Next Generation Sequencing for Microbial Analysis to Select Prophylactic Antibiotic Selection before Urologic Stone Surgery: A Culture Change

Background: This paper aims to determine if the combination of polymerase chain reaction (PCR) and next-generation sequencing (NGS) could identify bacteria in culture-negative urine that would alter prophylaxis management. Methods: We sent approximately 5 10 mL of a preoperative urine sample to MicrogenDx for PCR/NGS analysis performed after surgery (blind to the surgeon). The physician prescribed standard of care antibiotic prophylaxis. Cases modeling the hospital course of 3 random patients were reviewed by eight urologists after surgery to determine if NGS results would change their prophylaxis regimen. An infectious disease pharmacist reviewed the cases and provided the “ideal” regimen. Results: Urine cultures identified bacteria in 11% (2/18) of cases. Culture speciation results were consistent with NGS results. NGS detected a dominant bacteria in 56% (10/18) of negative cultures and targetable bacteria in all samples. There was a 15% (3/20) infection rate. In both cases, NGS results suggest inadequate prophylaxis. In response to the case scenarios, 100%, 88%, and 88% of the urologists reported they would change prophylaxis with NGS results. During a case scenario, physicians would tend to overprescribe antibiotics given PCR/NGS data for prophylaxis selection. Conclusion: NGS identifies a targetable bacterium in up to 80% of negative urine cultures before urologic stone surgery. Responses to case scenarios indicate that physicians would change management based on NGS results. Inter-professional (urologic and pharmacy) antibiotic selection with PCR/16S DNA testing may be helpful to improve antibiotic stewardship. *Co-First Authors. #Corresponding Author. How to cite this paper: Das, R., Tseng, T., Short, M.I., Reveles, K., Wheeler, A., Hudson, R., Fongang, B. and Liss, M.A. (2021) Next Generation Sequencing for Microbial Analysis to Select Prophylactic Antibiotic Selection before Urologic Stone Surgery: A Culture Change. Open Journal of Urology, 11, 289-304. https://doi.org/10.4236/oju.2021.117027 Received: April 6, 2021 Accepted: July 23, 2021 Published: July 26, 2021 Copyright © 2021 by author(s) and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access R. Das, T. Tseng et al. DOI: 10.4236/oju.2021.117027 290 Open Journal of Urology


Introduction
Ureteroscopy for urinary tract urolithiasis is a standard procedure with an estimated 9200 cases in the US per year [1]. Despite antibiotic prophylaxis, the incidence of significant infections after flexible ureteroscopic (URS) lithotripsy is on the rise [2]. To prevent infections, the standard of care is to obtain a negative urine culture before surgery. However, with the next-generation sequencing, we are now able to understand that the urine is not sterile and that pathogenic bacteria may be still present, undetected by urine culture.
One of the challenges when choosing a prophylactic agent is that preoperative urine cultures often show no growth for patients who later develop SIRS [3] [4] [5]. Singh et al. found no significant association between pelvic urine cultures or stone cultures and the occurrence of SIRS [6]. We hypothesize the stone or previously placed stent may allow bacteria to form biofilms containing a small number of essential bacteria not detected by standard cultures.
Next-generation sequencing (NGS) poses an alternative to the traditional culture model by using high-throughput sequencing of rapid PCR for resistance genes combined with 16S rRNA (a type of NGS) to detect specific bacterial strains and has been used for the detection of urinary tract infections [7]. With its improved sensitivity, NGS (PCR/16s rRNA) with resistance genes could be used to guide antibiotic therapy. In this study, we pilot this novel approach quantitatively to evaluate the NGS platform's ability to identify bacteria in culture-negative urine that would alter the choice of antibiotic prophylaxis for patients undergoing urologic stone procedures. We also evaluate this approach qualitatively to determine its usefulness to urologic surgeons to inform clinical trial design.

Population
After IRB approval (HSC20050234H), subjects were recruited from urology clinics at their preoperative appointment before their planned ureteroscopy (URS) surgery for urinary stone. We did not exclude patients with ureteral stents, including all patients undergoing percutaneous nephrolithotomy (PCNL) or URS within the next two weeks. We informed patients that the results were purely for research purposes, and no analysis performed until after surgery. We also described we would not provide additional information to their physician that could alter antibiotics.

Next-Generation Sequencing
We collected whole urine (approximately 30 -50 mL) utilizing special vaccutanors supplied by MicrogenDx. We sent around 5 -10 mL of urine taken for culture to MicroGen Diagnostics, a CAP-accredited and CLIA licensed clinical diagnostic lab, for analysis. MicrogenDx performed rapid PCR for common resistance genes in 21 subjects before URS. Each initial target bacterial or fungi DNA, whose concentration was measured to obtain an initial concentration (ng/uL), was diluted to obtain a six to eight-fold serial dilution series and run on the quantitative PCR (qPCR) panel assay on the Roche LightCycler 480 II instrument.
We sequenced the V1-V2 hypervariable bacterial regions with 16S RNA using Ion Torrent (Ion Torrent PGM). Physicians did not obtain results before surgery. The physician proceeded with surgical management and standard of care antibacterial prophylaxis.

Data Collection
We recorded data from physician notes and medical record review regarding the details of patient history, urologic stone procedure, postoperative course, and infection outcomes. We recorded all speciation results from the standard of care urine cultures. We then compared these results to speciation results in the PCR/ NGS (16s rRNA).

Case Study Creation
We perform a qualitative review in the form of case scenarios to determine if NGS would indeed alter antibiotic prophylaxis. Choosing from the cases enrolled, we randomly selected three cases to be reviewed by eight board-certified Urologists. One infectious disease pharmacists also reviewed the case reports and recommended the "ideal" antibiotic regimen for each case.

Surgical Methods
A variety of urologic stone procedures were represented in this cohort, with cases of PCNL and URS, laser lithotripsy, and basket retrieval, with and without

Bacterial Identification and Antibiotic Resistance
Of the 20 cases, only 2 SOC cultures had speciation results and 16s rRNA analysis identified the exact bacteria colonized on both accounts ( Figure 1). One patient had grown two bacterial species on culture, and NGS identified both species. In the 10 cases where SOC cultures resulted in no growth, NGS was able to identify targetable bacteria. In only 2 of those 10 cases with no growth, the dominant species detected by NGS (Citrobacter and Lactobacillus) was unlikely to cause postoperative infections. The other dominant species identified in these cases were E. coli and S. epidermidis, respectively. In nearly 80% of negative cultures, NGS provided dominant speciation data to consider when choosing antimicrobial prophylaxis before urologic stone procedures ( Figure 2).

Cases of Infection
Of the 18 patients who underwent their procedure, 2 cases (11%) developed infections postoperatively. In the first infection case, the patient did not supply a urine culture before URS. The surgeon chose cefazolin for antibiotic prophylaxis. NGS detected a high bacterial load (>107): 53% Morganella morganii and 45% E. coli. NGS was also able to identify resistance genes against beta-lactams and fluoroquinolones. Based on this data, we would recommend trimethoprimsulfamethoxazole or a 3 rd generation cephalosporin rather than a 1 st generation cephalosporin. In the second case of infection, three or more organisms present, each higher than 10,000 cu/mL, per SOC UC results. The surgeon chose cefazolin and piperacillin/tazobactam for prophylaxis. NGS detected Citrobacter (64%), Veillonella atypica (27%), and resistance genes to methicillin, beta-lactams, macrolides, and aminoglycosides. NGS also detected Candida albicans in its fungal We display our enrollment of 20 subjects through the study. The urine was split between culture and next-generation sequencing with PCR of plasmid genes. We then compared the results of the PCR to the culture specimen as ground truth. Figure 1. Consort diagram.
We display a diagram of urine culture and corresponding PCR results from subjects On the left are the positive urine cultures noting alignment between genetic and phenotypic results. Of note, Candida is not routinely cultured before urologic stone procedures. On the right are all of the negative urine culture results and their associated NGS/PCR results. Bright yellow represents those that had postoperative clinical infection with a red circle to indicate the causative organism. The orange squares were those subjects selected for the case reports in the supplemental figures. Figure 2. Results diagram.
screen as a standard part of the test. The patient developed funguria postoperatively with Candida albicans, indicating that NGS may be specific enough to recommend antifungal prophylaxis appropriately. These findings would be helpful in PCNL patient's that have had multiple antibiotics with high level of suspicion for candida

Case Scenario Data
Eight urologists responded to the three case scenarios, and in each of the 3 cases, 100%, 88%, and 88% of the physicians would have changed their prophylaxis have concerns about using this technology for antibiotic prophylaxis related primarily to implementation.

Discussion
We report several findings from our pilot study that include: NGS may provide actionable information above standard urine culture, urologists find the information useful, and concern for antibiotic escalation may benefit from interprofessional collaboration with pharmacists to select preoperative antibiotics in the context of PCR/NGS testing. Testing NGS is a practical extension to standard urinary culture in that the specimen can be split and sent for culture and NGS.
Results of NGS are available in 48 hours to allow for time for preoperative antibiotic selection well before surgery.
In our first question, we address the usefulness of a PCR/NGS based urine profile before URS in this pilot study. We found that NGS detects targetable bacteria and fungi in culture-negative urine and propose utilizing this data in a prospective trial to use NGS to determine preoperative antibiotic prophylaxis.
NGS may be more effective and specific than those predicated upon SOC preoperative urine cultures mostly negative before surgery. For example, we identified several culture-negative patients with dominant bacterial types that may influence a physician to prescribe a 1 st generation cephalosporin compared to a gram-negative dominant group that may need a fluoroquinolone or 3 rd generation cephalosporin. Importantly, urologists do not typically send a urine specimen specifically for fungi (candida sp.  [13]. Given the mortality rates and economic costs of urosepsis, SOC practices must be optimized to reduce the risk of infectious complications following urologic procedures.
Our qualitative questions address the usefulness of PCR/NGS testing to urologists before URS. We identified that 88% -100% of urologists reported that they would have changed their prophylactic regimen based on NGS results in the post case survey. However, during the cases, most urologists did not choose the ideal regimen recommended by infectious disease pharmacists and in many escalated antibiotic use to either more antibiotics or broader spectrum. The findings of this study suggest that interdisciplinary collaboration between physicians and pharmacists may prevent excessive use of aggressive therapy when utilizing highly sensitive pathogen detection modalities, such as NGS.
We have several limitations to our study. The sample size limits the power of this study; however, this is a pilot study to inform a larger clinical trial and is informative for planning. We decided to publish these results due to the novelty of the research and its potential to change the preoperative antibiotic selection for a

Conclusion
Infectious complications are the most common cause of death following urologic stone procedures. This study found that NGS can identify a targetable bacterium in up to 80% of negative urine cultures before urologic stone surgery. Responses to case scenarios indicate that physicians would change management based on NGS results. Using this data we have initiated a clinical trial using NGS to augment antibiotic section in urine culture negative patients prior to ureteroscopy for stone surgery (NCT04404855).