High-throughput ML-guided design of diverse single-domain antibodies against SARS-CoV-2 (preprint)

Treating rapidly evolving pathogenic diseases such as COVID-19 requires a therapeutic approach that accommodates the emergence of viral variants over time. Our machine learning (ML)-guided sequence design platform combines high-throughput experiments with ML to generate highly diverse single-domain antibodies (VHHs) that bind and neutralize SARS-CoV-1 and SARS-CoV-2. Crucially, the model, trained using binding data against early SARS-CoV variants, accurately captures the relationship between VHH sequence and binding activity across a broad swathe of sequence space. We discover ML-designed VHHs that exhibit considerable cross-reactivity and successfully neutralize targets not seen during training, including the Delta and Omicron BA.1 variants of SARS-CoV-2. Our ML-designed VHHs include thousands of variants 4-15 mutations from the parent sequence with significantly improved activity, demonstrating that ML-guided sequence design can successfully navigate vast regions of sequence space to unlock and future-proof potential therapeutics against rapidly evolving pathogens.

Early versus delayed ileocolic resection for complicated Crohn's disease: is "cooling off" necessary?

BACKGROUND: Ileal Crohn's disease (CD) complicated by intraabdominal abscess, phlegmon, fistula, and/or microperforation is commonly treated with antibiotics, bowel rest, and percutaneous drainage followed by interval ileocolic resection (ICR). This "cool off" strategy is intended to facilitate the safe completion of a one-stage resection using a minimally invasive approach and minimize perioperative complications. There is limited data evaluating the benefits of delayed versus early resection. METHODS: A retrospective review of a prospectively maintained inflammatory bowel disease (IBD) database at a tertiary center was queried from 2013-2020 to identify patients who underwent ICR for complicated ileal CD confirmed on preoperative imaging. ICR cohorts were classified as early (≤ 7 days) vs delayed (> 7 days) based on the interval from diagnostic imaging to surgery. Operative approach and 30-day postoperative morbidity were analyzed. RESULTS: Out of 474 patients who underwent ICR over the 7-year period, 112 patients had complicated ileal CD including 99 patients (88%) with intraabdominal abscess. Early ICR was performed in 52 patients (46%) at a median of 3 days (IQR 2, 5) from diagnostic imaging. Delayed ICR was performed in 60 patients (54%) following a median "cool off" period of 23 days of non-operative treatment (IQR 14, 44), including preoperative percutaneous abscess drainage in 17 patients (28%). A higher proportion of patients with intraabdominal abscess underwent delayed vs early ICR (57% vs 43%, p = 0.19). Overall, there were no significant differences in the rate of laparoscopy (96% vs 90%), conversion to open surgery (12% vs 17%), rates of extended bowel resection (8% vs 13%), additional concurrent procedures (44% vs 52%), or fecal diversion (10% vs 2%) in the early vs delayed ICR groups. The median postoperative length of stay was 5 days in both groups with an overall 25% vs 17% (p = 0.39) 30-day postoperative complication rate and a 6% vs 5% 30-day readmission rate in early vs delayed ICR groups, respectively. Overall median follow-up time was 14.3 months (IQR 1.2, 24.1) with no difference in the rate of subsequent CD-related intestinal resection (4% vs 5%) between the two groups. CONCLUSIONS: In this contemporary series, at a high-volume tertiary referral center, a "cool off" delayed resectional approach was not found to reduce perioperative complications in patients undergoing ICR for complicated ileal Crohn's disease. Laparoscopic ICR can be performed within one week of diagnosis with low rates of conversion and postoperative complications.