United States

Background
Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccination elicits a variable, often suboptimal immune response among solid organ transplant recipients (SOTRs), who are at increased risk of severe COVID-19 compared to immunocompetent individuals. An accurate prediction model for lack of immune response to SARS-CoV-2 vaccination could provide for a more tailored approach to antibody testing, risk stratification, and clinical decision making. The objective was to create a prediction model for failure to generate an antibody response to threedose
SARS-CoV-2 vaccination in SOTRs.
Methods
782 thrice-vaccinated COVID-naïve SOTRs were recruited for this national prospective observational study. Participants underwent serologic antibody testing (Roche Elecsys antireceptor binding domain [RBD] or Euroimmun anti-spike enzyme immunoassays) before and 1 month after receiving a third dose of vaccine (D3). A machine learning model for predicting antibody response was created using the extreme gradient boosting (XGBoost) algorithm and validated using bootstrap validation.
Results
634 (82.3%) SOTRs were seropositive for antibodies against the spike protein before and after D3. Shorter time since transplant, mycophenolate use, corticosteroid use, and older age were key risk factors, collectively attributing to 86.6% of the model’s predictive ability. In the bootstrap validation, we observed a C-statistic of 0.80 (95% CI; 0.75, 0.83) and Brier score of 0.14 (95% CI; 0.13, 0.15), representing a good prediction performance and an 11.7% (95% CI; 7.6%,
14.4%) improvement in calibration over the non-informative model. A second model predicting antibody response after the second vaccine was created and assessed on its ability to predict antibody response after the third dose; this model showed a similarly high C-statistic (0.78 [95% CI; 0.74, 0.82]), suggesting that SOTRs with higher risk of negative antibody response to two doses remain at high risk for non-response even after subsequent doses.
Conclusion
Failure to generate an antibody response to SARS-CoV-2 vaccination in SOTRs can be predicted accurately. We created an online calculator for predicting lack of antibody response to D3 based off this model, enabling transplant professionals to perform more personalized risk assessments, therefore helping to identify which SOTRs may warrant prioritization for additional vaccine doses, antibody monitoring, and other protective measures.

AMA Research Challenge 2022

A Machine Learning Model for Predicting Antibody Response After ThreeDoses of Vaccine Against SARS-CoV-2 in Solid Organ Transplant Recipients

poster

Welcome to the largest national, multi-specialty research event **featuring research from 750 medical students, residents and fellows, and international medical graduates.**


#### Event Details
The 2022 American Medical Association Research Challenge includes a poster symposium hall and a semifinals hall with the top scored research. Log in at any time throughout the event to:
* **View posters and presentations** on a variety of topics and specialties
* **Score posters in the semifinals** to help decide the top five finalists (AMA member exclusive opportunity)
* **Connect with participants** and provide feedback to help advance their research

At the conclusion of this event, the AMA will announce the top five finalists who will advance to the AMA Research Challenge finals, present their research to an elite panel of judges and compete to win a $10,000 grand prize sponsored by Laurel Road.
 

#### Attend Live Forums
**Oct. 20, 6 — 7 p.m. Central**
 **How to plan a career in academic medicine** 
 Learn how to plan a career in academic medicine and get your questions answered. [Access Here](https://underline.io/events/363/lounge)
 
 **Speakers**
![](https://assets.underline.io/uploads/markdown_image/1/image/d5adb133dfc58eb0a6d469e9cedcd5d4.jpg)
Ricardo Correa, MD, EdD
![](https://assets.underline.io/uploads/markdown_image/1/image/5f9437ab87861538bf02b9ca7c0d926e.jpg)
Stuart M. Greenstein, MD
![](https://assets.underline.io/uploads/markdown_image/1/image/05b1b637c99286d558c371f8e8b185ce.jpg)
Stephanie Strohbeen, MD
 
 **Oct. 22, 11 a.m. — Noon Central**
 **How to get published**
 Learn tips to get your research published and get your questions answered. [Access Here](https://underline.io/events/363/lounge)
 
 **Speakers**
![](https://assets.underline.io/uploads/markdown_image/1/image/8ce6235e957d31cb5e6f9097cc6f6972.jpg)
Jessica Li, MD/PhD Candidate
![](https://assets.underline.io/uploads/markdown_image/1/image/9950ef2e0237f54cb3d8280334d42491.jpg)
Charles Lopresto, DO

### Sponsor
AMA would like to thank Laurel Road for sponsoring a $10,000 grand prize for the winner of the AMA Research Challenge. Laurel Road is not involved in the judging or scoring process to select the winner. The winner will be announced on Dec. 7 during the AMA Research Challenge finals. Payment of the grand prize will be directly from Laurel Road. See the AMA Research Challenge Official Rules for terms and conditions.

![](https://assets.underline.io/uploads/markdown_image/1/image/611e331d4a0eba0920f240cfc04b4d77.png)

**There is no registration fee for this event! **

Register to explore research in a variety of topics. AMA members get exclusive access to score posters in the semifinals.

**Please use the link below to register for free if you have not already done so:**

https://www.ama-assn.org/about/events/ama-research-challenge

Registration Is Required

Welcome to the largest national, multi-specialty research event featuring research from 750 medical students, residents and fellows, and international medical graduates.


Objective
This study examines trends in racial and gender diversity of trainees within Complex General Surgical Oncology fellowships. Further, this study compares the racial and gender proportions of trainees across different fields to assess potential barriers to increasing diversity within surgical oncology training programs.

Methods
Deidentified Accredited Council for Graduate Medical Education data were queried to identify surgical trainees between 2013-2021. Trainees were identified based on self-reported race and gender and were stratified based on residency type (i.e., general surgery and the overall trainee workforce, composed of all trainees in every program) and fellowship program if applicable. Chi-squared tests were used to assess differences between groups and trends.

Results
A significantly lower proportion of individuals who are underrepresented in medicine (URMs) trained in surgical oncology fellowships (8.9%) compared with both the overall trainee pool (12.8%) and general surgery residency programs (13.1%) (p < 0.05). There was no significant increase in URM representation in surgical oncology fellowships across the study period. In addition, there was a significantly lower proportion of females training in surgical oncology fellowships (38.6%) compared with the overall trainee pool (45.6%) (p < 0.05). Despite a significant increase in female representation in general surgery residency and other surgical fellowships, there was no significant increase in female representation in surgical oncology fellowships across the study period.

Conclusions
This study identifies disparities in gender and racial minority representation within ACGME-accredited Complex General Surgical Oncology fellowship training programs. While steps have been taken to expand diversity, more needs to be done to combat the systemic barriers that both racial minorities and women face during their training.

Analysis of Racial/Ethnic Minority and Female Representation in theSurgical Oncology Training Pathway

Introduction: Risk calculators to predict post-operative surgical site infections (SSI) often rely on logistic regression (LR) analysis. Machine learning models are able incorporate a greater number of input variables by identifying non-linear relationships. Automated machine learning (AutoML) processes regularly outperform regular machine learning (ML) and LR methods for predictive accuracy outside of healthcare settings. AutoML systems have not yet been applied to predict post-operative surgical site infections. 

Methods: We used an AutoML system developed and released by Amazon in 2020, AutoGluon v0.3.1, to predict post-operative SSI in the 2019 ACS NSQIP database. A total of 3,049,617 patients and 79 pre-operative variables were included. Post-operative SSI was defined as a superficial, deep or organ-space infection that occurred within 30 days of the surgery and was not present prior to surgery. Models were trained for four hours to optimize performance on the Brier score, with lower being better. Validation of all performance metrics was done using the 2019 ACS NSQIP database. 

Results: 2.41% of the patients (n = 73,581) developed post-operative SSI. Brier scores were calculated for each model with the top performing model being an ensembled XGBoost model having a Brier score of 0.00287 on the validation set. The corresponding AUROC and AUC-PR was 0.768 and 0.110 respectively (Figure).

Conclusions: Automated machine learning models offer similar discriminatory characteristics to existing postoperative SSI calculators. Future post-operative SSI models may benefit from AutoML analysis.

Automated machine learning (AutoML) with AutoGluon to predict post-operative surgical site infection (SSI) using the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) database

In the United States, >30 million people (10.5% of the population) have diabetes, both diagnosed and undiagnosed. Many of these patients go on to develop diabetes related complications, such as peripheral neuropathy; these patients are also prone to foot injury. The purpose of this study is to determine the outcome of an isolated foot burn in the diabetic patient. A retrospective chart review of adult (>18yo) patients admitted to a major metropolitan burn center at a safety-net hospital from 2008-2020 with an isolated burn to the lower extremity and a diagnosis of diabetes mellitus. The primary outcome was Hospital length of stay, The secondary outcome was infection, number of operations, amputations, and mortality. Subgroup analysis was done to see how HbA1C affect patient’s outcome. A total of 136 patients were included for evaluation. 79% of the patient were male. 84% of the patients were <65yo and the mean age was 54.1yo and HbA1C of 8.99. Scald injury was most common (54%). The average burn size was 2.29% TBSA. 54.5% of the patients had neuropathy. The average hospital length of stay was 15 days ± 14.5 SD. Mean length of stay for all patients at our burn center is 6.55 days/% TBSA burned (internal data). 73.5% of the patients underwent surgical excision. 22% of the patients underwent an amputation and 3.7% were admitted to the intensive care unit with an average ICU length of stay of 7 Additionally, there was 1 inpatient mortality. The study revealed that scald burn was the most common mechanism of burn. Diabetic patients delay seeking medical attention and have extended hospital length of stay, High rate of infection and amputation. Most of these burn injuries and outcomes are preventable with improved education of diabetic patient on foot burn and glycemic control.

DFB: Tiny burn, big problem - The implications of a diabetic foot burn  

Abstract Title: Efficiency of IES device in facilitating distraction enterogenesis for treating short gut syndrome.

Background
Short gut syndrome (SGS) is intestinal failure because of inadequate intestinal length often encountered by pediatric surgeons. In SGS, gut adaptation and enteral absorption is low, often requiring long term parenteral nutrition for adequate development. This poses increased risk for complications including line sepsis, necrotizing enterocolitis, and liver failure. Surgical options like small bowel implantation pose significant risks and often provide limited gut expansion. We aim to experimentally evaluate the efficiency of the intestinal expansion sleeve (IES) in determining lengthening of the native bowel. We hypothesize that intestinal expansion commensurate to the IES compression can be achieved already at time of surgery.

Methods
IES devices were made using proprietary cylindrical material with helicoid trusses and isometric ends measuring 30mm in length. Ends of the device were coated with a barium-sealant mixture preventing distress and improved visualization after placement. IES devices were pre-contracted over a dilator, leading to diametric expansion. Mechanical characterization of IES devices and native small intestine segments were performed using an Instron 8874 Biaxial servo hydraulic fatigue testing system. IES devices were implanted in ex vivo intestinal segments of Sprague Dawley rats. Post-deployment gut expansion was measured using calipers. Lastly, histological analysis of the gut tissue was performed for native and experimental tissue following IES removal.

Results
IES devices deployed in rat small intestine immediately produced lengthening of 2.17 ± .17cm to 2.75 ± .22cm (p<.001, n=11). Mechanical characterization demonstrated IES exerted longitudinal forces of 0.13 ± 0.01 N to 0.22 ± 0.01 N upon expansion to post deployment length. Mechanical characterization of native small bowel demonstrated that forces greater than 1.88 ± .21 N produced rupturing of the tissue. The data proves IES devices provide significant expansion without risk of small intestinal failure. Histological examination showed significantly decreased thickness of serosal, muscular, and mucosal layers of the small intestine. 

Conclusion
This study shows that distraction enterogenesis and intestinal lengthening can be achieved using novel IES devices without the risk of mechanical failure. The findings of this study ex-vivo specimens need further corroboration in vivo for long term outcomes.

Efficiency of IES device in facilitating distraction enterogenesis for treating short gut syndrome.

Background: This study identifies factors associated with an extended hospital length of stay (LOS), exceeding the Center for Medicare and Medicaid Services (CMS) predicted LOS in pediatric transplant patients, based off of patients’ actual coded diagnostic related group (DRG) by hospital billing. DRG codes are not specific to pediatric patients. 

Methods: We conducted a retrospective chart review of 45 patients who received a solid organ transplant at a single academic pediatric hospital between January 2018- December 2021. All patients who underwent transplant surgeries were included regardless of the outcome. Clinical data was collected and analyzed.

Results: The average age was 10 years, 64% male and 44% white. Of the 45 patients, 27 were kidney recipients, 12 heart, and 6 liver. 81%(n=21) of kidney, 75%(n=9) of heart, and 67%(n=4) of liver transplant patients were noted to have exceeded their predicted LOS. The average number of exceeded days was 2.1 for kidney, 104.8 for heart, and 16.6 for liver recipients. 
Factors associated with exceeding LOS in kidney transplant patients included longer operative time (262.9-vs-220.6minutes,(p=0.01)) and greater warm ischemia time (42.1-vs-31.4minutes,(p=0.00)), cold ischemia time was not significantly different. Protective factors for discharge within the predicted timeframe included early transplantation from the time of admission (0-vs-0.27days after admission, (p=0.005)) and live organ transplantation (60%-vs-9%,(p=0.01)). However the relative significance of these factors may be influenced by possible discrepancies in hospital coding. 36.4% of kidney transplant patients who exceeded LOS were on hemodialysis with 31.8% on peritoneal dialysis however only 4.6% received the DRG code Kidney transplant with hemodialysis. DRG codes with hemodialysis have longer expected LOS than without.
Patients who underwent heart and liver transplantation showed minimal significant differences across the analyzed factors, likely due to the limited number of patients who underwent these procedures at our institution. 

Conclusion: Most solid organ transplant patients, regardless of organ transplanted, where noted to have exceeded the expected LOS for their procedure according to CMS billing. Clinical data from these patients’ hospital stay can be utilized to mitigate factors associated with exceeding LOS and should be used to adjust CMS LOS predictions to be specific for pediatric patients.

Factors Impacting Length of Hospital Stay in Pediatric Transplant Patients

Mucoepidermoid Carcinoma (MEC) in Remission with Gefitinib

Background: Acute cholecystitis is commonly treated with laparoscopic cholecystectomy, if feasible. However, critically ill patients can be managed with a percutaneous cholecystostomy tube (PCT) for biliary drainage. This is a temporizing measure and does not represent a final treatment. We aimed to compare the outcomes of performing an interval cholecystectomy vs. conservative management after PCT placement. 
Methods: We conducted a retrospective study of all patients with a PCT placed for acute calculous cholecystitis (ACC) at a single center from October 2010-March 2015. 
Results: Ninety-five patients (median age 79 years) underwent PCT placement for ACC. Only 26.3% of patients underwent interval cholecystectomy. The mortality rate within 6 weeks of ACC diagnosis was higher in the non-surgical group but this was not statistically significant. The number of re-admissions and complications of biliary disease was higher in the surgical group before undergoing interval cholecystectomy, specifically, choledocholithiasis (44% vs. 21.4%; p=0.03). Patients who had an ICU admission [ OR 0.21; 95% CI 0.07-0.66] or increased age [OR 0.94; 95% CI 0.89-0.98] were less likely to undergo interval cholecystectomy. Nine cholecystectomies were performed laparoscopic, 6 were converted to open and 10 were done in an open fashion. The complication rate in the first 30 days after cholecystectomy was 56%. Most complications occurred when surgery was performed in an open or converted to open technique (11/14=78.6%) vs. laparoscopic (3/14=21.4%).
Conclusion: The relatively high incidence of biliary complications in the period between PCT placement and cholecystectomy calls for timely cholecystectomy in patients treated with PCT. When feasible, patients with a history of ACC should undergo interval cholecystectomy. 


What happens after percutaneous cholecystostomy tube for acute calculous cholecystitis?

Background: Targeted radionuclide therapy (TRT) is an emerging treatment for metastatic disease that delivers systemic radiation via a therapeutic radionuclide-linked, tumor-targeting vector. We have previously shown that low dose radiation delivery with 90Y-NM600 improves tumor response to immune checkpoint inhibitors (ICIs). NM600 is selectively taken up and retained in murine and human tumors. Understanding the effect of different radionuclide physical properties (emission type, linear energy transfer (LET), half-life, and tissue range) on immunomodulation of immunologically cold cancers can improve clinical treatment and avoid immunosuppression encountered with systemic external beam radiation therapy (EBRT). Here, we study 90Y, 177Lu, and 225Ac in immunologically cold syngeneic murine tumor models, MOC2 head and neck squamous cell carcinoma and B78 melanoma, monitoring cGAS/STING mediated type 1 interferon (IFN1) responses. We hypothesized that the radionuclide’s physical properties would differentially impact immunomodulation.
Methods: 90Y, 177Lu, or 225Ac were added to culture media in activities estimated by the Medical Internal Radiation Dose method to deliver 12 Gy (MOC2) or 4 Gy (B78) to the cell monolayer by an infinite timepoint. qPCR was performed on cDNA from cells that were irradiated with EBRT, 90Y, 177Lu, or 225Ac, and harvested on days 1, 3, or 7. Mice bearing B78 WT or Tmem173 -/- CRISPR deletion B78 (STING KO) tumors were randomized to receive 1.5 Gy EBRT, an equivalent tumor dose of 90Y- or 177Lu-NM600 determined by the Monte Carlo-based RAPID platform, 0.5 μCi 225Ac-NM600, or no radiation on day 1. Tumors were harvested on days 4, 7, and 10 for RT-qPCR.
Results: Expression of immune susceptibility markers (Mhc1, Pdl1) and IFN1 response- associated genes (Ifnβ1, Mx1) was upregulated in MOC2 and B78 cells following EBRT/TRT, compared to controls. The timing and magnitude of these effects correlated with radionuclide half-life and LET. 
Conclusion: These studies show the capacity to deliver immunomodulatory radiation to tumors using γ-, β- or α-emitting TRT. Understanding TRT effects on the tumor microenvironment may pave integration of TRT and immunotherapies into clinical care to enhance anti-tumor immunity.

90Y-, 177Lu-, and 225Ac-NM600 dose rate effect on immune response in murine tumor models 

Background: Head and neck squamous cell carcinoma (HNSCC) is the most common cancer of its anatomical region. Cisplatin-based therapies are the gold standard of treatment for both HPV- associated and non-HPV-associated etiologies. The treatment regimen includes either 3 weeks of high-dose cisplatin or 6 weeks of low-dose cisplatin, both with concurrent radiation. Malnutrition, secondary to decreased swallowing function, poses a significant challenge to HNSCC treatment. Swallowing function is a significant marker for prognosis and outcome of patients being treated for HNSCC, but the effects of cisplatin dosage on swallowing function have not been studied. The present study sought to understand how each cisplatin regimen affected swallowing function.
Methods: A cohort of 237 patients who were treated for HNSCC at the University of South Dakota Sanford Medical Center in Sioux Falls, SD, between 2010 and 2019 were studied via retrospective electronic medical record review; 109 received high-dose and 128 received low- dose cisplatin. Gastrostomy tube dependence at various timepoints was used as the primary endpoint given that it serves as a proxy for swallowing function. Secondary endpoints included weight changes, esophageal stricture, lymphedema, hospital admission, and overall
survival. Descriptive statistics were used to summarize demographics and treatment variables. Bivariate associations between independent and dependent variables were analyzed. Chi-square tests were used for categorical variables and t-tests for continuous variables. Statistical tests were two-tailed and the significance level set at p <0.05.
Results: Both the primary and secondary outcomes were not statistically significant.
Ototoxicity and renal toxicity were significantly higher in the high-dose group. Ototoxity overall was significantly higher (p=0.0005), with audiology testing and referral for hearing aids both being significantly higher (p<0.001 and 0.0003, respectively). Renal toxicity was significantly higher at the 30-day and 1-year post-treatment timepoints (p=0.0005 and p=0.02, respectively).
Conclusion: This study suggests that swallowing function is not different in patients receiving high vs. low dose cisplatin for HNSCC treatment. Although the significant data on ototoxicty and nephrotoxicity supports already established conclusions regarding the toxicity of high-dose cisplatin, it strengthens our lack of statistical significance in the primary and secondary endpoints.

Assessment of Swallowing Function in Patients with Head and Neck Squamous Cell Carcinoma in High vs. Low Dose Cisplatin

Background: Historically, central nervous system (CNS) involvement in the setting of acute myeloid leukemia (AML) was identified on the basis of symptoms, including headaches, cranial nerve deficits, and mental status changes. In recent years, screening recommendations have been developed to include FLT3 gene mutations, monocytic features, hyperleukocytosis, and CNS symptoms. However, the incidence, prognosis, and optimal treatment strategy are not well described.

Methods: This study was conducted through a retrospective electronic medical record review at the University of Kansas Medical Center from years 2015-2020. The inclusion criteria were AML patients who received a diagnostic lumbar puncture. Data included the overall incidence of CNS involvement among adult AML patients, AML characteristics associated with CNS involvement, and number of intrathecal treatment doses required for CNS clearance. The protocol was approved by the University of Kansas Medical Center Human Subjects Committee. 

Results: Forty-eight patients met our inclusion criteria; 25 were CNS-positive and 23 were CNS-negative (overall screening detection rate of 52%). The incidence of CNS involvement among only newly diagnosed AML patients was 3.24%. Indications for lumbar puncture included 26 patients with CNS symptoms, 24 with a monocytic phenotype, 16 with the FLT3 mutation, and 14 with hyperleukocytosis, with many having more than one characteristic concerning for CNS involvement. Screening detection rates were 58% for patients with 3 characteristics suggestive of CNS involvement, 60% for 2 characteristics, 45% for CNS symptoms, and 30% for monocytic phenotype. The CNS-positive group consisted of 28% with 3 characteristics, 24% with 2 characteristics, 20% with CNS symptoms, 12% with monocytic phenotype, 4% with FLT3 mutation, and 4% with hyperleukocytosis. Seventy-two percent of the CNS-positive group achieved confirmed CNS clearance with 72% of those patients only requiring 1 or 2 doses of intrathecal chemotherapy.

Conclusion: This data indicates that current screening guidelines appear to properly identify a patient population at risk for CNS involvement. Due to the diverse patient presentations in this analysis, these results also emphasize the need for ongoing research to further delineate risk factors for CNS involvement in AML. Improved identification of risk factors will likely lead to earlier detection and improved patient outcomes.

Downloads

Next from AMA Research Challenge 2022

Analysis of Racial/Ethnic Minority and Female Representation in theSurgical Oncology Training Pathway

Similar lecture

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES