Predictive Factors of Survival Time of Breast Cancer in Kurdistan Province of Iran between 2006-2014: A Cox Regression Approach

Breast cancer is a major cause of morbidity and cancer related mortality among women. The prevalence of breast cancer was reported as increasing in most of the Asian countries (Aini abdullah et al., 2013; Najafi et al., 2013). Iran is also sharing the same experience. The etiology of breast cancer is largely unknown; therefore there is no established primary prevention strategy. So, the main strategy is the establishment of screening protocols and early detection programs, which at least theoretically can improve the survival rates. Despite the increasing incidence, the survival rates of breast cancer patients in many developed countries were substantially improved. According to national cancer registry project report, breast cancer is the most common cancer and also the most common cause of cancer-related deaths of female population in Iran (Mohagheghi et al., 2009). According to international agency for research on


Introduction
cancer, 4 million new cases of breast cancer has been diagnosed in 2004, and in contrast to the United States, the incidence rates is increasing in developing countries (IARC, 2006). Survival rate is the proportion of patients who survive for a specified period of time after diagnosis or treatment; it is a simple concept, but there are different ways to estimate it. For analysis of the survival of cancer patients during a specific period using population-based cancer registries data, relative survival rate as an estimate of net survival rate is a better measure than the observed survival and cause-specific survival (Esteve et al., 1990).
For cancer patients, survival rates have been accepted as the main criteria to measure the impact of treatment on cancer control (Ederer et al., 1961). Calculating mean lifetime in different groups of populations and assessment of effective factors in order to achieve healthy life is one of the challenges of scientists in various branches of biology and medicine. The population life time is a random statistical variable, therefore predicting it is not possible except through the use of statistical methods (Klein and Moeschberger, 1997;Hosmer and Lemeshow, 1999). Prognostic factors of breast cancer have already identified by using non-parametric survival methods such as Kaplan-Meier and Cox proportional hazard (PH) in many studies (Moran et al., 2008Akhsan andAryandono, 2010;Khodabakhshi et al., 2011), the latter is used when the effect of covariates on the hazard ratio is desired. Review of literature shows the extensive use of the Cox PH regression model for hazard rate or instantaneous risk of a given event (Orbe et al., 2002;Moran et al., 2008). However, the basis and the most important assumption underlying this model is the proportionality of hazard rates, which may not be held in some situations. Where PH assumption is not met, it is improper to use standard Cox PH model as it may entail serious bias and loss of power when estimating or making inference about the effect of a given prognostic factor on mortality (Moran et al., 2008). A review of survival analysis in cancer journals reveals that only 5% of all studies using Cox PH model considered the underlying assumption (Orbe et al., 2002).
The survival of breast cancer patients depends on factors such as genetic, age at diagnosis, access to care, stage of cancer, weight, phisical activity status, alcohol consumption, social, economic, enviromental factors and ethnicity (Rafiqullah khan et al., 2014). Identifying prognostic factors in patients with breast cancer plays an important role in treatment and care of patients. Although several studies have been conducted to determine those factors affecting survival and disease-free survival rates in patients with breast cancer, patients of the country participated in the studies had different features from patients of other countries; an issue that has paid little attention. Determinants in each study differed from many other studies, and the effect of such factors is discussed (Khodabakhshi et al., 2011).
Breast cancer patients' survival rate and affecting factors have not been evaluated in Kurdistan province; hazard function distribution is also not clear in Iran. The aim of this study was determination of effective factors on breast cancer survival rate in Kurdistan province, Iran during 2006-2014.

Materials and Methods
Through a retrospective study, data were sourced mainly from pathology report of the patients and hospital database record. Three hundred and thirteen patients with breast cancer who had been hospitalized at the Towhid Hospital, Sanandaj city, Kurdistan province, Iran were recruited. Inclusion criterion was definite diagnosis of breast cancer during a seven-year period from 2006 onward. Clinical data such as the stage of disease were obtained through a structured questionnaire and the patients' clinical record. Vital status and date of death were determined through phone call and also by official death certificates, with maximum follow-up of 96 months. Survival time (in month) was calculated from the date of diagnosis through the date of death or last follow-up (the end of March 2014). Patients who were alive at the end of the follow-up period were censored. Overall, 84patients were excluded from analyses according to exclusion criteria (5 cases were outlier, 79 patients due to indeterminate date of death or current status, most of them were Iraq subsidiary). Overall 229 patients were enrolled. If the date of death for survival analysis was unknown, the patient excluded from survival analysis, but remained for other assessments (descriptive variables). Clinical and pathologic variables were age of patient, clinical stage of disease( determined according to AJCC TNM staging system for breast cancer) (AJCC, 2002), date of diagnosis, type of surgery (MRM or lumpectomy), tumor laterality (right or left), Estrogen receptor status, Progesterone receptor status, P53 amplification, HER-2 positivity, metastasis of tumor, and the site of metastasis. Relapses were entered into parametric regression models for multivariate analysis in order to assess the relationship between the characteristics and prognostic factors for survival.

Statistical analysis:
Statistical analyses were performed using SPSS software for windows, version 16(SPSS Inc. Chicago, IL, USA). Kaplan-Meier curves were constructed for survival analysis and the Log rank test was used to determine the differences in survival. The survival of a patient is referred to as the number of months from the date of diagnosis to the date of death, or to the date of the end of the study for patients who were still alive or date of loss of follow-up. Cox multivariate analysis was also performed to analyze prognostic factors.
To evaluate the proportional hazard assumption three methods were used including: The Graphical (diagram Log (S) t vs time), and analytical (Time-Varying covariate method) (Sayehmiri et al., 2010). p<0.05 was considered to indicate a statistically significant difference.

Results
In this retrospective study, a total of 313 patients were enrolled; according to exclusion criteria for survival analysis, data of 229 patients were evaluated. Among all (n=229) patients 181 (79%) were alive and the rest were dead. Median follow-up time was 39 months, ranging from 6-96 months.
The mean age of patients at the time of diagnosis was 46.10±10.81 years (ranging from 22-76). About 207 (67.6%) were under 50 years. Based on Kaplan-Meier method, median survival was 81 months, and 5 year survival rate was 75%±0.43.The difference between these two age groups was not statistically significant (p=0.1).
One hundred and twenty four patients had stage III (48.2%) and 7 patients (2.7%) had stage IV disease. Clinical Stage of disease was not determined to be prognostic factor for survival (p=0.119), but after recoding, the patients that were at clinical stage III or IV of the disease had lower survival rates than those at stage I or II; this difference was statistically significant (p=0.03).
The median±(SD) survival rate for patients under 50 years at the time of diagnosis was 81±5.54 months and annual survival rate in this group was higher than patients over 50 years. Cox regression model was estimated as: h (t)=h0 (t) e 0.021 age shows that one year increasing in age rises 0/02 hazard of death. Patients who had brain metastasis had the lowest median survival rate (33±7.11) ( Table 3). The lowest 5-year survival rate was for patients who had stage IV of breast cancer (0%). This results emphasizes that early diagnosis of breast cancer (e.g. at lower stages) may be accompanied with better prognosis. The highest 5-year survival rate was for patients with non-metastatic disease (89%).
Univariate analysis showed that age of patient at diagnosis, type of surgery, Estrogen and Progesterone receptor status, HER-2 positivity, P53 status and the laterality of the tumor had no significant effect on survival rate of the patients. In contrast, having metastasis (p=0.0001), cancer relapse (p=0.0001), site of metastasis (p=0.0001), and clinical stage of disease (p=0.03) were significantly correlated to the survival rate (Table 2).
Prognostic factors for breast cancer survival were analyzed by Cox multivariate analysis. The analyzed factors were metastasis, relapse of cancer, clinical stage, and site of tumor metastasis. The results are shown in Table  3. Proportional assumption was met for all factors which entered to Cox regression. Based on our results the risk of death for patients with metastasis is 9.06 times higher than for other patients (HR=9.06; 95%CI=4.43-18.50). Also the risk of death for patients who had recurrent disease was 3.20 times higher than patients without that (HR=3.20; 95%CI=1.55-6.61). Patients with brain metastasis had the least probability of surviving and the highest risk of death (HR=16.56; 95%CI=7.01-39.09). Clinical stage of cancer is one of the most important prognostic factors for survival. Our data analysis showed a similar result: the risk of death for patients with stages III and IV disease is 2.30times higher than that for patients with stages 0 through II. Based on Kaplan-Meier curve the 5-year survival rates for patients with and without metastasis were 37% and 89%, respectively (Figure 1).  Cox regression model using time-varying covariate approach was estimated as: (h (t) =h0 (t0)*e 1.44±0.23h (t) ) the time-varying covariate model shows that the hazard of death in patients with tumor metastasis has had an increasing trend during follow-up time (HR=4.24; 95%CI=0.08-224.4).

Discussion
In present study, our objective was to determine prognostic factors for breast cancer patients' survival. Unfortunately, known prognostic factors are unable to accurately define the prognosis of a given patient, although it has been studied frequently. This study has investigated many factors of breast cancer patients. The strength of the study was using Cox regression after evaluating proportionality of hazard by three techniques graphical, analytical and time-varying approach on variables and calculation of five survival rate seperately for sub-groups analysis and as the hospital radiotherapy machine is equipped type, this study included patients with a variety of different features because breast cancer patients were reffered to the hospital from different west parts of Iran.
Breast cancer is uncommon in very young women (<35 years old), only accounting for fewer than 4% in Western countries (Wei et al., 2013). Like as previous studies in Iran, median age of breast cancer patients at the time of diagnosis was lower than in developed countries (Korizumi et al., 2010;Khodabakhshi et al., 2011;Thakur et al., 2014).
In this study, Kaplan-Meier method showed that for some variables such as: ER, PR, HER-2, P53, surgery (MRM, lumpectomy), age of patient, and tumor laterality were not prognostic factors for survival. But we found statistically significant relationship between metastasis (p=0.0001), site of metastasis (0.0001), relapse of cancer (0.001) and clinical stage (p=0.03) with survival. The risk of death for patients with tumor metastasis was very higher than for patients without it. Prognosis of patients with earlier stages was favorable compared to patients with advanced stages. Lower survival of high stages of breast cancer patients highlights the importance of screening programs for general population. The effect of clinical stage on survival confirmed by significant correlation on survival analysis; a result that is similar to some other studies (Tsan garis et al., 1992;Moradi Marjaneh et al., 2008;Akhsan and Aryandono, 2010;Khodabakhshi et al., 2011) There are a few studies that compared the survival rates of breast cancer patients with metastasis in different sites. The risk of death in patients with brain metastasis was 16.56 times higher than non-metastatic patients; for the rest of metastatic patients (except bone secondaries) the risk was 9.50 times higher than non-metastatics. The majority (n=24) of metastases was in bones. Brain metastasis accompanies with poor prognosis; without treatment, the patients will survive only about one month (Saha et al., 2013). Brain metastases were increased in recent years due to increased survival of cancer patients and widespread use of imaging modalities (Nicholas et al., 1766). In adults, lung cancer, breast cancer and melanoma are the most common causes of brain metastases respectively. In a study conducted in Yazd, central Iran, breast was the primary site for the majority of brain metastases (Akhavan et al., 2014).
We observed that the most important cause of shorter survival for breast cancer patients is the presence of distant metastasis. In a study conducted by (Kaviani et al., 2013) lymph node metastasis was not responsible for poorer survival in obese breast cancer patients. For metastasis to occur cancer cells have to attain a specific genotype and epigenotype, which allows them to disseminate from the primary tumor mass, survive, and then proliferate at secondary sites. Breast cancer cells have a distinct epigenomic DNA methylation profile which affects their metastatic potential (Thakur et al., 2014). Recent reports have shown that C-X-C chemokine receptor 4, TNF-α and IL-6 play an important role in metastasis (Tripsianis et al., 2013;Jiyang et al., 2014).
Comparing to western countries, the overall survival rates are still low in developing societies. Five-year survival rates of more than 80% have been reported in European countries (Berrino et al., 2007).
Our study is consistent with some previous studies   .15.19.8483 Predictive Factors of Survival of Breast Cancer in Kurdistan Province of Iran between 2006 which showed that expression of ER (p=0.3) was not a predictor of survival (Tsan garis et al., 1992;Wolberg et al., 1999;Horita et al., 2001), although there are some contradicting researches (Akbari et al., 2007;Korizumi et al., 2010). Our study were in accordance with many studies ( Dawood et al., 2008;Moradi Marjaneh et al., 2008) that revealed PR and HER-2 do not have effect on survival, but in some studies HER-2 was a prognostic factors (Khodabakhshi et al., 2011). Protein P53 was not prognostic factor for survival. This outcome agreed with a study in Mashhad (Moradi Marjaneh et al., 2008). Effects of breast cancer biomarkers on survival in different studies have had different results, probably because of ethnic factors and different features of breast cancer patients. ER, PR, HER-2 and P53are routinely checked in breast cancer specimens, are reliable, inexpensive and useful for treatment decision making; the results of these tests are recorded in cancer registries, allows population-based researches and makes them reasonable substitute for the more expensive molecular subtyping. It is suggested to conduct other studies with larger sample sizes to determine definite effects of biomarkers on survival.
Usually a sentinel lymph node biopsy is performed before surgery and if cancer cells are foundthen surgery is conducted. Surgery including radiotherapy is usally the standard choice of treatment (Rafiqhullah Khan et al., 2014). Like some other studies (Tewsend et al., 2004;Branicadi et al., 2005), regarding the operation technique, our study showed that there is no statistically significant difference between patients with MRM and breast-conserving surgeries. So, it is reasonable to offer lumpectomy rather than mastectomy for some patients for whom the less aggressive operation is technically suitable.
Non-metastatic breast cancer patients have significantly higher survival rates than those with systemic metastases, so, mass screening and early detection programs may help in improving treatment outcome by diagnosing the disease in less advanced stages.
In conclusion, cancer-related mortality of breast cancer patients is almost always because of metastatic spread, thus in order to achieve higher survival rates and better quality of life, it is very important to understand mechanisms of gaining metastatic potential, in addition to prevention and treatment of metastasis. Based on data of health care cancer data base of Iran, the age of appearing of the disease and median age of breast cancer patients are almost one decade earlier than other countries. This can make earlier starting of screening programs in this population reasonable. Early diagnosis of breast cancer and early treatment can increase survival and likely prevent metastasis and relapse.