Risk factors associated with suicide among kidney cancer patients: A Surveillance, Epidemiology, and End Results analysis

Abstract Background The suicide risk was higher in kidney cancer patients than in the general population. The purpose of this study was to characterize the suicide rates among kidney cancer patients and to identify the potential risk factors associated with suicide from the Surveillance, Epidemiology, and End Results (SEER) database. Methods Kidney cancer patients were identified from the SEER database during 1973‐2015. Suicide rates and standardized mortality ratios (SMRs) of this population were calculated, and the US general population during 1981‐2015 was chosen as a reference. Univariable and multivariable Cox regression were performed to find out potential risk factors of suicide. Results There were 207 suicides identified among 171 819 individuals with kidney cancer observed for 948 272 person‐years. The suicide rate was 21.83 per 100 000 person‐years, and SMR was 1.83 (95% CI: 1.59‐2.10). On Cox regression, diagnosis in early years (1973‐1982 vs 2003‐2015, HR: 2.03, 95% CI: 1.01‐4.11, P = 0.048; 1983‐1992 vs 2003‐2015, HR: 1.99, 95% CI: 1.18‐3.35, P = 0.010), male sex (vs female sex, HR: 4.43, 95% CI: 2.95‐6.65, P < 0.001), unmarried status (vs married status, HR: 2.54, 95% CI: 1.91‐3.38, P < 0.001), non‐black race (white race vs black race, HR: 4.47, 95% CI: 2.09‐9.58, P < 0.001; other races vs black race, HR: 3.01, 95% CI: 1.08‐8.37, P = 0.035), higher histologic grade (grade IV vs grade I, HR: 3.27, 95% CI: 1.50‐7.13, P = 0.003; grade III vs grade I, HR: 2.13, 95% CI: 1.19‐3.81, P = 0.011) and cancer‐directed surgery not performed (vs performed, HR: 2.78, 95% CI: 1.52‐5.11, P < 0.001) were independent risk factors of suicide among kidney cancer patients. Conclusions Diagnosis in early years, male sex, unmarried status, non‐black race, higher histologic grade, and cancer‐directed surgery not performed were significantly associated with suicide among kidney cancer patients. In order to prevent suicidal death, clinicians should pay more attention to patients with high‐risk factors of suicide.


| 5387
GUO et al 1 | BACKGROUND Suicide, a global public health problem, was a complex behavior affecting by physiological, psychological, social, environmental, and cultural factors. 1 World Health Organization reported that over 800 000 persons committed suicide every year and one person every 40 seconds. 1 Globally, 81 7000 persons died of suicide in 2016, comprising 1.49% of total deaths. 2 In the US, the suicide mortality rate was 15.3 per 100 000 in 2016, which was relatively higher in comparison with the surrounding countries. 3 In recent years, studies reported a notable association between depression and suicide, and the ssuicide rate among patients with depression was much higher than the usual population. [4][5][6] The patients diagnosed as diseases with poor prognosis, especially cancer, were more likely to feel hopeless, suffer from depression and subsequently commit suicide. 7,8 Several studies reported increased suicide ideation and attempts in patients with cancer and showed a high suicide rate among these patients. 7,9,10 Compared with the general population, the suicide rate of patients with cancer was almost twice in the US. 11 A current study by Zaorsky et al revealed that the cancer patients had a suicide SMR of 4.44 in comparison with the general population. 12 For the behavior of suicide could be potentially recognized and prevented, the identification of patients with high-risk factors of suicide was particularly important. The risk factors including male sex, white race, and unmarried status were demonstrated to be associated with a high incidence of suicide for some types of cancer, such as breast cancer, colorectal cancer, and pancreatic cancer. [13][14][15] Kidney cancer was the seventh and the ninth most common malignancy in men and women. 16 In the US, the number of patients with kidney cancer was estimated to be 65 340 in 2018, and the number of death cases was estimated to be 14 970. 17 Misono et al reported increased suicide rates of patients with cancer in the US and characterized this specific population. 11 However, the risk factors of suicide in patients with kidney cancer were not involved in the study. Klaassen et al selected the patients with kidney cancer through the SEER database and identified the risk factors of suicide. 18 However, papillary renal cell carcinoma was not included in this study, which was the second most frequency subtype in all renal cortical neoplasms. 19 Histologic subtype was an important prognostic factor for kidney cancer survival. 20 To our best knowledge, there was no previous study to investigate the association between histologic subtype and suicide among patients with kidney cancer. Additionally, few studies explored the risk factors associated with suicide among kidney cancer patients based on a large representative sample. Therefore, the purpose of our study was to characterize the suicide rates and SMRs in this specific cohort and to identify the potential risk factors relevant to suicide based on the SEER database.

| Data source
All patients with kidney cancer were selected from the SEER database . The SEER database reported cancerspecific outcomes from specific geographic areas covering 28% of the US population. The data in the SEER program were considered to be representative of the entire US population. 21 The SEER database could freely provide information including patient demographics, cancer incidence, and survival data to registered researchers. The permission of accessing the database was obtained after we signed and submitted a SEER Research Data Agreement form through email. The software of SEER*Stat (version 8.3.5) was used to identify the patients.

| Ascertainment of the outcome
The primary outcome of this study was suicidal death, which was ascertained by cause of death code (suicide and self-inflicted injury) in the software of SEER*Stat.

| Statistical analysis
The suicide rates between groups were compared with the chi-square test, and the Bonferroni-corrected P-value was used for multiple comparisons. SMR was calculated using the ratio of actual mortality with expected mortality. Because of the lack of suicide rates during 1973-1980 in the Centers for Disease Control and Prevention's Web-based Injury Statistics Query and Reporting System, the US general population between 1981 and 2015 was chosen as a reference, and a sensitive analysis with the kidney cancer patients diagnosed in 1981-2015 was performed to validate the robustness of SMRs. Ninety-five percent of CI of SMR was obtained through the method of Byar's approximation. 22 Univariable and multivariable Cox regression were performed to generate crude and adjusted HRs and 95% CI for finding out potential risk factors of suicide. The proportional hazards assumption of Cox regression modeling was evaluated with the use of Schoenfeld residuals. All statistical analysis was carried out using R software (version 3.5.2, http://www.r-proje ct.org/). All tests were two-sided, and the significance level was set at P < 0.05.

| Difference in suicide rates and SMRs
From 1973 to 2015, there were 207 suicides among 171 819 patients with kidney cancer observed for 948 272 personyears, yielding a suicide rate of 21.83 per 100 000 personyears. The suicide rate of US general population reported by the Centers for Disease Control and Prevention was 11.93 per 100 000 person-years between 1981 and 2015, 23 which was significantly lower than that of the kidney cancer patients in our cohort (P < 0.001). The results showed that higher suicide rates in patients with kidney cancer were associated with male sex (vs female, P < 0.001), unmarried status (vs married, Bonferroni-corrected P < 0.001), white race (vs black race, Bonferroni-corrected P = 0.002), and cancer-directed surgery not performed (vs performed, Bonferroni-corrected P < 0.001). The result of the chi-square test for linear trend showed that suicide rates among kidney cancer patients increased with histologic grade (P < 0.001) and disease stage (P < 0.001). Meanwhile, no statistical differences in suicide rates were found with respect to year of diagnosis, age at diagnosis, histologic type, radiotherapy, and time from diagnosis. SMRs were used to compare suicide mortality in the study population to that in the general population. A SMR of 1.83 (95% CI: 1.59-2.10) was noted between kidney cancer patients and the US general population, with 1.57 (95% CI:

| Associations of risk factors with suicide
The result of univariable analysis presented that significant associations with high risks of suicide were obtained with respect to diagnosis in 1983-1992 (vs 2003-2015 (1973)(1974)(1975)(1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983) were at significantly higher risk of suicide compared with those in recent years (2004)(2005)(2006)(2007)(2008)(2009)(2010). 28 In the present study, the suicide rate of males was 30.28 per 100 000 person-years, which was 2.9 times higher than that of females. Furthermore, males were more likely than females to commit suicide with an HR of 4.43, which was consistent with the findings in the general population and patients with other types of cancer. 22,28 Though it seemed that male and female cancer patients suffered similar pressure, male patients were more likely to react through self-directed violence. [28][29][30] Unmarried status was another significant risk factor of suicide with an HR of 2.54 in comparison with married status. Married patients usually had higher cancer-specific survival and lower mortality in comparison with unmarried patients, [31][32][33] which could be ascribed to better health, higher socioeconomic status, more emotional support and social attention of married persons. [34][35][36] The present study corroborated the finding that race had a significant impact on suicidal death. As was reported by the Centers for Disease Control and Prevention, the black race had the lowest suicide rate among all races in the US. 37 Our study showed that the suicide rates of the white race and other races were respectively 2.77 and 1.34 times higher than that of the black race. Furthermore, the black race was proved to be a protective factor of suicide, matching the findings in previous investigations. 13,38 In terms of the black race, this finding might be explained by religious beliefs, family support, and culture of rejecting suicide. 39 Some recent publications showed that older age was a predictor of suicide for cancer patients and the common population. 11,40 Interestingly, older breast cancer patients who fare better in terms of functional status than younger patients were less likely to commit suicide. 13,41 In our study, a tendency of increased suicide rate with age was also observed, although without statistical association.
As reported in previous studies, suicide risk among cancer patients varied by time after diagnosis, and elevated suicide rates could be found in the initial period after diagnosis. 27,38 Our findings showed significantly increased suicide rates among kidney cancer patients with the general population in the initial 9 years after diagnosis. A statistical trend of decreased suicide rates over time was not found, which might be attributed to relative better prognosis of kidney cancer in comparison with other types of cancer.
Regarding specific clinical variables of kidney cancer, several findings should be noteworthy. It was well known that low histologic grade meant well differentiated of cancer cells, predicting good prognosis and improved healthrelated quality of life. 42 The patients with higher histologic grade (Grade III and IV) in our study were found to have higher suicide risks than those with lower histologic grade (Grade I). However, this finding was contrary to the previous research by Gaitanidis  and lower tumor grade were associated with higher risks of committing suicide among breast cancer patients. This conflicting result might be partially attributed to the fact that all the patients with breast cancer enrolled in the study had passed away. 13 Increased suicide risks were commonly associated with cancer types of poor prognosis. 27 Recently, some investigations had reported that histologic subtype was not significantly associated with suicide risks among non-small cell lung cancer patients and colorectal cancer patients. 38,43 As we know, histologic subtype was a prognostic indicator for kidney cancer survival. 19 Compared with clear cell renal cell carcinoma, patients with sarcomatoid renal cell carcinoma and collecting duct renal cell carcinoma had worse overall survival and cancer-specific survival. 44,45 Conversely, patients with chromophobe histology had improved survival. 45 Based on our study, there was no significant difference in suicide rates among patients with different histologic subtypes, and no apparent association between histologic subtype and suicide was found by Cox regression modeling. However, it should be noted that the lack of suicide event for sarcomatoid and collecting duct renal cell carcinoma which was low incidence and poor prognosis might have an effect on this result.
As shown in Table 3, another factor associated with suicide was cancer-directed surgery. The patients underwent surgery were less likely to commit suicide than those who did not undergo surgery. Samawi et al identified the predictors of suicide in colorectal cancer patients based on the SEER database. Primary site surgery as an independent protective factor was reported in this study, probably attributing to the improvement of mortality after primary tumor resected. 14,46,47 Similarly, it was reported that patients with cancers of the digestive system who underwent surgery were less likely to commit suicide. 48 Conversely, cancer-directed surgery was associated with increased suicide rate in breast cancer patients, which might be related to the increased adverse psychological impact after mastectomy and the relative body image disturbance. 13,49,50 Zhou et al reported that cancer-directed surgery was a risk factor for committing suicide among patients with non-small cell lung cancer, probably due to the increased frailty and lack of dignity in the postoperative period. 38 Jayakrishnan and colleagues investigated the association between suicide and surgery. The patients with cancer who underwent high-morbidity surgeries (30-day overall postoperative morbidity > 30%) were more likely to commit suicide than those with low-morbidity surgeries performed. In addition, no clear association of suicide with the anatomic site of cancer was obtained. 40 Indeed, the patients usually suffered from general debility, and experienced depression, hopeless, and despair after surgery, which were risk factors of suicide. 40 However, there was no consensus regarding the association between cancer-directed surgery and suicide so far. Of note, suicide was a complicated phenomenon involving not only physiological but also psychological and social factors. Therefore, more factors should be taken into account to clarify this relationship.
Suicidal behavior in patients with cancer is affected by various factors. Compared with other causes of death such as accident, suicide is preventable. 1 It is different to predict suicide, and therefore, more efforts should be made to improve the situation. Based on our findings, we suggest that kidney cancer patients with high risks of suicide should be considered for psychiatric evaluation. At present, several validated tools can be used for identifying depression risk, including the Beck Depression Inventory and the National Comprehensive Cancer Network guidelines. 13,28 Second, the patients at risk should receive psychotherapeutic interventions as soon as possible. It has been widely proved that psychotherapeutic interventions to the depression population can reap benefits in terms of reducing the suicide rate and improving the quality of life, such as participation in cancer support groups, stress management, and integrate psychological support into cancer care. 13 for decreasing suicide need coordination and collaboration, including medical workers, family members even the whole society. Besides professional care, family communication and social support also play an important role in preventing suicides and suicide attempts. 1,51,52 There were some limitations in our study. Besides pathological factors, suicide was affected by psychological and social factors which could not be provided by the SEER database. Moreover, there was also a lack of sufficient medical information in the SEER database, such as pharmacotherapy and genetic factors. Additionally, our study was based on the SEER database, which only collected the corresponding data of cancer patients in the US. Therefore, further study coving more countries should be conducted. Finally, data of failed suicide could not be obtained and exhibited, which might result in an under evaluation to suicide risk.