Abstract
1.1: Background
Transarterial radioembolization (TARE) with 90Y is a primary therapy for hepatocellular carcinoma (HCC), including patients with portal vein thrombosis (PVT). Patient survival subgroups post TARE have not been clearly defined.
1.2. Aims
To identify prognostic factors in patients treated with TARE.
1.3. Methods
Non-surgical HCC patients (n=116) were treated with TARE and prospectively followed for survival and changes in blood parameters relating to inflammation and tumor indices.
1.4. Results
Baseline parameters with the highest hazard ratios (HR>2) for survival were C-reactive protein, alpha-fetoprotein (AFP), maximum tumor diameter (MTD) and PVT. A 2-parameter prognostic model was developed, consisting of serum AFP and C-reactive protein, which was especially significant in patients with PVT. Temporal changes post TARE in ALBI and serum gamma glutamyl transferase (GGT) levels occurred in patients with both high and low serum AFP. A network phenotyping strategy (NPS) showed that patient phenotypes also change after TARE. Post TARE reduction in AFP or GGT or increase in albumin levels were associated with improved survival.
1.5. Conclusions
Changes in serum CRP, AFP, GGT, albumin, ALBI, and NPS phenotype occurred early after TARE. Changes in AFP, GGT and albumin related to survival and may be useful in selecting patients for additional treatment.
Keywords: HCC, Radioembolization, Prognosis
2. Introduction
Transarterial radioembolization with (TARE) Yttrium-90 (90Y) has increasingly become part of the therapeutic armamentarium for locoregional therapy for patients with non-metastatic hepatocellular carcinoma (HCC) in recent years, owing to its effectiveness, relatively low toxicity [1–3] and comparable or better response and survival rates to transarterial chemoembolization (TACE) [3–5]. It is also relatively safe compared with TACE in patients with portal vein thrombosis (PVT) [6,7]. Despite the burgeoning literature on its use, precise knowledge is still lacking concerning its mechanisms, predictors of response and resistance, optimal uses and combinations with other treatment modalities. In countries such as Turkiye with a developing economy, many patients are too poor for repetitive transport to tertiary medical centers such as ours and the availability of cheap and useful HCC biomarkers would be beneficial in their treatment follow-up. In this paper, we have explored several potential seromarkers, especially for the large proportion of HCC patients with low baseline levels of alpha-fetoprotein (AFP).
3. Methods
One hundred and sixteen patients with unresectable and untransplantable hepatocellular carcinoma (HCC) were treated with TARE and followed for changes in serum HCC and inflammation parameters.
Demographic and clinical characteristics of the HCC patients analyzed for this study were: age, gender, baseline and last alpha-fetoprotein (AFP), liver function and complete blood count values, maximal baseline tumor diameter, and presence or absence of macroscopic portal vein invasion by tumor (PVT) on the CAT scan. ALBI grade was calculated by use of: www.mdcalc.com ALBI grade for HCC. NPS phenotype was calculated by use of the NPS web tool: https://apkatos.github.io/webpage_nps.
A diagnosis of HCC was established by percutaneous biopsy or noninvasively based on presence of a hepatic mass greater than 2 cm diameter with characteristic imaging findings in the setting of liver cirrhosis as per EASL guidelines. Treatment was performed on an inpatient basis, and patients were typically not discharged home for 24 hr after the procedure for post-TARE safety observations.
Inclusion criteria included patients with HCC who were considered unamenable to surgical treatment, age >18 years, absence of metastases, lung shunt <20%, prior informed consent for the radioembolization procedure and treatment; ECOG performance status <3 and adequate hematologic, renal and liver function tests for a safe interventional procedure.
4. Radioembolization Procedures
Patients were treated as per published guidelines (8–10) and underwent pre-treatment angiographic mapping of the abdominal aorta and hepatic arteries. Planar scans of the lung and liver area in anterior and posterior views were acquired after injection of 99mTc-labelled albumin macroaggregated albumin (99mTc-MAA) into selected arterial branches followed by SPECT (until 2006) or SPECT/CT scans. Radioembolization was delivered using Yttrium-90 resin microspheres (SIR-Spheres; Sirtex Medical Limited, North Sydney, NSW, Australia) or glass microspheres (Therasphere, Boston Scientific, Boston MA, USA). Dose calculation performed through personalized dosimetry planning, considering optimal absorbed doses by tumoral and non-tumoral volumes, was used by the treating interventional radiologist.
5. Statistical Methods
The Shapiro-Wilk test was used to assess the normality of the quantitative data. Comparisons between two independent groups were performed by the Mann-Whitney U test. Median, interquartile range, minimum, and maximum values were used to summarize the quantitative data. Distribution of the qualitative data was presented by count and percentage, and the continuity-corrected chi-square test was applied for comparisons. The Kaplan-Meier method was used for survival estimations, and the Log-Rank test was used for survival comparisons between groups. Univariate Cox regression analysis was used to obtain the Hazard Ratios. In all analyses, the two-sided significance level was considered as <0.05. IBM SPSS Statistics for Windows, Version 26.0 (Armonk, NY: IBM Corp.) was used for statistical analyses.
The study protocol was approved by the ethical committee of the Inonu University institutional review board (IRB) for non-interventional studies (2025/8125) for data collection and analysis, and conducted in adherence to the Declaration of Helsinki. Since this was a non-intervention retrospective study, patient informed consent was not required for the study evaluation, although each patient signed an informed consent for the procedure.
6. Results
6.1. Survival in The Total Cohort and in Parameter Subgroups
Survival in this TARE-treated patient cohort was calculated by Kaplan-Meier analysis (Table 1a). Mean survival was 19.95 +1.59 months and median survival was 13.93 + 3.24 months. We and others have previously reported that HCC prognosis depends on both tumor and liver factors, especially liver inflammation [11–12]. Accordingly, these factors were dichotomized as reported previously [13–15] and examined for their relationship to survival (Table 1b). The dichotomized parameters with the highest hazard ratios were, in descending order, C-reactive protein or CRP (HR 2.787), maximum tumor diamter or MTD (HR 2.512), alpha-fetoprotein or AFP (HR 2.149), portal vein thrombosis or PVT (HR 2.032), ALBI (HR 1.978) and albumin (HR 1.857). For each of these dichotomized parameter pairs, survival was 3–4 fold different.
Table 1:
Kaplan-Meier analysis and survival comparisons by parameter.
| Survival (mo.) Mean±SE | Survival (mo.) Median±SE | |
|---|---|---|
| All (n=116) | 19.95±1.59 | 13.93±3.24 |
| Survival (mo.) Mean±SE | Survival (mo.) Median±SE | Log-Rank p-value | HR (95% C.I.) | HR p-value | |
|---|---|---|---|---|---|
| Albumin≥3.5 (n=84) | 22.17±1.84 | 18.70±3.04 | 0.010 | reference | |
| Albumin<3.5 (n=32) | 13.27±2.57 | 5.33±0.61 | 1.857 (1.151–2.996) | 0.011 | |
| ALBI grade 1 (n=50) | 24.48±2.31 | 22.70±5.34 | 0.004 | reference | |
| ALBI grade ≥2 (n=66) | 16.01±1.95 | 7.40±2.96 | 1.978 (1.239–3.157) | 0.004 | |
| PVT− (n=60) | 24.07±2.15 | 20.87±6.79 | 0.002 | reference | |
| PVT+ (n=56) | 14.77±2.02 | 7.77±2.79 | 2.032 (1.288–3.206) | 0.002 | |
| AFP<100 (n=57) | 24.84±2.23 | 22.70±3.79 | 0.001 | reference | |
| AFP≥100 (n=55) | 14.27±2.03 | 7.77±2.35 | 2.149 (1.351–3.418) | 0.001 | |
| MTD≤5 cm (n=48) | 26.49±2.38 | 30.43±NA | <0.001 | reference | |
| MTD>5 cm (n=68) | 14.52±1.76 | 7.90±2.20 | 2.512 (1.541–4.095) | <0.001 | |
| CRP≤1 (n=49) | 23.60±2.24 | 22.70±4.52 | <0.001 | reference | |
| CRP>1 (n=32) | 11.35±2.49 | 5.53±0.81 | 2.787 (1.625–4.780) | <0.001 | |
| Hb≥13 (n=77) | 20.90±1.81 | 18.70±3.41 | 0.186 | reference | |
| Hb<13 (n=39) | 17.65±2.88 | 6.43±1.62 | 1.369 (0.858–2.184) | 0.188 | |
| Lymphs <1.5 (n=71) | 20.87±2.08 | 15.33±4.01 | 0.492 | reference | |
| Lymphs ≥1.5 (n=45) | 18.03±2.31 | 13.20±4.21 | 1.171 (0.745–1.843) | 0.493 | |
| NE≤3.5 (n=55) | 24.12±2.39 | 20.87±4.90 | 0.014 | reference | |
| NE>3.5 (n=61) | 16.09±1.91 | 9.90±3.96 | 1.750 (1.111–2.756) | 0.016 |
6.2. A Model for Survival Using Parameter Combinations.
A model was constructed for combinations of the single parameters with highest HRs (Table 2). The patient combination in the total cohort of CRP≤1 & MTD≤5cm & AFP<100 (low CRP plus low MTD plus low AFP) had a median survival of 32.03 months, whereas the combination of CRP>1 & MTD>5cm & AFP≥100 (high CRP plus high MTD plus high AFP) had a median survival of 3.83 +1.68 months (log rank p<0.001) and HR 10.3. Using only the 2 parameters of CRP and MTD, survival differences were also significantly different, with an HR of 4.3. In the clinically important group of poor prognosis patients with PVT, the 2 parameter combination of CRP plus AFP also was significantly associated with 2 prognostic groups, with median survival of 22.70±2.71 and 4.27±0.54 months respectively, for patients who had low serum CRP plus low AFP values versus high serum CRP plus high AFP values, log rank p<0.001 and HR of 8.658. This is potentially useful in selecting patients with PVT who might benefit from future definitive HCC surgery or other therapies.
Table 2:
Survival comparisons in dichotomized parameter combination groups.
| Survival (mo.) Mean±SE | Survival (mo.) Median±SE | Log-Rank p-value | HR (95% C.I.) | HR p-value | ||
|---|---|---|---|---|---|---|
| All patients | CRP≤1 & MTD≤5cm & AFP<100 (n=17) | 26.30±3.63 | 32.03±NA | <0.001 | reference | |
| CRP>1 & MTD>5cm & AFP≥100 (n=13) | 4.11±0.81 | 3.83±1.68 | 10.301 (3.201–33.149) | <0.001 | ||
| All patients | CRP≤1 & AFP<100 (n=28) | 23.09±2.85 | 22.70±3.67 | <0.001 | reference | |
| CRP>1 & AFP≥100 (n=18) | 6.43±1.36 | 4.27±0.78 | 4.309 (1.980–9.380) | <0.001 | ||
| PVT negative | CRP≤1 & AFP<100 (n=18) | 23.04±3.72 | 18.73±11.10 | 0.885 | reference | |
| CRP>1 & AFP≥100 (n=3) | 13.12±4.64 | NA | 0.858 (0.108–6.806) | 0.885 | ||
| PVT positive | CRP≤1 & AFP<100 (n=10) | 19.48±2.90 | 22.70±2.71 | <0.001 | reference | |
| CRP>1 & AFP≥100 (n=15) | 5.14±1.11 | 4.27±0.54 | 8.658 (2.397–31.274) | 0.001 | ||
| PVT positive | AFP<100 (n=24) | 22.74±3.14 | 20.87±4.62 | <0.001 | reference | |
| AFP≥100 (n=31) | 7.64±1.29 | 4.97±0.59 | 3.852 (1.902–7.800) | <0.001 | ||
| PVT positive | CRP≤1 (n=17) | 18.94±2.37 | 22.70±2.32 | <0.001 | reference | |
| CRP>1 (n=23) | 7.75±1.86 | 5.33±0.85 | 4.333 (1.908–9.836) | <0.001 |
Abbreviations: as in Table 1.
6.3. Time-Dependent Serum Parameter Changes Post TARE
Figure 1 shows changes in several parameters at 3 and 6 months post TARE. For the prognostically important ALBI score, 33.9% of patients had improved (lower) scores, 12.6% had worse scores (higher), while 53.4% of patients had unchanged scores at 3 months post TARE, compared with baseline levels. The findings were similar at 6 months. For total serum bilirubin, 20% of patients had higher (worse) serum levels, 18.6% had lower (improved) serum levels and 61.3% had stable serum levels at 3 months, compared with baseline. The changes at 6 months post TARE were similar. For serum albumin, 14.2% of patients had higher than baseline levels (improved), 30.3% of patients had lower than baseline levels (worsened) and 54.4% of patients had stable levels at 3 months, compared with baseline levels. Again, the results at 6 months post TARE were similar. Blood platelet levels were higher (impoved) compared with baseline values in 31.5% of patients, lower (worse) compared with baseline values in 23.2% of patients and unchanged in 45.3% of patients at 3 months post TARE. There were nearly identical findings at 6 months post TARE. For blood CRP levels, 36.8% of patients had higher (worse) values at 3 months compared with baseline, 36.8% of patients had lower (improved) values compared with baseline and 26.3% of patients had stable values. There were insufficient 6 month survivors with CRP data.
Figure 1:
Serum parameter changes with time post TARE.
6.4. Changes in ALBI, AFP, GGT, Albumin and NPS Phenotype Post TARE
Owing to the importance of the ALBI score in HCC prognostication [16,17], ALBI changes were examined separately in patients with high (>100 IU/mL) or low serum AFP levels (Table 3 [A1] and [A2]. For patients who had high serum AFP levels, ALBI was decreased from grade 3 to 2 or 1, or from grade 2 to 1 (improved) in 32.7% of patients had a decrease (improved), increased from grade 1 or 2 (worsened) in 12.7% of patients and stable in 54.5 % of patients at 3 months post TARE. Very similar findings were seen in patients with low serum AFP levels, showing that the ALBI score is useful in HCC patient groups with both high and low serum AFP levels.
Table 3:
(A-E). ALBI, AFP, GGT and NPS changes at 3 months post TARE.
| A (1). ALBI changes in high AFP patients n (%) | A (2). ALBI changes in low AFP patients n (%) | |
|---|---|---|
| Decreased | 18 (32.7) | 19 (39.6) |
| Increased | 7 (12.7) | 7 (14.6) |
| Stable | 30 (54.5) | 22 (45.8) |
| n (%) | |
|---|---|
| Decreased | 37 (67.3) |
| Increased | 12 (21.8) |
| Stable | 6 (10.9) |
| Changes in AFP levels | Survival (mo.) Mean±SE | Survival (mo.) Median±SE | Log-Rank p-value | HR (95% C.I.) | HR p-value |
|---|---|---|---|---|---|
| Decreased >50% (n=17) | 34.03±3.17 | NA | <0.001 | reference | |
| Decreased ≥20% – ≤50% (n=9) | 24.43±5.35 | NA | 2.157 (0.609–7.645) | 0.223 | |
| Decreased <20% – increased (n=11) | 10.17±2.58 | 5.53±0.95 | 6.855 (2.369–19.840) | 0.012 |
| Low baseline AFP (<100) | High baseline AFP (≥100) | |
|---|---|---|
| GGT changes n (%) | GGT changes n (%) | |
| Decreased | 15 (34.9) | 10 (32.3) |
| Increased | 21 (48.8) | 17 (54.8) |
| Stable | 7 (16.3) | 4 (12.9) |
| Changes in GGT levels | Survival (mo.) Mean±SE | Survival (mo.) Median±SE | Log-Rank p-value |
|---|---|---|---|
| Decreased or stable (n=22) | 29.72±2.98 | 22.7±NA | 0.063 |
| Increased (n=21) | 16.36±4.49 | 10.9±5.85 |
| Changes in Albumin levels | (.Survival (mo Mean±SE | (.Survival (mo Median±SE | Log-Rank p-value | HR (.C.I 95%) | HR p-value |
|---|---|---|---|---|---|
| (Increased (n=13 | 34.76±4.04 | NA | 0.001> | reference | |
| (Stable (n=32 | 31.59±2.56 | NA | (0.402–3.984) 1.265 | 0.688 | |
| (Decreased (n=22 | 16.38±2.74 | 7.77±7.84 | (1.537–13.894) 4.622 | 0.006 |
| NPS changes n (%)** | |
|---|---|
| Decreased | 12 (27.3) |
| Increased | 18 (40.9) |
| Stable | 14 (31.8) |
| Baseline NPS phenotype | Response NPS phenotype |
|---|---|
| 25 ---> | 10 |
| 23 ---> | 18 |
| 23 ---> | 14 |
| 20 ---> | 8 |
| 18 ---> | 14 |
| 17 ---> | 13 |
| 17 ---> | 5 |
| 12 ---> | 5 |
% of patients with changes at 3 mo post TARE compared with baseline values; for AFP, this relates only to patients with baseline elevated AFP values. AFP response was a decrease ≥50%. ALBI grade decrease was a change from grade 3 to 2 or 1, or from grade 2 to 1, and conversely for an increase. Albumin decrease was a change in Child-Pugh class >3.5g/dL vs. 2.8–3.5 g/dL vs. < 2.8g/dL.
% of patients with changes in phenotype category at 3 mo post TARE in relation to baseline values, with highest phenotype being the poorest survival category).
NPS phenotype number. Highest (25) has worst survival; lowest (1) has best survival
AFP measurements have been found useful in monitoring responses in HCC patients who were treated with TARE and the levels have been reported to correlate well with tumor measurements]18,19]. AFP levels at 3 months post TARE in our cohort were decreased (improved) in 67.3% of those patients who had elevated baseline AFP levels, increased (worsened) in 21.8% of the patients who had elevated baseline AFP levels and were stable in 10.9% of patients with elevated baseline AFP levels, as shown in Table 3B (i). The responses in AFP levels were associated with changes in survival, as shown in Table 3B (ii), with a 50% decrease in AFP level being associated with 34.03±3.17 months survival, whereas a decrease of <20% or an increase was associated with a significantly different mean 10.17±2.58 months survival.
However, AFP cannot be used as a serum biomarker in that substantial percentage of HCC patients who have normal serum AFP levels at baseline [20,21]. For these patients with low baseline serum AFP levels, it has previously been shown that changes in serum GGT levels can be a useful biomarker [22]. GGT changes were found at 3 months in the patients of this cohort with low baseline AFP levels. In patients in this cohort with low baseline serum AFP levels, 34.9% of patients had a decrease in GGT levels (improved) at 3 months post TARE compared with baseline, 48.8% of patients had an increase (worsened) in GGT levels at 3 months post TARE compared with baseline and 16.3% of patients had stable GGT levels, as shown in Table 3C (i). Similar GGT changes were found in patients with high serum AFP levels. Patients who had a decrease or stable serum GGT levels at 3 months post TARE had a median survival of 22.7 months, compared to 10.9±5.85 months in patients who had an increase in GGT levels. However, the result did not quite reach significance, p=0.063, as shown in Table 3C (ii).
Changes in ALBI in relation to survival could not be calculated, due to the small patient numbers with high ALBI grades. However, we found changes in serum albumin levels that significantly related to survival (Table 3D). Patients who had an increased albumin at 3 months post TARE had a mean survival of 34.76+4.04 months, compared to a mean survival of 31.59+2.56 months in patients with stable albumin, and compared to a mean survival of 16.38+2.74 months in patients with a decreased albumin, p<0.001. The network phenotype strategy (NPS) was developed to integrate 17 standard clinical parameters into 25 prognostically-related HCC phenotypes, 1–25 with phenotype 25 being the most aggressive with the worst prognosis and phenotype 1 having the best prognosis (23, 24). NPS was used here for the first time to evaluate changes of HCC phenotype over time, as observed after TARE therapy (Tables 3E). We found that 27.3% of patients had a decrease in phenotype (less aggressive), 40.9% of patients had an increase in phenotype (more aggressive) and 31.8% of patients had stable phenotype, as shown in Table 3E (i). There was a wide range in phenotype response, depicted in 8 patients who started with the highest baseline phenotype, as shown in Table 3E (ii).
6.5. Clinical Characteristics of the Dichotomized Subgroups
The clinical characteristics of the dichotomised subgroups for the 4 parameters with highest HR values, namely CRP, AFP, MTD and PVT are shown in Table 4. The 2 CRP subgroups (Tables 4a and 4b) were significantly different with respect to MTD, AFP, percent of patients with PVT, tumor focality, serum albumin and levels of gamma glutamyl transferase (GGT). The 2 AFP subgroups (Tables 4c and 4d) significantly differed only for MTD, platelets and neutrophils. The 2 MTD subgroups (Tables 4e and 4f) were significantly different with respect to AFP, percent of patients with PVT, tumor focality, GGT, ALKP, AST, platelets and neutrophils. The 2 PVT subgroups (Tables 4g and 4h) were significantly different with respect to MTD, AFP, tumor focality, total bilirubin, CRP, GGT, ALKP, AST, platelets and neutrophils. Thus, each dichotomized pair had a slightly different pattern of clinical characteristics. Interestingly, only the CRP pair was significantly different for all 4 tumor characteristics that were studied, namely MTD, AFP, PVT and tumor focality. Total serum bilirubin was significantly different only in the PVT dichotomized pairs and the AFP dichotomized pairs had significant differences only for MTD, platelets and neutrophils.
Table 4:
Baseline parameter dichotomizations and clinical characteristics.
| CRP≤1 mg/dL | CRP>1 mg/dL | ||||||
|---|---|---|---|---|---|---|---|
| n | Median (IQR) | (Min.-Max.) | n | Median (IQR) | (Min.-Max.) | p | |
| MTD (mm) | 49 | 58 (52) | (12–126) | 32 | 94 (82.5) | (25–169) | 0.003 |
| AFP | 48 | 48.74 (654.56) | (1.26–76057) | 30 | 1203.89 (5617.67) | (2.22–54000) | 0.017 |
| Albumin | 49 | 4.2 (0.9) | (2.5–5.2) | 32 | 3.65 (1.12) | (2.4–4.7) | <0.001 |
| T.Bilirubin | 49 | 0.9 (0.56) | (0.2–2.2) | 32 | 0.95 (0.65) | (0.4–7.7) | 0.985 |
| GGT | 49 | 62 (83) | (13–320) | 32 | 157 (186.75) | (41–715) | <0.001 |
| ALKP | 49 | 103 (65) | (38–293) | 32 | 148 (167.25) | (66–509) | <0.001 |
| AST | 49 | 35 (24.5) | (16–290) | 32 | 68 (70.5) | (15–407) | <0.001 |
| ALT | 49 | 27 (19) | (10–160) | 32 | 44 (51.25) | (10–672) | 0.029 |
| Hb | 49 | 14.2 (3) | (7.9–17.1) | 32 | 12.8 (1.9) | (9.7–17.5) | 0.025 |
| PLT | 49 | 152 (112.5) | (65–486) | 32 | 158.5 (157.75) | (30–688) | 0.935 |
| LY# | 49 | 1.47 (0.77) | (0.42–3.15) | 32 | 1.23 (0.7) | (0.27–12.9) | 0.129 |
| NE# | 49 | 3.62 (1.85) | (0.91–13.93) | 32 | 4.46 (5.07) | (2.2–12.05) | 0.025 |
| CRP≤1 n (%) | CRP>1 n (%) | p | ||
|---|---|---|---|---|
| Multifocality | No | 21 (42.9) | 5 (15.6) | 0.020 |
| Yes | 28 (57.1) | 27 (84.4) | ||
| PVT | No | 32 (65.3) | 9 (28.1) | 0.002 |
| Yes | 17 (34.7) | 23 (71.9) |
| AFP<100 IU/mL | AFP≥100 iU/mL | ||||||
|---|---|---|---|---|---|---|---|
| n | Median (IQR) | (Min.-Max.) | n | Median (IQR) | (Min.-Max.) | p | |
| MTD (mm) | 57 | 50 (57) | (12–149) | 55 | 77 (55) | (25–165) | 0.021 |
| Albumin | 57 | 3.8 (0.95) | (2.7–5.2) | 55 | 4 (0.9) | (2.4–5.2) | 0.958 |
| T.Bilirubin | 57 | 1 (0.55) | (0.4–7.7) | 55 | 0.9 (0.7) | (0.2–2.4) | 0.527 |
| CRP | 40 | 0.38 (1.19) | (0.30–6.93) | 38 | 0.60 (2.45) | (0.31–43.00) | 0.241 |
| GGT | 57 | 90 (108.5) | (13–634) | 55 | 120 (189) | (22–1328) | 0.156 |
| ALKP | 57 | 117 (83.5) | (38–509) | 55 | 124 (92) | (45–528) | 0.317 |
| AST | 57 | 43 (41.5) | (16–407) | 55 | 50 (43) | (15–303) | 0.555 |
| ALT | 57 | 35 (31) | (10–672) | 55 | 30 (30) | (13–230) | 0.648 |
| Hb | 57 | 14 (2.5) | (7.9–17.5) | 55 | 13.8 (2.9) | (6.7–17.7) | 0.988 |
| PLT | 57 | 120 (113) | (44–454) | 55 | 196 (111) | (30–688) | 0.002 |
| LY# | 57 | 1.3 (0.63) | (0.35–12.9) | 55 | 1.43 (0.83) | (0.39–3.62) | 0.162 |
| NE# | 57 | 3.09 (1.5) | (0.91–11.57) | 55 | 4.53 (4.22) | (1.65–13.93) | <0.001 |
| AFP<100 IU/mL n (%) | AFP≥100 IU/mL n (%) | p | ||
|---|---|---|---|---|
| Multifocality | No | 22 (38.6) | 13 (23.6) | 0.133 |
| Yes | 35 (61.4) | 42 (76.4) | ||
| PVT | No | 33 (57.9) | 24 (43.6) | 0.187 |
| Yes | 24 (42.1) | 31 (56.4) |
| MTD≤5 cm | MTD>5 cm | ||||||
|---|---|---|---|---|---|---|---|
| n | Median (IQR) | (Min.-Max.) | n | Median (IQR) | (Min.-Max.) | p | |
| AFP | 46 | 37.32 (635.02) | (1.26–22283) | 66 | 315.36 (3589.45) | (2.22–76057) | 0.009 |
| Albumin | 48 | 4 (0.87) | (2.7–5.2) | 68 | 3.8 (1) | (2.4–5.2) | 0.266 |
| T.Bilirubin | 48 | 0.99 (0.51) | (0.4–2.8) | 68 | 0.9 (0.62) | (0.2–7.7) | 0.290 |
| CRP | 34 | 0.34 (0.80) | (0.31–43.00) | 47 | 0.60 (2.42) | (0.30–20.00) | 0.211 |
| GGT | 48 | 89.5 (107) | (13–388) | 68 | 123 (179) | (22–1328) | 0.025 |
| ALKP | 48 | 99 (58.25) | (38–268) | 68 | 131 (105) | (45–528) | 0.006 |
| AST | 48 | 37 (36.75) | (15–290) | 68 | 51 (47.75) | (16–407) | 0.018 |
| ALT | 48 | 29.5 (29.25) | (10–160) | 68 | 34 (30.25) | (10–672) | 0.447 |
| Hb | 48 | 13.6 (2.82) | (6.7–17.7) | 68 | 13.9 (2.55) | (9–17.5) | 0.541 |
| PLT | 48 | 119.5 (83.25) | (30–378) | 68 | 190.5 (131.5) | (43–688) | <0.001 |
| LY# | 48 | 1.25 (0.87) | (0.27–3.62) | 68 | 1.4 (0.68) | (0.39–12.9) | 0.201 |
| NE# | 48 | 3.2 (1.95) | (0.91–8.11) | 68 | 3.96 (3.74) | (1.43–13.93) | 0.002 |
| MTD≤5cm n (%) | MTD>5cm n (%) | p | ||
|---|---|---|---|---|
| Multifocality | No | 23 (47.9) | 15 (22.1) | 0.006 |
| Yes | 25 (52.1) | 53 (77.9) | ||
| PVT | No | 42 (87.5) | 18 (26.5) | <0.001 |
| Yes | 6 (12.5) | 50 (73.5) | ||
| PVT negative | PVT positive | ||||||
|---|---|---|---|---|---|---|---|
| n | Median (IQR) | (Min.-Max.) | n | Median (IQR) | (Min.-Max.) | p | |
| MTD (mm) | 60 | 40 (29) | (12–116) | 56 | 94 (50) | (35–169) | <0.001 |
| AFP | 57 | 40.35 (765.96) | (1.26–37159) | 55 | 182.32 (6317.51) | (2.22–76057) | 0.049 |
| Albumin | 60 | 4 (1) | (2.7–5.2) | 56 | 3.8 (0.9) | (2.4–4.9) | 0.238 |
| T.Bilirubin | 60 | 1.07 (0.63) | (0.43–2.8) | 56 | 0.89 (0.56) | (0.2–7.7) | 0.031 |
| CRP | 41 | 0.33 (0.54) | (0.31–7.36) | 40 | 1.37 (2.47) | (0.30–43) | 0.001 |
| GGT | 60 | 80.5 (96) | (13–1328) | 56 | 136.5 (194.5) | (22–715) | 0.002 |
| ALKP | 60 | 105 (60.5) | (38–528) | 56 | 131.5 (143) | (50–509) | 0.008 |
| AST | 60 | 38 (39.5) | (15–380) | 56 | 51 (47) | (16–407) | 0.027 |
| ALT | 60 | 32.5 (30) | (10–672) | 56 | 30 (27) | (10–434) | 0.927 |
| Hb | 60 | 14.3 (2.2) | (6.7–17.7) | 56 | 13.8 (2.9) | (7.9–16.4) | 0.194 |
| PLT | 60 | 125 (94) | (30–378) | 56 | 208 (143) | (43–688) | <0.001 |
| LY# | 60 | 1.38 (0.65) | (0.27–3.62) | 56 | 1.39 (0.72) | (0.39–12.9) | 0.556 |
| NE# | 60 | 3.39 (2) | (1.15–9.68) | 56 | 3.96 (4.62) | (0.91–13.93) | 0.012 |
| PVT negative n (%) | PVT positive n (%) | p | ||
|---|---|---|---|---|
| Multifocality | No | 27 (45.0) | 11 (19.6) | 0.007 |
| Yes | 33 (55.0) | 45 (80.4) |
Abbreviations for Table 4: MTD, maximum tumor diameter (cm); PVT, portal vein thrombosis; Alb, albumin (g/dL); AFP, alpha-fetoprotein (IU/L); CRP, C-reactive protein (mg/dL); Hb, hemoglobin (g/dL); NE, neutrophils (x 109 cells/L); Lymphs, lymphocytes (x 109 cells/L); T. bilirubin, total bilirubin (mg/dL); GGT, gamma.glutamyltransferase (IU/mL); ALKP, alkaline phosphatase (IU/mL); AST, aspartate amino transferase (IU/mL); ALT, alanine amino transferase (IU/mL); PLT, platelets (x 109/L).
7. Discussion
The median survival (Table 1a) of 13.93±3.24 months for the total cohort (mean 19.53±1.59 months) is similar to survival reported by others and better than survival recently reported for randomized trials involving sorafenib (25, 26). Using several parameters used in other HCC studies and their cutoffs (14, 15), we found here that there were significant survival differences using tumor factors (AFP, MTD and PVT) and inflammation factors (albumin, ALBI and CRP) as shown in Table 1b.
A model was constructed for combinations of the parameters associated with the highest HR. The combination of low CRP, AFP and MTD values was associated with significantly much longer survival (median 32.03 months) than the combination of high CRP, AFP and MTD values (median 3.83+ 1.68 months), for an HR of 10.3. Just as interesting, was the parameter combination for PVT positive patients known to have poor prognosis. Using only baseline serum CRP or AFP levels (there were not enough PVT patients to analyze the triple parameter combination) there were significant survival differences in PVT positive patients who had low CRP or AFP levels, compared to patients with high levels. The high CRP plus high AFP combination patients had an HR of 8.658 compared to the low CRP plus AFP combination (Table 2) for these PVT positive patients. This identifies a subgroup of otherwise poor prognosis (PVT positive) HCC patients who had a 5-fold increase in survival compared with other PVT positive patients, providing a possible selection criterion for subsequent liver transplantation or even other therapies.
Many of our patients are poor and live at a distance from a tertiary center such as ours, so that it can be difficult for them to return for follow up scans, or sometimes even to geographically close laboratory follow up tests, once they learn that definitive surgical therapy is not being offered. It would thus be important for their follow up if simple and cheap laboratory tests could be shown to facilitate their post-TARE management and decision-making. For this reason, we were interested to examine non-radiological, laboratory test options. We examined several, including albumin (Fig 1), ALBI, AFP, GGT and a network phenotyping strategy (NPS) as shown in Fig 1 and Table 3. Baseline patient ALBI has been previously shown to be prognostically useful in HCC patients (16, 17, 27, 28, 29). We found that ALBI was similarly useful both in patients with elevated and in patients with low serum AFP levels (Table 3A). This is clinically useful, as there are currently no other validated HCC markers for patients without abnormal (high) AFP levels. This is similarly true for GGT measurements and responses in patients with low serum AFP levels (Table 3C) as reported in other HCC series treated without TARE (30, 31). Our patients who had either baseline elevated or normal serum AFP levels, had similar GGT responses, as shown in Table 3C (ii). This analysis shows that both AFP and GGT responses (decrease) post TARE were associated with increased survival compared with patients without these serum responses, as seen in Tables 2C (ii) and 3C (ii), although the survival differences for GGT did not quite reach significance. Thus, follow up decreased AFP levels in patients with baseline elevated levels, or follow up decreased GGT levels may be associated with improved survival. Although we did not have enough patients followed up with high CRP levels, the fact that baseline low serum CRP was associated with longer survival than for patients with high CRP levels, suggests that CRP responses at follow up might be similarly useful as were GGT and especially AFP responses.
Most current prognostication methods use one or a limited number of clinical parameters. However, this gives only a partial view of the complexities of clinical HCC parameters and their putative interactions, since prognosis is known to be affected by both tumor and by liver parameters. The network phenotyping strategy (NPS) was designed to integrate 17 standard clinical parameters into 25 prognostically-related HCC phenotypes, 1–25, with phenotype 25 having the most aggressive phenotype and worst prognosis and phenotype 1 having the least aggressive clinical characteristics and the best prognosis [23, 24, 32]. The results from this small series of TARE patients (Table 3D), show that 27.3% of patients had an improved (lower phenotype number) NPS phenotype post TARE, 40.9% of patients had a worse NPS phenotype post TARE and 31.8% had a stable phenotype. Of the 12 patients who had an improved phenotype, the 8 who started with the highest phenotype are depicted in Table 3D (ii) had a wide range of quantitative phenotype responses. Those lower (response) phenotypes reflect an improved survival [23]. This is a potentially useful clinical management tool, available at https://apkatos.github.io/webpage_nps [23, 24]. This was a retrospective study and prospective data collection with estimated changes in NPS phenotype is planned, to assess the relationship between change in NPS phenotype after TARE and survival. We examined the clinical profiles of the 4 dichotomized parameters with the highest hazard ratios, namely CRP, AFP, MTD and PVT to determine which associated parameters might be important for significant differences in survival amongst parameter pairs, shown in Table 1b. We found slightly different clinical parameter profiles (Table 4) for the 4 parameter pairs that had significant differences in survival. Thus, the profiles for the longer-survival patients (low CRP, low AFP, low MTD and absent PVT), revealed smaller tumors (MTD), lower AFP levels, lower percent of patients with PVT and less multifocality for dichotomized low CRP patients and for dichotomized low MTD patients (except MTD, by definition). Similarly for the dichotomized cohort with low percent of PVT (except PVT, by definition). Only AFP dichotomized groups were without significance for PVT or multifocality. These 2 AFP groups also showed few differences in any liver or other inflammation parameters (albumin), so the explanation for the dichotomized AFP survival differences remains unclear.
The significance of these parameter changes remain to be investigated. AFP levels are a measure of HCC aggressiveness and thus an AFP decrease represents a decrease in aggressiveness. This may also be true for levels of the HCC biomarker GGT. The reasons for an increase in ALBI or a decrease in albumin with growing HCC may reflect either the contribution of the microenvironment or its response to a growing HCC. Therefore effective HCC might be expected to result in a reversal of these. Regardless of their exact meaning, these parameter changes probably reflect the complex interactions of an HCC mass and its hepatic microenvironment. Some modern pivotal HCC treatment trials showed an increase in survival with minimal tumor size change [33].
The strengths of this study are the identification of 6 potential treatment response biomarkers, namely, ALBI, albumin, AFP, GGT, NPS and possibly CRP. Patient subsets with elevated baseline levels of all 5 of these parameters showed responses to TARE, while our follow-up data showed improved survival for patients with responses to AFP and GGT levels. Although were found patients with decreasing CRP and NPS values post TARE, larger patient numbers need follow up to determine whether these changes are associated with survival benefit. The limitations of this study are both the relatively small sample size and the loss of several patients to follow up who lived in distant places.
Conclusions.
Several baseline parameter values predicted survival. A new survival model is described for TARE-treated HCC patients, especially for those with PVT. Several serum markers can change after TARE, including AFP, CRP, ALBI, albumin, GGT, and the newly-described NPS profile composed of 17 clinical parameters. Decreased AFP or GGT or increased albumin was associated with improved survival.
Funding
This work was supported in part by NIH grant CA 82723 (BIC).
Abbreviations:
- HCC
Hepatocellular Carcinoma
- TARE
Transarterial Radioembolization
- 90Y
Radioactive 90Yttrium Microspheres
- AFP
Alpha-Fetoprotein
- MTD
Maximum Tumor Diameter
- PVT
Portal Vein Invasion
- CRP
C-Reactive Protein
- AlB
Albumin
- HB
Hemoglobin
- NE
Neutrophils
- Lymphs
Lymphocytes
- GGT
Gamma Glutamyl Transferase
- ALKP
alkaline phosphatase
- AST
Aspartate Amino Transferase
- ALT
Alanine Amino Transferase
- PLT
Platelets
- ALBI
Albumin-Bilirubin Grade
- NPS
Network Phenotyping Strategy
- CAT
Scan: Computed Axial Tomography Scan
References
- 1.Salem R, Thurston KG, Carr BI, Goin JE, Geschwind JF. Yttrium-90 microspheres: radiation therapy for unresectable liver cancer. J Vasc Interv Radiol. 2002;13(9 Pt 2): S223–9. [DOI] [PubMed] [Google Scholar]
- 2.Maleux G, Albrecht T, Arnold D, Bargellini I, Cianni R. CIRT Principal Investigators. Predictive Factors for Adverse Event Outcomes After Transarterial Radioembolization with Yttrium-90 Resin Microspheres in Europe: Results from the Prospective Observational CIRT Study. Cardiovasc Intervent Radiol. 2023; 46(7): 852–867. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Carr BI, Kondragunta V, Buch SC, Branch RA. Therapeutic equivalence in survival for hepatic arterial chemoembolization and yttrium 90 microsphere treatments in unresectable hepatocellular carcinoma: a two-cohort study. Cancer. 2010; 116(5): 1305–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Salem R, Lewandowski RJ, Kulik L, Wang E, Riaz A, Ryu RK. Radioembolization results in longer time-to-progression and reduced toxicity compared with chemoembolization in patients with hepatocellular carcinoma. Gastroenterology. 2011; 140(2): 497–507.e2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Lewandowski RJ, Kulik LM, Riaz A, Senthilnathan S. A comparative analysis of transarterial downstaging for hepatocellular carcinoma: chemoembolization versus radioembolization. Am J Transplant. 2009; 9(8): 1920–8. [DOI] [PubMed] [Google Scholar]
- 6.Kulik LM, Carr BI, Mulcahy MF, Lewandowski RJ. Safety and efficacy of 90Y radiotherapy for hepatocellular carcinoma with and without portal vein thrombosis. Hepatology. 2008; 47(1): 71–81. [DOI] [PubMed] [Google Scholar]
- 7.Martelletti C, Ricotti A, Gesualdo M, Carucci P, Gaia S, Rolle E. Radioembolization vs sorafenib in locally advanced hepatocellular carcinoma with portal vein tumor thrombosis: A propensity score and Bayesian analysis. J Dig Dis. 2021; 22(8): 496–502. [DOI] [PubMed] [Google Scholar]
- 8.Miller FH, Lopes Vendrami C, Gabr A, Horowitz JM, Kelahan LC. Evolution of Radioembolization in Treatment of Hepatocellular Carcinoma: A Pictorial Review. Radiographics. 2021; 41(6): 1802–1818. [DOI] [PubMed] [Google Scholar]
- 9.Kao YH, Tan EH, Ng CE, Goh SW. Clinical implications of the body surface area method versus partition model dosimetry for yttrium-90 radioembolization using resin microspheres: a technical review. Ann Nucl Med. 2011; 25(7): 455–61. [DOI] [PubMed] [Google Scholar]
- 10.Braat AJ, Huijbregts JE, Molenaar IQ, Borel Rinkes IH, van den Bosch MA, Lam MG. Hepatic radioembolization as a bridge to liver surgery. Front Oncol. 2014; 4: 199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Dhanasekaran R, Kooby DA, Staley CA, Kauh JS, Khanna V, Kim HS. Prognostic factors for survival in patients with unresectable hepatocellular carcinoma undergoing chemoembolization with doxorubicin drug-eluting beads: a preliminary study. HPB (Oxford). 2010; 12(3): 174–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kang SH, Kim DY, Jeon SM, Ahn SH, Park JY, Kim SU, Kim JK, Lee KS. Clinical characteristics and prognosis of hepatocellular carcinoma with different sets of serum AFP and PIVKA-II levels. Eur J Gastroenterol Hepatol. 2012; 24(7): 849–56. [DOI] [PubMed] [Google Scholar]
- 13.Giannone F, Slovic N, Pessaux P, Schuster C, Baumert TF, Lupberger J. Inflammation-related prognostic markers in resected hepatocellular carcinoma. Front Oncol. 2023; 13: 1267870. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Carr BI, Guerra V. Serum Inflammation Parameters and Survival in Hepatocellular Carcinoma Patients: Importance of Albumin and Gamma-Glutamyltranspeptidase. Oncology. 2023; 101(5): 313–320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Carr B, Guerra V, Karaogullarindan U, Akkiz H, Ince V, Isik B, Yilmaz S. Liver Inflammation Parameters in Relation to Survival in Patients with Hepatocellular Carcinoma Tumor of differing tumor sizes. J Cancer Sci Ther. 2022. [Google Scholar]
- 16.Gui B, Weiner AA, Nosher J. Assessment of the Albumin-Bilirubin (ALBI) Grade as a Prognostic Indicator for Hepatocellular Carcinoma Patients Treated with Radioembolization. Am J Clin Oncol. 2018; 41: 861–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Antkowiak M, Gabr A, Das A. Prognostic Role of Albumin, Bilirubin, and ALBI Scores: Analysis of 1000 Patients with Hepatocellular Carcinoma Undergoing Radioembolization. Cancers. 2019; 11: 879. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Memon K, Kulik L, Lewandowski RJ, Wang E, Ryu RK, Riaz A. Alpha-fetoprotein response correlates with EASL response and survival in solitary hepatocellular carcinoma treated with transarterial therapies: a subgroup analysis. J Hepatol. 2012; 56(5): 1112–1120. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Riaz A, Ryu RK, Kulik LM, Mulcahy MF, Lewandowski RJ, Minocha J. Alpha-fetoprotein response after locoregional therapy for hepatocellular carcinoma: oncologic marker of radiologic response, progression, and survival. J Clin Oncol. 2009; 27(34): 5734–42. [DOI] [PubMed] [Google Scholar]
- 20.Galle PR, Foerster F, Kudo M, Chan SL, Llovet JM. Biology and significance of alpha-fetoprotein in hepatocellular carcinoma. Liver Int. 2019; 39(12): 2214–2229. [DOI] [PubMed] [Google Scholar]
- 21.Carr BI, Akkiz H, Üsküdar O, Yalçın K, Guerra V, Kuran S, Karaoğullarından Ü. HCC with low- and normal-serum alpha-fetoprotein levels. Clin Pract (Lond). 2018; 15(1): 453–464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Carr BI, Akkiz H, Bag HG, Karaoğullarından U, Yalçın K, Ekin N. Serum levels of gamma-glutamyl transpeptidase in relation to HCC human biology and prognosis. J Transl Sci. 2021; 7(3): 10.15761. [Google Scholar]
- 23.Pancoska P, Sotáková P, Carr BI. A New Approach to Analysis of Clinical Data and Prognostication for Patients with Hepatocellular Carcinoma, Based Upon a Network Phenotyping Strategy (NPS) Computational Method. J Inonu Liver Transplant Inst. 2024; 2(3): 109–116. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Pancoska P, Carr BI, Ince V, Yilmaz S. A Network Phenotyping Strategy approach in a Turkish HCC dataset from Inonu University and comparison of patients selected for transplant and those who were not. J Inonu Liver Transplant Inst. 2025; 3(1): 31–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Pollock RF, Brennan VK, Shergill S, Colaone F. A systematic literature review and network meta-analysis of first-line treatments for unresectable hepatocellular carcinoma based on data from randomized controlled trials. Expert Rev Anticancer Ther. 2021; 21(3): 341–349. [DOI] [PubMed] [Google Scholar]
- 26.Csipak AR, da Fonseca LG, López RVM, Estevez-Diz MDP. Real-World, Observational, Retrospective Study to Evaluate the Effectiveness and Safety of Treatment with Sorafenib in Patients with Advanced Hepatocellular Carcinoma. Curr Oncol. 2024; 31(11): 6778–6790. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Toyoda H, Johnson PJ. The ALBI score: From liver function in patients with HCC to a general measure of liver function. JHEP Rep. 2022; 4(10): 100557. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Gelardi F, Rodari M, Pini C, Zanca R, Artesani A, Tosi G, Chiti A, Sollini M. ALBI grade for outcome prediction in patients affected by hepatocellular carcinoma treated with transarterial radioembolization. Front Nucl Med. 2022; 2: 934446. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Bayona Molano MDP, Kolber M, Barrera JV, Akram MR, Alnablsi MW. Prognostic Value of Liver Biomarkers in Hepatocellular Carcinoma Patients Undergoing Yttrium 90 Transarterial Radioembolization (TARE): A Retrospective Pilot Study. Cureus. 2024; 16(6): e61904. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Carr BI, Pancoska P, Branch RA. Significance of increased serum GGTP levels in HCC patients. Hepatogastroenterology. 2010; 57(101): 869–74. [PubMed] [Google Scholar]
- 31.Ince V, Carr BI, Bag HG, Koc C, Usta S, Ersan V, Baskiran A, Sahin TT, Yilmaz S. Gamma glutamyl transpeptidase as a prognostic biomarker in hepatocellular cancer patients especially with >5 cm tumors, treated by liver transplantation. Int J Biol Markers. 2020; 35(2): 91–95. [DOI] [PubMed] [Google Scholar]
- 32.Pančoška P, Skála L, Nešetřil J, Carr BI. Evaluation of total hepatocellular cancer lifespan, including both clinically evident and preclinical development, using combined network phenotyping strategy and fisher information analysis. Semin Oncol. 2015; 42(2): 339–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Shan M, Moscovici M, Voliotis D, Bruix J. SHARP Investigators Study Group. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med. 2008; 359(4): 378–90. [DOI] [PubMed] [Google Scholar]