Identification of 2 large size HCC phenotypes, with and without associated inflammation

Abstract

Background:

Large HCCs can often be associated with low levels of cirrhosis. However, inflammation is also regarded as a driver of HCC growth.

Objectives:

To compare patients with large >5 cm HCCs having high versus low serum inflammation parameters.

Materials and methods:

A Turkish patient HCC dataset with known survivals was retrospectively analyzed after dichotomization according to several clinical inflammation markers.

Results:

Amongst several parameters examined, only AST levels were significantly associated with elevated AFP levels and increased percent PVT and tumor multifocality. The dichotomization of the cohort according to high or low AST levels resulted in 2 subcohorts with a 5-fold difference in median survival. The 2 AST-dichotomised cohorts comprised patients with similar large-size HCCs, but which were significantly different with respect to serum AFP levels, percent PVT, and percent tumor multifocality.

Conclusions:

Two large-sized HCC phenotypes were identified. One had more aggressive HCC characteristics, higher inflammatory indices, and worse survival. The other had the opposite. Despite inflammation being important for the growth of some large tumors, others of a similar size likely have different growth mechanisms.

Introduction

The clinical presentation and survival of patients with Hepatocellular Carcinoma (HCC) depend in large part on the combination of both tumor factors (maximum tumor diameter or MTD, presence or absence of portal venous invasion or PVT, number of tumors or multifocality, serum levels of alpha-fetoprotein or AFP) and the tumor micro-environment or state of the underlying liver [1]. As with most tumors of solid organs in adult humans, HCC treatment options and prognosis is greatly influenced by tumor size at clinical presentation, with most options and best results occurring in patients with smaller tumors. Despite this, some patients with very large tumors have a good liver function and can be surgically treated [2], which has been ascribed in part to the growth of large tumors in the absence of cirrhosis or liver inflammation. To investigate large HCCs further, the current study focuses only on large MTD HCCs in relation to indices of inflammation and the associated clinical patient and tumor features. We found that dichotomization according to inflammatory parameters results in 2 identifiable large MTD patients and tumor phenotypes that also differed in their median survival.

Methods

Clinical

An HCC database was studied, that comprised 149 prospectively-accrued non-transplant HCC patients who had both survival data and baseline tumor parameter data, including CT scan information on HCC Maximum Tumor Diameter (MTD), number of tumor nodules, and presence or absence of macroscopic Portal Vein Thrombosis (PVT), and serum Alpha- Fetoprotein (AFP) levels; complete blood counts; and routine serum liver function tests, (total bilirubin, GGTP, ALKP, albumin, transaminases) plus CRP. The diagnosis was made either via tumor biopsy or according to international guidelines. Database management conformed to legislation on privacy and this study conforms to the ethical guidelines of the Declaration of Helsinki approval for this retrospective study on de-identified HCC patients was obtained from Cukurova University and Inonu University the Institutional Review Board (approval #2021/2572).

Statistical

The normality of the quantitative data was examined by the Shapiro-Wilk test. Median, first quartile (1stQ), and third quartile (3rdQ) were used as descriptive statistics. Comparisons were performed by the Mann-Whitney U test. Categorical variables were presented as counts and percentages, chi-square tests were applied for comparisons. Kaplan-Meier analysis and Log-rank test were used for survival analysis. The hazard ratio was estimated by univariate cox regression analysis. The significance level was considered as 0.05 in all analyses.

Results

Inflammation parameters and indices of HCC aggressiveness

A series of clinical serum parameters were examined, using previously-determined ROC cutoffs [3], in relation to 3 HCC aggressiveness indices, namely AFP, tumor multifocality, and PVT in large size MTD >5 cm HCC patients (TABLE 1). Several of the parameters correlated with 1 or more aggressiveness indices, but only AST was associated with differences in all 3 indices.

TABLE 1.

Inflammation markers for PVT, multifocality, and AFP in patients (n=149) with HCCs >5 cm diameter.

	PVT			Tumor Focality
	(%)			(%)			AFP (IU/mL)
	No	Yes	p	1	≥ 2	p	Median (min.- max)	p
AST ≤ 50	66.7	33.3		73.7	26.3		13.76 (1.84-1000)
			0.047			0.007		0.022
AST>50	48.2	51.8		50	50		67.9 (1.3-1000)
ALT ≤ 50	58.2	41.8		65.4	34.5		16 (1.5-1000)
			0.496			0.084		0.207
ALT>50	52.5	47.5		51.5	48.5		54.5 (1.3-1000)
ALKP ≤ 150	58.2	41.8		58.9	41.1		15 (1.47-1000)
			0.788			1		0.149
ALKP>150	54.3	45.7		58	42		58 (1.30-1000)
GGT ≤ 50	70.6	29.4		62.9	37.1		6.13 (1.47-1000)
			0.072			0.728		<0.001
GGT>50	51	49		57.7	42.3		71.45 (1.3-1000)
Albumin ≥ 3.5	61.5	38.5		76.3	23.7		8.43 (1.12-1000)
			0.573			0.103		<0.001
Albumin<3.5	57.8	42.2		65	35		46 (0.5-1000)
CRP ≤ 10	60	40		71	29		25.2 (0.5-1000)
			0.617			0.805		0.823
CRP>10	63.8	36.2		68.1	31.9		18.02 (0.7-1000)
ESR ≤ 15	64.3	35.7		60.7	39.3		12.7 (1.6-1000)
			0.685			0.898		0.045
ESR>15	57.5	42.5		57	43		74.5 (1.3-1000)
NLR ≤ 2.3	70.4	29.6		71.1	28.9		57 (1.3-1000)
			0.264			0.531		0.3
NLR>2.3	55	45		62.7	37.3		19.51 (1.47-1000)

Abbreviations and units: AFP: Alpha-Fetoprotein (IU/mL); PVT: Portal Vein Thrombosis; AST: Aspartate Amino Transferase (IU/mL); ALT: Alanine Aminotransferase (IU/mL); ALKP: Alkaline Phosphatase (IU/mL); GGT: Gamma Glutamyl Transferase (IU/mL); Albumin (g/dL); CRP: C-Reactive Protein (mg/dL); ESR: Erythrocyte Sedimentation Rate (mm/hour); NLR: Neutrophil to Lymphocyte Ratio

AST in relation to survival

The relationship of serum AST values to survival was next examined (TABLE 2). A statistically significant survival difference was found when the 2 AST groups were compared, with low AST patients having 5-fold the median survival of high AST patients.

TABLE 2.

Kaplan-Meier and Cox survival analysis according to serum AST levels (IU/mL) in patients with large HCCs.

		Kaplan-Meier		Univariate
		Analysis		Cox regression
	Survival time	Survival time	Log-Rank	HR	HR
	Mean ± SE	Median ± SE	p-value	(95% CI)	p-value
AST ≤ 50 (n=61)	40.90 ± 5 .99	21 ± 4.69	<0.001	reference
AST>50 (n=88)	9.70 ± 2.17	4 ± 0.50		3.27 (2.17-4.92)	<0.001

Tumor and clinical characteristics of AST cohorts

The clinical characteristics of the 2 AST patient cohorts were next compared (TABLE 3). The MTD of the 2 groups was similar, by design (TABLE 3). However, the AFP levels, percent of patients with both PVT and multifocality was significantly higher in the high AST group compared with the low AST group (TABLE 3).

TABLE 3.

Tumor and laboratory serum parameters of 2 groups of patients with large HCCs.

A.Tumor parameters	AST ≤ 50 IU/mL	AST>50 IU/mL	p
MTD (median)	10	9	0.938
AFP (median)	13.76	67.9	0.022
PVT+ve (%)	33.3	51.8	0.047
Tumor #>2 (%)	26.3	50	0.007

Abbreviations: AFP: Alpha-Fetoprotein; MTD: Maximum Tumor Diameter; PVT: Portal Vein Thrombosis.

The clinical characteristics also displayed several differences (TABLE 4). Patients in the high AST group had significantly higher levels of serum CRP, ALKP, GGT, ALT, and bilirubin, with significantly lower albumin levels and a higher percentage of patients with cirrhosis (81.2% vs. 60.3%) than patients in the low AST group.

TABLE 4.

Median laboratory serum parameter values^*.

Lab parameters	AST ≤ 50 IU/mL	AST>50 IU/mL	p
CRP	3.85	6.95	0.019
ESR	21	35	0.069
ALKP	143	196.5	<0.001
GGT	68	140	<0.001
ALT	26	71	<0.001
T. Bilirubin	0.9	1.71	<0.001
Albumin	3.25	2.8	0.001
Cholesterol	167.5	146.5	0.023
HDL	35.2	30.02	0.016
Platelets	232 (153-300)	189.5 (119-336)	0.471
Cirrhosis %	60.30%	81.20%	0.011

Abbreviations and units: as in Tables 1 and 2.

^*Significant parameters from Table 2.

Relationships of AST levels to tumor characteristics

An attempt was then made to examine semi-quantitatively, the relationships between serum AST levels and the 3 HCC aggressiveness indices of serum AFP levels, percent PVT, and percent multifocality (FIGURE 1). There was a significant relationship between AST levels and serum AFP levels (p<0.001) and percent multifocality (p<0.015) and approaching significance in relation to percent PVT (p<0.068).

Fig 1.

a, b, c. AST concentration (IU/mL) in relation to tumor parameter values.

Discussion

Tumor size is one of several important prognostic factors for HCC [1] and the larger MTD HCCs can often occur in noncirrhotic livers [2,4]. It was previously shown that characteristic differences of tumor biology relate to tumor size, with larger tumors having increased PVT and AFP levels, but lower bilirubin and higher albumin and platelet levels, than smaller tumors, suggesting differences in both HCC biology as well as microenvironment [5]. Massive HCCs also tend to have higher levels of AFP, ALKP, GGTP, and platelets [6]. Large tumor size usually has a negative prognosis for most tumortypes, but not necessarily for large single HCCs [7-11]. Large HCCs in patients with relatively normal liver function and thus less cirrhosis might have different biology or different drivers of their tumor growth (predominantly tumor factors such as oncogenes and other tumor growth regulators) compared with smaller or large size HCCs that arise on the basis of cirrhosis. Those HCCs might have both tumor drivers and cirrhosisassociated inflammatory microenvironmental tumorgrowth factors. Patients with high ALKP, GGTP, and AFP all have been shown to have in common, significantly higher mean bilirubin levels than patients with low levels of these parameters [5]. Both AFP and FIGURE 1c have been shown to reflect different HCC cell lineages. Some of these lineages have features of stem or progenitor cells, possibly with different HCC developmental pathways, and have unique molecular signatures and tumor biology [12,13]. Although AFP is known to be a strong negative prognostic factor, exactly how high AFP HCC cells behave differently and thus cause lower survival rates, is still unclear.

The tumor microenvironment is increasingly recognized as contributing to HCC growth biology [14,15]. Amongst the several tumor microenvironmental processes, inflammation has gained much recent attention and there is evidence for a role of inflammation in many cancers, including HCC [16,17]. Inflammationbased prognostic scores have been shown to be associated with survival [18-21], possibly via the effects of inflammation on various parameters of tumoraggressiveness [22]. Inflammation is also thought to induce a microenvironment that is involved in DNA damage, tumorgrowth, and angiogenesis. It seems to be a bidirectional process, in which inflammation can be seen as a response to growing tumor cells; yet is also involved in their growth and invasiveness [16,23,24].

In this work, therefore, investigated whether inflammation is necessarily associated with the development of large size HCC or whether, in view of the absence of cirrhosis in a proportion of large MTD patients, other pathways might possibly also be involved. We found that although several of the 8 inflammation parameters that were examined, correlated with one or more HCC aggressiveness characteristics, only AST correlated with all 3 of the aggressiveness characteristics that we studied, namely AFP levels, tumor focally, and PVT. MTD was not studied as large MTD patients were selected for the study. Despite allowing a large range of MTD (>5 cm HCCs), the 2 groups of patients with either high or low serum AST levels were fairly closely matched for MTD. The 2 groups of patients were significantly different with respect to the tumorcharacteristics of AFP levels and percent of patients with PVT and multifocality. There were also striking general differences. Liver enzymes were different between the 2 groups, as expected from the AST selection, with bilirubin levels being higher and albumin levels being lower in the high compared to the low AST group. Percent cirrhosis was also different, being higher in the high compared to the low AST group (81.2 vs. 60.3%). Unsurprisingly, there were semi-quantitative differences between AST levels and serum AFP, percent of patients with PVT, and percent with tumormultifocality (FIGURE 1). The 5-fold difference in median survival between the groups was striking.

Even though there were significant differences in percent cirrhosis between the 2 AST groups, at least 60% of the patients had cirrhosis, with the majority of our Turkish patients having hepatitis B-based HCC. We have previously noticed that albumin decreases as AFP increases since, in development, there is an increase in albumin as AFP (an oncofetal form of albumin) decreases.

Total bilirubin has previously been shown to increase with an increase in the size of large HCCs. This has been explained as due to parenchymal encroachment and destruction by the growing HCC. However, in this study, the HCCs in the 2 groups were of a similar size, so this explanation would seem to be inadequate, with the elevated bilirubin being in the high AST/inflammation group.

Conclusion

Inflammation has been shown to be associated with HCC growth, multifocality and portal venous invasion by tumor and several inflammation-associated effectors have been suggested to be involved in these processes.

However, the non-inflammation associated HCCs, as determined here by low serum AST levels (and associated low ALT, ALKP, and GGT) presumably have some different mechanisms of growth increase and tumor invasion. Presumably, these involve endogenous factors such as tumor oncogenes, vascularity factors and endogenous growth factors, and possibly immune mechanisms. The notable finding of such a large survival difference between the 2 groups of patients with large size HCCs suggests possible surgical implications. In this view, patients with large MTD HCCs with low serum levels of AST (or ALT and ALKP and GGT) may be predicted to have better survival post-liver transplant or resection, than patients with similar large MTD but an elevation of these serum parameters.

Abbreviations

HCC

Hepatocellular Carcinoma

WBC

White Blood Count

TBil

Total Bilirubin

AST

Aspartate Aminotransferase

ALT

Alanine Aminotransferase

ALKP

Alkaline Phosphatase

GGT

Gamma-Glutamyl Transferase

CRP

C-Reactive Protein

ESR

Erythrocyte Sedimentation Rate

NLR

Neutrophil-Lymphocyte Ratio

MTD

Maximum Tumor Diameter

PVT

Portal Vein Thrombosis

AFP

Alpha-Fetoprotein

CAT scan

Computed Axial Tomography

EpCAM

Epithelial Cellular Adhesion Molecule