Ramucirumab plus triplet chemotherapy as an alternative salvage treatment for patients with metastatic colorectal cancer.

BACKGROUND: Ramucirumab is indicated for salvage treatment after failure of first-line treatment for metastatic colorectal cancer (mCRC). However, the application of ramucirumab at later-line treatment in real-world practice has not received much discussion. METHODS: In this retrospective study, we enrolled 70 patients with mCRC who received ramucirumab plus chemotherapy at National Taiwan University Hospital between 2018 and 2019. RESULTS: Compared with those who received third- or later-line ramucirumab treatment, patients who received second-line ramucirumab treatment had significantly longer median time to treatment discontinuation (mTTD; 6.7 vs 3.6 months, P = .004) and median overall survival (mOS; not reached vs 7.6 months, P = .009). Multivariate analyses revealed that second-line ramucirumab and triplet chemotherapy backbone were the only independent predictive factors for long mTTD and mOS. Patients who received ramucirumab with triplet chemotherapy had a significantly longer mOS than did patients who received ramucirumab with doublet chemotherapy (not reached vs 5.6 months, P = .002). Among those receiving second-line ramucirumab treatment, combination with triplet chemotherapy led to a longer mTTD than did combination with doublet chemotherapy, but the difference was non-significant (not reached vs 4.4 months, P = .108). By contrast, in patients receiving fourth- or later-line ramucirumab, combination with triplet chemotherapy led to significantly longer mTTD than did combination with doublet chemotherapy (8.0 vs 2.9 months, P = .032). CONCLUSION: Ramucirumab plus triplet chemotherapy may be an alternative regimen in patients with mCRC, particularly as a later-line treatment modality.

Evaluating capabilities of large language models: Performance of GPT-4 on surgical knowledge assessments.

BACKGROUND: Artificial intelligence has the potential to dramatically alter health care by enhancing how we diagnose and treat disease. One promising artificial intelligence model is ChatGPT, a general-purpose large language model trained by OpenAI. ChatGPT has shown human-level performance on several professional and academic benchmarks. We sought to evaluate its performance on surgical knowledge questions and assess the stability of this performance on repeat queries. METHODS: We evaluated the performance of ChatGPT-4 on questions from the Surgical Council on Resident Education question bank and a second commonly used surgical knowledge assessment, referred to as Data-B. Questions were entered in 2 formats: open-ended and multiple-choice. ChatGPT outputs were assessed for accuracy and insights by surgeon evaluators. We categorized reasons for model errors and the stability of performance on repeat queries. RESULTS: A total of 167 Surgical Council on Resident Education and 112 Data-B questions were presented to the ChatGPT interface. ChatGPT correctly answered 71.3% and 67.9% of multiple choice and 47.9% and 66.1% of open-ended questions for Surgical Council on Resident Education and Data-B, respectively. For both open-ended and multiple-choice questions, approximately two-thirds of ChatGPT responses contained nonobvious insights. Common reasons for incorrect responses included inaccurate information in a complex question (n = 16, 36.4%), inaccurate information in a fact-based question (n = 11, 25.0%), and accurate information with circumstantial discrepancy (n = 6, 13.6%). Upon repeat query, the answer selected by ChatGPT varied for 36.4% of questions answered incorrectly on the first query; the response accuracy changed for 6/16 (37.5%) questions. CONCLUSION: Consistent with findings in other academic and professional domains, we demonstrate near or above human-level performance of ChatGPT on surgical knowledge questions from 2 widely used question banks. ChatGPT performed better on multiple-choice than open-ended questions, prompting questions regarding its potential for clinical application. Unique to this study, we demonstrate inconsistency in ChatGPT responses on repeat queries. This finding warrants future consideration including efforts at training large language models to provide the safe and consistent responses required for clinical application. Despite near or above human-level performance on question banks and given these observations, it is unclear whether large language models such as ChatGPT are able to safely assist clinicians in providing care.

Impact of preoperative factors and waiting time on post-appendectomy complications: a retrospective study.

BACKGROUND: Several factors are associated with increased postoperative complications after appendectomies. However, few studies combined these potential factors for comprehensive prediction of surgical outcomes. Whether high-risk patients benefit from a shorter waiting time for surgery remains unclear. This study aimed to explore the impact of surgical waiting time and potential risk factors on postoperative complications. METHODS: A total of 1343 patients diagnosed with acute appendicitis requiring an emergent appendectomy were included from 2013 to 2018. The preoperative risk factors associated with postoperative complications were selected and the probability of postoperative complications was calculated by multivariate logistic regression model. Patients were divided into four groups based on the risk (high & low) and time to surgery (> 12 & ≤12 hours). The odds ratios for complications were evaluated between groups. RESULTS: The selected risk factors included age, neutrophil-lymphocyte ratio, systemic inflammatory response syndrome and abdominal pain duration. Compared with low-risk patients with time to surgery ≤12 hours, high-risk patients with time to surgery > 12 hours had significant increased overall postoperative complication rate (16.85% vs. 8.16%, p = 0.002) and a trend toward increased surgical site infection rate (10.99% vs. 6.46%, p = 0.058). When operated within 12 hours, there was no difference in outcomes between high- and low-risk patients. On the other hand, time to surgery > 12 hours did not increase complication rate in low-risk patients. CONCLUSIONS: The surgical outcome may be affected by preoperative factors and time to surgery. It is suggested that high-risk patients receive appendectomy within 12 hours to avoid increased postoperative complications.

Intraoperative Detection of Surgical Gauze Using Deep Convolutional Neural Network.

During laparoscopic surgery, surgical gauze is usually inserted into the body cavity to help achieve hemostasis. Retention of surgical gauze in the body cavity may necessitate reoperation and increase surgical risk. Using deep learning technology, this study aimed to propose a neural network model for gauze detection from the surgical video to record the presence of the gauze. The model was trained by the training group using YOLO (You Only Look Once)v5x6, then applied to the testing group. Positive predicted value (PPV), sensitivity, and mean average precision (mAP) were calculated. Furthermore, a timeline of gauze presence in the video was drawn by the model as well as human annotation to evaluate the accuracy. After the model was well-trained, the PPV, sensitivity, and mAP in the testing group were 0.920, 0.828, and 0.881, respectively. The inference time was 11.3 ms per image. The average accuracy of the model adding a marking and filtering process was 0.899. In conclusion, surgical gauze can be successfully detected using deep learning in the surgical video. Our model provided a fast detection of surgical gauze, allowing further real-time gauze tracing in laparoscopic surgery that may help surgeons recall the location of the missing gauze.

Evaluating Capabilities of Large Language Models: Performance of GPT4 on Surgical Knowledge Assessments.

Background: Artificial intelligence (AI) has the potential to dramatically alter healthcare by enhancing how we diagnosis and treat disease. One promising AI model is ChatGPT, a large general-purpose language model trained by OpenAI. The chat interface has shown robust, human-level performance on several professional and academic benchmarks. We sought to probe its performance and stability over time on surgical case questions. Methods: We evaluated the performance of ChatGPT-4 on two surgical knowledge assessments: the Surgical Council on Resident Education (SCORE) and a second commonly used knowledge assessment, referred to as Data-B. Questions were entered in two formats: open-ended and multiple choice. ChatGPT output were assessed for accuracy and insights by surgeon evaluators. We categorized reasons for model errors and the stability of performance on repeat encounters. Results: A total of 167 SCORE and 112 Data-B questions were presented to the ChatGPT interface. ChatGPT correctly answered 71% and 68% of multiple-choice SCORE and Data-B questions, respectively. For both open-ended and multiple-choice questions, approximately two-thirds of ChatGPT responses contained non-obvious insights. Common reasons for inaccurate responses included: inaccurate information in a complex question (n=16, 36.4%); inaccurate information in fact-based question (n=11, 25.0%); and accurate information with circumstantial discrepancy (n=6, 13.6%). Upon repeat query, the answer selected by ChatGPT varied for 36.4% of inaccurate questions; the response accuracy changed for 6/16 questions. Conclusion: Consistent with prior findings, we demonstrate robust near or above human-level performance of ChatGPT within the surgical domain. Unique to this study, we demonstrate a substantial inconsistency in ChatGPT responses with repeat query. This finding warrants future consideration and presents an opportunity to further train these models to provide safe and consistent responses. Without mental and/or conceptual models, it is unclear whether language models such as ChatGPT would be able to safely assist clinicians in providing care.

Lipocalin-2 gene expression during liver regeneration after partial hepatectomy in rats.

BACKGROUND: Lipocalin-2 (Lcn2) is related to cell proliferation. We studied Lcn2 gene expression during liver regeneration after partial hepatectomy (PH). METHODS: Male Wistar rats were sacrificed before and 2, 4, 6, 12, 24, 72 h, 7 days after 70% or 40% PH. The remnant liver weight/body weight (RLW/BW) ratio, Lcn2 gene and mRNA expression in the remnant livers were measured. Hepatocytes and nonparenchymal cells were isolated from the remnant livers. Expression of Lcn2 related protein was detected by Western blot. RESULTS: The RLW/BW ratio increased to nearly 90% of the original liver 72 h after PH. Lcn2 gene expression showed upward curves from 4 to 72 h after PH in both 70% and 40% PH rats and peaked at 12 h (8 times vs 0 h). Lcn2 mRNA expression showed parallel upward curves from 2 to 72 h. The peak was significantly higher in 70% PH rats (2(7) times vs 0 h) than in 40% PH rats (2(5) times vs 0 h) 12 h after PH (p < 0.05). Lcn2 related protein in the isolated liver cells was markedly enhanced 24 h after PH, more in hepatocytes than in nonparenchymal cells. CONCLUSION: The expressions of Lcn2 gene and mRNA, and its related protein increased markedly after PH. Lcn2 might be important in the genetic regulation of liver regeneration after PH.

Combined cholecystectomy in gastric cancer surgery.

BACKGROUND: Many studies have described the risk factors of gallstone formation in gastric cancer patients after gastrectomy, but few studies focus on the management of asymptomatic gallstones. Our goal is to examine the rationale of simultaneous cholecystectomy during gastric cancer surgery, and influence of surgical mortality, morbidity and overall survival after combined cholecystectomy and gastrectomy. METHODS: We retrospectively reviewed 445 gastric cancer patients and the gallbladders evaluated by abdominal ultrasound or computed tomography preoperatively and postoperatively. Clinicopathologic factors, including surgical morbidity, mortality and overall survival of combined surgery, were compared between patients receiving gastrectomy with simultaneous cholecystectomy and patients receiving gastrectomy only. We also evaluated the risk factors of gallstone formation after gastrectomy and the probability of subsequent cholecystectomy after gastrectomy in gastric cancer patients with or without asymptomatic gallstones. RESULTS: Of 445 gastric cancer patients, 52 (11.7%) patients had asymptomatic gallstones upon diagnosis of gastric cancer. Among patients with healthy gallbladders, 15.2% developed gallstones after gastrectomy. Men and older patients (age over 60) had significantly higher risk of gallstone formation. Rate of subsequent cholecystectomy in patients with and without preoperative asymptomatic gallstones was 30.8% and 4.5%, respectively (p = 0.005). The rates of mortality and morbidity were not significantly different between combined surgery (3.4%, 24.2%) and gastrectomy only (3.1%, 22%). There was also no significant difference in 5-year survival between combined surgery (61%) and gastrectomy only (63%) groups. CONCLUSION: Combined cholecystectomy for asymptomatic gallstone in gastric cancer surgery may be considered. It was not associated with increased surgical morbidity or mortality, and had no significant effect on overall survival.

IL-6 regulates Mcl-1L expression through the JAK/PI3K/Akt/CREB signaling pathway in hepatocytes: implication of an anti-apoptotic role during liver regeneration.

AIMS: To investigate the role and the regulation of the long variant of myeloid cell leukemia-1 protein (Mcl-1L) during liver regeneration. BACKGROUND: Liver regeneration is an important phenomenon after liver injury. The rat partial hepatectomy (PH) model was used to characterize liver regeneration and Mcl-1L expression after PH. METHODS: Male Wistar rats were subjected to 70% PH. The expression of mcl-1L mRNA was determined by quantitative RT-PCR, and protein levels were analyzed by Western blot analysis and immunohistochemistry during liver regeneration. Functional evaluations of Mcl-1L were tested using chemical inhibition (flavopiridol), genetic inhibition (siRNA) of Mcl-1L production, and by assaying for annexin V levels and DNA ladder formation. Serum IL-6 levels were determined by enzyme immunoassays; signal transduction of IL-6-regulated Mcl-1L expression was verified by chemical inhibitors and decoy double-stranded oligodeoxynucleotides. RESULTS: High levels of Mcl-1L were observed in remnant tissue at 4 h after PH. Administration of flavopiridol decreased Mcl-1L accumulation and also inhibited liver regeneration. IL-6 administration promoted the accumulation of Mcl-1L in rat hepatocytes, an effect that was impaired by siRNA treatments that reduced Mcl-1L production. Chemical inhibition and decoy oligonucleotide competition demonstrated that IL-6-induced Mcl-1L production required signaling mediated by JAK kinase, phosphoinositide 3-kinase (PI3K), and cAMP response-element-binding (CREB) proteins. CONCLUSION: Mcl-1L is an anti-apoptotic protein induced during liver regeneration after PH in rats. The expression of Mcl-1L is induced by IL-6 through the JAK/PI3K/Akt/CREB signaling pathway. Chemotherapy drugs that depend on Mcl-1L- or IL-6-related signaling should be considered carefully before use in patients undergoing hepatectomy for malignant tumor resection.

Interleukin-6 mediates angiotensinogen gene expression during liver regeneration.

BACKGROUND: Angiotensinogen is the precursor of angiotensin II, which is associated with ischemia-reperfusion injury. Angiotensin II reduces liver regeneration after hepatectomy and causes dysfunction and failure of reduced-size liver transplants. However, the regulation of angiotensinogen during liver regeneration is still unclear. AIMS: To investigate the regulation of angiotensinogen during liver regeneration for preventing angiotensin II-related ischemia-reperfusion injury during liver regeneration. METHODS: A mouse in vitro partial hepatectomy animal model was used to evaluate the expression of interleukin-6 (IL-6) and angiotensinogen during liver regeneration. Serum IL-6 and angiotensinogen were detected by enzyme immunoassay (EIA). Angiotensinogen mRNA was detected by RT-PCR. Tissue levels of angiotensinogen protein were detected by Western blot analysis. Primary cultures of mouse hepatocytes were used to investigate IL-6-induced angiotensinogen. Chemical inhibitors were used to perturb signal transduction pathways. Synthetic double-stranded oligodeoxynucleotides (ODNs) were used as 'decoy' cis-elements to investigate transcription. Ki 67 staining and quantification were used to verify liver regeneration. RESULTS: In the in vivo model, the levels of serum IL-6 and angiotensinogen correlated. In the in vitro model, IL-6 transcriptionally regulated angiotensinogen expression. Additionally, IL-6 mediated angiotensinogen expression through the Janus kinase (JAK)/signal transducer and activator of transcription 3 (STAT3) and JAK/p38 signaling. Decoy ODN analyses revealed that STAT3 and nuclear factor-kB (NF-kB) also played critical roles in the transcriptional regulation of angiotensinogen by IL-6. IL-6-mediated signaling, JAK2, STAT3 and p38 inhibitors reduced angiotensinogen expression in the partially hepatectomized mice. CONCLUSION: During liver regeneration, IL-6-enhanced angiotensinogen expression is dependent on the JAK/STAT3 and JAK/p38/NF-kB signaling pathways. Interruption of the molecular mechanisms of angiotensinogen regulation may be applied as the basis of therapeutic strategies for preventing angiotensin II-related ischemia-reperfusion injury during liver regeneration.