The use of artificial intelligence in medicine is currently an issue of great interest, especially with regard to the diagnostic or predictive analysis of medical images. Adoption of an artificial ...intelligence tool in clinical practice requires careful confirmation of its clinical utility. Herein, the authors explain key methodology points involved in a clinical evaluation of artificial intelligence technology for use in medicine, especially high-dimensional or overparameterized diagnostic or predictive models in which artificial deep neural networks are used, mainly from the standpoints of clinical epidemiology and biostatistics. First, statistical methods for assessing the discrimination and calibration performances of a diagnostic or predictive model are summarized. Next, the effects of disease manifestation spectrum and disease prevalence on the performance results are explained, followed by a discussion of the difference between evaluating the performance with use of internal and external datasets, the importance of using an adequate external dataset obtained from a well-defined clinical cohort to avoid overestimating the clinical performance as a result of overfitting in high-dimensional or overparameterized classification model and spectrum bias, and the essentials for achieving a more robust clinical evaluation. Finally, the authors review the role of clinical trials and observational outcome studies for ultimate clinical verification of diagnostic or predictive artificial intelligence tools through patient outcomes, beyond performance metrics, and how to design such studies.
RSNA, 2018.
Background
Our previous randomized controlled trial revealed no difference in 2‐year overall survival (OS) between extended and standard resection for pancreatic adenocarcinoma. The present study ...evaluated the 5‐year OS and recurrence patterns according to the extent of pancreatectomy.
Methods
Between 2006 and 2009, 169 consecutive patients were prospectively enrolled and randomized to standard (n = 83) or extended resection (n = 86) groups to compare 5‐year OS rate, long‐term recurrence patterns and factors associated with long‐term survival.
Results
The surgical R0 rate was similar between the standard and extended groups (85.5 vs. 90.7%, P = 0.300). Five‐year OS (18.4 vs. 14.4%, P = 0.388), 5‐year disease‐free survival (14.8 vs. 14.0%, P = 0.531), and overall recurrence rates (74.7 vs. 69.9%, P = 0.497) were not significantly different between the two groups, although the incidence of peritoneal seeding was higher in the extended group (25 vs. 8.1%, P = 0.014).
Conclusions
Extended pancreatectomy does not have better short‐term and long‐term survival outcomes, and shows similar R0 rates and overall recurrence rates compared with standard pancreatectomy. Extended pancreatectomy does not have to be performed routinely for all cases of resectable pancreatic adenocarcinoma, especially considering its associated increased morbidity shown in our previous study.
Highlight
Jang and colleagues conducted a multicenter randomized controlled trial to evaluate long‐term overall survival and recurrence patterns according to the extent of pancreatectomy for pancreatic head cancer. Extended pancreatectomy did not demonstrate better short‐term and long‐term survival outcomes, and showed similar R0 rates and recurrence rates compared with standard pancreatectomy.
Full text
Available for:
FZAB, GIS, IJS, KILJ, NLZOH, NUK, OILJ, SAZU, SBCE, SBMB, UL, UM, UPUK
Background
The comprehensive complication index (CCI) integrates all complications of the Clavien–Dindo classification (CDC) and offers a metric approach to measure morbidity. The aim of this study ...was to evaluate the CCI at a high-volume center for gastric cancer surgery and to compare the CCI to the conventional CDC.
Methods
Clinical factors were collected from the prospective complication data of gastric cancer patients who underwent radical gastrectomy at Seoul National University Hospital from 2013 to 2014. CDC and CCI were calculated, and risk factors were investigated. Correlations and generalized linear models of hospital stay were compared between the CCI and CDC. The complication monitoring model with cumulative sum control-CCI (CUSUM-CCI) was displayed for individual surgeons, for comparisons between surgeons, and for the institution.
Results
From 1660 patients, 583 complications in 424 patients (25.5%) were identified. The rate of CDC grade IIIa or greater was 9.7%, and the overall CCI was 5.8 ± 11.7. Age, gender, Charlson score, combined resection, open method, and total gastrectomy were associated with increased CCI (
p
< 0.05). The CCI demonstrated a stronger relationship with hospital stay (
ρ
= 0.721,
p
< 0.001) than did the CDC (
ρ
= 0.634,
p
< 0.001). For prolonged hospital stays (≥30 days), only the CCI showed a moderate correlation (
ρ
= 0.544,
p
= 0.024), although the CDC did not. The CUSUM-CCI model displayed dynamic time–event differences in individual and comparison monitoring models. In the institution monitoring model, a gradual decrease in the CCI was observed.
Conclusions
The CCI is more strongly correlated with postoperative hospital stay than is the conventional CDC. The CUSUM-CCI model can be used for the continuous monitoring of surgical quality.
Full text
Available for:
EMUNI, FIS, FZAB, GEOZS, GIS, IJS, IMTLJ, KILJ, KISLJ, MFDPS, NLZOH, NUK, OBVAL, OILJ, PNG, SAZU, SBCE, SBJE, SBMB, SBNM, UKNU, UL, UM, UPUK, VKSCE, ZAGLJ
It is unclear whether laparoscopic distal gastrectomy for locally advanced gastric cancer is oncologically equivalent to open distal gastrectomy. The noninferiority of laparoscopic subtotal ...gastrectomy with D2 lymphadenectomy for locally advanced gastric cancer compared with open surgery in terms of 3-year relapse-free survival rate was evaluated.
A phase III, open-label, randomized controlled trial was conducted for patients with histologically proven locally advanced gastric adenocarcinoma suitable for distal subtotal gastrectomy. The primary end point was the 3-year relapse-free survival rate; the upper limit of the hazard ratio (HR) for noninferiority was 1.43 between the laparoscopic and open distal gastrectomy groups.
From November 2011 to April 2015, 1,050 patients were randomly assigned to laparoscopy (n = 524) or open surgery (n = 526). After exclusions, 492 patients underwent laparoscopic surgery and 482 underwent open surgery and were included in the analysis. The laparoscopy group, compared with the open surgery group, suffered fewer early complications (15.7%
23.4%, respectively;
= .0027) and late complications (4.7%
9.5%, respectively;
= .0038), particularly intestinal obstruction (2.0%
4.4%, respectively;
= .0447). The 3-year relapse-free survival rate was 80.3% (95% CI, 76.0% to 85.0%) for the laparoscopy group and 81.3% (95% CI, 77.0% to 85.0%; log-rank
= .726) for the open group. Cox regression analysis after stratification by the surgeon revealed an HR of 1.035 (95% CI, 0.762 to 1.406; log-rank
= .827;
for noninferiority = .039). When stratified by pathologic stage, the HR was 1.020 (95% CI, 0.751 to 1.385; log-rank
= .900;
for noninferiority = .030).
Laparoscopic distal gastrectomy with D2 lymphadenectomy was comparable to open surgery in terms of relapse-free survival for patients with locally advanced gastric cancer. Laparoscopic distal gastrectomy with D2 lymphadenectomy could be a potential standard treatment option for locally advanced gastric cancer.
Background
Spasmolytic polypeptide-expressing metaplasia (SPEM) is considered a precursor lesion of intestinal metaplasia and intestinal-type gastric cancer (GC), but little is known about microRNA ...alterations during metaplasia and GC developments. Here, we investigate miR-30a expression in gastric lesions and identify its novel target gene which is associated with the intestinal-type GC.
Methods
We conducted in situ hybridization and qRT-PCR to determine miR-30a expression in gastric tissues. miR-30a functions were determined through induction or inhibition of miR-30a in GC cell lines. A gene microarray was utilized to confirm miR-30a target genes in GC, and siRNA-mediated target gene suppression and immunostaining were performed. The Cancer Genome Atlas data were utilized to validate gene expressions.
Results
We found down-regulation of miR-30a during chief cell transdifferentiation into SPEM. MiR-30a level was also reduced in the early stage of GC, and its level was maintained in advanced GC. We identified a novel target gene of miR-30a and
ITGA2,
and our results showed that either ectopic expression of miR-30a or
ITGA2
knockdown suppressed GC cell proliferation, migration, and tumorigenesis. Levels of
ITGA2
inversely correlated with levels of miR-30a in human intestinal-type GC.
Conclusion
We found down-regulation of miR-30a in preneoplastic lesions and its tumor-suppressive functions by targeting
ITGA2
in GC. The level of
ITGA2
, which functions as an oncogene, was up-regulated in human GC. The results of this study suggest that coordination of the miR-30a-
ITGA2
axis may serve as an important mechanism in the development of gastric precancerous lesions and intestinal-type GC.
Full text
Available for:
EMUNI, FIS, FZAB, GEOZS, GIS, IJS, IMTLJ, KILJ, KISLJ, MFDPS, NLZOH, NUK, OBVAL, OILJ, PNG, SAZU, SBCE, SBJE, SBMB, SBNM, UKNU, UL, UM, UPUK, VKSCE, ZAGLJ
Background
With improved short-term surgical outcomes, laparoscopic distal gastrectomy has rapidly gained popularity. However, the safety and feasibility of laparoscopic total gastrectomy (LTG) has ...not yet been proven due to the difficulty of the technique. This single-arm prospective multi-center study was conducted to evaluate the use of LTG for clinical stage I gastric cancer.
Methods
Between October 2012 and January 2014, 170 patients with pathologically proven, clinical stage I gastric adenocarcinoma located at the proximal stomach were enrolled. Twenty-two experienced surgeons from 19 institutions participated in this clinical trial. The primary end point was the incidence of postoperative morbidity and mortality at postoperative 30 days. The severity of postoperative complications was categorized according to Clavien–Dindo classification, and the incidence of postoperative morbidity and mortality was compared with that in a historical control.
Results
Of the enrolled patients, 160 met criteria for inclusion in the full analysis set. Postoperative morbidity and mortality rates reached 20.6% (33/160) and 0.6% (1/160), respectively. Fifteen patients (9.4%) had grade III or higher complications, and three reoperations (1.9%) were performed. The incidence of morbidity after LTG in this trial did not significantly differ from that reported in a previous study for open total gastrectomy (18%).
Conclusions
LTG performed by experienced surgeons showed acceptable postoperative morbidity and mortality for patients with clinical stage I gastric cancer.
Full text
Available for:
EMUNI, FIS, FZAB, GEOZS, GIS, IJS, IMTLJ, KILJ, KISLJ, MFDPS, NLZOH, NUK, OBVAL, OILJ, PNG, SAZU, SBCE, SBJE, SBMB, SBNM, UKNU, UL, UM, UPUK, VKSCE, ZAGLJ
According to recent international consensus conferences, pure laparoscopic donor hepatectomy (PLDH), particularly right and left hepatectomies, is not yet considered a standard practice because ...studies remain limited. Thus, we present the outcomes of more than 500 PLDH patients, mostly receiving a right hepatectomy. To our knowledge, this is the largest study to date on PLDH. Data from all living liver donors who underwent PLDH at 5 centers in Korea until June 2018 were retrospectively analyzed. The outcomes of both donors and recipients were included. Pearson correlation analysis was used to explore the relationship between the duration of surgery and cumulative experience at each center, which reflects the learning curve. Overall, 545 PLDH cases were analyzed, including 481 right hepatectomies, 25 left hepatectomies, and 39 left lateral sectionectomies (LLS). The open conversion was necessary for 10 (1.8%) donors, and none of the donors died or experienced irreversible disability. Notably, there were 25 (4.6%) patients with major complications (higher than Clavien‐Dindo grade 3). All centers except one showed a significant decrease in surgery duration as the number of cases accumulated. Regarding recipient outcomes, there were 110 cases (20.2%) of early major complications and 177 cases (32.5%) of late major complications. This study shows the early and late postoperative outcomes of 545 donors and corresponding recipients, including 481 right hepatectomies and 25 left hepatectomies, from 5 experienced centers. Although the results are comparable to those of previously reported open donor hepatectomy series, further studies are needed to consider PLDH a new standard practice.
Full text
Available for:
BFBNIB, FZAB, GIS, IJS, KILJ, NLZOH, NUK, OILJ, SAZU, SBCE, SBMB, UL, UM, UPUK
Background
The purpose of this study was to evaluate the nationwide benefit and cost of the national cancer screening program (NCSP) for gastric cancer treatment.
Methods
For this nationwide, ...population‐based study, the Korean National Health Insurance Big Data Base, which included gastric cancer–related treatment information and the costs for all patients with gastric cancer who were 40 years old or older between 2004 and 2013, was restructured. Patients with gastric cancer who participated in the NCSP at least once (the screening group) were compared with those who did not participate in the NCSP (the nonscreening group).
Results
The screening group (n = 116,775) spent significantly less on medical care expenses than the nonscreening group (n = 74,927) during the 5 years since the initial treatment (P < .0001). The screening group presented a significantly better prognosis for 5 and 9 years than the nonscreening group (P < .0001). The screening group revealed a 41% decreased hazard ratio (P < .0001) for death in comparison with the nonscreening group; the prognostic benefit became more obvious when treatment was started within the first 4 months after screening. The age‐standardized mortality rate ratio of the screening group versus the nonscreening group was 0.62 (P < .0001). The NCSP for gastric cancer required an average of 22,169,769 Korean Republic won (US $20,309) for 1 life‐year saved, which was less than the average gross domestic product (GDP) per capita in Korea.
Conclusions
The screening group had significantly lower medical care expenses and showed a significantly better prognosis than the nonscreening group. On the basis of the GDP per capita, the NCSP for gastric cancer was cost‐effective for treatment prognosis.
The nationwide, population‐based health insurance big database in Korea, where a national cancer screening program for gastric cancer was introduced in 2002, reveals that a screening group has lower medical care expenses and presents with a better prognosis than a nonscreening group even after stratification of the treatment group. On the basis of medical care expenses and screening costs, the national cancer screening program for gastric cancer is cost‐effective.
Full text
Available for:
BFBNIB, FZAB, GIS, IJS, KILJ, NLZOH, NUK, OILJ, SAZU, SBCE, SBMB, UL, UM, UPUK