2021 Research Forum

Genomics has revealed that many microorganisms have far greater biosynthetic potential to produce natural products than was estimated from classic bioactivity screenings. However, this potential has been hampered because many specialized biosynthetic gene clusters (BGCs) are not expressed in laboratory cultures. Several pleiotropic approaches have been developed to induce the production of metabolites that are not expressed under conventional conditions. Among those, ribosome engineering has led to discoveries of novel medicinal compounds, including piperidamycins and coelimycins. This strategy involves selecting spontaneous mutations in genes encoding RNA polymerase (RNAP) and ribosomal proteins to affect metabolic profiles by stabilizing transcriptional and translational machinery. Indeed, inducing and selecting for mutations in the genes encoding ribosomal S12 (rpsL) and RNAP beta-subunit (rpoB) by using the ribosome-targeting antibiotics streptomycin and gentamicin, and the RNAP-targeting antibiotic rifampicin has led to the identification of eight antibacterial compounds produced by a Streptomyces. mauvecolor mutant. Here we used streptomycin and rifampicin to select drug-resistant mutations in Streptomyces sp. F181 (i.e., CMA wt). Combining UV mutagenesis, we generated four types of mutant strains that have distinct phenotypes. Along with the wildtype strain, three selected mutant strains were incubated in R2YE liquid cultures and extracted to obtain crude extract. Raw MS data of the crude extracts were obtained by Quadrupole Time-of-Flight (Q-TOF) LC/MS system, then analyzed by feature-based molecular networking provided by Global Natural Products Social Networking (GNPS). The molecular network was visualized by Cytoscape, revealing interestingly different chemical profiles of wildtype and mutant strains.

Linwei Yang

College of Pharmacy, University of Michigan-Ann Arbor

Yuanhuang Chen

Department of Clinical Pharmacy, University of Michigan College of Pharmacy

Zebedee Miller

Universtiy of Michigan, College of Pharmacy, Pharmaceutical Sciences

Jiayi Nie

Department of Pharmaceutical Science, University of Michigan

Yanling Xue

Department of Clinical Pharmacy, College of Pharmacy,University of Michigan

Xiaoyue Shi

Department of Pharmaceutical Sciences, College of Pharmacy

NAD(P)H:quinone oxidoreductase 1 (NQO1) is responsible for the reduction of several cytotoxic quinones and reactive oxygen species (ROS). However, it is found that NQO1 is overexpressed in many human cancers and the high level of expression is closely related with tumor progression and metastasis. Therefore, NQO1 becomes a potential chemotherapy target and it is critical to find the inhibitor of NQO1 with high affinity and high selectivity. Here we summarizes the existing inhibitors’ profile. Use docking software to simulate docking models for some novel inhibitors that didn’t have crystal structures with NQO1 and analyze the active sites. Also, use docking software to evaluate optimized compounds based on 6/7-amino-quinoline-5,8-dione to find promising candidates for next step study.

Zixian Zhu

University of Michigan, Department of Medicinal Chemistry, College of Pharmacy

Objective: The historical SPECT agent [131I]NP-59 was a useful diagnostic for determining the laterality of adrenal pathology. However it was limited in its application because of the high radiation dosimetry associated with the decay properties of iodine-131, as well as the prevalence of in vivo deiodination and thyroid accumulation. The objective of this work is to develop a PET analogue, [18F]FNP-59, for its likely superior dosimetry and image resolution, evaluate its use in preclinical animal imaging, and translate it into humans under an IND.

Methods: FNP-59 standard and precursor were synthesized over 8 and 6 steps respectively. The radiosynthesis of [18F]FNP-59 was achieved in 2 steps over 2 hours with a radiochemical yield of 44.1±18.6 mCi. Preclinical imaging was conducted in both female Sprague Dawley rat and female Watanabe heritable hyperlipidemic rabbit. Biodistribution experiments were conducted with 2 male and 2 female Sprague Dawley rats sacrificed at 5 timepoints and each of their organs weighed and measured for radioactivity. Dosimetry estimates were then calculated based on this biodistribution data using OLINDA/EXM 2.0. Finally, preliminary human imaging was undertaken under an IND approved by the FDA.

Results: Preclinical imaging and biodistribution in rats demonstrated high uptake in the adrenal glands and low interference from the liver and other abdominal organs. Imaging in rabbits, which have a gallbladder and significantly different lipid handling properties, indicated reduced adrenal uptake and higher liver and gallbladder uptake. Translation to humans demonstrated further reductions in adrenal uptake and high liver and gallbladder uptake. Studies utilizing alternate administration strategies to alter [18F]FNP-59 trafficking by altering its incorporation into lipoproteins are ongoing.

Wade Winton

1) 4th year BS Pharmaceutical Sciences, 2) Division of Nuclear Medicine, Department of Radiology, 3) Nuclear Medicine Service, Ann Arbor Veterans Administration, 4) The Interdepartmental Program in Medicinal Chemistry, 5) Research Advisor

Moroidin is a biologically active bicyclic plant octapeptide, which can inhibit tubulin polymerization and makes moroidin a potential compound for treating cancer. However, low yield from source plant and difficulty in organic synthesis hinder moroidin-based drug development. My research purpose is to solve this problem through biosynthetic method of moroidin, namely, heterologous production and diversification of moroidin in tobacco.

Jun Zhou

University of Michigan, College of Pharmacy

Secondary AML (sAML) represents up to 30% of total AML cases and is associated with a poor prognosis. Prior studies have shown that both fludarabine, high-dose cytarabine, and granulocyte colony-stimulating factor (FLAG) and HMA-based regimens (e.g. decitabine ± venetoclax) can provide favorable responses with lower treatment-related toxicities than traditional regimens such as 7+3 or CPX-351. This single-center retrospective cohort study aimed to compare response rates in patients with sAML receiving10-day decitabine ± venetoclax to those receiving FLAG. The median age was 73 (range, 37–80) in the decitabine ± venetoclax group and 67 (range, 27–82) in the FLAG group (p = 0.037). There was a numerically lower but statistically similar CR/CRi rate in patients treated with decitabine ± venetoclax compared to FLAG (40 vs. 62%, p = 0.076). More patients in the decitabine ± venetoclax group achieved MLFS (25% vs. 1%, p = 0.002). The median overall survival was comparable in both groups (11.1 months, decitabine ± venetoclax vs. 9.5 months, FLAG; p = 0.472). There was a higher incidence of documented infections (20% vs. 55%, p = 0.01) with FLAG, despite days to absolute neutrophil count recovery being longer with decitabine ± venetoclax (median, 34 vs. 19 days; p < 0.001). Induction-related mortality was low and similar among the two treatment regimens. Overall, response rates after FLAG and decitabine ± venetoclax are comparable and translate to similar overall survival in the sAML population in this real-world, single-center retrospective review. Larger studies are necessary to elucidate the benefits of the individual regimens in the sAML patient population.

Devon Stonerock

University of Michigan

Background: In hepatic cirrhosis, the progressive decline in the number of viable hepatocytes results in reduced conversion of nitrogenous wastes to creatine, the precursor to muscle-generated creatinine. Because of these resultant lower levels of circulating creatinine, equations using measured creatinine may often overestimate creatinine clearance in this patient population. When empirically dosing vancomycin using standard dose nomograms, this overestimation of creatinine clearance may result in “over-dosing”, thus resulting in supratherapeutic levels and a commensurately higher risk of nephrotoxicity.

Objective: Using a calculated “trough to dose ratio”, or TDR (trough level/mg/kg/24h), we wanted to compare vancomycin dosing requirements in cirrhotic patients with those of non-cirrhotic matched cohorts.

Methods: We conducted a single-centered, retrospective chart review of cirrhotic and non-cirrhotic control patients admitted to Michigan Medicine who received IV vancomycin between January 2016 and February 2020. All patients had to have serum creatinines less than or equal to 1.2 mcg/ml at baseline and at least one steady-state trough to qualify. For comparison, both groups were further allocated to three matched age groups (30-39; 40-49; 50-59).

Results: Among the 200 total patients (150 non-cirrhotic patients and 50 cirrhotic), 43 patients were between the ages of 30-39 (86% non-cirrhotic, 14% cirrhotic), 55 between the ages of 40-49 (78% non-cirrhotic, 22% cirrhotic), and 102 between the ages of 50-59 (69% non-cirrhotic, 31% cirrhotic). Sex was evenly distributed within each age group. Baseline characteristics such as serum creatinine (SCr), weight, and mean age were also similar. TDRs for cirrhotics were 50, 45, and 32% higher than their controls, respectively. However, due to the small sample size, only age 50-59 had statistically significantly higher TDR values than the non-cirrhotic patients (95% CI, 0.05-0.27; P<0.006). Supratherapeutic vancomycin troughs (>20 mg/L) also occurred more frequently in our cirrhotic cohort (22%) when compared to our control group (8%).

Conclusions: In this retrospective study comparing vancomycin dosing standardized by mg/kg/24h in cirrhotic patients and non-cirrhotic patients with similar baseline characteristics, patients with cirrhosis achieved roughly 30-50% higher trough levels when compared to non-cirrhotic cohorts receiving the same dose. These results suggest that making empiric dose reductions in cirrhotic patients when using standard vancomycin dosing nomograms is a reasonable approach.

Ryan Larson

University of Michigan, College of Pharmacy

Michelle Nguyen

University of Michigan College of Pharmacy

Age-appropriate pediatric formulations for oral administration can be challenging to formulate. The development of such formulations is often time-consuming, labor-intensive, and costly. The Biopharmaceutical Classification System (BCS), developed more than two decades ago, is used to develop suitable oral drug formulations for adult use. In theory, some of the same principles could be applied to formulate pediatric oral liquid dosage forms. However, the present BCS system was developed using adult gastrointestinal physiologic factors. Direct extrapolation of this method to develop pediatric oral dosage forms is inappropriate due to differences in adult and pediatric gastrointestinal physiologic differences during development. To date, age-appropriate BCS to guide pediatric oral liquid formulation development has not been developed for various pediatric subpopulations. The objective of this study was to provisionally classify oral liquid formulations of extemporaneously prepared drugs at our institution into an age-appropriate BCS class after elimination of any duplicate listing when matched with the most current World Health Organization's Essential Medicines List for Children available at the time of this study and other published studies that may have reported BCS classification of drugs used as extemporaneous oral liquid formulations in children to treat chronic or rare diseases. A total of 96 orally administered extemporaneously compounded liquid formulations were included in this classification. Dose numbers were calculated using age-appropriate initial gastric volume for neonates, 6-month-old infants, and children up to 6 years of age. Using age-appropriate initial gastric volumes and pediatric and neonatal Lexicomp® age-specific maximal dosing recommendations for calculation of dose numbers, the solubility classes shifted for 62.5% of the drugs studied. A significant number of currently used extemporaneously compounded oral liquid formulations for age groups of children included in this study may not provide formulations with predictable safety and efficacy. Factors used in the development of adult BCS cannot be applied directly to pediatric subpopulations.

Hannah Hammoud

University of Michigan College of Pharmacy

Expanded Access (EA) is a Food and Drug Administration (FDA) program that provides a process for the use of experimental drugs, biologics, and medical devices for patients who lack other therapeutic options or are ineligible for clinical trials.1 EA use has potential for collecting outcomes data for patients utilizing treatments EA and contributing to the ongoing movement to collect real world data (RWD) on EA use.2,3 Since 2009, over 1600 patients at Michigan Medicine (MM) have gained access to investigational drugs through EA. The Michigan Institute for Clinical & Health Research (MICHR) launched a new project in 2018, the Transforming Expanded Access to Maximize Support and Study (TEAMSS), to establish a national framework for more efficient, consistent, and widespread EA use. One of the main goals is to create a database to standardize EA data reporting and develop a body of RWD. To better understand the quality and utilization of EA at MM we conducted a retrospective review of patients with infectious diseases using medications obtained through EA.

1. Jarow JP, Lurie P, Ikenberry SC, Lemery S. Overview of FDA’s Expanded Access Program for Investigational Drugs. Ther Innov Regul Sci. 2017;51:177–9. 2. Polak TB, van Rosmalen J, Uyl-de Groot CA. Expanded Access as a source of real‐world data: An overview of FDA and EMA approvals. Br J Clin Pharmacol 2020;86:1819-26. 3. Reagan-Udall Foundation for the Food and Drug Administration. Public Meeting Report Leveraging Real-World Treatment Experience from Expanded Access Protocols. https://reaganudall.org/sites/default/files/2019-12/Leveraging%20Real-World%20Treatment%20Experience%20from%20Expanded%20Access%20Protocols%20Report.pdf. Accessed 30 November 2020.

Ibtihal Makki

Michigan Institute for Clinical & Health Research

Michelle Azar

Michigan Oncology Quality Consortium (BCBS)/University of Michigan College of Pharmacy

Psychiatric illness is a leading cause of death and disability, with mood disorders estimated to affect up to 1 in 4 people over their lifetimes. Medication therapy is a mainstay of treatment, however many patients experience multiple medication trial failures prior to finding an appropriate treatment regimen. The advent of pharmacogenetic testing (PGx) offers the potential to predict patients’ responses to psychiatric medications based on genetic information. Despite the promise of this treatment approach, real world studies gauging the impact of these tests is lacking. Here, we report results from a single-center, retrospective review of 90 patients treated at a large, academic medical center who had received pharmacogenetic testing to guide psychotropic medication dosing and prescription. 89/90 patients received PGx testing via GeneSight®, a commercially available test. Patients had a mean of 8.4 medication trials over the entirety of their care at our institution with significantly more medication trials prior to return of PGx testing results (6.1 ± 3.7 pre vs. 2.3 ±1.9 post, p < 0.001). Antidepressant prescribing congruence was significantly higher after return of PGx testing results X2(1, n = 221) = 12.1, p < 0.001 and antidepressant prescribing was more likely to be congruent after PGx testing (OR: 3.90, [95%CI]: 1.21-12.58, p = 0.023). Lack of clinical efficacy was the most common reason for medication switching and non-response occurred in 123/460 (27%) of medication trials across the duration of the study. PGx testing results were not uploaded into the electronic health record in 36/90 (40%) of patients. Our data represent the first independent report assessing the use of pharmacogenetic testing in a real world clinical setting. Among study findings, we identified a significant need for improved management of pharmacogenetic results. Further studies are needed to assess the impact of pharmacogenetic testing on clinical improvement in patients with mental illness.

David Schapiro <dschap@med.umich.edu>

University of Michigan College of Pharmacy

Audrey Yeiter

University of Michigan

Purpose: Statins are highly effective in preventing and treating cardiovascular disease, but statin associated muscle symptoms (SAMS) are a common adverse reaction. Coenzyme Q10 (CoQ10) is an attractive option for treating SAMS, since it is inexpensive, widely available, and allows for continued statin-based treatments at cardioprotective doses. The cholesterol treatment guidelines from the American Heart Association and American College of Cardiology currently do not recommend adding CoQ10 for SAMS. However, there are some meta-analyses showing CoQ10 effectiveness in treating SAMS. Due to contradicting evidence, our objective was to determine whether supplemental CoQ10 is associated with the improvement of SAMS. Methods: This retrospective analysis was based on questionnaires, which collected data on participants’ past and current use of cholesterol-lowering drug therapies, symptoms, comorbid conditions, and concomitant medications. The questionnaires for the original study were gathered from participants in six medical centers throughout the U.S. and Canada through 2004 and 2013. In this follow-up study, we analyzed data for all participants who experienced SAMS in the past 5 years, and the primary outcome was binary: whether or not they still currently had SAMS (i.e., at the time of the questionnaire). SAMS were defined as any symptom specific to muscle that participants associated with the onset of statin therapy and included muscle symptoms (e.g., pain, weakness, cramps) and/or elevated CK levels. Baseline characteristics were calculated overall and stratified by whether a participant had current SAMS and if a participant was on CoQ10. Continuous variables were compared using student’s t-test, while categorical variables were compared using chi-square tests or Fisher’s exact test when necessary. Univariate and multivariate logistic regression analyses (adjusted for SAMS risk factors and significant differences between CoQ10 users & non-users) were used to evaluate the association between CoQ10 use and current SAMS. Results: A total number of 511 participants were included in this study: mean (+sd) age 58±11 years, 54% males and 100% Caucasian race. Seventy percent of the participants had current SAMS, and 64 participants were on CoQ10 (12.5%). Upon completion of the univariate logistic regression analysis, there was no statistically significant difference in current SAMS between CoQ10 users vs. non-users (75% vs. 69.5%, respectively; OR=1.33 [95% CI=0.73-2.42] p=0.357). Smoking status, obesity, and family history of muscle disease between the groups who did or did not have current SAMS were shown to be statistically different (p < 0.05). Even when these baseline characteristics were included in the multivariable logistic regression, CoQ10 usage did not have a statistically significant association with decreasing SAMS (OR=1.22 [95% CI=0.66-2.25] p=0.522). Lastly, a third multivariable logistic regression analysis was performed with baseline variables that were statistically significant in the original study (hypertension, smoking status, obesity, past cholesterol lowering drug type, and family history of heart disease):(OR=1.15 [95% CI=0.62-2.14] p=0.659). Conclusion: In this large, observational study of Caucasians, the addition of CoQ10 was not significantly associated with decreasing SAMS. Future studies should include more racially and ethnically diverse participants, larger sample sizes, and dose of CoQ10.

Wilson Chen

University of Michigan, College of Pharmacy

A single-centered prospective genotype panel study was performed to determine the impact of individual genetic variation on the activation and pharmacokinetics of an ACEI prodrug, trandolapril (Mavik®). Healthy and non-smoking male and female (non-pregnant) volunteers aged 18 to 55 were enrolled in the clinical studies participant program at the University of Michigan. They were not taking any medications and did not have abnormal physical or laboratory findings nor significant diseases. After identifying and categorizing each individual’s genotype into two subgroups, non-carrier (143G/G) or carrier (143G/E or 143E/E), a single trandolapril 1 mg with a glass of water (240mL) was administered by each participant on Day 1. Blood was drawn, and blood pressure was measured every 30 minutes for the initial 4 hours, every hour up to hour 6, and the 2 hours thereafter, followed by 24 hours, 48 hours, and 72 hours after the drug administration. The mean of an area under the plasma concentration-time curve from 0 h to 72 hours (AUC0–72hr) of trandolapril and trandolaprilat for each genotype group was assessed. The non-carrier group had a higher mean value of AUC0–72hrs of trandolapril and trandolaprilat; however, it was not statistically different from carrier group values. The difference in the mean of Cmax and Tmax between the two groups was not statistically significant. Taken together, the CES1 gene variation on G143E does not significantly impact trandolapril metabolism in healthy volunteers.

Hee Jae Choi

University of Michigan, College of Pharmacy

Tara Bracken

University of Michigan College of Pharmacy

Krestina Bednarz

University of Michigan

CDI is associated with high morbidity and mortality. Although it is well-known that antimicrobial use is a risk factor for CDI development, different antimicrobial classes and administration routes are associated with varying degrees of risk. Acne is a common skin conditions frequently treated with antimicrobial agents, both topically and orally, and it is unclear if these agents, particularly topical clindamycin and oral tetracyclines, are associated with significant risk of developing CDI. Among a small population of acne patients at Michigan Medicine, those who developed CDI were more likely to have used topical clindamycin than control patients who more often used a combination of topical clindamycin and oral tetracyclines. Additionally, controls were more likely to use doxycycline (or minocycline), supporting existing literature that suggests oral tetracyclines may protect against CDI. However, considering the small size of this cohort and the additional risk factors that contributed to CDI in cases, it remains unclear if topical clindamycin or doxycycline (or minocycline) use alone was a risk factor for CDI development.

Sydney VanDorf

University of Michigan College of Pharmacy

| Background | Nontuberculous mycobacteria (NTM) are increasing in global prevalence as a cause of pulmonary infections and are difficult to treat with traditional antibiotics. Here, we study the repurposing of clofazimine (CFZ) to treat NTM using expanded access in a single health system. Our main objectives are to describe the feasibility of accessing and analyzing expanded access data and to generate hypotheses regarding CFZ use in NTM treatment. | Methods | A retrospective chart review was performed on patients within a single health system who had been approved for expanded access of clofazimine or who received it through an outside hospital for treatment of NTM disease. Data were collected on patients’ baseline demographics, details of their NTM infection, concomitant therapies, and results as of 30 June 2020. Descriptive statistics were performed in Microsoft Excel (2019). | Results | A total of 47 patients were identified upon initial review as potentially receiving CFZ for NTM infection. After excluding 8 patients who did not initiate CFZ, data from the remaining 39 patients were collected and summarized. The median age at which patients were diagnosed with NTM was 53 years old, with a median BMI of 21.6 kg/m^2. Patients were more likely to be female (71.8%), have a baseline lung disease (69.2%), and 47.2% were current or former smokers at the time of their diagnosis. The most common species isolated was M. avium complex (48.7%), with the most common site of infection being the lung (79.5%). Almost half of the patients presented with productive cough, and a majority presented with bronchiectasis and nodules present on radiograph. Aminoglycosides and macrolides were the most common antibiotics combined with CFZ (53.9% and 66.7%, respectively), which is reflective of prescribing patterns. Only 20.5% achieved culture conversion with CFZ, with another 20.5% still on therapy by the end of data collection. Of the 8 patients who had discontinued CFZ secondary to negative culture conversion, 7 (88%) had been infected with M. abscessus and 7 (88%) had NTM pulmonary infections. In addition, none of them had a history of smoking. | Conclusions | This study demonstrated the difficulty of collecting retrospective data on symptoms, side effects, and radiography from patients who obtained a drug through expanded access. Based on the findings of this study, we recommend further research into the potential use of CFZ in non-smoking patients with M. abscessus pulmonary infections.

Madeleine Davies

University of Michigan, College of Pharmacy, Department of Medicinal Chemistry

Purpose: Interprofessional education (IPE) has been implemented in recent years due to increased quality of patient care. While IPE has expanded, assessment of these implementations in the curriculum have yet to be studied extensively. The objective of this study was to evaluate the impact of an intentional one-time IPE interaction between pharmacy and medical students.

Method: Participants included 2nd year pharmacy students and 3rd year medical students enrolled in University of Michigan during the 2018-2019 academic year. Students were required to have an in-person interaction for a geriatric patient case. Separate patient interviews were conducted by each discipline and the gathered information was compared during the interaction. Students completed a post-interaction survey. Qualitative data from this instrument for these IPE experiences was de-identified and quantified into different categories to allow measurable outcomes which were then prospectively analyzed. Overall response rates and discipline-specific response rates were calculated and compared by theme.

Results : During the 2018-2019 academic year, 138 medical students and 86 pharmacy students participated in this study for a total of 224 participants. The positive impact of the interaction on patient care was reported by 166 students of the 224 participants. Learning about the other profession's training and education was most surprising to students, especially so for medical students compared to pharmacy students (33% vs 24%). Future interactions between the two disciplines seemed more likely to happen with 43% of medical students more likely to consult pharmacists. Of note, 41% of pharmacy students showed a desire to increase interprofessional practice either through increased communication with patients and doctors or more collaboration with health professionals compared to 22% of medical students.

Conclusion: Overall, implementing a one-time intentional IPE experience to healthcare students increased their IPE knowledge, enhanced their understanding of the other discipline’s role and training, and created a positive perception of working together in the future.

Christina Tang

University of Michigan, College of Pharmacy

Chad Compagner

University of Michigan College of Pharmacy

High-dose chemotherapy followed by autologous stem cell transplantation (ASCT) is the standard of care treatment option for relapsed/refractory (R/R) DLBCL patients fit enough for intensive therapy. Recently, two chimeric antigen receptor (CAR) T-cell therapies - axicabtagene ciloleucel (Yescarta) and tisagenlecleucel (Kymriah) - were FDA approved for treatment of R/R DLBCL. Pivotal trials leading to FDA approval of these CAR T therapies were not randomized and included some patients who had previously received ASCT, leaving unanswered questions with regard to their optimal place in therapy. Although outcomes with ASCT are best among patients who exhibit complete response (CR) to salvage therapy, those with partial response (PR) still derive significant benefit. While trials comparing CAR T to ASCT are underway, outcomes are not expected to be reported for several years. In the meantime, assessment of existing data is needed to help guide present therapy decisions until data from these trials are reported.

To address this question, we performed a single-center, retrospective, propensity-score matched cohort study of R/R DLBCL patients treated at the University of Michigan who exhibited PR to salvage therapy (assessed via PET-CT) and then proceeded to ASCT or CAR T therapy from January 2015 to December 2020. Cohort 1 includes patients who went to ASCT after salvage, while Cohort 2 includes patients who received CAR T therapy. Exclusion criteria include patients < 18 years old and those with insufficient records. Data related to demographics, upfront treatment, salvage treatment, toxicity, time to ASCT/CAR T infusion, and survival outcomes will be collected. The primary endpoint is progression-free survival, defined as time from treatment until relapse or death. Time-to event analyses were assessed via the Kaplan-Meier method and log-rank test.

Brian Bazzell

Michigan Medicine

Background: Improvements in the management of Multiple Myeloma (MM) have drastically increased 5-year survival to over 50%. Historically, infections were a major cause of morbidity and mortality, with infection risk being highest in the first 3 months after diagnosis. Data characterizing this early infection risk is based on outdated treatment regimens. At this time, the true risk of bacterial infection in NDMM patients undergoing current standard of care induction with lenalidomide, bortezomib, and dexamethasone (RVd) is unknown and randomized controlled trials to guide antibiotic prophylaxis are limited. Mayo Clinic MSmart and NCCN updated their recommendations to include bacterial prophylaxis based on the results of the TEAMM trial which assessed the benefit of 12 weeks of levofloxacin prophylaxis in NDMM patients. Levofloxacin significantly reduced febrile episodes and deaths without increasing health care-associated infections. However, this trial has significant limitations including outdated treatment regimens, loss of mortality benefit over the entire 12-month study duration, and a higher proportion of deaths with progressive disease in those receiving levofloxacin prophylaxis. This analysis will inform the true risk of bacterial infection for NDMM patients undergoing standard of care induction.

Objective: To characterize the risk of bacterial infection and identify risk factors for bacterial infection in NDMM patients receiving standard therapy with RVd. Methods: Patients with NDMM undergoing induction with RVd at Michigan Medicine from January 2014 until January 2020 will be included. Patients will be followed for at least 12 weeks during induction therapy until the time of maintenance therapy or autologous-stem cell transplant (ASCT). Patients that start with Cyclophosphamide + Bortezomib + Dexamethasone for 1-2 cycles and switch to RVd before transplant or maintenance will also be included. Patients who are less than 18 years of age, receiving an investigational agent in combination with RVd, or undergoing treatment with RVd but switch to another treatment before ASCT will be excluded. Patients will be divided into two cohorts based on if they do or do not experience a bacterial infection during induction. Data that will be collected include: age, gender, eGFR, treatment regimen, comorbidities, ISS/R-ISS staging, cytogenetics risk at diagnosis, M protein, Serum β-2-microglobulin, IgG, IgA, IgM, absolute neutrophil count, previous infection in 30 days preceding, baseline bone disease, infectious prophylaxis received. The primary outcome will be to establish risk factors for early bacterial infection in NDMM patients undergoing induction with RVd. Secondary outcome includes assessing early mortality rates.

Results/Conclusions: TEAMM trial demonstrated febrile episodes in 19% of patients receiving levofloxacin and in 27% of patients receiving placebo. The rate of febrile episodes was much lower in our study, even in those who experienced a true bacterial infection. Our study found a higher incidence of febrile episodes in patients that experienced a bacterial infection (P=0.003) but no difference in the incidence of deaths from infectious complications (p=1.000). Overall incidence of true bacterial infection in NDMM patients receiving induction with RVd was low despite many patients receiving no bacterial prophylaxis. The most common types of infections included pneumonia and UTI and there was a very low incidence of sepsis/bacteremia. A multivariate analysis will identify risk factors for bacterial infection which will help assess if antibiotic prophylaxis should be considered for certain high-risk patients. Our study has many limitations including its retrospective design, small sample size, missing data on bacterial infections if treated at an outside facility and selection bias. Future directions include a multivariate analysis and a multicenter trial with an increased patient population.

Anisa Bici

Michigan Medicine

Kirsten Carstarphen

VA Ann Arbor Healthcare System

Background: While the prevalence and management of cancer-related pain, in those with solid tumors and advanced disease is well documented this information is lacking for patients with acute leukemia. Recommendations from The National Comprehensive Cancer Network (NCCN) clinical practice guidelines are based on the management of cancer pain in solid tumors, where there is a well-established role of opioids. However, the role of opioids for the management of pain in acute leukemias is unknown. Non-opioid therapies may be avoided in acute leukemias due to fear of masking a fever with acetaminophen or thrombocytopenia with a non-steroidal anti-inflammatory agent. Using the NCCN "one size fits all" approach, may lead to inappropriately managing pain, unnecessary utilization of opioid agents, and increased risk of opioid dependence.

Objective: Evaluate opioid prescribing for patients with acute leukemias at the University of Michigan Health System to determine the overall burden and impact of long-term opioid use in acute leukemia patients initially naïve to opioids.

Methods: This was a single center, retrospective matched cohort study of adult patients undergoing induction therapy for newly diagnosed acute leukemias from June 2014 to June 2020. This study is IRB exempt approved by the Michigan Medicine IRB. Patients were excluded if patient record is incomplete or for opioid use prior to admission, defined as any opioid on their home medication list. Data was collected characterizing demographics, comorbidities, hematologic disease, induction chemotherapy received, pain regimen utilized outside of treatment plan (if applicable), and opioid duration post-induction therapy (if applicable). The primary endpoint was the percentage of opioid-naive patients on chronic opioid therapy, defined as opioids prescribed 3 months after acute leukemia induction therapy. Secondary endpoints include proportion of opioid naïve patients prescribed opioids for pain, risk factors for long-term opioid use, percentage of patients on opioid therapy at 3, 6, and/or 9 months after induction therapy.

Results/Conclusions: non-applicable, research in progress

Kaylee Clark

University of Michigan | Michigan Medicine

Taylor Coe-Eisenberg

Michigan Medicine Department of Pharmacy

Sydnee Cohen

University of Michigan Health System, Department of Pharmacy Services

Purpose: Underdiagnosed or undertreatment of cancer pain can lead to poor outcomes, treatment interruptions, increase in healthcare costs and overall poor quality of life. In order to improve the management of cancer related pain across all care settings and providers, the National Comprehensive Cancer Network (NCCN) publishes the Adult Cancer Pain guideline on an annual basis. The guideline provides a comprehensive outline of pain management initiation and maintenance. Evidence has shown that guideline-based recommendations reduced pain scores in patients with cancer, however, there is a lack of evidence surrounding its impact on length of stay. The purpose of this study is to assess the impact of NCCN guideline adherence to a patient’s length of stay.

Methods: This study will be a single-center, retrospective chart review of adult patients with active cancer and admitted to Michigan Medicine for pain. Patients must receive at least one dose of opioids within 24 hours upon admission. Patients admitted into the intensive care unit, an admission stay less than 24 hours, expired during the encounter, or with a current or past diagnosis of opioid use disorder will be excluded. Data collected from medical records will include demographics, primary cancer site, presence of metastases, type of opioid and dose ordered, daily average of opioid utilization in morphine milligram equivalents, presence of nausea, vomiting and/or constipation, pain scores, and length of stay. The primary outcome will be the percentage of orders that are guideline adherent and the length of stay measured in days. Secondary outcomes will include pain scores and incidence of opioid related adverse events.

Results/Conclusions: Data is currently being collected.

Jacqueline Dela Pena

Michigan Medicine

Purpose: Perioperative management is a complex process that requires close coordination between patient, anticoagulation provider, and proceduralist. Appropriate interruption of anticoagulation therapy with or without bridging is necessary to minimize patients’ risk for both bleeding and thrombosis. The purpose of this study is to assess the effectiveness of pharmacists’ involvement in the perioperative process for anticoagulated patients undergoing a catheterization procedure and to describe the current process as there is no published data on perioperative management by pharmacists on a system-wide level. Methods: Patients who underwent cardiac catheterization at Michigan Medicine and were referred to the anticoagulation clinic for perioperative management from January 2016-December 2020 will be included. The study population and data were retrieved via use of electronic medical record. Data collected includes: patient demographics, procedure type, indication for anticoagulation, anticoagulant drug, bleed risk, thrombotic risk, proposed anticoagulation plan, final anticoagulation plan, and any bleeding or thrombosis after procedure. The primary outcome is rate of pharmacist adherence to established perioperative guideline. Secondary outcomes will include rate of agreement between pharmacist recommended plan and case request, rate of acceptance of pharmacist perioperative anticoagulation plan by provider, and proportion of cases where pharmacist intervention prevented unnecessary bridging or admission. Summary of preliminary results: Data collection and analysis are ongoing. Conclusions: Results will be used to characterize pharmacist interventions, determine predictors for cases requiring intervention and to describe the pharmacists’ role and impact in perioperative management of anticoagulation and services provided at Michigan Medicine.

Mary Jo Elder

Michigan Medicine

Purpose: Michigan Medicine (MM) averages caring for 95 aortic dissections per year. Type A dissections make up 72.7% of these dissections with type B dissections constituting the other 27.3%. The high mortality risk and emergent nature of dissections, specifically type A, demonstrate the necessity of proper management. Currently, surgical management is the gold standard for type A dissections with medical management the standard for type B dissections without complications. Beta blockers are the first line treatment to reduce the elevated blood pressure to a goal systolic pressure of between 100-120 mmHg and goal heart rate of 60 beats per minute. At MM, esmolol and fenoldopam have been used as first line treatment for aortic dissection. In January, 2019, MM implemented an aortic dissection protocol and computer order entry order set to provide guidance for the use of intravenous continuous ant-hypertensive agents for both type A and type B aortic dissection. The purpose of this study is to assess the adherence and cost implications of the aortic dissection protocol. Study Design: This is a single center, retrospective, chart review of all patients with urgent and emergent aortic dissections at MM between January 1st, 2018 and January 31st, 2020. Methods: All patients seen at MM for emergent aortic dissection management who were older than 18 years were included. Data collected included choice of medication administered for blood pressure control (infusions monitored were clevidipine, esmolol, fenoldopam, labetalol, nicardipine, nitroglycerin, and nitroprusside), patient demographics, surgery date, surgeon, dissection type, and order set used. The primary outcome was esmolol cost pre-protocol (January 1st, 2018 – December 31st, 2018) versus post-protocol (January 1st, 2019 – January 31st, 2020). The secondary outcome was adherence to the aortic dissection protocol and hospital length of stay pre and post-protocol implementation. Preliminary Results: There were 166 emergent dissections included in the study (70 in 2018 and 96 in 2019). Of the total number of dissections, 135 (81.3%) were type A and 6 were type B (18.7%). More data collection and analysis is ongoing, and final results and conclusions will be presented at 2021 Great Lakes Pharmacy Resident Conference.

Jalen Evans

University of Michigan Medicine

Colin Finley

Michigan Medicine

Purpose: Pediatric intestinal failure (IF) is defined as the reduced capability of the intestines to absorb nutrients and fluids to sustain adequate growth in children. While enteral nutrition is the preferred method to provide essential nutrients in patients with IF, it is not always possible to wean patients off of parenteral nutrition based on the amount and location of the remaining bowel. Parental nutrition is delivered through a central venous access device (CVAD) which increases the patients’ risk for central line-associated bloodstream infections (CLABSI). To prevent line infections, daily 70% ethanol lock solutions are used due to their proven efficacy and safety in reducing rates of CVAD infections. The cost of ethanol vials has recently increased in the past year by 7,525%. Due to this dramatic increase in cost, which is not covered by insurance, Michigan Medicine has made the executive decision to decrease the use of 70% ethanol lock solutions from daily to three times weekly in patients. The primary outcome of this study is to determine if CLABSI rates will increase in pediatric IF patients using a reduced frequency of 70% ethanol lock solution three times weekly as compared to daily. Secondary aims include a cost analysis related to this change and to compare the rates of catheter repair, catheter replacement, and use of a thrombolytic agent for suspected catheter-related thromboses between groups.

Methods: This longitudinal chart review is a single group cohort non-inferior study utilizing a pre-post intervention design of 13 pediatric patients followed by providers from the Children’s Intestinal Rehabilitation Program (ChIRP) at Michigan Medicine. Pre- and post-intervention data will include the rates of CLABSI per 1000 catheter days, the microorganisms causing the CLABSI, and reported CVAD complications. From a previous a previous study, the basis of absolute difference in rates of CLABSI per 1000 catheter days with one-weekly ethanol lock solution therapy versus placebo, the non-inferiority margin for the difference in rates of CLABSI per 1000 catheter days was defined as 73%.

Results: The mean baseline infection rate when our patients were using daily ethanol lock solution was 1.01 CLABSI per 1000 catheter days. This baseline increased to a mean of 1.06 CLABSI per 1000 catheter days when these patients were switched over to three times weekly ethanol lock solution. The absolute mean difference of -0.05 or 5% increase in CLABSI per 1000 days (95% CI -2.26 to 2.15, p = 0.47) was non-significant demonstrating non-equivalence between three-times-weekly and daily ethanol lock solution therapy. There was, also, no significant difference in catheter repair or replacement between the two regimens (p = 0.11 and 0.41, respectively).

Conclusion: As a pilot assessment of a change in protocol focusing on a single group cohort, we were unable to conclude that three times weekly is non-inferior to daily ethanol lock solution therapy. A study design with a larger sample size over an extended time period beyond 6 months may help demonstrate if times weekly is equivalent to daily ethanol lock solution therapy.

Mon-Yee Fung

University of Michigan

Purpose: Hepatitis C is a viral infection primarily transmitted through intravenous (IV) drug use. Acute infections may last weeks to months, but usually present with non-specific and/or minimal signs or symptoms. However, over a period of 10 to 20 years or more, chronic HCV may progress to cirrhosis, hepatocellular cancer, or death. Patients at risk may warrant treatment with one of the newer HCV agents: Harvoni®(ledipasvir and sofosbuvir), Epclusa®(sofosbuvir and velpatasvir), Mavyret® (glecaprevir and pibrentasvir), Zepatier® (elbasvir and grazoprevir), or Vosevi® (sofosbuvir, velpatasvir, and voxilaprevir) . Currently, there is minimal research comparing these medications in head-to-head “real world” settings.

Methods: This study was an IRB approved, retrospective chart review of patients 18 years of age or older who have been diagnosed with HCV. In addition, these patients were prescribed one of the five HCV agents from July 1, 2019 through June 30, 2020 in the outpatient setting at Michigan Medicine (MM). This report was generated through EPIC MiChart and contained pertinent information including patient name, medication ordered, order date and time, pharmacy, and the primary payer. The primary outcome was the successful completion rates (SCR) between the agents studied. This was defined as patients who had an HCV Polymerase Chain Reaction (PCR) Quantitative test done during therapy and the lab resulted as "Target Not Detected." Secondary outcomes include: patients who experienced a side effect and SCR amongst those who utilized MM Specialty Pharmacy which were both analyzed utilizing a Chi Square analysis. Other outcomes included: distribution of the use of the agents, and when assessed, characterize time to clearance of HCV. Since these agents have been found to be relatively well tolerated in previous studies, the hypothesis is there are no differences in successful completion rates between the agents studied.

Results/Conclusions: Data collection and analysis are currently ongoing. Results will be presented at the the virtual poster session.

Lisa Giangardella

University of Michigan

Introduction: Within the neonatal intensive care unit (NICU), probability of early-onset sepsis is evaluated using maternal risk factors and the neonate's clinical presentation. If antibiotics are initiated based on the Kaiser Permanente probability, inborn neonates receive 36 hours of antibiotic coverage with ampicillin and gentamicin based on our internal guidelines. Antimicrobial stewardship is highly important within this population due to the risks associated with prolonged antibiotic administration such as antibiotic resistance, development of necrotizing enterocolitis, and increased incidence of death in low birth weight infants. Our institution completed a quality improvement (QI) project in 2016 targeted at improving adherence to antibiotic stop times for patients receiving ampicillin and gentamicin for early-onset sepsis. Antibiotic stop times were verbally discussed during daily management system huddle, however, since the cessation of this daily reminder, it was hypothesized that there has been regression to previous trends of inappropriate antibiotic continuation past 36 hours. Methods: The primary aim of this study is to quantify the antibiotic exposure time of patients pre and post implementation of a stop time alert which would be predicted to have a more lasting effect when compared to the previous QI project. The secondary aims are to compare how many patients required extension of their antibiotic course due to positive cultures, if any patients missed antibiotics as a result of the implementation of a stop time alert, and the number of gentamicin levels obtained pre and post implementation of a stop time alert. An antibiotic stop time alert (page to primary provider) will be implemented within the electronic medical record. Six months of data will be collected pre-implementation and 3 months post-implementation which will be analyzed to evaluate the effectiveness of this intervention. Descriptive statistics will be used to characterize and depict patients included in the study population and their outcomes with respect to baseline demographics, maternal characteristics, and adverse effects. The primary outcome will be evaluated utilizing the following statistical tests: t-test. Secondary outcomes will be evaluated using Chi squared test and T-tests as appropriate.

Kaitlynn Hughes

Michigan Medicine

Purpose: Healthcare systems across the United States are facing shortages of trained and experienced pharmacy technicians. Recruiting and retaining pharmacy technicians is a multi-factorial challenge. If not addressed these shortages may lead to increased risk to patients, increased hiring and training costs, and service disruptions. The purpose of this project is to develop and implement strategies to improve pharmacy technician recruitment and retention across Michigan Medicine’s Pharmacy Department. Methods: This quasi-experimental, interventional, non-randomized, non-controlled, one-group, pre-post study will measure the pharmacy technician turnover rate, first year turnover rate, and vacancy rate on a monthly basis from strategy implementation until June 2023. A key stakeholders group composed of pharmacy technicians, pharmacists, pharmacy managers, and human resource representatives brainstormed strategies to improve pharmacy technician recruitment and retention. The strategies were stratified across five domains: recruitment, retention incentives, training/development, schedules/shifts, culture/environment. An electronic feasibility survey was developed to quantitatively prioritize the strategies based on effort, time frame, and estimated cost to implement. This feasibility survey was distributed to a Michigan Medicine pharmacy management email distribution list. A second electronic survey was developed to quantitatively collect pharmacy technician preferences for recruitment and retention strategies. The survey was distributed to a Michigan Medicine Pharmacy Technician email distribution list. Results/Conclusions: Pre-implementation results show that from 2010 to 2020 the average pharmacy technician turnover rate is 15% and average first year turnover rate is 28%. The average vacancy rate of total pharmacy technician positions at Michigan Medicine is 4.77%. Implementation of strategies and data collection of pharmacy technician vacancy rate and first year turnover is ongoing. Available results and conclusions will be presented at the Great Lakes Residency Conference.

David Jackson

Michigan Medicine, Pharmacy Department

At Michigan Medicine, therapeutic drug monitoring (TDM) of tacrolimus was managed by individual providers without a standardized protocol. When a tacrolimus level resulted, nurses needed to contact providers before communicating dose adjustments to renal transplant recipients (RTRs). In September 2019, a standardized nurse-managed tacrolimus TDM protocol was implemented. This single-center, retrospective study evaluated the effectiveness of the protocol compared to the provider-managed TDM in outpatient RTRs.

Stable RTRs who were at least three years post-transplant and taking immediate-release tacrolimus were eligible. Patients were excluded if they had fewer than three tacrolimus levels within 8 months from protocol implementation. Primary outcome was the time to intervene on tacrolimus levels outside the goal range 4-8 ng/ml. Secondary outcomes included tacrolimus time in therapeutic range (TTR), coefficient of variance (CV), and protocol adherence.

Of the 1712 levels (from 174 RTRs), 259 levels (15.1%) were out-of-range. The average time to intervention was significantly different between the Pre-Arm and the Post-Arm (58.8±56.6 vs. 45.6±50.6 hours, respectively; P=0.048). Excluding 50 provider interventions in the Post-Arm, nurses took 42.3±48.9 hours to intervention (P=0.027), -16.5 hours (95%CI: -31.0-1.9 hours) shorter than the provider interventions in the Post-Arm. The TTR was 89.3±17.6% vs 89±19.4% (P=0.816) and CV was 0.20±0.16 vs 0.18±0.11 (P=0.358) in the Pre-Arm and Post-Arm, respectively. Nurses adhered to the protocol 56.2% of the time.

Implementation of a standardized TDM protocol improved efficiency for stable RTRs in the outpatient setting by reducing time to respond to tacrolimus trough levels by 16.5 hours. The difference can also be considered clinically significant, given the twice-daily dosing of tacrolimus. The intrapatient variability in tacrolimus levels was not negatively affected by the TDM protocol. Further analysis is needed to improve the protocol adherence.

Trung Ky <tky@med.umich.edu>

University of Michigan, Michigan Medicine

Lauren McCulley

Veterans Affairs Ann Arbor Healthcare System (VAAAHS)

Background/Rationale: The optimal therapy for first relapse or refractory (R/R) Philadelphia chromosome (Ph-) negative ALL is unknown. The National Comprehensive Cancer Network (NCCN) ALL guideline version 1.2020 includes blinatumomab and inotuzumab as preferred regimens for Ph- R/R ALL and lists a number of chemotherapy regimens as other recommended options, without providing recommendations on when to utilize one regimen over another. Patients with R/R ALL consist of a very heterogeneous population, depending on whether patients are primary refractory or relapsed early versus late following remission. Novel therapies such as blinatumomab and inotuzumab have been studied in Phase III randomized controlled trials in the TOWER and INNOVATE studies, respectively. Despite their inclusion in the current NCCN guidelines, the control arms were sub-standard comparative regimens, as a number of prior studies have demonstrated significantly higher CR rates with chemotherapy regimens, particularly in first salvage with CR rates ranging from 40% to 68%. In the TOWER study, the CR rate was 34% with blinatumomab, compared to 16% with chemotherapy. In INOVATE, the CR/CRi rate was 29.4% with chemotherapy, compared to 80.7% with inotuzumab. Despite the use of suboptimal chemotherapy comparisons, the overall survival difference was only 1 month with inotuzumab versus chemotherapy. There are few published studies comparing chemotherapy versus novel therapies in the relapsed/refractory setting. Furthermore, there are limited data describing outcomes in patients receiving salvage chemotherapy or identifying factors that predict chemosensitivity in the R/R setting.

Objective(s): Compare the overall survival of chemotherapy versus novel therapies in patients with R/R Ph- ALL.

Methods: This study will be a multi-center (U of M, MSK, and UNC), retrospective propensity score matched cohort study of adult patients receiving 2nd line therapy for first relapse or primary refractory B-cell, Ph-negative ALL from January 1, 2012 date to August 1, 2020. Cohort 1 will consist of patients with R/R Philadelphia chromosome negative ALL who received chemotherapy re-induction (ie: hyper-cvad, MOAD, Larson/CALGB-9511, etc). Cohort 2 will consist of patients who received novel therapies (blinatumomab or inotuzumab). Exclusion criteria include Ph+ ALL, patients who only received HyperCVAD part A prior to 2nd line therapy, and patient with incomplete records. Data will be collected on demographics, disease related characteristics, initial therapy, re-induction therapy, toxicity, post re-induction therapy, and outcomes. The primary endpoint is overall survival. Time-to-event analyses (EFS and OS) will be compared using the Kaplan-Meier method and long-rank test.

Results: Non-applicable, research in progress.

Conclusions/Discussion: Non-applicable, research in progress.

Madeleine Ochs

Michigan Medicine

Carl Okerberg

Michigan Medicine

Emily Sahagian

VA Ann Arbor Healthcare System

Purpose: To standardize hyperkalemia management, a pediatric hyperkalemia guideline was developed and an order set was built into the electronic medical record system. The purpose of this study was to determine the impact of implementing a hyperkalemia protocol on hyperkalemia management in pediatric patients. Methods: This is a retrospective, single center cohort study comparing the efficacy of hyperkalemia management pre-protocol verses post-protocol in pediatric patients at C.W. Mott Children’s Hospital. Patients with a serum potassium ≥ 5.5 mmol/L who received a medication of interest, such as calcium, insulin, furosemide, sodium bicarbonate, or sodium polystyrene sulfonate between December 2018 to December 2020 were screened for inclusion in this study. Those with a hemolyzed sample, admitted to the NICU, received dialysis, or missing a serum potassium level within 6 hours after treatment were excluded. The primary outcome was the median reduction in serum potassium level from baseline to last recorded potassium value within 6 hours post treatment. Results/Conclusion: Data collection is ongoing and finalized results will be presented at the UM COP Research Forum.

Andrea Setiawan

Michigan Medicine

Jessica Sharkey

Michigan Medicine

Andrew Smith

Michigan Medicine

Purpose: Clinical decision support (CDS) systems have been shown to improve outcomes in patients whose medications require renal dose adjustments. Literature has also shown that CDS systems can result in a significant improvement in medication safety. At Michigan Medicine, a novel renal scoring tool was implemented that included flowsheet data integration, resulting in an automatic and customizable method of tracking clinical outcomes. To maximize utilization of this renal scoring tool, a new renal dosing workflow was implemented in November 2020 with a Pharmacy and Therapeutics committee approved renal dosing policy for a subset of medications that pharmacists can dose adjust without contacting the provider.

Methods: The Institutional Review Board at Michigan Medicine approved this observational prospective study. Patients at least 18 years of age and receiving at least one of the pre-specified subset of medications on the renal dose policy while admitted from September 2020 to January 2021 were included. The scoring tool fires if any of the following occurs: patient’s serum creatinine or calculated creatinine clearance crosses a dosing threshold on the renal dosing policy or an order is signed or verified for a medication on the renal dose policy and the patient has abnormal renal function. Each time the scoring tool is triggered, the patient’s current creatinine clearance, serum creatinine, and medication related information are populated onto a flowsheet. The primary outcome will be the incidence of medication verification with non-concordant renal dosing pre- and post-implementation of the new renal workflow. Secondary outcomes will include the duration of non-concordant medication orders, dose related adverse drug events for non-optimized patients, and reported medication errors within the study time frame.

Steven Stettner

Michigan Medicine, Department of Pharmacy

Patients diagnosed with acute leukemias have an underlying susceptibility to bloodstream infections (BSI) due to impaired immune function, myelosuppressive treatment regimens, and needs for invasive intravenous access.1,2 The estimated prevalence of BSIs in these patients is about 11-38%, representing a crude mortality rate of 12-42%.3 The level of antibiotic use and frequency of prolonged hospitalization represent significant risks of exposure and colonization with multi-drug resistant organisms, enhancing risk of mortality.4

The guidelines regarding the treatment and prevention of bacteremia in leukemia patients lack substantial data to support their recommended antibiotic treatment durations and seem to be a result of consensus opinion. In studies of other patient populations presenting with bacteremias, antibiotic courses less than ten days compared to standard courses of antibiotics evidenced no significant differences in microbiological cure or mortality.

The objective of this study is to assess the outcomes of shorter courses of antibiotic therapy compared to standard courses in leukemia patients presenting with a bacteremia from an enterobacteriaceae pathogen.

Ryan Sutherlan

University of Michigan, College of Pharmacy

Carolanne Wartman

Michigan Medicine

Lauren Williams

Veterans Affairs Ann Arbor Healthcare System

Roberta Wilson

Michigan Medicine

Jiashan (Jenni) Xu

Michigan Medicine

Caitlyn Young

University of Michigan College of Pharmacy