The incidence and prognosis of nasopharyngeal carcinoma patients with family history

Purpose Epidemiological data showed that nasopharyngeal carcinoma (NPC) was a regional malignancy. It suggested that genetic factor may play an important role in tumorigenesis of NPC. The aim was to investigate the incidence and the prognosis of NPC patients with family history. Methods The clinical data of patients with NPC treated in Fudan University Shanghai Cancer Center from January 2008 to December 2012 were reviewed, and the patients with family history were selected. The prognosis of patients with family history was follow-up. The 5-year overall survival (OS), local recurrence-free survival (LRFS), and distant metastasis-free survival (DMFS) were analyzed by Kaplan-Meier and log-rank test. Cox proportional hazard model was used for multivariate analysis. Results There were 3.64% (135/3706) NPC patients with family history of NPC. Eighty-three percent (112/135) patients had only one family member suffering from NPC previously, and 74.1% (100/135) patients who had family history only in first-degree family members. Excluding five patients lost to follow-up, 130 patients were eventually used to analyze the prognosis. The 5-year OS, LRFS, and DMFS rates of all patients with family history were 84.1%, 83.4%, and 83.8%, respectively. There were no significant differences of OS, LRFS and DMFS between one relative group and at least two relatives group. In addition, the degree of NPC had no association with OS, LRFS and DMFS, respectively. Conclusion Our results showed that there was an incidence rate of 3.64% NPC patients with family history. These patients had a satisfied prognosis, and the prognosis of NPC patients with family history in different degree or numbers of relatives had no significant differences.


INTRODUCTION
Nasopharyngeal carcinoma (NPC) is one of the most common cancers in areas of Southeast Asia and North Africa, especially in Southern China. The incidence of Southeast Asia was 6.4 per 100,000 for males and 2.0 per 100,000 for females, which is the highest in the world. In 2012, an estimates total of 86,700 cases added to NPC and 50,800 to death [1,2].
Similar studies reported a significant association between neoplastic prognosis and positive family history, which found better prognosis in breast cancer, colon cancer and gastric cancer [15][16][17][18]. In addition, the high incidence of NPC among Southern Chinese and population of Southern Chinese descent suggest a component of genetic susceptibility, and genetic factor may play an important role in tumorigenesis of NPC. The epidemiological investigation also showed that there were many families with high incidence of NPC in endemic areas. However, there were few studies to focus on the incidence and the prognosis for the NPC patients with family history, especially in non-endemic areas. In present study, the aim was to investigate the incidence and the prognosis of NPC patients with family history.

The incidence of NPC patients with family history
There were 3706 patients with NPC treated in our hospital during five years. A total of 135 patients were found to have family history of NPC, and the incidence rate was 3.64%. The median age of these 135 patients was 50 years old (ranged: 18 to 77 years). The ratio of male to female was 2.97:1. There were 83.0% (112/135) patients who had only one family member suffering from NPC previously. Seventeen percent patients (23/135) had more than one family members suffering from NPC previously. There were 74.1% (100/135) patients with family history in first-degree relatives and 20.0% (27/135) patients with it in second-or third-degree members. Eight patients (5.9%) had family history in both first-and second-/third-family members, respectively.

The prognosis of NPC patients with family history
Excluding five patients lost to follow-up, 130 patients were eventually used to analyze the prognosis. The main characteristics of these patients were summarized in Table 1. All patients received radiotherapy, and most of patients (93.1%) received intensity modulated radiotherapy (IMRT). Of the 130 patients, 36 (27.7%) patients were treated with radiotherapy alone. The median radiation doses of the planning target volume were 70 Gy (59.4-76 Gy) to the primary tumor, 66 Gy (54-74.8 Gy) to the regional lymph nodes, 60 Gy (54-74 Gy) to the highrisk clinical target volume and 54 Gy (0-64 Gy) to the low-risk clinical target volume. Ninety-four patients (72.3%) received platinum-based chemotherapy, Both first-and second-/third-degree 8 (6.2) including induction chemotherapy (n=28, 29.8%), concomitant chemoradiotherapy (n=19), induction chemotherapy followed by adjuvant chemotherapy (n=7), induction chemotherapy followed by concomitant chemoradiotherapy (n=34), concomitant chemoradiotherapy followed by adjuvant chemotherapy (n=5) and induction chemotherapy followed by concomitant chemoradiotherapy followed by adjuvant chemotherapy (n=1). In statistics analysis, we put the patients into four subgroups: first-degree group, second-/third-degree group, one family member group and at least two family members group. First-degree group included the patients with first-degree relatives and both first-and second-/third-degree relatives.
Second-/third-degree group enrolled the patients with only second-/third-degree relatives. The main characteristics in various subgroups were listed in Table 2.
With a median follow-up of 71 months (range: 8-97 months), there were 28 deaths and 50 treatment relapse. Distant metastasis (n=28, 44%) was the major failure pattern, followed by local and regional relapse (n=10, 20%), isolated local relapse (n=10, 20%) and isolated regional relapse (n=8, 16%). The 5-year OS, LRFS, and DMFS rates of all patients with family history were 84.1%, 83.4%, and 83.8%, respectively. Of these, one person afflicted with malignant melanoma during the follow-up after one year. However he was not discovered any recurrence and metastasis about secondary neoplasia.
The univariate analysis results by Kaplan-Meier and log-rank test were listed in Table 3. Stage I-II or T1-T2 was associated with improved OS, LRFS and DMFS. The patients in stage N0-N1 had a reduction of distant metastasis. The locoregional control of IMRT was better than conventional radiotherapy. Male patients and the patients received chemotherapy seemed to have a tendency of distant metastasis. However, there were no significant differences of OS, LRFS and DMFS between one relative group and at least two relatives group. Similarly, the degree of NPC members had no association with OS, LRFS and DMFS, respectively ( Figure 1). Multivariate analysis showed that T stage was an independence prognostic factor for OS and LRFS.

DISCUSSION
Epidemiological data showed that a few of NPC patients had family history, and it suggested that genetic factor may play an important role in tumorigenesis of NPC. Studies reported that HLA was a significant gene for NPC. Three alleles including HLA-A * 1101, HLA-A * 0207 and HLA-B * 5801 have been confirmed to have a connection with NPC incidence in southeastern China [19][20][21]. A meta-analysis reported that rs29232 made a contribution to the geographical differences in NPC incidence and suggested that rs29232 might be more associated with NPC in moderate-incidence regions than in high-incidence regions among Han Chinese people. There may be three reasons. First, the heterogeneity of rs29232 was reduced in different regions. Second, rs29232 was independent of HLA-A gene in moderate-incidence regions instead of high-incidence regions. Finally, the odd ratios of rs29232 was higher in moderate-incidence regions than in high-incidence regions [22].
Several studies had reported the incidence of NPC patients with family history in various endemic areas. The incidence rate of familial NPC patients was 9.9%-14.6% in Southern China, including Guangdong and Guangxi province [9,10,23,24]. Another study from Taiwan revealed that 11.0% NPC patients had family history [11]. Furthermore, there were 11.7% NPC patients who had family history with first-degree relatives [24]. In present study, we respectively collected the clinical data of 3706 patients with NPC, and found that 3.64% patients (135/3706) had family history and 2.91% (108/3706) patients had family history with first-degree relatives. The incidence rate of NPC patients with family history in present study was significantly lower compared with the previous studies in endemic areas. The main cause may be that most NPC patients in present study came from Eastern China, a non-endemic area of NPC. These results also suggested that the prevalence of NPC patients with family history was more common in endemic areas than in non-endemic areas.
There were several studies to investigate the prognosis of NPC patients with family history. Cao et al. found that the 5-year OS, LRFS and DMFS rates were 67%, 70% and 76% of sporadic patients, 70%, 83% and 77% of low-frequency familial patients, and 61%, 87% and 64% of high-frequency familial patients. The 5-year LRFS showed significant difference between sporadic and familial NPC patients, whereas the 5-year OS and DMFS revealed no difference [23]. Ouyang et al. suggested that patients with first-degree family history obviously improved OS, LRFS and DMFS. In addition, there was a trend of improving OS, LRFS and DMFS with increasing number of family members [24]. Our study considered that the number of positive relatives and kinship were not associated with OS, LRFS and DMFS, which was contrary to the published researches. One reason is that the patients in the other studies came from endemic areas, while our study was focused on the incidence and the prognosis for familial NPC patients in non-endemic areas. Another reason may be the limit to the small sample.
A limitation of our study is that we didn't make a case control study. However we compared with the prognosis of 869 NPC patients with IMRT during the same period in our hospital [25]. The 5-year OS, LRFS, and DMFS rates were 84.0%, 89.7%, and 85.6%, which were similar to our study.
Another limitation is the skewed distribution in the subgroups of radiation technique and chemotherapy. The result suggested that the difference between IMRT and conventional radiotherapy was statistically significant.
In spite of the result that chemotherapy was associated with DMFS in univariate analysis, it was not statistically significant in multivariate analysis. The main reason that patients received chemotherapy had a lower trend of DMFS than those without chemotherapy might be that most patients with chemotherapy belonged in advanced stage. Another limitation is the skewed distribution in the subgroups of radiation technique and chemotherapy, due to a small number of patients. In spite of the result that chemotherapy was associated with DMFS in univariate analysis, it was not statistically significant in multivariate analysis. The main reason is that most patients with chemotherapy belonged in advanced stage.

CONCLUSION
Our results showed that there was an incidence rate of 3.64% NPC patients with family history. These patients had a satisfied prognosis after IMRT with or without chemotherapy. The prognosis of NPC patients with family history in different degree or numbers of relatives had no significant differences. However, the major limitation of the present study is that we did not compare the prognosis between patients with and without family history.

Screening of NPC patients with family history
The clinical data of patients with NPC treated in Fudan University Shanghai Cancer Center from January 2008 to December 2012 were reviewed, and the patients with family history were selected. Inclusion criteria were previously untreated, pathologically confirmed NPC, including distant metastasis or not, and receiving whole course of radiotherapy as planed. All investigational sites had Institutional Review Board approval (IRB), and all patients provided voluntary informed consent to participate in the study. The collected data of patients with family history were recorded, including age, gender, pathologic diagnose, clinical stage, therapeutic modalities. All patients were re-staged according to the seventh www.impactjournals.com/oncotarget edition of the UICC/AJCC staging system [26]. The pathology of all the patients enrolled was nonkeratinizing carcinoma (WHO type II, ICD-O 8072/3 [27]). The family members diagnosed with NPC previously were divided into three degrees: first-degree family members (parent, sibling and offspring), second-degree family members (aunt, uncle, nephew, niece and grandparent) and thirddegree family members (cousin).

Follow-up of NPC patients with family history
The follow-up results of these patients were assessed by clinical physical examinations and imaging methods, mainly including nasopharynx and neck MRI or CT scanning, chest CT scanning, abdominal ultrasound and bone scanning. Neoplasm recurrence is defined as local relapse more than six months after first course radiotherapy or re-radiotherapy. Persistent disease is defined as uncontrolled leisions or relapse after first course radiotherapy within six months. The observing end-points were 5-year OS, LRFS and DMFS. The survival duration was defined from the day of treatment to the day of death or last follow-up.

Statistical analysis
Statistical analysis was performed by using SPSS 19.0. The different survival rates were analyzed by the Kaplan-Meier method and log-rank test. Cox proportional hazard model was used for multivariate analysis by the estimate of hazard ratio (HR) and 95% confidence interval (CI). For all tests, a two-sided p< 0.05 was considered to be significant.

Author contributions
The first two authors contributed equally to this work.