Value of a nomogram model in early recurrence of pancreatic ductal adenocarcinoma after laparoscopic pancreaticoduodenectomy / 临床肝胆病杂志

ABSTRACT

ObjectiveTo investigate the risk factors for early tumor recurrence after laparoscopic pancreaticoduodenectomy （LPD） in patients with pancreatic ductal adenocarcinoma （PDAC）， and to establish a predictive model. MethodsA retrospective analysis was performed for the clinical data of 240 PDAC patients who underwent LPD in The First Hospital of Jilin University from April 2016 to July 2022， with early postoperative tumor recurrence （time to recurrence ≤12 months） as the study outcome. The patients were randomly divided into training group with 168 patients and validation group with 72 patients at a ratio of 7∶3. In the training group， there were 70 patients （41.67%） with early postoperative recurrence and 98 （58.33%） without early recurrence， and in the validation group， there were 32 （44.44%） with early postoperative recurrence and 40 （55.56%） without early recurrence. The chi-square test or the Fisher’s exact test was used for comparison of categorical data between groups； a logistic regression analysis was used to investigate the risk factors for early postoperative recurrence； the receiver operating characteristic （ROC） curve and the area under the ROC curve （AUC） were used to evaluate the discriminatory ability of the model， with AUC>0.75 indicating that the model had adequate discriminatory ability. The Bootstrap resampling method was used for validation after 1 000 times of random sampling， and the model was validated again in the validation group. The calibration curve and the Hosmer-Lemeshow goodness-of-fit test were used to evaluate the degree of calibration， and the decision curve analysis was used to evaluate clinical practicability. ResultsThe univariate and multivariate analyses showed that preoperative CA19-9 level≥37 U/mL （odds ratio ［OR］=6.265， 95% confidence interval ［CI］： 1.938‍ ‍—‍ ‍20.249， P<0.05）， maximum tumor diameter >3 cm （OR=10.878， 95%CI： 4.090‍ ‍—‍ ‍28.932， P<0.05）， poor tumor differentiation （OR=3.679， 95%CI： 1.435‍ ‍—‍ ‍9.433， P<0.05）， lymph node metastasis （OR=0.209， 95%CI： 0.080‍ ‍—‍ ‍0.551， P<0.05）， and absence of adjuvant chemotherapy after surgery （OR=0.167， 95%CI： 0.058‍ ‍—‍ ‍0.480， P<0.05）. A nomogram model was constructed based on these factors； the ROC curve analysis showed that the model had an AUC of 0.895 （95%CI： 0.846‍ ‍—‍ ‍0.943， P<0.001）， and the calibration curve and the Hosmer-Lemeshow test showed that the model had a good degree of calibration （P=0.173）. The decision curve analysis showed that the nomogram had a good clinical application value. ConclusionPreoperative CA19-9 level ≥37 U/mL， maximum tumor diameter >3 cm， poor tumor differentiation， lymph node metastasis， and absence of adjuvant chemotherapy after surgery are independent risk factors for the early recurrence of PDAC after LPD， and the nomogram model established based on these factors can effectively predict early postoperative recurrence.