As shown in figure 1, a total of 45117 patients met the eligibility criteria, and among them, there were 32506(72.05%) patients in Alive group, 3087 (6.84%) in CCD group, 3212 (7.12%) in KCD group and 6312 (13.99%) in OE group.
As Table 1 showed, in CCD group, the majority patients were married (n=1645, 53.29%), white race (n=2482, 80.40%), male (n=1928, 62.46%) and diagnosed with KC at age≥75 (n=1400, 45.35%). There were more patients stayed as AJCC T1 (n=2668, 86.43%), N0 status (n=3063, 99.22%) and with tumor size＞3cm (n=2030, 65.76%). As for treatment situation, the minority of patients received radiation therapy (n=23, 0.75%), chemotherapy (n=16, 0.52%) and scope lymph node (n=177, 5.73%), but most of them performed surgery (n=2570, 83.25%). In KCD group, patients’ characteristics were similar to the patients in CCD group. There were 1093(34.03%) patients diagnosing at age>75, 1890 (58.84%) married, 2670 (83.13%) White people and 2035 (63.36%) males. There were 2039 (63.48%) patients in AJCC T1 status, 2992 (93.15%) in N0 status. As for treatment situation, there were 86 (2.68%) patients received radiation therapy, 164 (5.11%) received chemotherapy, 453 (14.10%) received scoping lymph node and 2606 (81.13%) operated surgery. In Alive group, the patients’ characteristics were different from KCD and CCD group. There were 3280 (10.09%) patients diagnosed at age >75, 22042 (67.81%) married, 26784 (82.40%) White people and 19568 males. There were 28286 (87.02%) patients in AJCC T1 status, 32389 (99.64%) in N0 status. As for treatment situation, there were 49 (0.15%) patients received radiation therapy, 139 (0.43%) received chemotherapy and 2252 (6.93%) received scoping lymph node and 31733 (97.62%) operated surgery.
Cumulative incidence function of CCD, KCD and OE
We estimated the cumulative incidence ratios of each outcomes via competing risks model. As figure 2 showed, KCD and CCD represented almost the same cumulative incidences in patients after diagnosed with T1/2 KC. On 36 months after diagnosed, the cumulative incidences of CCD, KCD and OE were 2.50%, 3.50% and 4.83%, respectively. On 60 months after diagnosed, the cumulative incidences were 4.03 %, 5.08% and 7.97%, respectively. On 120 months after diagnosed, the cumulative incidences were 8.13%, 8.04% and 16.60%, respectively (Table 2).
Univariate and multivariate analysis by Fine-Gray’s competing risk model
Firstly, we used the univariate Fine-Gray’s competing risk model to analysis all factors listed above. In CCD group, as Table 3 showed, except American Indian/Alaska native, sex and AJCC N status, all factors were significant associated with CCD (p<0.05). Then we estimated those factors by multivariate Fine-Gray’s competing risk model, and it was found that age at diagnosis, race, marital status, AJCC T status, chemotherapy and surgery of primary site stayed as significant risk factors of CCD. In detail, elder patients showed a higher risk of CCD, the respective HRs were 2.117 (95% CI:1.756-2.552), 3.200 (95% CI:2.672-3.832), 4.981 (95% CI:4.187-5.925) and 9.525 (95% CI:8.049-11.273) at age 51-58, 59-65, 66-73 and over 75 versus 1-50. Black patients faced higher risk (black versus white: HR=1.475, 95% CI:1.334-1.632), but Asian/Pacific Islander patients had less risk of CCD(Asian/Pacific Islander versus white: HR=0.826, 95% CI:0.683-0.998), married patients also showed a lower risk (married versus single/divorce/widow: HR=0.677, 95% CI:0.628-0.730). As for AJCC T status, T2 status had less risks of CCD than T1 (HR=0.847, 95% CI:0.758-0.946) , tumor size>3 cm group had higher risk (HR=1.111, 95% CI:0.1.023-1.206); patients received chemotherapy and surgery of primary site also reduced risks of CCD, in detail, the respective HRs were 0.574 (95% CI:0.347-0.949) with chemotherapy versus without; 0.631 (95% CI:0.532-0.747), 0.526(95% CI:0.466-0.594), 0.607 (95% CI:0.543-0.680) with local tumor destruction/excision, with partly nephrectomy and with radical nephrectomy versus no operation.
When it came to KCD group, as Table 4 showed, it was found that age at diagnosis, marital status, sex, AJCC TN status, tumor size, radiation therapy, chemotherapy, scope lymph node and surgery of primary site were significantly associated with this outcome by univariate competing risk model. Then we estimated these factors by multivariate competing risk model, these factors were still significant associated with KCD. In detail, elder patients still showed a higher risk of KCD, the respective HRs were 1.578 (95% CI:1.376-1.810), 1.933 (95% CI:1.688-2.215), 2.176 (95% CI:1.899-2.494) and 3.205 (95% CI:2.814-3.150) at age 51-58, 59-65, 66-73 and over 75 versus 1-50. Married patients showed a lower risk of KCD (married versus single/divorce/widow: HR=0.867, 95% CI:804-0.935). As for AJCC T status, T2 status had more risks of KCD than T1 (HR=2.259, 95% CI:2.081-2.451), advanced N status had more risks of KCD, the respective HRs were 3.347 (95% CI:2.698-4.152), 4.004 (95% CI: 2.837-5.650) stayed as N1, N2 versus N0 status; tumor size>3 cm group had higher risk (HR=2.319, 95% CI:2.086-2.579). Patients received chemotherapy, radiation therapy and scope of lymph node still had high risks of KCD, the respective HRs were 2.896 (95% CI:2.342-3.581) with chemotherapy versus without; 2.552 (95% CI:1.946-3.346) with radiation therapy versus without, 1.378(95% CI:1.206-1.575) with 1-3 regional lymph nodes removed, 1.230 (95% CI:1.022-1.480), 4 or more regional lymph nodes removed than no scoping of lymph node. Patients received surgery of primary site faced lower risks of KCD, the respective HRs were 0.356 (95% CI:0.286-0.443), 0.275 (95% CI:0.242-0.312) and 0.393 (95% CI: 0.352-0.438) with local tumor destruction/excision, with partly nephrectomy and with radical nephrectomy versus no operation.
Forest plots of the risk factors and hazard ratios of CCD and KCD
We summarized the multivariate competing risk analysis results of CCD and KCD in patients with T1/2 kidney cancer, and curved forest plots to visualized these results, which were shown in Figure 3 and 4. Comparing these results, we found that there were some differences in the risk factors that significant associated with CCD or KCD. For CCD, age was the biggest risk factor that represented the highest HRs, race was also associated with CCD, but N status and scope of lymph node didn’t show the relationship with CCD. However, for KCD, age didn’t represent the comparatively high HRs to KCD, race was not associated with KCD. Moreover, N status was the biggest risk factors associated with KCD, and T status showed the opposite HRs to KCD comparing with CCD. Radiation therapy, chemotherapy and scope of lymph node also showed different HRs to CCR and KCD, but surgery of primary site represented the similar HRs.