A total of 180 patients with nasopharyngeal carcinoma were included in the analysis. The median survival time of OS, LPFS, and DMFS were 60 months, 55 months, and 55 months, respectively. The rates of 3-year OS, 3-year LPFS, and 3-year DMFS were 91%, 93%, and 91%, respectively. Among the 180 patients, 11 had fatal events, 13 had recurrence and metastasis, 2 had simple local recurrence in the nasopharynx, 5 had simple neck lymph node recurrence, 1 had nasopharyngeal and neck recurrence, and 2 had local recurrence accompanied by distant metastasis. Distant metastases occurred in 12 patients: 2 liver metastases, 8 lung metastases, 6 bone metastases, 1 mediastinal metastasis, and 1 abdominal lymph node metastasis.
Factors analyzed in the study: General clinical factors and PET-related parameters, including age, gender, T stage, N stage, clinical stage, pathological type, T-SUVmax, LN-SUVmax, and NTR.
Determination of the cut-off value of each parameter: With OS as the main survival end point, the boundary value of each parameter was determined according to the ROC curve. The best cut-off value of NTR related to OS was 0.95 (AUC of 0.710, P = 0.044, 95% CI = 0.560–0.860, sensitivity = 0.750, and specificity = 0.674) (Figure 3). The patients were divided into two groups: the high NTR group (NTR > 0.95; n = 80) and the low NTR group (NTR < 0.95; n = 100). The cut-off values of the other two parameters were T-SUVmax of 12.35 (AUC = 0.582, P = 0.361, 95% CI = 0.395–0.770, sensitivity = 0.636, specificity = 0.586) and LN-SUVmax of 9.32 (AUC = 0.765, P = 0.003, 95%CI = 0.650–0.879, sensitivity = 0.909, specificity = 0.521). A total of 93 patients had LN-SUVmax > 9.32 and 87 had LN-SUVmax < 9.32.The results of univariate prognostic analysis of different survival endpoints are shown in Table 1. A univariate analysis of OS, LPFS, and DMFS of the whole group of patients was carried out. The observation targets were OS mainly, LPFS, and DMFS. Univariate analysis showed that age, LN-SUVmax, and NTR were associated with OS (P = 0.013, 0.001, and 0.001, respectively); LN-SUVmax and NTR with LPFS (P = 0.005 and 0.020, respectively), and clinical stage, LN-SUVmax, and NTR with DMFS (P = 0.023, 0.007, and 0.032, respectively). Kaplan–Meier results showed that the 3-year OS (97.0% vs. 85.4%, c2 = 11.25, P = 0.001), 3-year LPFS (91.3% vs. 82.1%, c2 = 4.035, P = 0.045), and 3-year DMFS (92.3% vs. 87.9%, c2 = 4.576, P = 0.032) of patients with NTR < 0.95 were higher than those with NTR > 0.95 (Figure 4–8). In addition, patients with LN-SUVmax < 9.32 had better 3-year OS (100% vs. 88.0%, c2 = 10.565, P = 0.001), 3-year LPFS (96.6% vs. 93.4%, c2 = 7.715, P = 0.005), and 3-year DMFS (97.8% vs. 82.5%, c2 = 7.387, P = 0.007) than patients with LN-SUVmax > 9.32.
Multivariate analysis showed that NTR was a significant influencing factor for OS (HR = 3.00, 95%CI = 1.06–8.48, P = 0.038) and LPFS (HR = 3.08, 95%CI = 1.27–7.50, P = 0.013), while only marginally significant for DMFS (HR = 1.84, 95% CI = 0.99–3.42, P = 0.054) Table 2. In addition, N stage (HR = 3.32, 95%CI = 1.04–10.66, P = 0.044) and clinical stage (HR = 9.78, 95%CI = 1.91–50.15, P = 0.006) were also influencing factors of LPFS, while clinical stage was a marginally significant influencing factor of DMFS (HR = 2.12, 95% CI = 0.99–4.55, P = 0.053).
Table 1
Univariate prognostic analysis of different survival endpoints in 180 patients with nasopharyngeal carcinoma
Variables
|
|
OS
|
|
|
LPFS
|
|
DMFS
|
|
|
|
|
|
|
|
|
|
n
|
c2
|
P value
|
c2
|
P value
|
c2
|
P value
|
Sex
|
|
3.009
|
0.083
|
3.817
|
0.051
|
2.949
|
0.086
|
male
|
121
|
|
|
|
|
|
|
female
|
59
|
|
|
|
|
|
|
Age
|
|
6.119
|
0.013
|
3.733
|
0.053
|
3.682
|
0.055
|
>59.5
|
41
|
|
|
|
|
|
|
<59.5
|
139
|
|
|
|
|
|
|
T stage
|
|
0.064
|
0.800
|
1.712
|
0.191
|
3.646
|
0.056
|
T1
|
27
|
|
|
|
|
|
|
T2
|
51
|
|
|
|
|
|
|
T3
|
61
|
|
|
|
|
|
|
T4
|
41
|
|
|
|
|
|
|
N stage
|
|
3.425
|
0.064
|
0.545
|
0.460
|
1.857
|
0.173
|
N1
|
64
|
|
|
|
|
|
|
N2
|
73
|
|
|
|
|
|
|
N3
|
43
|
|
|
|
|
|
|
Clinical stage
|
|
3.011
|
0.083
|
0.758
|
0.384
|
5.175
|
0.023
|
II
|
24
|
|
|
|
|
|
|
III
|
78
|
|
|
|
|
|
|
IV
|
43
|
|
|
|
|
|
|
Pathological type
|
|
0.088
|
0.767
|
0.088
|
0.767
|
0.049
|
0.824
|
Undifferentiated keratinizing carcinoma
|
179
|
|
|
|
|
|
|
Differentiated non-keratinizing carcinoma
|
1
|
|
|
|
|
|
|
T-SUVmax
|
|
1.877
|
0.171
|
0.049
|
0.825
|
0.076
|
0.783
|
<12.35
|
103
|
|
|
|
|
|
|
>12.35
|
77
|
|
|
|
|
|
|
LN-SUVmax
|
|
10.565
|
0.001
|
7.715
|
0.005
|
7.387
|
0.007
|
<9.32
|
87
|
|
|
|
|
|
|
>9.32
|
93
|
|
|
|
|
|
|
NTR
|
|
11.25
|
0.001
|
5.385
|
0.020
|
4.576
|
0.032
|
<0.95
|
100
|
|
|
|
|
|
|
>0.95
|
80
|
|
|
|
|
|
|
Abbreviation T-SUVmax: SUVmax of primary nasopharyngeal carcinoma, LN-SUVmax: SUVmax of lymph node metastases, NTR: ratio of LN-SUVmax to T-SUVmax
Table 2
Cox multivariate analysis
|
OS
|
|
LPFS
|
|
DMFS
|
|
|
HR(95%CI)
|
P
|
HR(95%CI)
|
P
|
HR(95%CI)
|
P
|
Sex
male vs female
|
|
|
|
|
|
|
Age
>59.5 vs <59.5
|
|
|
|
|
1.74(1.01-3.01)
|
0.046
|
T stage
T1+T2 vs T3+T4
|
|
|
|
|
|
|
N stage
N1+N2 vs N3
|
|
|
3.32(1.04-10.66)
|
0.044
|
|
|
Clinical stage
II+III vs IV
|
|
|
9.78(1.91-50.15)
|
0.006
|
2.12(0.99-4.55)
|
0.053
|
T-SUVmax
<12.35 vs >12.35
|
|
|
|
|
|
|
LN-SUVmax
<9.32 vs >9.32
|
|
|
|
|
|
|
NTR
<0.95 vs >0.95
|
3.00(1.06-8.48)
|
0.038
|
3.08(1.27-7.50)
|
0.013
|
1.84(0.99-3.42)
|
0.054
|
Abbreviation T-SUVmax: SUVmax of primary nasopharyngeal carcinoma, LN-SUVmax: SUVmax of lymph node metastases, NTR: ratio of LN-SUVmax to T-SUVmax.,HR:hazard ratio,CI,:confidence interval.