Which factors are associated with adverse prognosis in multiple myeloma patients after surgery? - preliminary establishment and validation of the nomogram

Abstract

Background: To investigate the prognosis of patients with Multiple Myeloma (MM) after surgery, analyze the risk factors leading to adverse postoperative outcomes, and establish a nomogram.

Methods: Clinical data from 154 patients with MM who underwent surgery at our institution between 2007 and 2019 were retrospectively analyzed. Assessing and comparing patients' pain levels, quality of life, and functional status before and after surgery (P < 0.05) were considered statistically significant. The Kaplan-Meier survival curve was used to estimate the median survival time. Adverse postoperative outcomes were defined as worsened symptoms, lesion recurrence, complication grade ≥ 2, or a postoperative survival period < 1 year. Logistic regression analysis was used to determine the prognostic factors. Based on the logistic regression results, a nomogram predictive model was developed and calibrated.

Results: Postoperative pain was significantly alleviated in patients with MM, and there were significant improvements in the quality of life and functional status (P < 0.05). The median postoperative survival was 41 months. Forty-nine patients (31.8%) experienced adverse postoperative outcomes. Multivariate logistic regression analysis identified patient age, duration of MM, International Staging System, preoperative Karnofsky Performance Status, and Hb < 90 g/L as independent factors influencing patient prognosis. Based on these results, a nomogram was constructed, with a C-index of 0.812. The calibration curve demonstrated similarity between the predicted and actual survival curves. Decision curve analysis favored the predictive value of the model at high-risk thresholds from 10% to-69%.

Conclusion: This study developed a nomogram risk prediction model to assist in providing quantifiable assessment indicators for preoperative evaluation of surgical risk.

Keywords: Multiple myeloma; Nomogram; Prognosis; Surgery; Survival analysis.

Fig. 1

Spinal cord compression due to plasmacytoma. (A-B) Preoperative MRI shows a tumor compressing the spinal cord at the T2-4 levels, with epidural spinal cord compression graded as level 3. (C-D) Postoperative MRI reveals cerebrospinal fluid filling around the spinal cord without obvious compression, with pedicle screws well-placed and spinal sequences stable. (E-F) Intraoperatively, fish-like tumor tissue is visible, and laminectomy and tumor resection are performed to decompress the spinal cord

Fig. 2

Right humeral pathological fracture in the upper segment. X-rays taken 5 months (A), 4 months (B), and 1 week (C) preoperatively show local thinning of the bone cortex and discontinuity of part of the bone cortex. (D-F) MRI taken 3 months preoperatively shows a soft tissue mass, approximately 2.0 × 1.9 × 4.8 cm, with unclear boundaries. (G) Exposure of the fracture and reveals callus formation. (H) Radiofrequency ablation. (I) Bone drilling, opening of bone window, and removal of tumor from the marrow cavity. (J) Bone cement filling, fixation with plate and screws. (K) X-ray taken 3 days postoperatively

Fig. 3

(A) Kaplan-Meier survival curve showing 12-month survival was 82.5%, and Kaplan-Meier estimate of median overall survival for all patients was 41.00 months. (B) Receiver operating characteristic curve resulting from the multivariate logistic regression. AUC, area under the curve

Fig. 4

Nomogram for predicting individual prognosis based on data from 154 MM patients underwent surgery. Anemia is defined as Hb < 90 g/L

Fig. 5

(A) The calibration curve of the nomogram model. (B) Receiver operating characteristic curve to assess predictive accuracy of the nomogram model. AUC, area under the curve. (C) Decision curve analysis of the nomogram model