• Korean Journal of Artificial Intelligence
• /
• v.6 no.2
• /
• pp.11-15
• /
• 2018

Breast cancer is one of the leading causes of cancer related death among women. So prediction of overall survival status is important into decided in adjuvant treatment. Deep belief network is a kind of artificial intelligence (AI). We intended to construct prediction model by deep belief network using associated clinicopathologic factors. 103881 cases were found in the Korean Breast Cancer Registry. After preprocessing of data, a total of 15733 cases were enrolled in this study. The median follow-up period was 82.4 months. In univariate analysis for overall survival (OS), the patients with advanced AJCC stage showed relatively high HR (HR=1.216 95% CI: 0.011-289.331, p=0.001). Based on results of univariate and multivariate analysis, input variables for learning model included 17 variables associated with overall survival rate. output was presented in one of two states: event or cencored. Individual sensitivity of training set and test set for predicting overall survival status were 89.6% and 91.2% respectively. And specificity of that were 49.4% and 48.9% respectively. So the accuracy of our study for predicting overall survival status was 82.78%. Prediction model based on Deep belief network appears to be effective in predicting overall survival status and, in particular, is expected to be applicable to decide on adjuvant treatment after surgical treatment.

• Radiation Oncology Journal
• /
• v.36 no.4
• /
• pp.317-324
• /
• 2018

Purpose: This study aimed to evaluate the long-term survival outcomes and prognostic factors that affect the clinical outcomes of patients who underwent surgery and postoperative radiotherapy for major salivary gland mucoepidermoid carcinoma (MEC). Materials and Methods: We retrospectively reviewed the clinical data of 44 patients who underwent surgery followed by radiotherapy for primary MEC of the major salivary glands between 1991 and 2014. The median follow-up period was 9.8 years (range, 0.8 to 23.8 years). Results: The overall outcomes at 5 and 10 years were 81.5% and 78.0% for overall survival (OS), 86.2% and 83.4% for disease-free survival, 90.6% and 87.6% for locoregional recurrence-free survival, and both 90.5% for distant metastasis-free survival (DMFS). Histologic grade was the only independent predictor of OS (low vs. intermediate vs. high; hazard ratio = 3.699; p = 0.041) in multivariate analysis. A poorer survival was observed among patients with high-grade tumors compared with those with non-high-grade tumors (5-year OS, 37.5% vs. 91.7%, p < 0.001; 5-year DMFS, 46.9% vs. 100%, p < 0.001). Conclusion: Surgery and postoperative radiotherapy resulted in excellent survival outcomes for patients with major salivary gland MEC. However, high-grade tumors contributed to poor DMFS and OS. Additional aggressive strategies for improving survival outcomes should be developed for high-grade MEC.

• Asian Pacific Journal of Cancer Prevention
• /
• v.15 no.5
• /
• pp.2195-2199
• /
• 2014

Background: Despite being rare in incidence, malignant tumors of major salivary glands show diverse histological variation. There are limited data on major salivary gland tumor management and outcome from Pakistan. The objective of this study was to share our experience with management of malignant tumors of major salivary glands. Materials and Methods: Patients who received treatment at Shaukat Khanum Cancer Hospital and Research Center from July 2002 to June 2011 with an underlying diagnosis of a major salivary gland malignancy were included. Patient characteristics and treatment modalities were assessed. Local, regional and distant failures were determined. Disease free survival (DFS) and overall survival (OS) were calculated using Kaplan Meier curves and the Log rank test was used to determine statistical significance. Univariate and multivariate analyses were performed using Cox proportional hazard regression. Results: The parotid gland was the primary site of origin in 104 (80%) patients. Mucoepidermoid carcinoma (43%) and adenoid cystic carcinoma (24%) were the most common histological types. Surgery followed by adjuvant radiation remained the mainstay treatment modality with 81 (62%) patients. Nineteen (15%) patients were treated with surgery alone and 30 (23%) patients with locally advanced surgically inoperable tumors received radiation only. Forty one (32%) patients failed the treatment (local 12, regional 11, locoregional 5, distant 13). The expected 5 year DFS and OS were 65% and 74% respectively. On multivariate analysis, grade was the only independent predictor of DFS and nodal involvement was the only independent predictor of overall survival. Conclusions: Employing existing standards of treatment, comparable survival can be achieved in Pakistani population with major salivary gland malignancies as elsehwere in the world.

• Radiation Oncology Journal
• /
• v.31 no.1
• /
• pp.25-33
• /
• 2013

Purpose: We reviewed the treatment outcomes and prognostic factors for patients with anal canal carcinoma who were treated with curative intent chemoradiotherapy (CRT) at Severance Hospital from 2005 to 2011. Materials and Methods: Data for 38 eligible patients treated during this period were reviewed. All patients were treated with curative intent using radiotherapy (RT) with (n = 35) or without concomitant chemotherapy (n = 3). Among 35 patients who received CRT, most of the chemotherapeutic regimens were either 5-fluorouracil (5-FU) plus mitomycin C (23 patients) or 5-FU plus cisplatin (10 patients). Recurrence-free survival (RFS), colostomy-free survival (CFS), overall survival (OS), and locoregional control (LRC) rates were calculated using the Kaplan-Meier method and survival between subgroups were compared using the log-rank test. Cox's proportional hazard model was used for multivariate analysis. Results: Over a median follow-up period of 44 months (range, 11 to 96 months), 3-year RFS, CFS, OS, and LRC were 80%, 79%, 85%, and 92%, respectively. In multivariate analysis, tumor size >4 cm was an independent predicting factor for poorer RFS (hazard ratio [HR], 6.35; 95% confidence interval [CI], 1.42 to 28.5; p = 0.006) and CFS (HR, 6.25; 95% CI, 1.39-28.0; p = 0.017), while the presence of external iliac lymph node metastasis was an independent prognosticator for poorer OS (HR, 9.32; 95% CI, 1.24 to 70.3; p = 0.030). No treatment-related colostomies or deaths occurred during or after treatment. Conclusion: Curative intent CRT resulted in excellent outcomes that were comparable to outcomes in previous randomized trials. No severe treatment-related toxicities were observed.

• Journal of Gastric Cancer
• /
• v.4 no.2
• /
• pp.89-94
• /
• 2004

Purpose: There is no established treatment-related prognostic factor for gastric cancer except a curative tumor resection. This study was done to clarify the prognostic value of early postoperative intraperitoneal chemotherapy (EPIC) in patients with serosa-positive gastric cancer. Materials and Methods: We analyzed retrospectively the postoperative survival data of 209 patients with serosapositive gastric cancer treated by surgery and chemotherapy. The survival period for patients was calculated from the date of resection until cancer-related death or the last date of follow-up; Kaplan-Meier survival curves were plotted and compared by using the log-rank test. A multivariate analysis was done by using the Cox proportional hazards model. Results: Statistically significant differences in survival rates were noted based on gender, depth of invasion, lymph node metastasis, distant metastasis, stage, location of tumor, macroscopic type, extent of gastric resection, curability of surgery, and adjuvant chemotherapy. Five-year survival rates of patients who received EPIC and systemic chemotherapy were 49 per cent and 25 per cent, respectively (P=0.009). A multivariate analysis revealed that invasion of an adjacent organ, lymph node metastasis, total gastrectomy, and palliative surgery were poor independent prognostic factors. Also, EPIC had a marginal prognostic value (P=0.056). Conclusion: Perioperative intraperitoneal chemotherapy can possibly be one of the independent prognostic indicators in case of serosa-positive gastric cancer. (J Korean Gastric Cancer Assoc 2004;4:89-94)

• Asian Pacific Journal of Cancer Prevention
• /
• v.17 no.7
• /
• pp.3515-3519
• /
• 2016

Background: In Thailand, a national treatment protocol for childhood leukemia and lymphoma (LL) was implemented in 2006. Access to treatment has also improved with the National Health Security system. Since these innovations, survival of childhood LL has not been fully described. Materials and Methods: Trends and survival of children under 15 with childhood cancers diagnosed between 1993 and 2012 were investigated using the hospital-based data from the Khon Kaen Cancer Registry, Srinagarind Hospital, Faculty of Medicine, Khon Kaen University, Thailand. Childhood cancers were classified into 12 diagnostic groups, according to the ICCC based on the histology of the cancer. Survival rates were described by period, depending on the treatment protocol. For leukemias and lymphomas, survival was assessed for 3 periods (1993-99, 2000-5, 2006-12) while for solid tumors it was for 2 periods (before and after 2000). The impacts of sex, age, use of the national protocol, and catchment area on leukemia and lymphoma were evaluated. Overall survival was calculated using the Kaplan-Meier method while the Cox proportional hazard model was used for multivariate analysis. Trends were calculated using the R program. Results: A total of 2,343 childhood cancer cases were included. Survival for acute lymphoblastic leukemia (ALL) from 1993-9, 2000-5, and 2006-12 improved significantly (43.7%, 64.6%, and 69.9%). This was to a lesser extent true for acute non-lymphoblastic leukemia (ANLL) (28.1%, 42.0%, and 42.2%). Survival of non-Hodgkin lymphoma (NHL) also improved significantly (44%, 65.5%, and 86.8%) but not for Hodgkin disease (HD) (30.1%, 66.1%, and 70.6%). According to multivariate analysis, significant risk factors associated with poor survival in the ALL group were age under 1 and over 10 years, while not using the national protocol had hazard ratios (HR) of 1.6, 1.3, and 2.3 respectively. In NHL, only non-use of national protocols was a risk factor (HR 3.9). In ANLL and HD, none of the factors influenced survival. Survival of solid tumors (liver tumors, retinoblastomas) were significantly increased compared to after and before 2000 while survival for CNS tumors, neuroblastoma and bone tumors was not changed. Conclusions: The survival of childhood cancer in Thailand has markedly improved. Since implementation of national protocols, this is particularly the case for ALL and NHL. These results may be generalizable for the whole country.

• Asian Pacific Journal of Cancer Prevention
• /
• v.17 no.7
• /
• pp.3575-3581
• /
• 2016

Background: Cancer survival analysis is an essential indicator for effective early detection and improvements in cancer treatment. This study was undertaken to document colorectal cancer survival and associated prognostic factors in Malaysians. Materials and Methods: All data were retrieved from the National Cancer Patient Registry-Colorectal Cancer. Only cases with confirmed diagnosis through histology between the year 2008 and 2009 were included. Retrieved data include socio-demographic information, pathological features and treatment received. Survival curves were plotted using the Kaplan-Meier method. Univariate analysis of all variables was then made using the Log-rank test. All significant factors that influenced survival of patients were further analysed in a multivariate analysis using Cox' regression. Results: Total of 1,214 patients were included in the study. The overall 3- and 5-year survival rates were 59.1% and 48.7%, respectively. Patients with localized tumours had better prognosis compared to those with advanced stage cancer. In univariate analysis, staging at diagnosis (p<0.001), primary tumour size (p<0.001), involvement of lymph nodes (p<0.001) and treatment modalities (p=0.001) were found to be predictors of survival. None of the socio-demographic characteristics were found to exert any influence. In Cox regression analysis, staging at diagnosis (p<0.001), primary tumour size (p<0.001), involvement of lymph nodes (p<0.001) and treatment modalities (p<0.001) were determined as independent prognostic factors of survival after adjusted for age, gender and ethnicity. Conclusions: The overall survival rate for colorectal cancer patients in Malaysia is similar to those in other Asian countries, with staging at diagnosis, primary tumor size, involvement of lymph node and treatment modalities having significant effects. More efforts are needed to improve national survival rates in future.

• Asian Pacific Journal of Cancer Prevention
• /
• v.16 no.6
• /
• pp.2237-2243
• /
• 2015

Background: Adjuvant chemotherapy improves survival in Dukes C colon cancers post-curative resection. However, the evidence for a role with Dukes B lesions remains unproven despite frequent use for disease characterized by poor prognostic features. In view of limited Asia-specific data, this study aimed to determine survival outcomes and identify prognostic factors in a tertiary teaching hospital in Malaysia. Materials and Methods: A total of 116 subjects who underwent curative surgery with and without adjuvant chemotherapy for Duke B and C primary colon adenocarcinomas diagnosed from 2004-2009 were recruited and data were collected retrospectively. Five-year overall survival (OS) and disease free survival (DFS) were analysed using Kaplan-Meier survival analysis and log-rank (Mantel-Cox) test. Prognostic factors were determined using Cox proportional hazards regression with both univariate and multivariate analyses. Results: The survival analysis demonstrated a 5-year OS of 74.0% for all patients, with 74.9% for Dukes C subjects receiving chemotherapy compared to 28.6% in those not receiving chemotherapy (p=0.001). For Dukes B disease, the 5-year survival rate was 82.6% compared to 75.0% for subjects receiving and not receiving chemotherapy, respectively (p=0.17). Independent prognostic factors identified included a CEA level more than 3.5 ng/ml (hazard ratio (HR)=4.78; p=0.008), serosal involvement (HR=3.75; p=0.028) and completion of chemotherapy (HR= 0.20; p=0.007). Conclusions: In a regional context, this study supports current evidence from the West that adjuvant chemotherapy improves survival in Dukes C colon cancers post curative surgery. However, although a clear benefit has yet to be proven for Dukes B disease, our results suggest survival improvement in selected cases.

• Journal of the Korean Data and Information Science Society
• /
• v.17 no.3
• /
• pp.999-1007
• /
• 2006

The semiparametric gamma frailty models have been often used for multivariate survival analysis because they give an explicit marginal likelihood. The commonly used estimation procedure is the profile likelihood method based on marginal likelihood, which provides the same parameter estimates as the EM algorithm. In this paper we show in finite samples the standard profile-likelihood method can lead to an underestimation of parameters, particularly for the frailty parameter. To overcome this problem, we propose an adjusted profile-likelihood method. For the illustration a numerical example and a small-sample simulation study are presented.

• The Korean Journal of Applied Statistics
• /
• v.29 no.7
• /
• pp.1459-1473
• /
• 2016

Both longitudinal data and survival data are collected simultaneously in longitudinal data which are observed throughout the passage of time. In this case, the effect of the independent variable becomes biased (provided that sole use of longitudinal data analysis does not consider the relation between both data used) if the missing that occurred in the longitudinal data is non-ignorable because it is caused by a correlation with the survival data. A joint model of longitudinal data and survival data was studied as a solution for such problem in order to obtain an unbiased result by considering the survival model for the cause of missing. In this paper, a joint model of the longitudinal zero-inflated count data and survival data is studied by replacing the longitudinal part with zero-inflated count data. A hurdle model and proportional hazards model were used for each longitudinal zero inflated count data and survival data; in addition, both sub-models were linked based on the assumption that the random effect of sub-models follow the multivariate normal distribution. We used the EM algorithm for the maximum likelihood estimator of parameters and estimated standard errors of parameters were calculated using the profile likelihood method. In simulation, we observed a better performance of the joint model in bias and coverage probability compared to the separate model.