Characteristics and prognostic factors of adult patients with osteosarcoma from the SEER database

Osteosarcoma is the most common bone malignancy. There are many studies on the prognostic factors of children and adolescents, but the characteristics and prognostic factors of adult osteosarcoma are rarely studied. The aim of this study was to construct a nomogram for predicting the prognosis of adult osteosarcoma. Information on all osteosarcoma patients aged ≥ 18 years from 2004 to 2015 was downloaded from the surveillance, epidemiology and end results database. A total of 70% of the patients were included in the training set and 30% of the patients were included in the validation set. Univariate log-rank analysis and multivariate cox regression analysis were used to screen independent risk factors affecting the prognosis of adult osteosarcoma. These risk factors were used to construct a nomogram to predict 3-year and 5-year prognosis in adult osteosarcoma. Multivariate cox regression analysis yielded 6 clinicopathological features (age, primary site, tumor size, grade, American Joint Committee on Cancer stage, and surgery) for the prognosis of adult osteosarcoma patients in the training cohort. A nomogram was constructed based on these predictors to assess the prognosis of adult patients with osteosarcoma. Concordance index, receiver operating characteristic and calibration curves analyses also showed satisfactory performance of the nomogram in predicting prognosis. The constructed nomogram is a helpful tool for exactly predicting the prognosis of adult patients with osteosarcoma, which could enable patients to be more accurately managed in clinical practice.


Introduction
Osteosarcoma is the most common primary bone malignancy, especially in children and adolescents, [1][2][3] and there have been many studies of osteosarcoma in children and adolescents. [4,5]owever few reports describe the characteristics and prognostic factors of adult patients with osteosarcoma.Duffaud F [6] studied the efficacy and safety of regorafenib in adult patients with metastatic osteosarcoma.Stefano Testa [7] studied prognosis and prognostic factors in adult and pediatric patients with osteosarcoma.10][11] Therefore, identifying adult patients with osteosarcoma at high risk of mortality can ensure appropriate treatment and can have a major impact on prognosis.As a statistical prediction model, the nomogram represents a graphical pattern in which variables are given scores so that it is easy to obtain event probabilities for individual patients compared to traditional assessment criteria.14][15][16] Due to the different clinical and prognostic characteristics between the different age groups, no studies were available to develop a prognostic nomogram for osteosarcoma in adult patients.Therefore, in this study, we aimed to construct and validate a survival nomogram incorporating available clinical characteristics to improve the prognosis of adult osteosarcoma patients in clinical practice.

Study population
The information was downloaded from the surveillance, epidemiology, and end results (SEER) database for all adult patients (age ≥ 18 years) with osteosarcoma from 2004 to 2015 (using the International The authors have no funding and conflicts of interest to disclose.
The datasets generated during and/or analyzed during the current study are publicly available.
Since the datasets generated and/or analyzed in this study are publicly available, the study protocol does not need to be submitted to an ethics review committee for consideration and approval.Histological Classification of Tumor Disorders, Third Edition (IDO-O-3) code 9180 to 9187 and 9192 to 9195 for identification).

Data collection
The following variables were identified from the dataset: age at diagnosis (18-39, 40-59, or ≥ 60 years), sex (male or female), race (white, black, or others), marital status (married or other), primary site (Limb or other), tumor size (<50 mm, 50-99 mm or ≥ 100 mm), tumor number (1 primary only or other), grade (I, II, III or IV), American Joint Committee on Cancer (AJCC) stage (I, II, III or IV), tumor stage (local/regional or distant), primary site surgery (yes or no), radiation therapy (yes or no), chemotherapy (yes or no).We further excluded patients with survival of <1 month and unknown variables such as race, marital status, primary site, grade, AJCC stage, and tumor size.We used the caret package in R software (version 4.2.1) to randomize 983 patients in a 7:3 ratio into a training set (70%) and a validation set (30%).

Establishment of cox regression models
Univariate log-rank analysis and multivariate cox regression analysis were performed using SPSS version 26 (IBM, Armonk, NY) to identify independent prognostic factors associated with prognosis in adult osteosarcoma patients.The nomogram was drawn and Concordance index (C-index), receiver operating characteristic (ROC), and calibration curves were calculated using the rms, foreign and survival packages in R version 4.2.1 (R development core team, Vienna, Austria) to evaluate the predictive performance of the nomogram.

Statistical analysis
Counts and percentages were used to describe categorical measures.Chi-square tests were used to compare categorical measures.P < .05, the difference was statistically significant.

Demographic baseline characteristics
According to the determined inclusion and exclusion criteria, 983 adult patients with osteosarcoma were included from SEER database.We randomly assigned patients to tow sets in R software, 70% of the patients to the training set (n = 691) and 30% to the validation set (n = 292) (Fig. 1).The majority of patients were aged between 18 to 39 (53.3%, 54.1%), White (75.4%, 77.4%), undergone the therapies surgery (87.4%, 90.8%) and Chemotherapy (72.4%, 71.2%), in the training and validation sets, respectively.Table 1 showed the clinicopathologic characteristics of all patients in the training and validation sets.

Nomogram construction
The nomogram is a visual regression model.The scoring criteria are set according to the regression coefficients of each influencing factor and the scores of the respective variables are added to obtain the total score of each patient and then calculate the prognosis and survival rate.We included all independent risk factors affecting the prognosis of patients with osteosarcoma, including age, primary site, tumor size, grade, AJCC stage, and surgery, and obtained a rank line that could predict the 3-year and 5-year overall survival (OS) of patients with osteosarcoma picture.The nomogram is mainly composed of variable names and tick marks.The length of the tick marks can reflect the contribution of the influencing factor to the outcome event.The total score can be obtained by adding up the individual scores corresponding to each variable under different values.According to the downward projection of the total score, the survival rate of the corresponding year of the patient can be obtained, as shown in Figure 3.

Validation of the nomogram
The C-index of the training set is 0.774, and the C-index of the validation set is 0.751, indicating that the model has good accuracy.The areas under the 3-year and 5-year ROC curves of the training set were 0.822 and 0.822, respectively.The areas under the 3-year and 5-year ROC curves of the validation set were 0.822 and 0.822, respectively, as shown in Figure 4.The verification of the calibration curve shows that the model predicts the 3-year and 5-year OS and the actual OS has a good consistency, as shown in Figure 5. Therefore, the established nomogram model is verified by the C-index, ROC curve, and calibration curve to have a good prediction effect.

Discussion
Prognostic evaluation is of great value for the treatment, monitoring and follow-up of patients with various tumors.19][20] For patients with osteosarcoma, several predictive models have been constructed in previous studies; [11,13,14] however, less attention has been paid to adult osteosarcoma.A study by Wenhao Chen [21] concluded that age, tumor site, historical grade, surgery, AJCC T/N, and M were independent prognostic factors for survival in patients with osteosarcoma.However, based on univariate and multivariate cox regression analysis, we found 6 informative variables (age, primary site, tumor size, grade, AJCC stage, and surgery) as independent prognostic factors for osteosarcoma in adults.Age is a well-known prognostic factor for many tumors factor. [22]This study came to the same conclusion that elderly patients have a poorer prognosis.[25][26][27][28] This study shows that tumor location affects survival in patients with osteosarcoma.][31] Several previous studies have reported poorer prognosis and reduced survival in patients with larger tumors.This study suggests that tumor size is an important factor affecting the healing of patients.][37] Consistent with these findings, we found that patients with osteosarcoma with distant metastases had a higher risk of death.[40][41][42] Nishida et al [4] showed that the 5-year overall survival rate was 53.2%Although the predictive nomograms in this study showed good predictive power, there are still some limitations that need to be considered.First, the current study only includes clinical data from patients diagnosed with osteosarcoma from 2004 to 2015 in the SEER database, not from all osteosarcoma patients.Second, because our study was retrospective, it was inevitable that some patient data were missing.Third, the current study used internal validation and lacked external validation.Therefore, it is necessary to collect further external data to validate the accuracy and reliability of the prediction model.As a next step, we will conduct further prospective studies at our institution to further validate the accuracy of this prediction model from external sources for clinical application.
In summary, age, primary site, tumor size, grade, AJCC stage and surgery were identified as independent prognostic variables for adult osteosarcoma patients.Our nomogram model provides an applicable tool with good discrimination

Figure 4 .
Figure 4. performance of the survival nomogram reflected by ROC curves.ROC curves for the 3-year and 5-year osteosarcoma in patients in the training cohort (A-B) and in the SEER validation cohort (C-D).ROC = receiver operating characteristic, SEER = surveillance, epidemiology, and end results.

Figure 5 .
Figure 5.The calibration curves for predicting osteosarcoma in the training and validation sets.Calibration plots of 3-year and 5-year osteosarcoma in the training cohort (A-B) and in the SEER validation cohort (C-D).SEER = surveillance, epidemiology, and end results.

Table 1
Demographics, tumor characteristics, and treatment characteristics of patients with osteosarcoma.

Table 2
Univariate and multivariate analyses of osteosarcoma.
AJCC = American Joint Committee on Cancer, HR = hazard ratio.