorectal excision (ME) with ME and lateral lymph node dissection (LLND), with a primary endpoint of recurrence-free survival (RFS). The planned primary analysis at 5 years failed to confirm the non-inferiority of ME alone compared with ME and LLND. The present study aimed to compare ME alone and ME with LLND using long-term follow-up data from JCOG0212.

METHODS

Patients with clinical stage II-III rectal cancer below the peritoneal reflection and no lateral pelvic lymph node enlargement were included in this study. After surgeons confirmed R0 resection by ME, patients were randomized to receive ME alone or ME with LLND. The primary endpoint was RFS.

RESULTS

A total of 701 patients from 33 institutions were assigned to ME with LLND (351) or ME alone (350) between June 2003 and August 2010. The 7-year RFS rate was 71.1 per cent for ME with LLND and 70·7 per cent for ME alone (hazard ratio (HR) 1·09, 95 per cent c.i. 0·84 to 1·42; non-inferiority P = 0·064). Subgroup analysis showed improved RFS among patients with clinical stage III disease who underwent ME with LLND compared with ME alone (HR 1·49, 1·02 to 2·17).

CONCLUSION

Long-term follow-up data did not support the non-inferiority of ME alone compared with ME and LLND. ME with LLND is recommended for patients with clinical stage III disease, whereas LLND could be omitted in those with clinical stage II tumours.

[Risk of Euglycemic Diabetic Ketoacidosis Due to Low-carbohydrate Diet While Taking Empagliflozin: a Case Report]

Empagliflozin reduces blood glucose levels independently of insulin secretion by reducing glucose reabsorption in the proximal renal tubules through inhibition of sodium-glucose cotransporter 2 (SGLT2). Because SGLT2 inhibitors have a different mechanism of action to conventional antidiabetic drugs, recommendations have been issued about the management of specific side effect such as ketoacidosis, urinary tract infection, and genital infection. There have been some reports of SGLT2 inhibitor-associated euglycemic diabetic ketoacidosis (euDKA), but there have been few reports about euDKA in patients with type 2 diabetes using SGLT2 inhibitors while on a low-carbohydrate diet. Here we report a patient who developed euDKA after starting a very low-carbohydrate diet while taking empagliflozin. A 51-year-old man was hospitalized with nausea and vomiting, and investigations revealed metabolic acidosis. euDKA was diagnosed from the information about medications in his drug notebook and a history of eating a low-carbohydrate diet (1900 kcal, consisting of 5.7% carbohydrate, 21.1% protein, 47.3% fat and 25.9% alcohol) for 4 d. The patient improved after infusion of acetated Ringer's solution with 5% glucose and administration of regular insulin. It is necessary for physicians and pharmacists to thoroughly inform patients about the side effects of SGLT2 inhibitors such as ketoacidosis, urinary tract infection, and genital infection. Patients should also be advised about the higher risk of euDKA associated with a low-carbohydrate diet while taking SGLT2 inhibitors.

Recurrence of rectal anastomotic leakage following stoma closure: assessment of risk factors

AIM

In patients with a previous history of rectal anastomotic leakage (AL), the surgical indications and timing for closure of a diverting stoma have to be carefully judged. Even if AL has apparently healed before stoma closure, re-leakage may occur after closure. The aim of this study was to determine the incidence and risk factors for recurrent AL following stoma closure. We also examined the treatment strategies aiming to minimize the risk of recurrent AL.

METHODS

From January 2009 to December 2016, 1008 patients underwent sphincter-saving surgery [low anterior resection, all-sphincter-preserving rectal resection with hand-sewn coloanal anastomosis (CAA) and intersphincteric resection (ISR)] for primary rectal cancer with curative intent at our hospital. A total of 69 patients with AL with a Clavien-Dindo Grade III or more who subsequently underwent closure of a diverting stoma were retrospectively reviewed for this study.

RESULTS

The incidence of recurrent leakage after stoma closure in this series was 13% overall with an incidence of 25% in the CAA/ISR group and 5% in the low anterior resection group. Significant risk factors included hand-sewn anastomosis (P = 0.0257) compared to stapled anastomosis, ischaemia at the anastomotic site as the cause of initial AL (P < 0.001) and a shorter interval between confirmation of healing and stoma closure (P = 0.00952).

CONCLUSION

Ischaemia at the anastomotic site was the main risk factor for recurrent leakage, particularly after CAA/ISR. Additional treatment options before stoma closure should be considered to avoid re-leakage in such cases.