Prognostic value of albumin-to-alkaline phosphatase ratio and CONUT score in rectal cancer patients undergoing XELOX-based chemotherapy: development of a nomogram-based predictive model

Abstract

Objective: To evaluate the prognostic significance of the albumin-to-alkaline phosphatase ratio (AAPR) and the Controlling Nutritional Status (CONUT) score in rectal cancer (RC) patients receiving XELOX-based chemotherapy, and to develop a nomogram for predicting recurrence risk.

Methods: This retrospective study included 389 RC patients treated at the First Affiliated Hospital of Chongqing Medical University, along with an independent validation cohort of 120 patients. Clinical variables, including AAPR and CONUT were analyzed using Cox regression and cumulative incidence function curves. A nomogram was constructed and validated using calibration plots and time-dependent receiver operating characteristic (ROC) curves.

Results: Both AAPR (HR = 0.073, P<0.001) and CONUT score (HR = 1.497, P<0.001) were identified as independent predictors of recurrence. Additional factors significantly associated with increased recurrence risk included TNM stage III, tumor size ≥5 cm, vascular invasion, and carcinoembryonic antigen (CEA) level ≥5 ng/ml. The nomogram demonstrated strong predictive performance with a C-index of 0.860 in the training cohort, and 0.835 in the validation cohort. Calibration plots showed excellent agreement between predicted and observed recurrence probabilities.

Conclusions: AAPR and CONUT score are independent prognostic indicators for recurrence in RC patients treated with XELOX-based chemotherapy. The proposed nomogram, incorporating these variables, provides a reliable tool for individualized risk prediction and may support personalized treatment decision-making.

Keywords: Albumin-to-alkaline phosphatase ratio; CONUT score; XELOX-based chemotherapy; nomogram; prognostic model; rectal cancer.

Figure 1

Study flow chart.

Figure 2

Comparison of AAPR and CONUT scores between patients with PR and SD after chemotherapy. A. Distribution of AAPR in PR and SD patients. B. Distribution of CONUT scores in PR and SD patients. Note: AAPR, albumin-to-alkaline phosphatase ratio; CONUT, Controlling Nutritional Status; PR, partial response; SD, stable disease.

Figure 3

AAPR and CONUT scores in patients with or without adverse reactions. A. Comparison of AAPR and CONUT scores between patients with and without bone marrow suppression. AAPR was lower in patients with bone marrow suppression, while CONUT scores showed no difference. B. Comparison of AAPR and CONUT scores between patients with and without gastrointestinal reactions. No differences were observed in either AAPR or CONUT scores. C. Comparison of AAPR and CONUT scores between patients with and without hand-foot syndrome. Both AAPR and CONUT scores showed no notable differences. Note: AAPR, albumin-to-alkaline phosphatase ratio; CONUT, Controlling Nutritional Status.

Figure 4

CIF curves for prognostic factors in rectal cancer recurrence. A. The CIF curve for age shows higher 3-year recurrence in patients aged ≥60 years compared to those aged <60 years. B. The CIF curve for TNM stage shows higher 3-year recurrence in stage III patients compared to stage II patients. C. The CIF curve for tumor size indicates higher recurrence in patients with a tumor size ≥5 cm. D. The CIF curve for vascular invasion demonstrates higher recurrence in patients with vascular invasion. E. The CIF curve for perineural infiltration reveals higher recurrence in patients with perineural infiltration. F. The CIF curve for histologic subtyping suggests higher recurrence in poorly differentiated tumors. G. The CIF curve for CEA shows higher recurrence in patients with a CEA level ≥5 ng/ml. H. The CIF curve for AAPR indicates a lower recurrence risk in patients with higher AAPR values. I. The CIF curve for CONUT score determines higher recurrence in patients with higher CONUT scores. Note: CIF, Cumulative Incidence Function; TNM, tumor-node-metastasis; CEA, carcinoembryonic antigen; AAPR, albumin-to-alkaline phosphatase ratio; CONUT, Controlling Nutritional Status.

Figure 5

Nomogram for predicting survival probabilities in rectal cancer patients based on Cox Regression. Note: TNM, tumor-node-metastasis; AAPR, albumin-to-alkaline phosphatase ratio; CONUT, Controlling Nutritional Status.

Figure 6

CIF curves for rectal cancer recurrence. A. The overall CIF curve for rectal cancer recurrence, showing the model’s predictive performance in the entire dataset. B. The CIF curve for rectal cancer recurrence in the training group, highlighting the model’s performance in the training dataset. C. The CIF curve for rectal cancer recurrence in the validation group, assessing the model’s predictive accuracy in the validation dataset. Note: CIF, Cumulative Incidence Function.

Figure 7

Calibration curves for rectal cancer prognosis. A. The overall calibration curve, showing the agreement between predicted and observed survival probabilities for the entire dataset. B. The calibration curve for rectal cancer recurrence in the training group, displaying the predictive accuracy of the model in the training dataset. C. The calibration curve for rectal cancer recurrence in the validation group, assessing the model’s predictive accuracy in the validation dataset.

Figure 8

Time-dependent ROC curves for rectal cancer prognosis. A-C. ROC curves for the 1-year survival probability for the overall, training, and validation groups. D-F. ROC curves for the 2-year survival probability for the overall, training, and validation groups. G-I. ROC curves for the 3-year survival probability for the overall, training, and validation groups. Note: ROC, receiver operating characteristic.