Microsurgical management of midbrain cavernous malformations: does lesion depth influence the outcome?

Background The purpose of this study was to clarify whether the intrinsic depth of midbrain cavernous malformations (MCMs) influenced the surgical outcome. Methods The authors conducted a retrospective study of 76 consecutive patients who underwent microsurgical resection of a MCM. The vascular lesions were categorized into 4 distinct groups based on how these lesions had altered the brainstem surface. Additionally, it was verified whether the actual aspect of the brainstem surface could be predicted only by evaluating the pertinent preoperative MRI slices. Clinical outcome was assessed by determining the modified Rankin Scale Score (mRS) before and after surgery. Results Twenty-three MCMs (30.3%) were located deeply within the midbrain. The overlying midbrain surface appeared to be normal (group nl). In 33 patients (43.4%), the midbrain surface showed only a yellowish discoloration (group yw). In another 14 individuals (18.4%), the midbrain surface was distorted by the underlying MCM and bulging out while the vascular lesion still remained covered by a thin parenchymal layer (group bg). In the smallest group comprising 6 patients (7.9%), the exophytic MCM had disrupted the midbrain surface and was clearly visible at microsurgical exposure (group ex). The mean mRS decreased in the group nl from 1.43 preoperatively to 0.61 at follow-up. Conclusion This study demonstrates in a large patient population that a deep intrinsic MCM location is not necessarily associated with an unfavorable clinical outcome after microsurgical lesionectomy. Predicting the aspect of the midbrain surface by evaluating preoperative MR images alone was not sufficiently reliable.

A Large Retrospective Assessment of Voriconazole Exposure in Patients Treated with Extracorporeal Membrane Oxygenation

Background: Voriconazole is one of the first-line therapies for invasive pulmonary aspergillosis. Drug concentrations might be significantly influenced by the use of extracorporeal membrane oxygenation (ECMO). We aimed to assess the effect of ECMO on voriconazole exposure in a large patient population. Methods: Critically ill patients from eight centers in four countries treated with voriconazole during ECMO support were included in this retrospective study. Voriconazole concentrations were collected in a period on ECMO and before/after ECMO treatment. Multivariate analyses were performed to evaluate the effect of ECMO on voriconazole exposure and to assess the impact of possible saturation of the circuit’s binding sites over time. Results: Sixty-nine patients and 337 samples (190 during and 147 before/after ECMO) were analyzed. Subtherapeutic concentrations (<2 mg/L) were observed in 56% of the samples during ECMO and 39% without ECMO (p = 0.80). The median trough concentration, for a similar daily dose, was 2.4 (1.2–4.7) mg/L under ECMO and 2.5 (1.4–3.9) mg/L without ECMO (p = 0.58). Extensive inter-and intrasubject variability were observed. Neither ECMO nor squared day of ECMO (saturation) were retained as significant covariates on voriconazole exposure. Conclusions: No significant ECMO-effect was observed on voriconazole exposure. A large proportion of patients had voriconazole subtherapeutic concentrations.

Transmission of severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2) through infant feeding and early care practices: A systematic review

BACKGROUND: Perinatal practices such as breast-feeding, kangaroo mother care, rooming-in, and delayed cord clamping have varied by institution during the COVID-19 pandemic. The goal of this systematic review was to examine the success of different practices in preventing viral transmission between SARS-CoV-2 positive mothers and their infants. METHODS: Electronic searches were performed in the Ovid MEDLINE, Ovid Embase, Cochrane Library, EBSCOhost CINAHL Plus, Web of Science, and Scopus databases. Studies involving pregnant or breastfeeding patients who tested positive for SARS-CoV-2 by RT-PCR were included. Infants tested within 48 hours of birth who had two tests before hospital discharge were included. Infants older than one week with a single test were also included. RESULTS: Twenty eight studies were included. In the aggregated data, among 190 breastfeeding infants, 22 tested positive for SARS-CoV-2 (11.5%), while 4 of 152 (2.63%) among bottle-fed (Fisher’s exact test p = 0.0006). The positivity rates for roomed in infants (20/103, 19.4%) were significantly higher than those isolated (5/300, 1.67%) (P < 0.0001). There was no significant difference in positivity rate among infants who received kangaroo care (25%vs 9%, p = 0.2170), or delayed cord clamping (3.62%vs 0.9%, p = 0.1116). CONCLUSIONS: Lack of robust studies involving large patient population does not allow meaningful conclusions from this systematic review. Aggregated data showed increased positivity rates of SARS-CoV-2 among infants who were breast fed and roomed-in. There were no differences in SARS-CoV-2 positivity rates in infants received skin to skin care or delayed cord clamping.

Characterization of the Immune Cell Infiltration Profile in Pancreatic Carcinoma to Aid in Immunotherapy

The tumor microenvironment (TME) is comprised of tumor cells, infiltrating immune cells, and stroma. Multiple reports suggest that the immune cell infiltration (ICI) in TME is strongly associated with responsiveness to immunotherapy and prognosis of certain cancers. Thus far, the ICI profile of pancreatic carcinoma (PC) remains unclear. Here, we employed two algorithms to characterize the ICI profile of PC patients. Based on our results, we identified 2 ICI patterns and calculated the ICI score by using principal component analysis. Furthermore, we revealed that patients with low ICI scores had a better prognosis, compared to high ICI scores. Moreover, we discovered that a low tumor mutation burden (TMB) offered better overall survival (OS), relative to high TMB. In this study, a high ICI score referred to elevated PD-L1/TGF-β levels, increased activation of cell cycle pathway and DNA repair pathway, as well as reduced expression of immune-activation-related genes. We also demonstrated that three metabolic pathways were suppressed in the low ICI score group. These data may explain why a high ICI score equates to a poor prognosis. Based on our analysis, the ICI score can be used as an effective predictor of PC prognosis. Hence, establishing an ICI profile, based on a large patient population, will not only enhance our knowledge of TME but also aid in the development of immunotherapies specific to PC.

EPEN-47. PEDS: PEDIATRIC EPENDYMOMA DISCOVERY STUDY

The prognosis for pediatric ependymoma remains unaffected by recent discovery. Upfront therapy is maximal surgical resection followed by radiation and the utility of histologic diagnosis remains unreliable. Nine molecular subgroups and possible genetic drivers of ependymoma have been identified, but the implementation of these findings into targeted therapy and stratified clinical trials has not occurred. It is imperative that researchers work collaboratively to move discovery towards clinical testing. Heterogeneity of ependymoma requires that we collect a large amount of data; progress in the field is dependent on deep analysis of this information. As we further subclassify ependymoma, it will be important to have a large patient population for enrollment onto clinical trials, which will maximize data collection and the amount of materials available for experimentation and analysis. Researchers in the United States, Europe, and Japan propose an international ependymoma research collaborative which aims to synthesize research across sites, foster drug discovery, and prove strategies to integrate clinical and molecular diagnostics into biology-based therapy. Our goal is to maximize information and materials from existing bio and data repositories and not to ‘re-create the wheel’. We envision PEDS as an open science platform and present this concept at ISPNO to invite our colleagues to harmonize efforts towards pediatric ependymoma discovery.

Abstract 14510: Post-traumatic Stress Disorder and Traumatic Brian Injury Are Associated With Increased Incidence of Myocardial Infarction in US Veterans

Background: We have recently shown that PTSD/TBI takes an especially devastating toll on America’s veterans and their families with mental, physical and financial ramifications and poor clinical outcomes (Thakur et al. Epidemiology (Sunnyvale) 2018; 8:353). Patients with PTSD were found to have higher rates of cardiovascular symptoms, hospitalizations and mortality. Whereas electrocardiographic abnormalities, arrhythmias and ventricular dysfunction have been reported in patients with TBI. Association of PTSD/TBI with Myocardial Infarction (MI) is not well studied; the objective of this study was to examine the relationship between PTSD/TBI and MI in a large patient population. Methods: This retrospective study used national data from Veterans Administrations Informatics and Computing Infrastructure (VINCI). Propensity score analysis was performed to obtain matched pairs of (1) Non-PTSD/TBI vs. PTSD (2) Non-PTSD/TBI vs. TBI. Confounding variables used for matching were patients’ demographics and co-morbidities. Cox proportional hazard regression analysis was conducted on matched pairs to determine the relationship between PTSD, TBI and incidence of MI. Kaplan-Meier curves were plotted to compare the time to development of incident MI among the groups. Relationships between variables were examined using chi-square tests, t-tests and non-parametric tests. Results: The risk of new-onset MI was significantly higher for patients with PTSD/TBI when compared with their counterparts without PTSD/TBI. Risk of MI with TBI (HR=2.687, CI 2.546-2.835, P<0.0001) was higher than with PTSD (HR=1.618, CI 1.574-1.662, P<0.0001). Conclusion: There is a significant increase in the incidence of MI in patients with a diagnosis of PTSD or TBI. Further studies are needed to formulate clinical guidelines to risk stratify and intervene to help reduce the incidence of MI in this patient population.

Hearing Outcomes and Complications in Stapes Surgery for Otosclerosis Performed Under General or Local Anesthesia

Objective Stapes surgery is highly successful in reducing or eliminating the audiometric air-bone gap (ABG) related to otosclerosis, and it can be performed under general anesthesia or local anesthesia with sedation. Literature on the relative outcomes of these 2 modalities is lacking. The purpose of this study was to compare hearing outcomes for these 2 modalities in a large patient population. Study Design Retrospective review. Setting Large otology referral center. Methods Patients undergoing primary stapes surgery for otosclerosis from 2005 to 2017 were grouped by anesthetic modality and their cases reviewed. Pre- and postoperative ABGs were primary outcomes. Results A total of 580 patients undergoing stapes surgery were included: 46% received local anesthesia and 54% received general anesthesia. These 2 groups were similar in demographic and disease characteristics. Mean preoperative ABGs were 25.6 and 26.6 dB for patients undergoing local and general anesthesia, respectively ( P = .2); mean postoperative ABGs were 9.5 and 9.7 dB ( P = .9). There were no significant differences in the rates of complications, the need for revision surgery, or the need to abort surgery intraoperatively between local and general anesthesia. Conclusion Consistent with limited prior data, in this cohort stapes surgery yielded similar hearing outcomes whether performed under general anesthesia or local anesthesia with sedation. While we report the largest sample size to date, this study, like previous work, carries the potential for sampling bias. Prospective study comparing local and general anesthesia for stapes surgery is needed.

Vitamin D Levels Are Associated with a Dichotomous Distribution of Hematological Parameters in a Large Cohort of Otherwise Healthy Subjects

Introduction Vitamin D deficiency is generally associated with a slowed hematopoiesis and the final outcome is varying degrees of peripheral blood cytopenia(s). The net effect of increased levels of Vitamin D is not well defined. In this study we tried to document the effect of varying levels of Vitamin D on hematological parameters. Material and Methods Whole patient registry of Medipol Complexes screened and a total of 12709 patients who met the pre-defined inclusion criteria enrolled in the study. Two datasets formed, first including all patients (Dataset A), second including patients with no CBC abnormality and not having any nutritional deficiency (Iron, Vitamin B12 and Folic acid). Results There was a statistically significant correlation between major CBC indices showing a maximum for the second pre-defined (according to quartiles 1st group included patients with a Vitamin D below 10 ng/ml, 2nd group included Vitamin D between 10-20 ng/ml, 3rd group included vitamin D between 20-30 ng/ml and the 4th group included patients who had a Vitamin d level above 30 ng/ml) Vitamin D group. This trend was statistically significant for major hematological parameters and post-hoc analysis also revealed a positive trend through Group 2 (10-20 ng/ml). Regression analysis also documented that patients who were classified under group 2 possessed a more potent hematopoiesis when compared to others even when adjusted for age, gender and baseline CRP values (Figure 1). Conclusion Our study reports a statistically significant correlation between major hematological parameters and Vitamin D levels in a particularly large patient population who lacks a significant confounder like chronic illnesses or conditions which can contribute and potentially serves as a bias. Even more, the group who has a maximal capacity of hematopoiesis are the ones harboring relatively normal but not high levels of Vitamin D. Figure Disclosures No relevant conflicts of interest to declare.

Electronic patient-reported outcomes (ePRO) platform engagement in cancer patients during COVID-19.

172 Background: Tennessee Oncology partnered with an ePRO platform solution to support patients during their cancer care journey. This cloud-based ePRO platform is designed to assist in improving the management of symptoms. Providing two core pieces of functionality allow both the patient and care teams to retrieve information quickly and communicate effectively. The patient portal is patient input driven and allows the patient to communicate with their care team, track symptoms, and access their health records via website or mobile app. The clinician portal provides multiple care teams the ability to manage and prioritize patient needs as well as communicate directly with patients. In March 2020, due to the pandemic, patients needed a convenient and remote way to communicate with the care team. Our communication plan had to be nimble and provide immediate updates to our large patient population. We leveraged our ePRO platform to meet this need. Methods: We focused efforts on increasing patient engagement by educating them on the benefits of this communication platform. We utilized secure messaging to send appointment details and for Telehealth visits a link to the visit was sent. We were able to provide weekly updates outlining our latest information regarding our safety protocols. Results: We noted an increase in the activation of patient accounts and patient-initiated messages in our ePRO platform. We saw an average of 1,000 new patient accounts activated each month during March, April and May. We saw that patient-initiated messages through the platform showed a 15% increase from February to March. The response rate for patients completing post-treatment questionnaires increased 8% from February to May. Conclusions: By providing patients with a single communication platform to contact their care teams outside of their office visits, patients become an active part of their care journey. As an organization, we continue to identify ways to connect our patients to their care team in a meaningful way through technology. Whether during normal business hours or after-hours, patients need a simple, reliable and consistent way to engage with their care team.

In the Eye of the BBHolder: A Look into Recent Trends in Blood Bank Hold Order Utilization

The transfusion service at a large academic institution provides transfusion support to the regional cancer center in inpatient and outpatient settings. This center has a large patient population many of whom require frequent blood count monitoring. To reduce the number of blood draws and port accesses, it is standard of care to collect a blood bank hold (BBHold) tube simultaneously with routine CBC samples. The BBHold tube is routed to the transfusion services lab and stored in case a type and crossmatch (TXM) is needed in preparation for transfusion; it is valid for 72 hours. While advantages to this system are eliminating the need for a secondary blood draw and ready access to a sample for transfusion testing, disadvantages include drawing an additional tube, increased cost and storage, and increased specimen mislabeling in the conversion of a BBHold to a TXM. Our objectives were to analyze trends in BBHold usage, make recommendations to reduce unnecessary utilization, and create an algorithm to predict which patients would need transfusion. We collected data from March 2016 to June 2019 including hemoglobin values, BBHold orders, and TXM orders, and used the statistical programming language R for data analysis. Approximately 1,000 BBHold samples were collected monthly, with 64.2% taking place in the outpatient setting. Our analysis focused on this outpatient population. A total of 31.1% of BBHold orders were unnecessary, as they were re-collected within the 72-hour window of sample viability. Of all the BBHold orders, including the previously mentioned unnecessary orders, only 32.7% were converted to a TXM, which was used as a surrogate marker for transfusion. Unexpectedly, 28.2% of the last hemoglobin values collected prior to a BBHold (either concurrently with the BBHold or up to 3 days prior) were ≥10 g/dL. Notably, only 44.6% of these hemoglobin orders occurred concurrently with a BBHold, indicating that over 55% of BBHold orders were placed after the hemoglobin order. Once a hemoglobin value is resulted, a BBHold has minimal use since the hemoglobin value would determine if a transfusion is indicated. If transfusion is indicated, a TXM order should be placed directly, rather than a BBHold. Next, the average hemoglobin values were compared between patients who did and did not receive a transfusion within 7 days, and these were 8.8 and 9.1 g/dL, respectively. In conclusion, the trends illustrated here indicate overutilization of BBHold orders and highlight potential areas for optimization. Hemoglobin values within 7 days of a BBHold may not be enough to design a robust algorithm to stratify patients based on need for transfusion. Recommendations moving forward may include improving the EHR interface to display the expiration date of the current BBHold sample or displaying the patient’s last hemoglobin if ≥10 g/dL.