Patient Cohort and Clinical Characteristics
Between 2015 and 2019, 604 consecutive patients were diagnosed with rectal cancer. Seventy-three patients were aged 50 years or less, and 531 were over 50 years of age (Fig.1). The clinicopathological characteristics of the study population are summarized in Table 1.
The clinicopathological characteristics of the study population after matching are summarized in Table 2. Young-onset patients had a significantly higher proportion of family cancer history (28.8%) than older patients (P<0.001). Three (4.3%) young-onset patients had a clear germline genetic mutation, including MLH1, MSH2, MSH6, and PMS2. However, among young-onset patients, 21 patients had a clear family cancer history, of which 19 (90.5%) had a first-degree relative (FDR) diagnosed with CRC. Regarding postoperative pathology, almost half of young-onset patients (46.6%) showed vascular invasion, and young-onset patients were more likely to have worse pathological stages than those in older patients. There was no significant difference in the primary site (P=0.368); however, synchronous and multiple primary tumors were more likely to occur among young-onset patients (P=0.079).
Young-onset patients were more likely to be treated aggressively than older patients with neoadjuvant therapy (32.9% vs. 17.8%) and adjuvant therapy (74.0% vs. 42.5%). However, young-onset patients did not achieve better oncologic outcomes with more aggressive treatment (Fig.2).
Table 1 Comparison of demographics and clinicopathological data
|
Total
(n=604)
|
≤ 50 years
(n=73)
|
>50 years (n=531)
|
P*
|
Sex, n (%)
|
|
|
|
0.249
|
Male
|
385(63.7)
|
42(57.5)
|
343(64.6)
|
|
Female
|
219(36.3)
|
31(42.5)
|
188(35.4)
|
|
Clinical stage, n (%)
|
|
|
|
0.898
|
I
|
80(13.2)
|
9(12.3)
|
71(13.4)
|
|
II
|
154(25.5)
|
18(24.7)
|
136(25.6)
|
|
III
|
324(53.6)
|
39(53.4)
|
285(53.7)
|
|
IV
|
46(7.6)
|
7(9.6)
|
39(7.3)
|
|
Pathological stage, n (%)
|
|
|
|
|
(y)pT
|
|
|
|
0.002
|
T0/Tis
|
23(3.9)
|
8(11.0)
|
15(2.8)
|
|
T1
|
67(11.1)
|
3(4.1)
|
64(12.1)
|
|
T2
|
108(17.9)
|
11(15.1)
|
97(18.3)
|
|
T3
|
378(62.6)
|
48(65.8)
|
330(62.1)
|
|
T4
|
28(4.6)
|
3(4.1)
|
25(4.7)
|
|
(y)pN
|
|
|
|
0.012
|
N0
|
364(55.7)
|
35(47.9)
|
329(62.0)
|
|
N1
|
163(29.7)
|
21(30.9)
|
142(26.7)
|
|
N2
|
77(14.6)
|
17(21.2)
|
60(11.3)
|
|
(y)pTNM
|
|
|
|
0.001
|
0
|
22(3.6)
|
8(11.0)
|
14(2.6)
|
|
I
|
147(24.3)
|
9(12.3)
|
138(26.0)
|
|
II
|
183(30.3)
|
17(23.3)
|
166(31.3)
|
|
III
|
204(33.8)
|
32(43.8)
|
172(32.4)
|
|
IV
|
48(7.9)
|
7(9.6)
|
41(7.7)
|
|
Values in parentheses are percentages
* χ2 or Fisher’s exact test
Table 2 Comparison of demographics and clinicopathological data (after matching)
|
Total
(n=219)
|
≤ 50 years
(n=73)
|
>50 years (n=146)
|
P*
|
Sex, n (%)
|
|
|
|
1.000
|
Male
|
126(57.5)
|
42(57.5)
|
84(57.5)
|
|
Female
|
93(42.5)
|
31(42.5)
|
62(42.5)
|
|
Family cancer history, n (%)
|
|
|
|
<0.001
|
Yes
|
32(14.6)
|
21(28.8)
|
11(7.5)
|
|
No
|
187(85.4)
|
52(71.2)
|
135(92.5)
|
|
BMI, mean ± SD
|
23.8±3.2
|
24.1±3.2
|
23.7±3.3
|
0.458
|
Clinical stage, n (%)
|
|
|
|
1.000
|
I
|
27(12.3)
|
9(12.3)
|
18(12.3)
|
|
II
|
54(24.7)
|
18(24.7)
|
36(24.7)
|
|
III
|
117(53.4)
|
39(53.4)
|
78(53.4)
|
|
IV
|
21(9.6)
|
7(9.6)
|
14(9.6)
|
|
Primary site, n (%)
|
|
|
|
0.368
|
Higha
|
53(24.2)
|
17(23.3)
|
36(24.7)
|
|
Middleb
|
107(48.9)
|
32(43.8)
|
75(51.4)
|
|
Lowc
|
59(26.9)
|
24(32.9)
|
35(24.0)
|
|
Synchronous and Multiple Primary, n, (%)
|
|
|
|
0.079
|
Yes
|
5(2.3)
|
4(5.5)
|
1(0.7)
|
|
No
|
214(97.7)
|
69(94.5)
|
145(99.3)
|
|
Neoadjuvant therapy, n (%)
|
|
|
|
0.012
|
Yes
|
50(22.8)
|
24(32.9)
|
26(17.8)
|
|
No
|
169(77.2)
|
49(67.1)
|
120(82.2)
|
|
Pathological stage, n (%)
|
|
|
|
|
(y)pT
|
|
|
|
0.039
|
T0/Tis
|
12(5.5)
|
8(11.0)
|
4(2.8)
|
|
T1
|
19(8.7)
|
3(4.1)
|
16(11.0)
|
|
T2
|
41(18.7)
|
11(15.1)
|
30(20.5)
|
|
T3
|
138(63.0)
|
48(65.8)
|
90(61.6)
|
|
T4
|
9(4.1)
|
3(4.1)
|
6(4.1)
|
|
(y)pN
|
|
|
|
0.036
|
N0
|
122(55.7)
|
35(47.9)
|
87(59.6)
|
|
N1
|
65(29.7)
|
21(30.9)
|
44(30.1)
|
|
N2
|
32(14.6)
|
17(21.2)
|
15(10.3)
|
|
(y)pTNM
|
|
|
|
0.011
|
0
|
12(5.5)
|
8(11.0)
|
4(2.7)
|
|
I
|
51(23.3)
|
9(12.3)
|
42(28.8)
|
|
II
|
53(24.2)
|
17(23.3)
|
36(24.7)
|
|
III
|
82(37.4)
|
32(43.8)
|
50(34.2)
|
|
IV
|
21(9.6)
|
7(9.6)
|
14(9.6)
|
|
Vascular invasion, n (%)
|
|
|
|
0.024
|
Yes
|
78(35.6)
|
34(46.6)
|
44(30.1)
|
|
No
|
141(64.4)
|
39(53.4)
|
102(69.9)
|
|
IHC, n (%)
|
|
|
|
0.135
|
dMMR
|
4(2.1)
|
3(4.3)
|
1(0.8)
|
|
MLH1(-)
|
2(1)
|
1(1.4)
|
1(0.8)
|
|
MSH2(-)
|
2(1)
|
2(2.9)
|
0(0)
|
|
MSH6(-)
|
2(1)
|
2(2.9)
|
0(0)
|
|
PMS2(-)
|
1(0.5)
|
1(1.4)
|
0(0)
|
|
pMMR
|
190(97.9)
|
67(95.7)
|
123(99.2)
|
|
Unknown
|
25
|
3
|
22
|
|
Adjuvant therapy, n (%)
|
|
|
|
<0.001
|
Yes
|
116(53.0)
|
54(74.0)
|
62(42.5)
|
|
No
|
103(47.0)
|
19(26.0)
|
84(57.5)
|
|
Values in parentheses are percentages
MMR, mismatch repair; IHC, Immunohistochemistry
* χ2 or Fisher’s exact test or t test
a 10-15cm from anal verge based on MRI
b 5-10cm from anal verge based on MRI
c -5cm from anal verge based on MRI
Survival
The overall median follow-up was 64.1 (range, 3–84.5) months. Among young-onset patients, the median OS was 36.7 (3.0–83.6) months, with 1-, 3-, and 5-year OS rates of 96%, 84%, and 78%, respectively. The equivalent values in the older group were 68 (7.2–84.5) months and 97%, 90%, and 82%, respectively (P=0.14) (Fig.2). Median PFS was 35.1 (2.5–83.6) months among young-onset patients and 66.2 (2.5–84.1) months among older patients. The PFS rates at 1, 3, and 5 years in the young-onset patients were 89%, 70%, and 70%, respectively, compared to 87%, 75%, and 70%, respectively, in the older patients (P=0.57) (Fig.2).
Young-onset female patients showed surprisingly worse outcomes than older females (P=0.008), and women had worse outcomes than men in the young-onset subgroup analysis (P=0.033). Among young-onset female patients, the median OS was 39.1 (4.1–83.6) months, with 1-, 3-, and 5-year OS rates of 94%, 77%, and 64.5%, respectively. The median PFS was 32.4 (2.5–83.6) months, with 1-, 3-, and 5-year PFS rates of 81%, 58%, and 58%, respectively (Fig.3). Young-onset patients had a higher proportion of vascular invasion (P=0.024) and a worse prognosis than those in older patients (Fig.3).
Prognostic factors
In the univariate analysis, the variables associated with PFS were BMI, cT4 status, cN2 status, cTNM IV stage, (y)pN1 status, (y)pN2 status, (y)pTNM IV stage, vascular invasion, and adjuvant therapy. In the univariate analysis, the variables associated with OS were cN1 status, cTNM IV stage, (y)pN1 status, (y)pN2 status, and (y)pTNM IV stage. In the multivariable analysis, (y)pN2 status (hazard ratio [HR] 4·505, 95% CI 1·141–17·784; P=0.032) was associated with PFS.
In the univariate analysis of the young-onset group, the variables associated with OS were sex, BMI, cTNM stage IV, and (y)pN2 status (Table 3). In the multivariable analysis of the young-onset group, sex (HR 3·924, 95% CI 1·252 12·294; P=0.019) was associated with OS (Table 4).
Table 3 Univariable logistic regression of factors predicting OS (≤50 years)
|
Hazard ratio
|
P
|
Age
|
0.989 (0.911, 1.074)
|
0.799
|
Sex
|
|
|
Male
|
1
|
Ref
|
Female
|
2.990 (1.037, 8.617)
|
0.043
|
Family cancer history
|
|
|
Negative
|
1
|
Ref
|
Positive
|
0.953 (0.328, 2.765)
|
0.929
|
BMI
|
0.847 (0.722, 0.993)
|
0.040
|
Clinical stage
|
|
|
cTNM I
|
1
|
Ref
|
cTNM II
|
2.290 (0.267, 19.604)
|
0.449
|
cTNM III
|
1.356 (0.163, 11.272)
|
0.778
|
cTNM IV
|
9.202 (1.020, 83.014)
|
0.048
|
Primary site
|
|
|
Higha
|
1
|
Ref
|
Middleb
|
0.305 (0.056, 1.674)
|
0.172
|
Lowc
|
1.511 (0.473, 4.828)
|
0.486
|
Neoadjuvant therapy
|
|
|
No
|
1
|
Ref
|
Yes
|
0.480 (0.136, 1.686)
|
0.252
|
(y)pT status
|
|
|
(y)pT0/pTis
|
1
|
Ref
|
(y)pT1+T2
|
0.507 (0.032, 8.129)
|
0.632
|
(y)pT3
|
2.464 (0.322, 18.874)
|
0.385
|
(y)pT4
|
3.144 (0.196, 50.403)
|
0.418
|
(y)pN status
|
|
|
(y)pN0
|
1
|
Ref
|
(y)pN1
|
0.729 (0.141, 3.776)
|
0.706
|
(y)pN2
|
4.372 (1.460, 13.096)
|
0.008
|
(y)Pathological stage
|
|
|
(y)pTNM 0
|
1
|
Ref
|
(y)pTNM I
|
0.781 (0.049, 12.525)
|
0.861
|
(y)pTNM II
|
1.499 (0.156, 14.428)
|
0.726
|
(y)pTNM III
|
1.827 (0.224, 14.890)
|
0.573
|
(y)pTNM IV
|
8.631 (0.985, 77.759)
|
0.055
|
Vascular invasion
|
|
|
Negative
|
1
|
Ref
|
Positive
|
2.387 (0.863, 6.601)
|
0.094
|
Adjuvant therapy
|
|
|
No
|
1
|
Ref
|
Yes
|
0.755 (0.262, 2.176)
|
0.602
|
Values in parentheses are 95 percent confidence intervals
a 10-15cm from anal verge based on MRI
b 5-10cm from anal verge based on MRI
c -5cm from anal verge based on MRI
Table 4 Multivariable logistic regression of factors predicting OS (≤50 years)
|
Hazard ratio
|
P
|
Sex
|
|
|
Male
|
1
|
Ref
|
Female
|
3.924 (1.252, 12.294)
|
0.019
|
BMI
|
0.867 (0.727, 1.035)
|
0.114
|
Clinical stage
|
|
|
cTNM I
|
1
|
Ref
|
cTNM II
|
2.577(0.259, 25.640)
|
0.419
|
cTNM III
|
1.438(0.139, 14.841)
|
0.760
|
cTNM IV
|
7.729 (0.542, 110.144)
|
0.131
|
(y)pN status
|
|
|
(y)pN0
|
1
|
Ref
|
(y)pN1
|
0.840(0.150, 4.711)
|
0.843
|
(y)pN2
|
2.499 (0.648, 9.645)
|
0.184
|
Values in parentheses are 95 percent confidence intervals