Demographic characteristics
In the systematic review, 20 articles revealed no significant association between age and the retrieval of LNs in colorectal cancer.21–23 26 27 30 32 35 37–41 47–49 54 55 64 65 However, 18 articles indicated that more LNs were obtained from young patients than elderly patients (online supplemental table 6).13 14 25 29 31 33 34 42 44 46 50 52 60–63 67 68 Five articles involving 16 184 patients with colorectal cancer included in the meta-analysis showed that fewer LNs were retrieved from elderly patients (aged≥65 years) than young patients (aged<65 years) (OR=0.70, 95% CI (0.54, 0.90), p=0.005, I2=43.96%) (figure 2 and online supplemental figure 1).12 51 53 59 66
The forest of factors associated with the number of lymph node yield in colorectal cancer. BMI, body mass index.
The analysis of gender and retrieved LNs in colorectal cancer using systematic review included ten colorectal cancer,21 26 27 31 32 36 37 39 41 51 five colon cancer22 30 35 44 47 and five rectal cancer studies.23 38 42 54 65 These studies showed that gender had no significant association with the LN yield. However, two colorectal cancer studies,29 33 one colon cancer study25 and one rectal cancer study45 showed that more LNs were retrieved from females than males (online supplemental table 6). One study revealed that gender was associated with the LN yield but did not indicate whether the LN yield was higher in females or males.49 Furthermore, the meta-analysis including seven colon cancer studies with 110 147 patients showed that more LNs were obtained from females than males (OR=1.12, 95% CI (1.00 to 1.26), p=0.045, I2=80.30%) (figure 3 and online supplemental figure 2).12 14 52 55 56 60 66 Finally, 16 colorectal cancer studies that contained 125 269 patients also revealed that more LNs were obtained from females than males (OR=1.15, 95% CI (1.04 to 1.28), p=0.006, I2=67.06%) (figure 2 and online supplemental figure 3).12–14 48 51–53 55 56 58 60 63 64 66–68
The forest of factors associated with the number of lymph node yield in colon cancer.
Although eight articles in the systematic review showed that BMI was not associated with the number of retrieved LNs,20 22 23 28 31 35 55 64 the other two articles revealed that high BMI was associated with fewer retrieved LNs (online supplemental table 6).26 30 Furthermore, the meta-analysis including 7 studies with 34 351 patients revealed no significant differences in the overweight and obesity groups compared with the normal-weight group in colorectal cancer (OR=0.93, 95% CI (0.87 to 1.00), p=0.054, I2=0.00%) (figure 2 and online supplemental figure 4).12 14 48 51 66–68
Three articles included in the meta-analysis revealed no significant difference in the LN yield between black and white patients (n=46 983, OR=1.09, 95% CI (0.94 to 1.26), p=0.245, I2=29.23%), as well as between other ethnic groups and white patients (n=52 136, OR=0.99, 95% CI (0.98 to 1.00), p=0.059, I2=0.00%) (figure 2 and online supplemental figures 5 and 6).14 55 62
Tumour-related characteristics
In the systematic review, six articles showed that tumour location was not significantly associated with the LN yield in colorectal cancer25 26 35 40 64 65; however, four articles revealed association, without explaining the specific association.5 34 36 63 Ten articles revealed that more LNs could be obtained from the right colon than the other locations.21 22 25 29 31 37 39 41 44 48 As per the location of the tumour in the rectum, although three studies suggested that more LNs could be obtained from the upper rectum than the lower rectum,31 62 68 three studies found no significant association between rectal location and LN yield (online supplemental table 6).23 38 42 The meta-analysis of eight studies containing 47 706 patients with colon cancer showed that fewer LNs were obtained from tumours located in the left colon than in the right colon (OR=0.43, 95% CI (0.33 to 0.56), p<0.001, I2=88.27%) (figure 3 and online supplemental figure 7).12–14 49 56 58 60 66 Likewise, the same eight studies that included 52 893 patients with colorectal cancer showed that fewer LNs were obtained from tumours located in other locations than in the right colon (OR=0.51, 95% CI (0.45 to 0.57, p<0.001, I2=38.47%) (figure 2 and online supplemental figure 8). Four studies with 6839 patients showed that fewer LNs were obtained from rectal cancer than colon cancer (OR=0.56, 95% CI (0.47 to 0.67), p<0.001, I2=7.58%) (figure 2 and online supplemental figure 9).13 51 61 63 Studies have indicated that neoadjuvant treatment led to LN shrinkage, making their detection and recognition more challenging.69 70 Three of the aforementioned four studies excluded patients with neoadjuvant treatment and one did not. Considering the widespread use of neoadjuvant treatment for rectal cancer, sensitivity analysis was performed by excluding the one study, and the results revealed stable findings (OR=0.59, 95% CI (0.41 to 0.85), p=0.005) (online supplemental figure 10) Therefore, neoadjuvant treatment had no impact on the outcome that fewer LNs were obtained from rectal cancer than colon cancer.
In the systematic review of association between the T stage and LN yield in colorectal cancer, 12 studies mentioned that the T stage had no significant association with the LN yield,23 26 29 30 32 35 39 40 42 44 47 54 but eight articles revealed that more LNs were obtained from the advanced T stage than early T stage.21 25 27 28 37 41 51 62 Moreover, four articles mentioned that the T stage was associated with the harvested LNs,19 24 36 48 without explaining the specific association. In the meta-analysis, seven studies containing 23 297 patients with colorectal cancer revealed higher LN yield in the T3+T4 group than the T1+T2 group (OR=1.57, 95% CI (1.25 to 1.97), p<0.001, I2=70.46%) (figure 2 and online supplemental figure 11).13 56–59 66 67 Likewise, four studies with 21 293 patients with colon cancer showed higher LN yield in the T3+T4 group than in the T1+T2 group (OR=1.41, 95% CI (1.09 to 1.82), p=0.008, I2=68.36%) (figure 3 and online supplemental figure 12).56 57 59 66 Five studies containing 22 254 patients with colorectal cancer showed that more LNs were obtained from the T2 group than the T1 group (OR=1.88, 95% CI (1.26 to 2.79), p=0.002, I2=95.09%) (figure 2 and online supplemental figure 13).12 14 50 61 63 In addition, 4 studies containing 44 563 patients with colorectal cancer showed that more LNs were obtained from the T3 group than the T1 group (OR=2.96, 95% CI (1.50 to 5.85), p=0.002, I2=98.26%) (figure 2 and online supplemental figure 14),12 14 50 63 and 4 studies containing 10 211 patients with colorectal cancer showed that more LNs were obtained from the T4 group than the T1 group (OR=3.95, 95% CI (2.14 to 7.27), p=0.002, I2=92.50%) (figure 2 and online supplemental figure 15).12 14 49 63 Three studies containing 10 211 patients with colon cancer revealed more LN yield in the T4 group than in the T1 group (OR=4.26, 95% CI (2.09 to 8.68), p<0.001, I2=94.94%) (figure 3 and online supplemental figure 16),12 14 49 similar to that observed in patients with colorectal cancer.
In the systematic review of association between the N stage and LN yield in colorectal cancer, six studies mentioned that the N stage had no significant association with the LN yield,21 28 35 38 39 57 whereas the other six studies indicated that the N status played an essential role in the number of retrieved LNs (online supplemental table 6).23 26 33 36 41 48 In the meta-analysis, 5 studies containing 21 508 patients showed that more LNs could be obtained from the N2 group than the N0 group (OR=1.32, 95% CI (1.15 to 1.51), p<0.001, I2=0.72%) (figure 2 and online supplemental figure 17).14 54 55 61 63 However, four studies containing 45 429 patients showed that the LN yield was not significantly different between the N1 and N0 groups (OR=0.94, 95% CI (0.75 to 1.16), p=0.557, I2=83.40%) (figure 2 and online supplemental figure 18),14 55 62 63 and nine studies containing 31 629 patients showed no significant difference between the N1+N2 and N0 groups (OR=1.42, 95% CI (0.99 to 2.04), p=0.060, I2=92.64%) (figure 2 and online supplemental figure 19).13 14 51 53 55 56 58 63 67
In the meta-analysis, 4 studies containing 29 013 patients with colorectal cancer showed that the LN yield was not significantly different between the M1 and M0 groups (OR=1.04, 95% CI (0.76 to 1.43), p=0.790, I2=56.55%) (figure 2 and online supplemental figure 20).13 55 57 63
In the systematic review, 4 studies indicated that the tumour size was not significantly associated with the LN yield in colorectal cancer,38 39 51 66 whereas 13 studies suggested that larger tumour size was associated with more LN yield.21 25 26 29 33 41 45 56 58–60 63 Moreover, four studies showed that the tumour size was associated with the LN yield but did not specify the association (online supplemental table 6).23 34 35 37
Five studies in the systematic review suggested that the specimen length was not associated with the LN yield in colorectal cancer,30 38 51 53 58 whereas five studies showed that longer specimens were associated with higher LN yield.21 39 43 56 66 Moreover, five studies indicated that the specimen length was associated with the LN yield but did not specify the association (online supplemental table 6).27 29 35 37 45
There were no consistent conclusions in the systematic review with respect to the association between tumour histology and LN yield in colorectal cancer. Two studies showed that mucinous adenocarcinoma and adenocarcinoma were not significantly different with regard to LN yield,59 62 whereas one study showed that more LNs were obtained from mucinous adenocarcinoma.42 Two studies revealed no significant difference in the LN yield between signet ring type adenocarcinoma and adenocarcinoma,42 59 whereas another study showed that more LNs could be retrieved from signet ring type adenocarcinoma than adenocarcinoma.62 Finally, one study showed that more LNs were obtained from poorly differentiated, mucous and signet ring carcinoma than adenocarcinoma (online supplemental table 6).63
In the systematic review, nine studies indicated no significant association between the degree of tumour differentiation and number of retrieved LNs.21 29 39 41 47 53 54 59 65 However, three studies suggested that more LNs could be obtained from poorly differentiated cancer.43 50 63 Moreover, four studies suggested that tumour differentiation was associated with the LN yield but did not specify the association (online supplemental table 6).19 22 33 37 In the meta-analysis, 5 studies involving 1937 patients showed no significant difference between the G3 (high-grade) and G1+G2 (low-grade) groups with regard to the LN yield (OR=1.76, 95% CI (0.53 to 5.85), p=0.357, I2=90.89%) (figure 2 and online supplemental figure 21).51 53 58 66 67
In the systematic review, one study mentioned that more LNs were obtained from positive lymph–vascular invasion group than the negative group in colorectal cancer,22 whereas two studies revealed the opposite result.25 31 Five studies mentioned that lymph–vascular invasion had no significant association with the LN yield in colorectal cancer (online supplemental table 6).21 36 39 41 47 In the meta-analysis of 8 studies with 3813 patients, the LN yield was not significantly different between the positive and negative groups (OR=1.13, 95% CI (0.84 to 1.53), p=0.419, I2=36.98%) (figure 2 and online supplemental figure 22).51 53 58 63 65–68
This post was originally published on https://bmjopen.bmj.com