Scientific Abstract Presentations: Research & Technology 2 Thursday, October 26 Thu, Oct 26 3:00-4:30 p.m. CDT

Background Matching into an integrated plastic surgery residency program is highly competitive, with a limited number of available seats and a highly qualified applicant pool. Many medical students elect to complete research fellowships to increase their research productivity and strengthen their residency applications. This study aimed to understand the process of identifying opportunities, financial considerations, and overall satisfaction with research fellowships.

Methods A national survey was distributed to integrated plastic surgery residency applicants in the 2020-2021 and 2021-2022 application cycles. The survey elicited information regarding demographics, resources utilized in identifying positions, as well as motivators and deterrents for pursuing research fellowships. Questions related to the productiveness of the fellowship as well as perceived benefit of the experience were also included.

Results Five hundred thirty-four integrated plastic surgery applicants were identified from our institution's records and included in our study. Our preliminary results revealed that out of 52 respondents, 19 completed research fellowships (37%). The average fellowship length was 11.6 months, with most completing their research at another institution (63%) between the third and fourth year of medical school (84%). Respondents utilized mentors (83%), previous applicants (48%), and other medical students (35%) to help with the decision to pursue or forgo additional time for research. All respondents, whether they chose to complete a research fellowship or not, considered publications (88%), opportunities for letters of recommendation (63%) and networking (52%) as the top motivational factors. Conversely, aspects that adversely impacted applicants' decisions included an additional year of education (83%), funding concerns (63%), and a delay in generating income (40%).

The most common avenues to identify available research positions were from online listings (42%), word of mouth (37%), and suggestions from mentors (37%). Over half of research fellows received full funding (58%), with the remainder receiving partial (16%) or no funding (26%). The average salary for a research fellowship was $27,900, with the main sources of funding being the institution where the fellowship was completed. Productivity was assessed by the average number of accepted or published publications (11.8 papers), oral presentations (6 presentations), and poster presentations (2.9 presentations) produced during dedicated research time. Overall, most respondents who completed a research fellowship found that a fellowship was worthwhile (83%), made them a stronger applicant overall (94%), and would recommend this opportunity to prospective applicants (72%).

Conclusion As the applicant pool for plastic surgery continues to expand, many candidates have been motivated to pursue dedicated research fellowships in attempts to make their residency applications stand out. Overall, almost all who completed research fellowships found their experiences worthwhile and recommended them to future applicants. However, the majority of prospective trainees have concerns about funding these endeavors and, for many applicants, the drawbacks still outweigh the benefits.

Purpose: Despite substantial technical advances that make vascularized composite allotransplantation (VCA) viable for patients with devastating soft tissue injuries, acute rejection remains a major morbidity, with approximately 85% of patients experiencing a rejection episode within the first postoperative year(1). The molecular mechanisms of VCA rejection have not yet been fully elucidated, including its potential variation across anatomical sites. While there is some data to suggest disparate histopathological findings between oral mucosa and skin(2,3), differences between the face and hand remain unexplored.

Methods: NanoString RNA sequencing (RNAseq) was performed on 27 FFPE skin biopsies (16 face, 11 hand) from 3 VCA recipients; 7 were categorized as nonrejection (NR) and 20 as acute rejection (AR) by clinical status. Data analysis and visualization was performed using the NanoTube and ggplot2 packages in R, in addition to Metascape for pathway analysis.

Results: Upon principal components analysis (PCA), AR samples grouped separately from NR samples, without discernable clustering by patient or anatomical site. These findings were confirmed with hierarchical heatmap clustering of gene fold changes. We then compared fold changes from AR face biopsies to AR hand biopsies, which yielded a strong positive correlation across anatomical site (r≥0.38, p<0.05). Indeed, AR face and AR hand biopsies each exhibited 151 DEGs when compared to NR samples, with 127 of these DEGs overlapping between the face and hand. With such similarities in rejection between anatomical sites, we aggregated face and hand samples for downstream differential expression analysis. We found that differentially expressed genes in AR were highly enriched for adaptive immune response, leukocyte activation, and chemokine pathways (p<0.05). Finally, using a regularized regression model, we determined 7 genes that most significantly correlated with clinical rejection: CCL5, NKG7, CD8A, CD38, FASLG, SH2D1A, and KLF2.

Conclusions: Our results demonstrate that the molecular mechanism of acute rejection in VCA appears to be conserved across different recipient anatomical sites. Antigen-activated T cells and NK cells mediate AR through production of cytokines and recruitment to graft tissue. In particular, cytotoxic and apoptotic markers such as CD8A, NKG7, and FASLG are highly diagnostic for clinical evidence of graft damage. Longitudinal sampling for RNAseq will identify key genes driving the mechanism of VCA rejection and inform targeted therapies.

References: (1) Petruzzo P, Lanzetta M, Dubernard JM, et al. The International Registry on Hand and Composite Tissue Transplantation. Transplantation. 2010 Dec 27;90(12):1590–4 (2) Bergfeld W, Klimczak A, Stratton JS, et al. A Four-Year Pathology Review of the Near Total Face Transplant. Am J Transplant. 2013;13(10):2750–64. (3) Kanitakis J, Badet L, Petruzzo P, et al. Clinicopathologic monitoring of the skin and oral mucosa of the first human face allograft: Report on the first eight months. Transplantation. 2006 Dec 27;82(12):1610–5

Introduction

Resident salaries have remained stagnant. The 2022 Medscape Report states a 3% increase from 2017-2020, well below the 5.6% increase in Consumer Price Index. The Medscape report suggests most residents feel inadequately compensated, stating that compensation neither reflects the number of hours worked nor meets residents' local cost of living. During residency, many trainees face financial stressors, decreasing quality of life, and increasing burnout and attrition. Furthermore, applicants to plastic surgery cite resident compensation as a core factor in rank-order. As such, it is necessary to investigate the true value of resident stipends as compared to local costs of living.

Methods

Resident stipend information for PGY levels 1-6 was collected directly from plastic surgery residency program websites or their institutional GME stipend pages. Programs without available data were excluded from analysis. Stipends were adjusted by the cost of living index (COLI) of each program's city, yielding an adjusted effective stipend. Differences between raw and adjusted resident salary were calculated and descriptive and statistical analysis was performed. Data was grouped and analyzed by US Census Bureau geographic region (Northeast, South, Midwest, and West). Statistical analysis was performed using Python and significance was set to p < 0.05.

Results

Most residency programs were included in analysis (85 of 88); compensation data was not available for three programs. The mean annual stipends for PGY 1-6 were $63,235, $65,682, $68,224, $71,064, $73,960, and $76,806, in increasing PGY-level order. The average COLI-adjusted effective salary was $60,065, $62,330, $64,663, $67,314, $70,020, and $72,683, respectively. In the Northeast (p = 0.044) and West (p < 0.001) regions, the COLI-adjusted effective stipends were significantly lower than the nominal value of the stipend, suggesting reduced purchasing power for residents in these regions. In contrast, for the Midwest region (p = 0.003), the COLI-adjusted effective stipends were significantly greater than the nominal value of the stipend, suggesting increased purchasing power for residents in these regions.

Residents in programs in the West received the lowest COLI-adjusted average stipends, at $46,342, $47,965, $49,847, $51,976, $54,175, and $56,444 respectively. The difference between unadjusted and COLI-adjusted stipends was -28.8%. Residents in the Midwest received the greatest COLI-adjusted average stipends, at $68,610, $71,162, $73,475, $76,288, $79,319, and $82,376 respectively. The difference between unadjusted and COLI-adjusted stipends was +10.5%. The Midwest was the only region in which residents had an increase in effective stipend based on COLI.

Conclusion

Residency stipends do not currently account for COLI, leading to large discrepancies in resident compensation. After adjusting for COLI, resident stipends in the Northeast and especially the West significantly lose value, while stipends in the Midwest gain value, suggesting that residents in Northeast or West are underpaid. This may lead to financial distress and burden for these residents, especially regarding student loan debt. Future research should investigate COLI-adjusted under-compensation versus resident wellness/burnout, quality of life, and financial burden by region. Such information may allow future applicants to more appropriately plan their rank-lists, and may allow programs to advocate for greater resident reimbursement.

Purpose

Of the procedures that make up facial feminization surgery, frontal sinus setback has a particularly high impact on gender perception. Mixed reality (MR) is a nascent technology that holds significant promise in plastic and reconstructive surgery. MR allows a user to view and manipulate three-dimensional patient images while superimposing them on the patient. This method allows for direct visualization of deep structures, improving a surgeons understanding of vital patient anatomy in real time. To the best of our knowledge, this is the first usage and evaluation of this technology inside of a plastic surgery operating room in the United States.

Methods

The Medivis SurgicalAR system was used in conjunction with the Microsoft HoloLens, an MR headset with a see-through visor. CT imaging was uploaded to the SurgicalAR system and a three-dimensional hologram was projected onto the display of the HoloLens. The CT was registered to the patient using a point-to-point framework that relied on bony fiducials identified intra-operatively, namely the supraorbital notches, nasion, and glabella, matched to virtual counterparts. The system relies on a localizing wand with an affixed optical code that the 2D RGB camera on the headset tracks. Time measures and discrepancy from our standard-of-care 3D cutting guide were measured along with survey of the operating surgeon.

Results

Qualitative descriptions demonstrated that 3-dimensional visualization of deep structures improved surgeon confidence and operative decision making. The process of matching the hologram to the patient and cropping to see intended structures took three minutes and twelve seconds. Tracing of the frontal sinuses based on the hologram took 61 seconds. Maximum discrepancy from the 3D cutting guide was 5mm and minimum was exactly the same. In addition, the workflow that was established was both efficient and intuitive.

Conclusion

Mixed reality was shown to be accurate in superimposing a patient's CT on top of their actual skull during surgery, allowing for a tracing of the frontal sinuses. This rapidly developing technology demonstrates promise for being a viable intraoperative image guidance technology and may provide a faster and more effective method of anatomical identification than the current standard of care.

Background: ChatGPT is an artificial intelligence (AI) model that can generate text and provide responses to questions, including answers that include medical knowledge and advice. Prior studies have established that text written by the AI model is often indistinguishable from human generated text. The purpose of this study was to gauge public opinion on responses to a medical question written by actual medical professionals versus simply generated by ChatGPT.

Methods: Amazon's Mechanical Turk crowdsourcing service and REDCap's survey manager were used to recruit survey participants and collect responses. An anonymous 13-question survey was distributed that provided two different responses to a plastic surgery-related question about which patients are appropriate candidates for liposuction. One response was written by a medical professional while the other was generated by artificial intelligence (AI) machine ChatGPT. Responses were graded for various qualities and participants chose if they thought the response was written by AI or a medical professional. Five-point Likert scales were converted to binary variables for tabulation and a logistic regression analysis was performed.

Results: A total of 578 participants were included for analysis. All participants were in the United States and had a mean age of 35.3 years. When assessing both responses on warmth, conciseness, thoroughness and clarity, there were no statistically significant differences between medical advice provided by AI writers versus medical professionals (p>0.05). Overall, only 41% of the public was correctly able to identify ChatGPT's response as having been written by AI. The public was better able to identify that the medical professionals' response was human generated (70% vs. 41%, p<0.0001). Further, as respondent age increased, there was a significant increase in the ability to correctly identify AI-written text (p = 0.00082). Oppositely, as age increased, there was also a significant decrease in the ability to correctly identify that a response was written from a medical professional (p = 0.00082).

Conclusion: When assessing responses to a plastic surgery-related question, respondents felt that the answers from both a medical professional and from ChatGPT were equally warm, concise, thorough and clear. Likewise, less than half of the public could identify a response as AI generated. However, significant limitations to ChatGPT have been made apparent, including the possibility for incorrect or misleading information. Therefore, considering its ability to produce convincing responses, increased vigilance by healthcare professionals is necessary to ensure that medical information relayed to the public is accurate.

Introduction: Autologous fat grafting is a common technique used to enhance aesthetic outcomes in post-mastectomy breast reconstruction patients. Adipokines are hormones secreted by adipose tissue that play a critical role in regulating metabolic processes and the immune system.1-2 However, dysregulated adipokine secretion and signaling can contribute to the development and progression of cancer by promoting angiogenesis, altering the immune response, and inducing the epithelial mesenchymal transition, which allows breast cancer cells to subsequently metastasize and become endocrine-resistant.1-3 Past meta-analyses on the oncologic safety of fat grafting did not show significant effects on loco-regional recurrence or patient disease free survival.4 We aimed to assess how breast cancer cells behave in conditioned media derived from fat grafting lipoaspirates, and gain a better understanding of the potential interactions that may occur within the tumor microenvironment.

Methods: A prospective randomized control trial (ClinicalTrails.gov identifier: NCT04891510) that enrolled post-mastectomy breast reconstruction patients is conducted at our center. Subjects are randomized into one of three fat grafting methods (active filtration, low pressure decantation, and standard suction decantation) in a 1:1:1 ratio. The extracted lipoaspirate from each patient is incubated in starving media for 24 hours. Conditioned media is then created using 20% of the secretome and 80% starving media. MCF-7, a human ER/PR+ breast cancer cell line, is plated at a density of 2000 cells per well and allowed to proliferate for four days. CyQUANT Cell Proliferation Assay, which utilized fluorescence binding to nucleic acid, is performed for growth quantification. Statistical analysis is done using One-Way ANOVA.

Results: MCF-7 cells incubated in lipoaspirates processed using active filtration, low pressure decantation, and standard decantation did not show significant differences in their proliferation patterns. Across all treatments, the breast cancer cells incubated in conditioned media showed similar growth trends as those in complete media, which is enriched to contain the growth requirements of the cell line. MCF-7 cell behavior in conditioned media differed significantly from their proliferation patterns when serum starved in 100% starving media (p < 0.05).

Conclusion: Existing clinical series with or without control group offered conflicting evidence on whether fat grafting increases the risk of loco-regional recurrence and distal metastases, and it is unclear whether the interplay of autocrine and paracrine factors produced by transplanted lipoaspirate is linked to cancer recurrence.5 Based on the in-vitro assessment of luminal breast cancer cell behavior in conditioned media, it seems that the microenvironment can support cell proliferation. This does not necessarily indicate comparable patterns will occur in-vivo. However, it could influence the ideal time point for autologous lipid transfer, and potentially suggest that revisions should be postponed until after the complete course of breast cancer treatment. Additionally, unlike mastectomy patients, who achieve immediate regional disease control after resection, breast conservation therapy (i.e., lumpectomy) patients may not be optimal candidates for fat transplantation before the completion of their treatment course.

References 1. Kang C, LeRoith D, Gallagher EJ. Diabetes, Obesity, and Breast Cancer. Endocrinology. 2018 Nov 1;159(11):3801-3812. doi: 10.1210/en.2018-00574. PMID: 30215698; PMCID: PMC6202853. 2. Bhardwaj P, Au CC, Benito-Martin A, Ladumor H, Oshchepkova S, Moges R, Brown KA. Estrogens and breast cancer: Mechanisms involved in obesity-related development, growth and progression. J Steroid Biochem Mol Biol. 2019 May;189:161-170. doi: 10.1016/j.jsbmb.2019.03.002. Epub 2019 Mar 6. PMID: 30851382; PMCID: PMC6502693. 3. Hass R, Otte A. Mesenchymal stem cells as all-round supporters in a normal and neoplastic microenvironment. Cell Commun Signal. 2012 Sep 3;10(1):26. doi: 10.1186/1478-811X-10-26. PMID: 22943670; PMCID: PMC3444900. 4. Sun J, Liang H, Lin D, Han B, Zhang T, Gao J. Oncological safety of reconstruction with autologous fat grafting in breast cancer patients: a systematic review and meta-analysis. Int J Clin Oncol. 2022 Sep;27(9):1379-1385. doi: 10.1007/s10147-022-02207-8. Epub 2022 Jul 5. PMID: 35790652.

Purpose: Much of our understanding of acute rejection in vascularized composite allotransplantation (VCA) is drawn from principles in solid organ transplantation. However, there remains much speculation given the unique immunogenicity of skin and involvement of multiple tissue types. Current studies suggest the implication of T cells in VCA rejection, though the exact molecular mechanisms have not yet been elucidated.

Methods: Single cell RNA sequencing (scRNAseq) was performed on 10 fresh skin biopsies (5 face, 5 hand) from one VCA recipient; 2 were categorized as nonrejection (NR) and 8 as acute rejection (AR) by clinical status. Spatial transcriptomics was performed on 8 FFPE skin biopsies (7 face, 1 hand) from 3 VCA patients; 3 were NR and 5 were AR. Analysis and visualization was performed using the Seurat and Monocle3 packages in R.

Results: We identified 13 clusters on scRNAseq corresponding to 14 distinct cell types. Spatial transcriptomics demonstrated colocalization of macrophages, T cells, and NK cells near the basement membrane during a rejection episode. We then focused on the T and NK cell cluster, further subsetting these cells by relative expression of key genes in the literature. Specifically, we identified 4 subsets of CD8 T cells: GZMKlow/GZMBhigh, GZMKhigh/GZMBlow, GZMBhigh/CX3CR1high, and "transitional" CD8 T cells, named for their lack of strongly expressed markers and indistinct clustering. To better understand the dynamic states of CD8 T cells in VCA rejection, we performed trajectory analysis starting at GZMKhigh/GZMBlow T cells, which did not express the exhaustion marker LAG3 or cytotoxic markers. Trajectory analysis revealed a shared path through transitional CD8 T cells which then diverged into 2 distinct endpoints: GZMKlow/GZMBhigh and GZMBhigh/CX3CR1high cells.

Conclusions: Our analyses demonstrate colocalization of T cells and macrophages near the basement membrane and the involvement of dynamic CD8 T cell states in VCA rejection. Cytotoxic markers such as GZMB and CX3CR1 are seen, however the role of GZMKhigh/GZMBlow-expressing cells is largely unknown. Future studies aim at clarifying CD8 T cell differentiation and their interaction with other cell types during AR.

PURPOSE Characterizing tumor cell behavior within patient-specific in vitro models of the breast milieu is critical to advancing therapeutics, particularly with the advent of increasingly personalized treatment options. Challenges to developing such platforms include successfully isolating and co-culturing the many primary cell types that constitute the tumor microenvironment and creating analytic methods that allow for reliable, accurate characterization of cell behavior over time. While prior studies have cultured tumor cells alongside patient-derived breast cells in a 3D extracellular matrix, we describe an engineered platform with a high fidelity to the tumor microenvironment due to its inclusion of mature patient derived adipocytes, an important contributor to tumor angiogenesis. We further characterize a novel machine learning-based image analysis approach.

METHODS & MATERIALS Stromal vascular fraction (SVF), organoids and mature adipocytes were isolated from breast tissue obtained from healthy female patients undergoing reduction mammoplasty. Isolated cells were embedded in non-ribosylated collagen, creating a biomimetic milieu that approximates the patient-specific in vivo cellular environment ("biomimetic collagen"). 3D collagen constructs consisting of a bottom layer of RFP-tagged MDA-231 tumor cell-embedded collagen followed by a layer of plain or biomimetic collagen were plated in triplicate on a 96-well plate. GFP-tagged human umbilical vascular endothelial (HUVEC) cells were plated in a monolayer layer on top of the collagen to mimic the endothelial barrier. Constructs were cultured at 37°C and 5% O2 and underwent confocal imaging on days 0 and 7. Vertical cell movement was measured in the z-axis; cells were considered "migrated" when z-axis position was more than one standard deviation above the mean z-axis position. Confocal image analysis was completed in Imaris; migration analysis and statistics were completed in RStudio. P-values of less than 0.05 were considered significant.

RESULTS Confocal microscopy at day 0 revealed ~800μm thick constructs with distinct MDA-231 and HUVEC cell layers. The custom-built imaging and machine learning algorithm accurately parsed fluorescent cells from surrounding background collagen and SVF autofluorescence, allowing for precise measurement of migration distances between timepoints. Analysis of confocal images at day 7 revealed MDA-231 in all collagen platforms, with significantly greater tumor cell migration in plain collagen groups than biomimetic groups (76.78μm versus 71.33μm, p < 0.05). Discrete migrating MDA-231 cell clusters associated with migrating HUVECs were observed in plain collagen and biomimetic collagen construct with average migration distances of 98.65 and 68.14, respectively (p < 0.05).

CONCLUSION This tissue engineered biomimetic platform design and image analysis allows for reliable, reproducible and precise characterization of tumor and endothelial cell interactions in the tumor microenvironment. Findings of cell migration in both plain and biomimetic collagen groups indicates that the angiogenic capacity of tumor cells preserved in each condition. Increased migration in the plain collagen groups indicates that the biomimetic milieu alters the dynamics of tumor cell invasion; this relationship will be further elucidated in future studies. Importantly, this platform's utilization of patient-derived SVF and mature adipocytes makes it inherently personalized, a vital component of any future clinically-relevant in vitro cancer platforms.

PURPOSE: The advent of the internet has significantly increased the amount of medical information that is available to doctors, scientists, and researchers who are seeking information about specific topics. Unfortunately, this information is increasingly scattered in disparate locations online. This leads to challenges in identifying and accessing the accumulated knowledge about a given topic, and thereby limits the additional research that can be performed. As proof-of-concept for a new approach in developing topic-specific research mini-databases, the software program DataTables was used to create a comprehensive electronic database for all publications regarding deep inferior epigastric perforator (DIEP) flap breast reconstruction.

METHODS: Three databases (Medline via both the PubMed and OVID interfaces, SCOPUS, and CINAHL) were exhaustively searched for every publication that referenced DIEP flap breast reconstruction. Full text articles were obtained and transported to Covidence, where inclusion and exclusion criteria were applied using a two-person reference review method. Included publications were exported to Mendeley which allowed further classification of these references. DataTables software was used to create a customized, user-friendly user-interface by which the mini-database can be accessed, sorted, and queried.

RESULTS: The first mini-database has been developed, contains 2,635 publications, and is freely accessible as a research tool at www.DIEPDIVE.org. Users are provided detailed information about each article, a single-click link to the article, a search tool, a filter tool, and multiple sorting capabilities. The mini-database is continuously maintained in order to stay up-to-date.

CONCLUSION: This topic-specific mini-database serves as a proof-of-concept of a novel type of research tool for studying and conducting research on specific topics. With further refinement and expansion of the concept, it can serve as a template by which countless other topics can be better studied in plastic surgery, and beyond.

Introduction: The cancellation of clinical rotations (CRs) and implementation of virtual interviews (VIs) during the COVID-19 pandemic profoundly affected the 2020-2021 residency application cycle. Many educators have raised concerns regarding the potential impact that the aforementioned changes had on the subjective characteristics (such as personality and program fit) and clinical skills of the 2021 intern class, however, no outcomes data has been published leaving much of this discussion left to conjecture. The aim of this study is to describe the impact that the cancellation of CRs and implementation of VIs had on the clinical skills and subjective characteristics exhibited by the 2021 intern class from the perspectives of program directors (PDs) of integrated plastic surgery residency programs.

Methods: A 13-question survey was administered to PDs of ACGME-accredited integrated plastic surgery residency programs with a "continued-accreditation" status using REDCap© (Vanderbilt University, Nashville, TN) from March 2022 through May 2022. Questions were designed to collect information regarding program characteristics along with PDs' outlooks on their 2021 intern class's subjective characteristics and clinical skills relative to previous resident classes, the relationship between cancelled CRs and incoming residents' clinical competency, and the utility of VIs for evaluating applicant subjective characteristics. An optional free-response text box was included for respondents to provide additional comments. Responses were collected in a deidentified fashion.

Results: A total of 32 (43.8%) PDs of included integrated plastic surgery residency programs responded to our survey. Programs in our sample offered a median of 2 [Interquartile Range (IQR), 2-3] training positions during the 2020-2021 residency application cycle. Twenty-six (81.3%) programs offered more than one training position with 52.4% and 19.0% of said programs taking approximately the same number or more home medical students in the 2021 NRMP Match, respectively. The majority of respondents observed no differences in their 2021 intern class's fit with their program (68.8%), communication skills (65.6%), responsiveness to clinical instruction and feedback (68.8%), work ethic (75.0%), and rotation evaluations (78.1%) when compared to previous resident classes. Five (15.6%) responding PDs reported that they received more complaints about their intern class that required disciplinary action, however, no increases in resident attrition rates were observed. Twenty-five percent of PDs reported that VIs negatively impacted their program's ability to accurately assess applicants' subjective characteristics. Interestingly, 37.5% of PDs reported that the cancellation of CRs in 2020 negatively affected their 2021 intern class's clinical competency at the start of residency; however, only 21.9% of respondents observed that their 2021 intern class exhibited poorer clinical skills when compared to previous resident classes one year following the MATCH 2021.

Conclusions: The outcomes of MATCH 2021 suggest that VIs limited selection committees' ability to accurately assess applicant's subjective characteristics to a lesser degree than previously described in the literature. Cancellation of CRs had a significant impact on the 2021 intern class's clinical skills at the start of residency; however, many residents rectified these shortcomings during their first year of clinical training.

Background Cell injury caused by ischemia and ischemia-reperfusion injury (IRI) affects most reconstructive surgical procedures including flap surgery, revascularization and replantation of composite tissues and vascularized composite allotransplantation. Compared to muscle and nerve, fewer studies have focused on pathological skin changes during ischemia and/or reperfusion. The aim of this study was to review the nature and timeline of pathological, molecular and genetic changes during skin ischemia and subsequent reperfusion to provide a guide for selection of appropriate markers in studies on skin ischemia and reperfusion injury. Methods A systematic review was performed in accordance with PRISMA guidelines. Studies investigating skin ischemia were included. Model type, experimental intervention, ischemia method and duration, reperfusion duration, biopsy location and time point, staining method, gene and protein expression (PCR and western blot) were collected. Results Ninety-nine articles were included. Hematoxylin and eosin (H&E) was the most used assessment, with three categories of changes identified: inflammatory, structural, and vascular. H&E showed inflammatory infiltration in early responses (12-24 hours), with structural modifications (3-14 days) and neovascularization (5-14 days) as delayed responses. Immunohistochemistry (IHC) was used to evaluate for angiogenesis and apoptosis. Cluster of differentiation 31 (CD31), cluster of differentiation (CD34), and vascular endothelial growth factor (VEGF) were used as angiogenic markers, while terminal deoxynucleotidyl transferase dUTP Nick-End Labeling (TUNEL), caspase-3, Bcl-2 associated x protein (Bax), and B-cell lymphoma 2 (Bcl-2) were used as apoptosis markers. Gene (PCR) and protein expression (western blot) were used to detect (1) inflammation and apoptosis; (2) Endoplasmic reticulum stress/oxidative stress and hypoxia; and (3) neovascularization. Gene and protein inflammatory markers were elevated with tumor necrosis factor α (TNF-α, increased 6 hours – 7 days) the most common, followed by Interleukin 6 (IL-6, increased 12 hours – 7 days) and Interleukin 1β (IL-1β, increased 1 – 7 days). The most common gene and protein apoptosis markers were Caspase-3 (increased 24 hours – 7 days), apoptosis signal-regulating kinase 1 (ASK-1, increased 12 – 14 hours), and the Bax/Bcl-2 ratio (increased 24 hours – 7 days). Hypoxia and neovascularization were evaluated with VEGF (increased 6 hours – 10 days) and HIF-1α (increased 4 hours – 7 days). Oxidative stress kits measured malondialdehyde (MDA), glutathione (GSH), glutathione peroxidase (GSH-Px), catalase (CAT), and superoxide dismutase (SOD) with changes observable after 24 hours of reperfusion. Conclusion Compared to other components of the composite organs, skin is less susceptible to ischemia due to low metabolic demands. H&E histology may show inflammatory cell infiltration (12-24 hours) in reperfusion injury, however structural and neovascular changes are delayed responses (3-14 days). While IHC may detect apoptosis (6 hours – 7 days) and angiogenesis (5-14 days), gene and protein expression of inflammatory, hypoxic, and apoptotic biomarkers may provide more quantitative and earlier insight into immediate cellular-level responses in skin ischemia.

Introduction: Presentations at national meetings are an important means of knowledge generation. Eventual publication of these studies is important for dissemination of findings beyond meeting attendees. Given growing emphasis on research productivity, a longitudinal evaluation of abstracts at Plastic Surgery The Meeting (PSTM) is warranted with critical attention on factors that improve odds of successful abstract conversion. We analyzed a 10-year sample of presented abstracts at PSTM and describe factors that improve rate and speed of conversion to peer-reviewed publication. Methods: Abstracts presented at PSTM from 2010-2019 were sourced from the ASPS Abstract Archive. A random sample of 100 abstracts from each year was evaluated. If fewer than 100 abstracts were available for a given year, all were evaluated. Demographics of primary presenter, academic affiliation, and subspecialty track were collected. Doximity research ranking was recorded for US-academic-affiliated institutions. The title of each abstract was used to find a corresponding published paper on PubMed, Google Scholar, and Google. Time to publication, journal, and impact factor at time of publication were recorded. Data were analyzed for trends over time and factors affecting conversion rate. Results: 983 presented abstracts were included. Overall conversion rate was 54.1%. 14 months was median time to publication. Median impact factor was 2.72, and this did not significantly change during the study period (p=0.09). Most common geographic region of presenters was the northeast (25.1%). Affiliation of the primary presenter changed over the study period, with an increasing number of private practice and international presenters (31% vs 52%) (p<0.001). Residents and fellows were the largest proportion of presenters (38.4%); however, there was an increase in the number of medical student and research fellow presenters during the study period (7.0% vs 21%) (p<0.001). Conversion rate was not affected by the Doximity research ranking of a presenter's affiliated institution (β=1.001, p=0.89), geographic location (p=0.60), or subspecialty tract (p=0.73). US-academic programs had a higher conversion rate (61.8%) than US-non-academic (32.7%) and international abstracts (47.1%) (p<0.001). Conversion rate differed by level of training of the primary presenter (p<0.001). Medical students had the highest conversion rate (65.6%). Attendings had the lowest rate (45.0). Level of training affected time to publication (p=0.007). Studies presented by research fellows had the shortest average time to publication (11.6+/-10.6 months), followed by Non-MD Researchers (14.65+/-7.72), and Medical Students (14.68+/-12.49). Residents had the longest average time (18.8 months). There was no difference in journal impact factor based on the presenter's level of training (p=0.25) Conclusions: Over half of presented abstracts included in our 10-year representative sample were published. Trends highlight an increasing proportion of students and trainees presenting at PSTM. Medical students as primary presenters and US-academic affiliation increased the rate and speed of successful publication. This increase in medical student publication, with maintained quality of research based on similar journal impact factor to other presenters, highlights the growing demand for high-quality research at the student level.