Abstract P631: Association Between Suboptimal Sleep Duration and Cardiovascular Disease Prevalence Among Asian Americans

Importance: Cardiovascular disease (CVD) risk and prevalence is widely varied among Asian Americans. The American Heart Association recently added healthy sleep to its metrics used to define ideal cardiovascular health. Little is known about the association between sleep and CVD prevalence among Asian subgroups.

Hypothesis: The association between suboptimal sleep duration and CVD risk prevalence will differ by Asian American subgroup.

Methods: We analyzed results from the National Health Interview Survey (2012-2018, N=234,126), a nationally representative sample. Participants (age>40 years old) included 214,956 non-Hispanic Whites (NHWs), 3,043 Asian Indians, 3,993 Chinese, 4,719 Filipinos, and 7,603 other Asians. Suboptimal sleep was self-reported and defined as <7 or >9 hours. CVD was defined as self-reported stroke, heart attack, coronary artery disease, or angina. We used multivariable logistic regression models to ascertain the association. All analyses included complex survey design considerations.

Results: Suboptimal sleep duration was associated with higher CVD prevalence for NHWs and all Asians, but in Asian subgroup stratified analyses, suboptimal sleep was only associated with higher CVD prevalence among Other Asians.

Conclusions: Our results show heterogeneity in the association between sleep duration and CVD prevalence among Asian subgroups, however further research is needed to better understand these relationships.

Correlation of Inguinal Lymph Node Number and Volume with Lower Extremity Lymphedema Severity

There are very limited studies on the relationship of inguinal lymph node number and volume correlated with lower extremity lymphedema severity. In this IRB-approved retrospective study, patients who obtained an MRI for lower extremity lymphedema and who did not have lymph node resection or biopsy were identified. The MRI images were used to determine the number and volume of inguinal lymph nodes for each limb in addition to fat and fluid-based scoring using a validated grading system. Wilcoxon signed-rank tests were used to compare the greater-affected limbs with the lesser-affected limbs. The spear-man-rank correlation was performed on a 'per limb' basis for MRI-based scoring and clinical parameters with ipsilateral lymph node number and volume and for differences between the limbs. A total of 32 patients were included. The greater-affected limb had higher MRI fluid scores (median (interquartile range) = 3 (3 - 3) vs. 0 (0 - 1), (p < 0.01) relative to the contra-lateral limb and had a median fat asymmetry score of 2 (1 - 3). On the per-limb analysis, lymph node number and volume inversely correlated with total MRI scores (ρ = -0.47, p < 0.01 for node number and volume). The difference of lymph node number and volume correlated with MRI score difference (node number: ρ = -0.66, p < 0.01; node volume: ρ = -0.64, p < 0.01) and perometer difference (node number: ρ = -0.58, p < 0.01; node volume: ρ = -0.59, p < 0.01). Inguinal lymph node number and volume inversely correlate with lower extremity edema presence and severity.

Concomitant left atrial appendage closure with left ventricular assist device surgery can reduce ischemic cerebrovascular accidents

Background

It remains unknown if concomitant left atrial appendage closure (LAAC) at the time of left ventricular assist device (LVAD) surgery can reduce ischemic cerebrovascular accidents.

Purpose

The purpose of this study is to assess the impact of LAAC at LVAD surgery on the incidence of ischemic cerebrovascular accidents.

Methods

Between January 2012 and November 2021, 310 patients underwent LVAD surgery with HeartMate II or III. Out of 310 patients, 98 patients (31.6%) underwent concomitant LAAC. The cohort was divided into two groups: patients with LAAC (Group A, n=98) and without LAAC (Group B, n=212). To minimize device bias, LVAD surgery with HeartWare HVAD device was excluded. The ischemic cerebrovascular accident was defined as ischemic stroke, hemorrhagic stroke or transient ischemic attack. We reviewed early and long-term clinical outcomes. The incidence of ischemic cerebrovascular accidents was compared between two groups using the Kaplan-Meier method. We also investigated if LAAC was associated with ischemic cerebrovascular accidents by Cox proportional hazards analysis.

Results

There were no significant differences in baseline characteristics between two groups including age (Group A: 55.0±12.3 years old, Group B: 56.9±14.1 years old, p=0.26), preoperative CHADS2 score (Group A: 2.40±1.1, Group B: 2.58±1.1, p=0.19) and history of atrial fibrillation (Group A: 42.9%, Group B: 42.5%, p=0.95). In-hospital mortality was not significantly different between the two groups (Group A: 7.1%, Group B: 12.3%, p=0.16). In terms of postoperative complications, there were no significant differences between two groups in requiring extracorporeal membrane oxygenation, re-exploration for bleeding and newly required hemodialysis. Median follow up period was 474 days. Thirty-five patients (11.2%) developed ischemic cerebrovascular accidents (5 patients in Group A and 30 patients in Group B). The rate of freedom from ischemic cerebrovascular accidents in Group A (94.1% at 500 days and 94.1% at 1500 days) was significantly higher than that in Group B (88.2% at 500 days and 77.4% at 1500 days; log rank=0.024). In a Cox proportional hazards regression analysis including LAAC, age, history of atrial fibrillation, diabetes mellitus and Heartmate 3 device implantation, LAAC was associated with reducing the incidence of ischemic cerebrovascular accidents (hazard ratio 0.37, 95% CI 0.13–0.89, p=0.02).

Conclusion

Concomitant LAAC at the time of LVAD surgery can reduce ischemic cerebrovascular accidents without increasing perioperative mortality and complications.

Funding Acknowledgement

Type of funding sources: None.

Comparison of surgical interventions for the treatment of early-onset scoliosis: a systematic review and meta-analysis

OBJECTIVE

Several growth-preserving surgical techniques are employed in the management of early-onset scoliosis (EOS). The authors' objective was to compare the use of traditional growing rods (TGRs), magnetically controlled growing rods (MCGRs), Shilla growth guidance techniques, and vertically expanding prosthetic titanium ribs (VEPTRs) for the management of EOS.

METHODS

A systematic review of electronic databases, including Ovid MEDLINE and Cochrane, was performed. Outcomes of interest included correction of Cobb angle, T1-S1 distance, and complication rate, including alignment, hardware failure and infection, and planned and unplanned reoperation rates. The percent changes and 95% CIs were pooled across studies using random-effects meta-analysis.

RESULTS

A total of 67 studies were identified, which included 2021 patients. Of these, 1169 (57.8%) patients underwent operations with TGR, 178 (8.8%) Shilla growth guidance system, 448 (22.2%) MCGR, and 226 (11.1%) VEPTR system. The mean ± SD age of the cohort was 6.9 ± 1.2 years. The authors found that the Shilla technique provided the most significant improvement in coronal Cobb angle immediately after surgery (mean [95% CI] 64.3% [61.4%-67.2%]), whereas VEPTR (27.6% [22.7%-33.6%]) performed significantly worse. VEPTR also performed significantly worse than the other techniques at final follow-up. The techniques also provided comparable gains in T1-S1 height immediately postoperatively (mean [95% CI] 10.7% [8.4%-13.0%]); however, TGR performed better at final follow-up (21.4% [18.7%-24.1%]). Complications were not significantly different among the patients who underwent the Shilla, TGR, MCGR, and VEPTR techniques, except for the rate of infections. The TGR technique had the lowest rate of unplanned reoperations (mean [95% CI] 15% [10%-23%] vs 24% [19%-29%]) but the highest number of planned reoperations per patient (5.31 [4.83-5.82]). The overall certainty was also low, with a high risk of bias across studies.

CONCLUSIONS

This analysis suggested that the Shilla technique was associated with a greater early coronal Cobb angle correction, whereas use of VEPTR was associated with a lower correction rate at any time point. TGR offered the most significant height gain at final follow-up. The complication rates were comparable across all surgical techniques. The optimal surgical approach should be tailored to individual patients, taking into consideration the strengths and limitations of each option.

The influence of anastomotic techniques on postoperative anastomotic complications: Results of the Oesophago-Gastric Anastomosis Audit

BACKGROUND

The optimal anastomotic techniques in esophagectomy to minimize rates of anastomotic leakage and conduit necrosis are not known. The aim of this study was to assess whether the anastomotic technique was associated with anastomotic failure after esophagectomy in the international Oesophago-Gastric Anastomosis Audit cohort.

METHODS

This prospective observational multicenter cohort study included patients undergoing esophagectomy for esophageal cancer over 9 months during 2018. The primary exposure was the anastomotic technique, classified as handsewn, linear stapled, or circular stapled. The primary outcome was anastomotic failure, namely a composite of anastomotic leakage and conduit necrosis, as defined by the Esophageal Complications Consensus Group. Multivariable logistic regression modeling was used to identify the association between anastomotic techniques and anastomotic failure, after adjustment for confounders.

RESULTS

Of the 2238 esophagectomies, the anastomosis was handsewn in 27.1%, linear stapled in 21.0%, and circular stapled in 51.9%. Anastomotic techniques differed significantly by the anastomosis sites (P < .001), with the majority of neck anastomoses being handsewn (69.9%), whereas most chest anastomoses were stapled (66.3% circular stapled and 19.3% linear stapled). Rates of anastomotic failure differed significantly among the anastomotic techniques (P < .001), from 19.3% in handsewn anastomoses, to 14.0% in linear stapled anastomoses, and 12.1% in circular stapled anastomoses. This effect remained significant after adjustment for confounding factors on multivariable analysis, with an odds ratio of 0.63 (95% CI, 0.46-0.86; P = .004) for circular stapled versus handsewn anastomosis. However, subgroup analysis by anastomosis site suggested that this effect was predominantly present in neck anastomoses, with anastomotic failure rates of 23.2% versus 14.6% versus 5.9% for handsewn versus linear stapled anastomoses versus circular stapled neck anastomoses, compared with 13.7% versus 13.8% versus 12.2% for chest anastomoses.

CONCLUSIONS

Handsewn anastomoses appear to be independently associated with higher rates of anastomotic failure compared with stapled anastomoses. However, this effect seems to be largely confined to neck anastomoses, with minimal differences between techniques observed for chest anastomoses. Further research into standardization of anastomotic approach and techniques may further improve outcomes.

Pathogenic Extracellular Vesicle (EV) Signaling in Amyotrophic Lateral Sclerosis (ALS)

Extracellular vesicles (EVs), once considered a pathway for cells to remove waste, have now emerged as an important mechanism for intercellular communication. EVs are particularly appealing in understanding the central nervous system (CNS) communication, given that there are very diverse cell types in the CNS and constant communications among various cells to respond to the frequently changing environment. While they are heterogeneous and new vesicles are continuously to be discovered, EVs are primarily classified as plasma membrane-derived microvesicles (MVs) and endosome-derived exosomes. Secretion of EVs has been shown from all CNS cell types in vitro and intercellular EV signaling has been implicated in neural development, axon integrity, neuron to glia communication, and propagation of protein aggregates formed by disease pathogenic proteins. However, significant hurdles remain to be tackled in understanding their physiological and pathological roles as well as how they can be developed as biomarkers or new therapeutics. Here we provide our summary on the known cell biology of EVs and discuss opportunities and challenges in understanding EV biology in the CNS and particularly their involvement in ALS pathogenesis.

POS0859 C-REACTIVE PROTEIN AND INTERLEUKIN-6: POTENTIAL BIOMARKERS OF DISEASE ACTIVITY AND TREATMENT RESPONSE IN SYSTEMIC SCLEROSIS-INTERSTITIAL LUNG DISEASE

Background

Clinically feasible and valid biomarkers of systemic sclerosis-interstitial lung disease (SSc-ILD) are needed. While promising candidate biomarkers are under investigation (e.g., KL-6, CCL-8), clinical assays for these proteins are not currently available. C-reactive protein (CRP) measurements are feasible, cost-effective, and have been shown to predict mortality in SSc.1 Measuring interleukin (IL)-6, a proinflammatory cytokine implicated in SSc-ILD pathogenesis, is also feasible in most clinical settings.

Objectives

(1) To investigate whether CRP and IL-6 levels change in response to treatment with immunosuppression in SSc-ILD; (2) To explore whether the change in CRP and IL-6 predict the future course of forced vital capacity (FVC).

Methods

CRP and IL-6 levels were measured in serum at baseline and after 12 months in participants of Scleroderma Lung Study (SLS) II (patients with active SSc-ILD receiving 24 months of mycophenolate or 12 months cyclophosphamide followed by 12 months of placebo2). Measured values were log-transformed to remove skewness. The FVC%-predicted was measured every 3 months over the course of 24 months. Spearman’s correlations evaluated the relationship between baseline CRP or IL-6 measurements and other patient parameters. Paired t-tests were used to compare the change in individual CRP and IL-6 measurements from baseline to 12 months. Linear mixed effects models were used to examine the relationship between the change in CRP and IL-6 (baseline to 12 months) and the subsequent course of the FVC (12 to 24 months). All analysis were performed for the entire cohort and separately by treatment arm.

Results

Of the 142 participants of SLS II, 101 had CRP and IL-6 measurements at baseline and 12 months. Baseline CRP and IL-6 levels correlated significantly with higher modified Rodnan skin score (CRP: r=0.3, P=0.005; IL-6: r=0.2, P=0.01) and shorter disease duration (CRP: r=-0.3, P=0.005; IL-6: r=-0.2, P=0.01) and were higher in patients with diffuse SSc (CRP: P=0.007; IL-6: P=0.01). Relationships to baseline FVC and DLCO were not observed. CRP decreased significantly from baseline to 12 months in the whole group (P=0.01), but the decrease was slightly greater in patients randomized to mycophenolate versus cyclophosphamide (-0.47 vs. -0.33 ug/mL). IL-6 also decreased from baseline to 12 months in the whole group with a trend towards significance (P=0.10) (Figure 1). The mean decrease in IL-6 was again slightly greater in patients randomized to mycophenolate versus cyclophosphamide (-0.31 vs. -0.10 pg/mL). After controlling for baseline FVC and treatment arm in the mixed effects model, there was a relationship between the decrease in CRP from baseline to 12 months and an improved course of FVC 12 to 24 months (Estimate -0.64), but this did not reach significance (P=0.14). However, after controlling for baseline FVC and treatment arm, a greater decrease in IL-6 from baseline to 12 months was significantly associated with a greater improvement in FVC from 12 to 24 months (Estimate -1.28; P=0.01).

Figure 1.

Change in CRP (A) and IL-6 (B) from baseline to 12 months by treatment arm in SLS II. CYC=cyclophosphamide (Blue), MMF=mycophenolate (red)

Conclusion

Patients with active SSc-ILD receiving one year of immunosuppressive therapy in the SLS II study experienced reductions in their CRP and IL-6 levels over this interval. The magnitude of the decrease in CRP and IL-6 over the first year also correlated with the course of FVC over the ensuing 12 months. These findings suggest a dynamic relationship between CRP and IL-6 measurements and the course of SSc-ILD in patients on immunosuppressive therapy. Further investigation of these findings is warranted.

References

[1]Liu et al. Arthritis Care Res 2013.

[2]Tashkin et al. Lancet Resp Med 2016.

Disclosure of Interests

Elizabeth Volkmann Speakers bureau: Boehringer Ingelheim, Consultant of: Boehringer Ingelheim, Grant/research support from: Forbius, Kadmon, Horizon, Boehringer Ingelheim, Donald Tashkin: None declared, Holly Wilhalme: None declared, Marka Lyons: None declared, Grace Kim: None declared, Jonathan Goldin: None declared, Michael Roth Grant/research support from: Genentech, Shervin Assassi Consultant of: Boehringer Ingelheim

POS0852 SYMPTOMS OF GASTROESOPHAGEAL REFLUX ARE A BETTER PREDICTOR OF SYSTEMIC SCLEROSIS-RELATED INTERSTITIAL LUNG DISEASE PROGRESSION THAN QUANTITATIVE RADIOGRAPHIC ASSESSMENT OF ESOPHAGEAL PARAMETERS

Background

Esophageal dysfunction affects the majority of patients with systemic sclerosis (SSc). Gastroesophageal reflux (GERD)-associated microaspiration may serve as an inciting and/or exacerbating factor in the pathogenesis of interstitial lung disease (ILD) in SSc. Few studies have investigated the relationship between objective measures of esophageal involvement in SSc and the progression of SSc-ILD.

Objectives

(1) To investigate whether patient-reported symptoms of GERD predict SSc-ILD severity and progression; (2) To explore whether quantitative radiographic measurements of esophageal parameters predict SSc-ILD severity and progression.

Methods

Participants of Scleroderma Lung Study (SLS) II (24 months of mycophenolate versus 12 months cyclophosphamide followed by 12 months of placebo in patients with active SSc-ILD) were included.1 The UCLA SCTC GIT 2.0 was used to assess the severity of reflux at baseline. Quantitative image analysis was used to calculate the diameter and area of the esophagus in the area of maximum dilation. Univariate and multivariable linear regression analyses were used to evaluate the relationship between baseline reflux scores/esophageal parameters and the severity and progression of SSc-ILD based on the quantitative radiographic extent of ILD (QILD) and fibrosis (QLF) in the lobe of maximum involvement (LM). All multivariable models controlled for treatment arm and baseline ILD severity, as well as proton pump inhibitor (PPI) use.

Results

Among the 141 of 142 SLS II participants with patient-reported reflux scores at baseline, the mean score was 0.57, indicating moderate reflux. The mean maximal esophageal diameter and area at baseline were 22 mm and 242 mm2, respectively. There were no significant correlations between reflux scores, esophageal diameter, esophageal area and QILD-LM and QLF-LM at baseline. However, in the 96 participants with HRCT scans available at 24 months, increased reflux scores were significantly associated with worsening of QLF-LM and QILD-LM (Figure 1). Even after controlling for disease severity at baseline, treatment arm and PPI use, reflux scores remained significantly associated with change in QLF-LM (P-value 0.05) and QILD-LM (P-value 0.003). Neither the maximum esophageal area, nor the maximum esophageal diameter was associated with the change in QILD-LM or QLF-LM in univariate or multivariate analyses.

Figure 1.

Increased reflux scores at baseline are associated with increased progression (worsening) of QILD (A) and QLF (B) in the lobe of maximum involvement in patients with SSc-ILD.

Conclusion

Increased self-reported reflux was independently associated with worsening radiographic fibrosis and ILD in patients receiving treatment for SSc-ILD. The maximum esophageal diameter and esophageal area did not correlate with reflux scores and did not predict severity or progression of SSc-ILD. These findings illustrate the importance of eliciting a careful patient history of reflux symptoms in patients with ILD and further highlight the need for improved objective measures of esophageal dysfunction in SSc.

References

[1]Tashkin et al. Lancet Resp Med 2016.

Disclosure of Interests

Elizabeth Volkmann Speakers bureau: Boehringer Ingelheim, Consultant of: Boehringer Ingelheim, Grant/research support from: Forbius, Horizon, Kadmon, Boehringer Ingelheim, Donald Tashkin: None declared, Mei Leng: None declared, Grace Kim: None declared, Jonathan Goldin: None declared, Michael Roth Grant/research support from: Genentech

POS0068 CHANGES IN QUANTITATIVE INTERSTITIAL LUNG DISEASE SCORES ON HIGH-RESOLUTION CT IN IDIOPATHIC INFLAMMATORY MYOSITIS

Background

The idiopathic inflammatory myopathies (IIM) are autoimmune connective tissue diseases affecting skeletal muscle, skin and other organ systems. IIM-related interstitial lung disease (IIM-ILD) is the most common extra-muscular manifestation, being the leading cause of morbidity and mortality. Several studies have suggested that ILD pattern based on chest high-resolution computed tomography (HRCT) can be related to disease course and treatment response, but the results vary considerably. Moreover, the clinical impact of the quantitative ILD (QILD) score, a validated computer-aided scoring system in assessing ILD severity from HRCT, and its longitudinal changes have not yet been evaluated in IIM-ILD.

Objectives

This study aims to investigate ILD patterns and QILD scores in patients with IIM-ILD, to identify their clinical impact, and to delineate longitudinal changes of QILD measurement.

Methods

A total of 80 patients with IIM (polymyositis 22, and dermatomyositis 58) who underwent at least 2 times of serial HRCT scans were included. Visual ILD patterns were assessed by multiple thoracic radiologists. Quantitative analysis of HRCT was presented as total extent of QILD scores (%) in whole lung and most severe zone. Individual time-estimated ΔQILD score between first 2 visits was derived using a linear approximation of yearly change, where the duration of median (IQR) was 1.0 (0.4-1.6) years in the first 2 HRCT scans.

Results

The median (IQR) age of the patients was 52.0 (43.5-58.5) years and 60 (75.0%) were women. Baseline median score of whole lung-QILD and most severe zone-QILD were 28.1% (19.1-43.8) and 68.0% (45.5-81.8), respectively, and QILD score showed significant correlations with pulmonary function tests (r=-0.349, p=0.002 for % predicted forced vital capacity; and r=-0.381, p=0.001 for % predicted diffusing capacity for carbon monoxide). The individual time-estimated yearly ΔQILD score between first 2 visits presented that approximately half of the patients showed improvement or stability in QILD scores; however, when patients were sorted by visual assessment in ILD subtype on HRCT, approximately two-thirds of the patients with usual interstitial pneumonia (UIP) pattern were aggravated in QILD scores and less than half of subjects with nonspecific interstitial pneumonia and organizing pneumonia were aggravated (Figure 1, 80% for UIP vs. 44.4% for non-UIP, p=0.013). There was no immunosuppressive drugs related to meaningful improvement in QILD scores during first 2 visits. Notably, we observed significant aggravation of QILD scores in tacrolimus users (n=7, median time-estimated whole lung-yarly ΔQILD 20.3 (2.7-38.4)) compared with tacrolimus non-users (n=73, median time estimated whole lung-yearly ΔQILD -1.2 (-8.3-6.5)). Among 80 patients, 6 (7.5%) were died due to various lung complications. Higher baseline QILD scores were noted in deaths (median whole lung-QILD 45.4 (32.9-56.5)) than in survivors (median whole lung-QILD 26.9 (19.0-42.4)), albeit not significant (p=0.084). Poor survival rate was observed in patients with high grade of ground glass opacity by visual assessment in right upper lobe (log-rank test, p=0.042). Among subgroup of patients with 3 serial HRCT scans (n=41), dynamic changes of four distinct patterns (improving, worsening, convex, and concave) were observed.

Figure 1.

Cleveland dot plot of individual time-estimated yearly ΔQILD during fist 2 visits.

Conclusion

The changes in QILD score in IIM-ILD are dynamic and present different by visual assessment. QILD score has the potential for evaluation of the severity changes, prognosis and medication response in patients with IIM-ILD.

References

[1]Tashkin DP, et al. Ann Rheum Dis 2016;75(2):374-81.

7 truncated values in the graph A. NSIP: nonspecific interstitial pneumonia; OP: organizing pneumonia; UIP: usual interstitial pneumonia.

Disclosure of Interests

None declared

Textbook outcome following oesophagectomy for cancer: international cohort study

Background

Textbook outcome has been proposed as a tool for the assessment of oncological surgical care. However, an international assessment in patients undergoing oesophagectomy for oesophageal cancer has not been reported. This study aimed to assess textbook outcome in an international setting.

Methods

Patients undergoing curative resection for oesophageal cancer were identified from the international Oesophagogastric Anastomosis Audit (OGAA) from April 2018 to December 2018. Textbook outcome was defined as the percentage of patients who underwent a complete tumour resection with at least 15 lymph nodes in the resected specimen and an uneventful postoperative course, without hospital readmission. A multivariable binary logistic regression model was used to identify factors independently associated with textbook outcome, and results are presented as odds ratio (OR) and 95 per cent confidence intervals (95 per cent c.i.).

Results

Of 2159 patients with oesophageal cancer, 39.7 per cent achieved a textbook outcome. The outcome parameter ‘no major postoperative complication’ had the greatest negative impact on a textbook outcome for patients with oesophageal cancer, compared to other textbook outcome parameters. Multivariable analysis identified male gender and increasing Charlson comorbidity index with a significantly lower likelihood of textbook outcome. Presence of 24-hour on-call rota for oesophageal surgeons (OR 2.05, 95 per cent c.i. 1.30 to 3.22; P = 0.002) and radiology (OR 1.54, 95 per cent c.i. 1.05 to 2.24; P = 0.027), total minimally invasive oesophagectomies (OR 1.63, 95 per cent c.i. 1.27 to 2.08; P &lt; 0.001), and chest anastomosis above azygous (OR 2.17, 95 per cent c.i. 1.58 to 2.98; P &lt; 0.001) were independently associated with a significantly increased likelihood of textbook outcome.

Conclusion

Textbook outcome is achieved in less than 40 per cent of patients having oesophagectomy for cancer. Improvements in centralization, hospital resources, access to minimal access surgery, and adoption of newer techniques for improving lymph node yield could improve textbook outcome.

Textbook outcome following oesophagectomy for cancer: international cohort study

BACKGROUND

Textbook outcome has been proposed as a tool for the assessment of oncological surgical care. However, an international assessment in patients undergoing oesophagectomy for oesophageal cancer has not been reported. This study aimed to assess textbook outcome in an international setting.

METHODS

Patients undergoing curative resection for oesophageal cancer were identified from the international Oesophagogastric Anastomosis Audit (OGAA) from April 2018 to December 2018. Textbook outcome was defined as the percentage of patients who underwent a complete tumour resection with at least 15 lymph nodes in the resected specimen and an uneventful postoperative course, without hospital readmission. A multivariable binary logistic regression model was used to identify factors independently associated with textbook outcome, and results are presented as odds ratio (OR) and 95 per cent confidence intervals (95 per cent c.i.).

RESULTS

Of 2159 patients with oesophageal cancer, 39.7 per cent achieved a textbook outcome. The outcome parameter 'no major postoperative complication' had the greatest negative impact on a textbook outcome for patients with oesophageal cancer, compared to other textbook outcome parameters. Multivariable analysis identified male gender and increasing Charlson comorbidity index with a significantly lower likelihood of textbook outcome. Presence of 24-hour on-call rota for oesophageal surgeons (OR 2.05, 95 per cent c.i. 1.30 to 3.22; P = 0.002) and radiology (OR 1.54, 95 per cent c.i. 1.05 to 2.24; P = 0.027), total minimally invasive oesophagectomies (OR 1.63, 95 per cent c.i. 1.27 to 2.08; P < 0.001), and chest anastomosis above azygous (OR 2.17, 95 per cent c.i. 1.58 to 2.98; P < 0.001) were independently associated with a significantly increased likelihood of textbook outcome.

CONCLUSION

Textbook outcome is achieved in less than 40 per cent of patients having oesophagectomy for cancer. Improvements in centralization, hospital resources, access to minimal access surgery, and adoption of newer techniques for improving lymph node yield could improve textbook outcome.