Baseline Characteristics
A total of 189 consecutive patients underwent surgery between 2002 and 2019. In 95 patients, liver resection was performed with curative intention, whereas 94 (49%) patients underwent an explorative laparotomy due to peritoneal carcinosis, reduced liver function, or major tumor extension.
The baseline characteristics of our patients are summarized in Table 2. The median age of the study population was 64.9 ± 10.2 years, and 54.7% (n = 52) of the patients were male. The mean preoperative total bilirubin levels before operation were 44.2 ± 11.1 µmol/l. Preoperative biliary drainage was performed in 73 patients (77%), with endoscopic biliary drainage as the most common procedure (n = 64; 67%). Preoperative portal vein embolization was performed in 21 patients (n = 22.1%).
Major hepatectomy was performed in 91 patients (96%), with left hepatectomy in 11 (12%), right hepatectomy in 10 (11%), and a left and right trisectionectomy in 16 (17%) and 54 (57%), respectively. Portal vein resection was performed in 64 patients (67.3%), whereas a caudate lobe resection was performed in 67 patients (70.5%).
The most common types of Bismuth-Corlette classification were Type IV (n = 49; 51.6%), but 6 patients (6.3%) had type I, 8 patients (8.4%) had type II, and 23 patients (24.2%) had type III. In total, 9 patients (9.5%) received neoadjuvant chemotherapy and 26 patients (27%) received adjuvant chemotherapy.
Pathologic Findings
All resected specimens were confirmed as adenocarcinomas. Negative resection margins (R0) were achieved in 68 patients (71.6%) (Table 2). Microvascular invasion was found in 24 (25.3%) patients and perineural invasion in 59 (62.1%) patients. Invasion of small lymphatic vessels was observed in 68 patients (71.6%).
Pathology revealed poorly differentiated tumors in most patients (n = 48, 50.5%) and either well differentiated (n = 5, 5.3%) or moderate tumors (n = 42, 44.2%) in the remainder. Almost one out of three patients had lymph node metastases (n = 27, 28.4%) and approximately one in ten patients had metastatic disease at diagnosis (n = 8, 8.8%).
The median number of lymph nodes harvested was 4 (IQR 3-8), and the median number of metastatic lymph nodes was 1 (IQR 0-2).
Stage transitions
The 6th edition of the AJCC staging categorized 5 (5.3%) patients as TNM stage Ia, 18 (18.9%) as stage Ib, 38 (40.4%) as stage IIa, 20 (21.1%) as stage IIb, 6 (6.3%) as stage III, and 8 (8.4%) patients as stage IV.
The 7th edition of the AJCC staging categorized 5 (5.3%) patients as TNM stage I, 52 (54.7%) as stage II, 4 (4.2%) as stage IIIa, 18 (18.9%) as stage IIIb, 5 (5.3%) as stage IVa, and 11 (11.5%) patients as stage IVb.
The 8th edition of the AJCC staging categorized 5 (5.3%) patients as stage I, 52 (54.7%) patients as stage II, 4 (4.2%) as stage IIIa, 3 (3.1%) as stage IIIb, 18 (18.9%) as stage IIIc, 5 (5.3%) as stage IVA, and 8 (8.4%) patients as stage IVB.
Table 3 shows a cross-tabulation of stage distributions and transitions for the AJCC stages for the 6th, 7th and 8th editions. Comparing the 6th edition with the 7th edition, stage migration was observed in a total of 45 (47.4%) patients, based on changes in major group definitions (Table 3). Comparing the 7th with the 8th edition, 25 patients (26.3%) were reclassified when considering substages (e.g., stage IIIa and IIIb) and 23 patients (25%) when considering only the major stages (i.e., stage I, II, III, or IV).
Staging according to the 8th edition upstaged 18 patients (18.9%) and downstaged 6 patients (6.3%) when compared with the 7th edition. A total of 18 patients (18.9%) with N1 disease (stage IIIb) in the 7th edition were upstaged to IIIc (N = 15 patients; 1–3 positive lymph nodes positive) or IVa (n = 3 patients; >4 or more positive lymph nodes) in the 8th edition. Most patients with T4 disease (n = 5 patients; stage IVa) in the 7th edition were downstaged to IIIb (n = 3 patients) or IIIc (n = 1 patient) in the 8th edition.
Univariate and multivariate analysis of factors associated with OS after curative resection
The median follow-up time of patients after liver resection was 4.3 ± 2.9 years.
The overall median survival time of the entire patient cohort was 21 months (95% CI 8.1–33.9) and the survival rates for 1, 3, and 5 years were 60.4%, 46.1%, and 36.2%, respectively. Tumors recurred in 34 (37.4%) patients after a median disease-free survival time of 35 (95% CI: 25.9–54.4) months.
We tested several clinical and pathological factors on OS. As shown in Table 4, OS was significantly longer in female patients (p = 0.02), patients with no extrahepatic distant metastases (p < 0.001), those with negative resection margins (p = 0.003), those aged < 65 years (p = 0.03), those with good/moderate tumor differentiation (p = 0.01), and patients with no microvascular invasion (p < 0.01).
Lymph node status was strongly associated with OS. Patients who were lymph node negative had a 5-year overall survival of 33.3% compared with 24.3% and 0% for patients with 1–3 and >3 metastatic lymph nodes (p < 0.01).
Portal vein resection was not associated with significantly better overall survival (p > 0.05), but patients who underwent caudate resection had a 5-year OS of 28.7% compared with 17.1% for those who did not undergo caudate lobe resection (p = 0.004).
The multivariate analysis (Table 4) revealed that gender (male versus female; HR = 2.6, 95% CI: 1.4–4.9; p < 0.01), age (> 65 years versus <65 years; HR = 1.9, 95% CI: 1.2–3.2; p = 0.04), margin status (R1 versus R0; HR = 2.1, 95% CI: 1.2–3.5; p = 0.01), distant metastases (M1 versus M0; HR = 4.2, 95% CI: 1.1–10.8; p = 0.002), metastatic lymph nodes (1-3 MLN versus 0MLN, HR = 1.1, 95% CI: 0.5–2.1; p = 0.04; > 3MLN versus 0 MLN; HR = 3.5, 95% CI: 1.4–9.1; p < 0.01), tumor differentiation (G3 versus G1; HR = 7.9, 95% CI: 1.1.-58.1; p = 0.04) and caudate lobe resection (no versus yes; HR 2.2, 95% CI: 1.2–4.1; p < 0.01) remained associated as independent predictors of OS.
Survival Across Stages
Overall, an unbalanced distribution was observed that favored stage IIa (40%) in the 6th edition and stage II (54.7%) in 7th and 8th editions. Table 6 present stage-specific median OS values for 1-, 3- and 5-year survival in patients staged according to the 6th, 7th and 8th editions of the AJCC staging system.
The median OS for patients staged according to the 6th, 7th and 8th edition per TNM classification were as follows: stage I (6th, 7th, and 8th edition: 58 months [95% CI: 48.8–67.2] versus 62 months [95% CI: 34.4–89.6] versus 62 months [95% CI: 34.4–89.6]), stage II (6th, 7th and 8th edition: 37 months [95% CI: 21.4–52.6] versus 43 months [95% CI: 19.5–66.5] versus 43 months [95% CI: 19.5–66.5]), stage III (6th, 7th, and 8th edition: 28 month [95% CI: 8.8–47.2] versus 37 [95% CI: 1.6–72.4] versus 38 [95% CI: 24.6–51.4]) and stage IV (6th, 7th and 8th edition: 12 months [95% CI: 0–24.5] versus 22 [95% CI: 7.6–36.4] versus 16 [95% CI: 10.1–21.9]), respectively (p-values between stages in the 6th edition = 0.039 vs. in the 7th = 0.010 versus in the 8th = 0.002).
Figures 1 and 2 show the Kaplan-Meier curves for OS for the main stages and sub-stages of the 6th, 7th, and 8th TNM editions. According to the 6th AJCC edition, we found that patients in stage IA and IB had 5-year OS rates of 80% and 37.5%, respectively, whereas patients in stage IIA and IIB had similar 5-year OS rates of 32.6% and 25%, respectively. Notably, patients in stage III had a 5-year OS rate of 16.7%. None of the patients in stage IV survived 5 years after operation.
Using the AJCC 7th edition, we found that patients in stage I, II, and IVa had a 5-year OS rates of 80%, 33.2%, and 20.0%, respectively. Notably, patients in stage IIIa and IIIb had 5-years OS rates of 50% and 14.3%, respectively (Table 6). No patients in stage IVb survived 5 years after operation.
Use of the AJCC 8th edition revealed 5-year OS rates for patients in stages I and II of 80% and 33.2%, respectively, which were similar to the rates for the 7th AJCC edition. Patients in stages IIIa, IIIb, and IIIc had 5-year OS rates of 50%, 33.3%, and 28%, respectively. No patients in stage IVa or IVb survived 5 years after operation
Use of the 6th edition resulted in no differences in risk of death for patients in stages Ib, IIIa, IIb, and III (both P > .05) when compared with patients in stage Ia. By contrast, when compared with patients in stage Ia, patients in stage IV had the greatest risk of death (stage IV versus stage Ia, HR 8.41, 95% CI 1.74–40.57, p < 0.01).
Use of the AJCC 7th edition revealed no differences in risk of death between patients in stage I and those in stages II, IIIa, IIIb and IVa (p > 0.05; Table 6). However, when compared with patients in stage I, those in stages IVb had the greatest risk of death (stage IVb versus stage I, HR 7.05, 95% CI 1.54–32.36, p = 0.01).
Use of the AJCC 8th edition revealed no differences in risk of death for patients in stages II, IIIa, IIIb, and IIIc (p > 0.05) when compared with patients in stage I. However, when compared with patients in stage I, the risk of death was greater for patients in stage IVa and IVb (stage IVa versus stage I, HR 6.36, 95% CI 1.22–33.24, P = 0.028; stage IVb versus stage I, HR 8.66, 95% CI 1.79–41.83, p < 0.01).
Prognostic accuracy
Table 7 presents the ROC analysis and Hoshmer- Lemeshow test (HLT) for discrimination and calibration of all the evaluated staging systems. The prognostic accuracy was remarkably similar across the three analyzed AJCC staging systems. The prognostic accuracy of the main stages was higher for the 8th edition than for the 7th edition (AUC: 8th edition versus 7th edition: 0.66 versus 0.61) but the difference did not reach statistical significance. Expanding the 7th edition to include substages slightly diminished its prognostic accuracy (AUC: from 0.61 to 0.58), whereas, expansion of the 8th edition with substages increased the prognostic accuracy (AUC main stages versus substages: 0.66 to 0.69) and a statistical trend was observed (p = 0.09). With regard to calibration of the evaluated stagings systems, the HLT shows acceptable calibration and fitness of the analyzed 7th and 8th staging systems (Table 8).