Circulating tumor DNA (ctDNA) as a biomarker of response to therapy in advanced Hepatocellular carcinoma treated with Nivolumab

Abstract

BACKGROUND:

Circulating tumor DNA (ctDNA) is a promising non-invasive marker for detection, diagnosis, treatment selection, and prognosis of hepatocellular carcinoma (HCC).

OBJECTIVE:

This study aimed to examine the utility of ctDNA as a prognostic and predictive tool in HCC patients treated with nivolumab.

METHODS:

We analyzed pre-treatment ctDNA from 44 HCC patients using comprehensive genomic testing on a commercially available platform. We utilized log rank test and univariate Cox models to correlate overall survival (OS) and progression-free survival (PFS) with ctDNA expressions.

RESULTS:

Of 44 patients, 77.3% were men with median age of 67 years. All but 3 patients had at least one alteration identified, and TP53 was the most commonly altered gene (52.3%). Median OS was 17.5 months (95% CI: 12.7, NA). Mutations involving PIK3CA, BRCA1, and CCND1 amplification were associated with shorter OS (P 0.0001, 0.0001 and 0.01, respectively). Median PFS time was 4.01 months (95% CI: 3.06, 9.33). Mutations involving KIT and PIK3CA were associated with shorter PFS (P 0.0001 and 0.0004, respectively), while mutation involving CTNNB1 were associated with longer PFS ( $p=$ 0.04).

CONCLUSIONS:

ctDNA profiling may provide a benefit for prediction of survival and progression of HCC patients treated with nivolumab. Future studies are needed for confirmation.

1. Introduction

Hepatocellular carcinoma (HCC) is one of the most lethal malignancies and a leading cause of cancer related mortality worldwide [1, 2]. Despite the recent advances in treatment of advanced HCC; The prognosis remains poor compared to patients diagnosed and treated at early stage. In addition, higher frequency of recurrence could occur after both local and systemic treatment [3].

Such poor prognosis for HCC patients represents a serious clinical problem, which could be attributed to the absence of specific symptoms in early stages, lack of accurate markers and tools for early diagnosis, treatment selection and outcome prediction; which leads to most patients often diagnosed at an advanced stage and suffering poor outcome. Thus, early diagnosis of HCC and accurate treatment strategies are highly important to improve HCC prognosis [4, 5].

Circulating tumor DNA (ctDNA) is the fraction of cell-free DNA(cfDNA) derived from primary or metastatic tumors; it has emerged as a potential noninvasive markers and is starting to be adopted in clinical practice to detect mutations and to monitor disease course for several major cancer types [6].

In HCC, ctDNA has proven beneficial for tracking traces of tumors in high risk population, detecting early stage and detection of genomic changes in advanced HCC patients [7, 8]. Studies have shown that ctDNA levels correlate with tumor burden and disease progression, making it a valuable tool for prognosis. ctDNA has been found to be inversely correlated with poor prognosis and shorter overall survival (OS) [9]. Given the increasing application of immune checkpoint inhibitors such as nivolumab in advanced HCC [10], understanding ctDNA alterations can provide valuable insights into treatment outcomes and guide therapeutic decisions. Our study aims to investigate whether pre-treatment ctDNA alterations can serve as prognostic biomarkers in HCC patients treated with nivolumab, thus potentially aiding in patient stratification and personalized treatment planning.

Recent studies have demonstrated that changes in ctDNA levels are associated with treatment outcomes in various cancers, including HCC. Specifically, a reduction in ctDNA levels has been correlated with improved progression-free survival (PFS) and overall survival (OS) in patients receiving immune checkpoint inhibitors. This highlights the potential of ctDNA as a non-invasive biomarker for monitoring and predicting patient responses to immunotherapy [11, 12].

We specifically selected a population of HCC patients treated with nivolumab due to its growing use and promising results in managing advanced HCC. Nivolumab, an anti-PD-1 antibody, has shown significant efficacy and manageable safety in advanced HCC, making it a critical component of current treatment strategies [13]. By focusing on this patient population, we aim to provide insights into how ctDNA alterations can inform and potentially predict responses to immunotherapy, addressing a significant clinical need for predictive biomarkers in this context.

In the current study, we assessed the prognostic significance of baseline ctDNA in patients with HCC treated with nivolumab (anti-PD-1). This investigation represents the first prospective study of pre-treatment ctDNA association with OS and progression free survival (PFS) among patients with HCC treated with nivolumab.

2. Patients and methods

The study was approved by the University of Texas MD Anderson Cancer Center’s Institutional Review Board, and informed consent was obtained from all patients. We prospectively collected Blood samples from 44 HCC patients who were treated with nivolumab and followed up until progression and/or death and analyzed correlation with pretreatment ctDNA expressions. Adult patients with pathologically or radiologically confirmed HCC, as defined by the American Association for the Study of Liver Diseases, who were treated at MD Anderson Cancer Center (MDACC) from December 2017 to May 2020 and had ctDNA results available were included in the study. We accessed medical records for research purposes of these patients between December 2017 and July 2021 for information regarding medical conditions, clinical parameters, including pretreatment of ctDNA expressions, and survival outcomes.

Table 1

Patient demographics and characteristics

Characteristics	N (%)
Age, y, median	67
Range	36–81
Sex
Female	10	(22.7%)
Male	34	(77.3%)
Race
White or Caucasian	27	(61.4%)
Black or African American	4	(9.1%)
Asian	7	(15.9%)
Other	5	(11.4%)
Unknown	1	(2.3%)
Ethnicity
Hispanic or Latino	6	(13.6%)
Non Hispanic or Latino	36	(81.8%)
Unknown	2	(4.5%)
Number of prior systemic therapies
1	44	(93.2%)
2	3	(6.8%)
Number of mutations
0	3	(6.8%)
1	13	(29.5%)
2	14	(31.8%)
3	8	(18.2%)
5	3	(6.8%)
6	2	(4.5%)
8	1	(2.3%)
Cirrhosis
No	3	(6.8%)
Present	41	(93.2%)
Etiology
HBV	5	(11.4%)
HCV	14	(31.8%)
HCV/HBV	3	(6.8%)
NASH	13	(29.5%)
NASH/hemochromatosis	1	(2.3%)
NASH/ALD	1	(2.3%)
ALD	2	(4.6%)
Hemochromatosis	1	(2.3%)
N/A	4	(9.1%)
Child_Pugh_group
A	33	(73.3%)
B	11	(24.4%)
Child_Pugh_Score
5	25	(10.2%)
6	8	(16.3%)
7	6	(12.2%)
8	4	(8.2%)
9	1	(2%)
Pathology
Hepatocellular carcinoma	41	(93.2%)
Not available	3	(6.8%)
Differentiation
Not available	8	(18.2%)
Well differentiated	6	(13.6%)
Well to mod differentiated	1	(2.3%)
Mod differentiated	21	(47.7%)
Mod to poorly differentiated	2	(4.5%)
-+Poorly differentiated	6	(13.6%)
Nodularity
Unimodular	3	(6.8%)
Multinodular	41	(93.2%)

Table 1, continued
Characteristics	N (%)
Vascular_invasion
None	18	(40.9%)
Present	26	(59.1%)
Metastasis
None	19	(43.2%)
Present	25	(56.8%)
HCV
Negative	26	(59.1%)
positive	17	(38.6%)
Not available	1	(2.3%)
HBsAg
Negative	36	(81.8%)
positive	8	(18.2%)
AntiHBc
Negative	35	(79.5%)
positive	9	(20.5%)
HIV
Negative	44	(100%)
Diabetes
Negative	24	(54.5%)
Positive	20	(45.5%)
Alcohol
Negative	30	(68.2%)
Positive	14	(31.8%)
Image response
No	31	(70.5%)
Yes	9	(20.5%)
Not available	4	(9.1%)
Disease_control
No	28	(63.6%)
Yes	14	(31.8%)
Not available	2	(4.5%)

Abbreviations: Y: Year, N/A: not applicable, ALD: Alcoholic liver disease, NASH: nonalcoholic steatohepatitis.

Patients’ blood samples and epidemiologic and clinical data were collected, and blood samples were analyzed retrospectively for ctDNA expressions [14, 15]. Clinical and epidemiological data were retrieved from medical records. PFS was calculated from the date that Nivolumab treatment began to the date of disease progression or death, whichever occurred first. OS was calculated from the date that Nivolumab treatment began to the date of death or to the date of the last follow-up visit. The Kaplan-Meier method was used to calculate the time to event outcomes (i.e., OS and PFS) with Log rank test to compare OS or PFS between subgroups [16]. This study was approved by MD Anderson Cancer Center’s Institutional Review Board.

2.1 ctDNA analysis

Blood-samples were shipped to a Clinical Laboratory Improvement Act (CLIA)-certified, College of American Pathologists-accredited laboratory (Guardant Health, Redwood City, California) and where ctDNA was analyzed. The ctDNA was extracted and analyzed using a comprehensive genomic testing platform. This platform provided detailed genomic profiling, focusing on genes frequently mutated in HCC, including TP53, PIK3CA, BRCA1, CCND1, and CTNNB1. The selected panel was chosen due to its relevance in cancer biology and its potential to provide insights for patient.

2.2 Statistical analysis

Continuous patient characteristics were summarized using descriptive statistics, categorical patient characteristics were tabulate with frequency and percentage. Fisher’s exact test and Wilcoxon rank sum test were used to evaluate the association between response and ctDNA expressions [17, 18]. Log rank test and univariate Cox models were used to evaluate the association between OS or PFS and ctDNA [19].

3. Results

Table 2

ctDNA detection and mutation number

Mutation	Mutation detected (Yes/No)	Frequency (%)
CTNNB1	No	28	(63.6%)
	Yes	16	(36.4%)
TP53	No	21	(47.7%)
	Yes	23	(52.3%)
TERT	No	32	(72.7%)
	Yes	12	(27.3%)
KRAS	No	41	(93.2%)
	Yes	3	(6.8%)
GNAS	No	43	(97.7%)
	Yes	1	(2.3%)
CCND1_AMPL	No	41	(93.2%)
	Yes	3	(6.8%)
CCNE1_AMPL	No	43	(97.7%)
	Yes	1	(2.3%)
NFE2L2	No	42	(95.5%)
	Yes	2	(4.5%)
KIT	No	43	(97.7%)
	Yes	1	(2.3%)
PIK3CA	No	43	(97.7%)
	Yes	1	(2.3%)
EGFR	No	42	(95.5%)
	Yes	2	(4.5%)
RAF1	No	43	(97.7%)
	Yes	1	(2.3%)
RB1	No	38	(86.4%)
	Yes	6	(13.6%)
ALK	No	43	(97.7%)
	Yes	1	(2.3%)
NF1	No	41	(93.2%)
	Yes	3	(6.8%)
BRAF	No	43	(97.7%)
	Yes	1	(2.3%)
NTRK1	No	43	(97.7%)
	Yes	1	(2.3%)
APC	No	43	(97.7%)
	Yes	1	(2.3%)
FGFR1	No	43	(97.7%)
	Yes	1	(2.3%)
FGFR2	No	43	(97.7%)
	Yes	1	(2.3%)
FGFR3	No	43	(97.7%)
	Yes	1	(2.3%)
ARIDA1A	No	41	(93.2%)
	Yes	3	(6.8%)
BRCA2	No	42	(95.5%)
	Yes	2	(4.5%)
MET	No	38	(86.4%)
	Yes	6	(13.6%)
HIF1A	No	43	(97.7%)
	Yes	1	(2.3%)
PDGFRA	No	42	(95.5%)
	Yes	2	(4.5%)
CDKN2A	No	43	(97.7%)
	Yes	1	(2.3%)
FBXW7	No	43	(97.7%)
	Yes	1	(2.3%)
NOTCH1	No	43	(97.7%)
	Yes	1	(2.3%)
BRCA1	No	43	(97.7%)
	Yes	1	(2.3%)
MTOR	No	42	(95.5%)
	Yes	2	(4.5%)
ESR	No	43	(97.7%)
	Yes	1	(2.3%)

Table 3

Log rank test to evaluate the association between ctDNA categorical data and OS

Variable name	Level	N	Event	Median OS (95%CI) (M)		OS Rate at 1 Year (95%CI)		$P$ -value
	All patients	44	17	17.51	(12.71, NA)	0.67	(0.53, 0.84)
CTNNB1	No	28	13	17.51	(11.53, NA)	0.62	(0.45, 0.85)	0.37
	Yes	16	4	NA	(12.71, NA)	0.75	(0.54, 1)
TP53	No	21	8	18.81	(12.71, NA)	0.7	(0.51, 0.97)	0.89
	Yes	23	9	13.93	(11.53, NA)	0.64	(0.46, 0.9)
TERT	No	32	13	20.11	(11.53, NA)	0.65	(0.49, 0.87)	0.66
	Yes	12	4	17.51	(17.51, NA)	0.73	(0.51, 1)
KRAS	No	41	15	20.11	(12.71, NA)	0.67	(0.52, 0.85)	0.63
	Yes	3	2	17.51	(7.06, NA)	0.67	(0.3, 1)
GNAS	No	43	16	20.11	(13.83, NA)	0.69	(0.56, 0.87)	0.32
	Yes	1	1	11.53	(NA, NA)	NA
NFE2L2	No	42	16	17.51	(12.71, NA)	0.68	(0.53, 0.86)	0.59
	Yes	2	1	2.99	(2.99, NA)	0.5	(0.13, 1)
PIK3CA	No	43	16	17.51	(13.83, NA)	0.69	(0.54, 0.86)	$<$ 0.0001
	Yes	1	1	3.88	(NA, NA)	NA
RAF1	No	43	16	20.11	(13.83, NA)	0.69	(0.55, 0.86)	0.2
	Yes	1	1	9.89	(NA, NA)	NA
RB1	No	38	14	20.11	(12.71, NA)	0.69	(0.54, 0.87)	0.31
	Yes	6	3	13.93	(10.78, NA)	0.53	(0.21, 1)
MET	No	38	15	17.51	(12.71, NA)	0.68	(0.54, 0.87)	0.4
	Yes	6	2	NA	(7.06, NA)	0.56	(0.23, 1)
BRCA1	No	43	16	17.51	(13.83, NA)	0.68	(0.54, 0.86)	$<$ 0.0001
	Yes	1	1	2.99	(NA, NA)	NA
MTOR	No	42	16	17.51	(13.83, NA)	0.68	(0.54, 0.86)	0.45
	Yes	2	1	7.85	(NA, NA)	NA
CCND1_AMPL	No	41	15	20.11	(13.83, NA)	0.71	(0.57, 0.88)	0.01
	Yes	3	2	11.53	(2.99, NA)	NA

Abbreviations: N: number; OS: overall survival; N/A: not applicable.

The study included 44 patients with advanced HCC who received nivolumab at MD Anderson Cancer Center and had ctDNA available prior to the start of treatment. The demographic characteristics of these patients (Table 1) were documented at the time of nivolumab administration; Median age at the time of diagnosis was 67 years (range: 36–81), 34 patients were male patients, 10 were female patients, 27 patients were white, 4 were African American, and 7 were Asian. Thirty-three (73.3%) patients were in Child-Pugh A stage and 11 (24.2%) were in Child-Pugh B. Fourteen patients tested positive for HCV, 5 tested HBV positive, and 3 tested positive for both HCV and HBV. ctDNA analysis identified at least 1 alteration in 41/44 (93.2%) of the patients. The median number of alterations/patient was 2 (range, 0–8). TP53 was the most common altered gene ( $n=$ 23) followed by CTNBB1 ( $n=$ 19), TERT ( $n=$ 12), MET ( $n=$ 6) and RB1 ( $n=$ 6). Other mutations were infrequently present: KRAS, EGFR, ARIDA1A, NF1 and CCND1 amplification in three tumors; BRCA2, NFE2L2, EGFR and PDGFRA in two tumors, and GNAS, CCNE1, KIT, PIK3CA, ALK, BRAF, NTRK1, APC and FGFR1 in only one tumor among others. Table 2.

The median OS was 17.5 months (95% CI: 12.7, NA The estimated median follow-up time was 14.7 months (95% CI: 12.7, 19.0). Mutations involving PIK3CA, BRCA1, and CCND1 amplification were associated with shorter OS (P 0.0001, 0.0001, 0.01, respectively. Table 3).

Table 4

Log rank test to evaluate the association between ctDNA categorical data and PFS

Varname	Level	N	Event	Median PFS (95%CI) (M)		PFS Rate at 6 m (95%CI)		$P$ -value
	All patients	43	37	4.01	(3.06, 9.33)	0.39	(0.27, 0.57)
CTNNB1	No	27	25	3.45	(2.92, 8.61)	0.28	(0.15, 0.52)	0.04
	Yes	16	12	7.42	(3.38, NA)	0.56	(0.37, 0.87)
TP53	No	21	19	3.48	(2.92, 10.84)	0.29	(0.15, 0.56)	0.32
	Yes	22	18	5.09	(3.06, 14.13)	0.49	(0.32, 0.76)
TERT	No	31	27	3.68	(3.06, 10.84)	0.38	(0.24, 0.6)	0.78
	Yes	12	10	4.62	(2.3, NA)	0.42	(0.21, 0.81)
KRAS	No	40	35	3.68	(3.06, 9.33)	0.39	(0.27, 0.58)	0.95
	Yes	3	2	4.14	(2.3, NA)	0.33	(0.07, 1)
GNAS	No	42	36	4.01	(3.06, 9.33)	0.4	(0.27, 0.58)	0.55
	Yes	1	1	3.45	(NA, NA)	NA
NFE2L2	No	41	36	4.01	(3.22, 9.33)	0.38	(0.26, 0.57)	0.76
	Yes	2	1	2.99	(2.99, NA)	0.5	(0.13, 1)
KIT	No	42	36	4.01	(3.22, 9.33)	0.4	(0.27, 0.58)	$<$ 0.0001
	Yes	1	1	1.84	(NA, NA)	NA
PIK3CA	No	42	36	4.01	(3.22, 9.33)	0.4	(0.27, 0.58)	$P=$ 0.0004
	Yes	1	1	1.91	(NA, NA)	NA
EGFR	No	41	35	3.68	(3.06, 9.33)	0.36	(0.24, 0.54)	0.83
	Yes	2	2	9.07	(8.61, NA)	1	(1, 1)
RAF1	No	42	36	4.01	(3.06, 9.33)	0.4	(0.27, 0.58)	0.61
	Yes	1	1	3.48	(NA, NA)	NA
RB1	No	38	32	3.68	(2.99, 9.53)	0.41	(0.28, 0.61)	0.63
	Yes	5	5	4.01	(3.22, NA)	0.2	(0.03, 1)
ALK	No	42	36	4.01	(3.22, 9.33)	0.4	(0.27, 0.58)	0.4
	Yes	1	1	3.06	(NA, NA)	NA
NF1	No	40	34	4.01	(3.06, 9.33)	0.39	(0.27, 0.58)	0.6
	Yes	3	3	3.38	(2.73, NA)	0.33	(0.07, 1)
BRAF	No	42	36	4.01	(3.06, 9.33)	0.4	(0.27, 0.58)	0.5
	Yes	1	1	3.38	(NA, NA)	NA
APC	No	42	36	3.68	(3.06, 9.53)	0.37	(0.25, 0.55)	0.88
	Yes	1	1	9.33	(NA, NA)	1	(1, 1)
FGFR2	No	42	36	3.68	(3.06, 9.53)	0.37	(0.25, 0.55)	0.88
	Yes	1	1	9.33	(NA, NA)	1	(1, 1)
FGFR3	No	42	36	3.68	(3.06, 9.53)	0.37	(0.25, 0.55)	0.88
	Yes	1	1	9.33	(NA, NA)	1	(1, 1)
ARIDA1A	No	40	35	3.68	(3.06, 8.61)	0.37	(0.24, 0.55)	0.77
	Yes	3	2	9.33	(2.76, NA)	0.67	(0.3, 1)
BRCA2	No	41	35	4.01	(3.22, 9.53)	0.38	(0.26, 0.57)	0.44
	Yes	2	2	5.58	(1.84, NA)	0.5	(0.13, 1)
MET	No	37	31	4.14	(3.45, 10.84)	0.42	(0.29, 0.62)	0.06
	Yes	6	6	3.02	(2.3, NA)	0.17	(0.03, 1)
HIF1A	No	42	36	3.68	(3.06, 9.53)	0.37	(0.25, 0.55)	0.98
	Yes	1	1	7.42	(NA, NA)	1	(1, 1)
PDGFRA	No	41	35	3.68	(3.06, 9.33)	0.36	(0.24, 0.54)	0.12
	Yes	2	2	64.77	(8.61, NA)	1	(1, 1)
CDKN2A	No	42	36	4.01	(3.22, 9.33)	0.4	(0.27, 0.58)	0.35
	Yes	1	1	2.99	(NA, NA)	NA
FBXW7	No	42	36	4.01	(3.22, 9.33)	0.4	(0.27, 0.58)	0.35
	Yes	1	1	2.99	(NA, NA)	NA
NOTCH1	No	42	36	4.01	(3.22, 9.33)	0.4	(0.27, 0.58)	0.35
	Yes	1	1	2.99	(NA, NA)	NA
BRCA1	No	42	36	4.01	(3.22, 9.33)	0.4	(0.27, 0.58)	0.35
	Yes	1	1	2.99	(NA, NA)	NA
MTOR	No	41	36	4.01	(3.22, 9.33)	0.38	(0.26, 0.57)	0.97
	Yes	2	1	1.84	(1.84, NA)	0.5	(0.13, 1)
CCND1_AMPL	No	40	34	4.14	(3.22, 9.53)	0.42	(0.29, 0.6)	0.17
	Yes	3	3	3.06	(2.99, NA)	NA
CCNE1_AMPL	No	42	36	3.68	(3.06, 9.53)	0.37	(0.25, 0.55)	0.88
	Yes	1	1	9.33	(NA, NA)	1	(1, 1)

Abbreviations: N: number; PFS: Progression-free survival; N/A: not applicable.

The median PFS time was 4.01 months (95% CI: 3.06, 9.33). Forty-three patients were available for PFS analysis, and 37 of the 43 patients had PFS events (death or PD whichever occurred first). Mutations involving KIT and PIK3CA were associated with shorter PFS (P 0.0001 and 0.0004, respectively), while mutations involving CTNNB1 were associated with longer PFS ( $p=$ 0.04). No significant differences in OS or PFS was observed for other alterations (Table 4).

4. Discussion

In this study, we investigated baseline ctDNA as a prognostic biomarker in advanced HCC patients treated with nivolumab. We assessed pretreatment ctDNA plasma samples from 44 patients. Whereas the small cohort size is considered a limitation of this study, we have demonstrated the predictive value and the utility of ctDNA as a clinical biomarker.

Nivolumab, an anti–programmed cell death 1 (anti-PD-1) checkpoint inhibitors nivolumab, has been approved for the treatment of HCC patients. In CheckMate 040, nivolumab monotherapy demonstrated manageable safety, objective response rate (ORR) of 14%, duration of response of at least 12 months in 59% of patients, and promising long-term median survival of 15.1 months in patients with advanced HCC treated with sorafenib [20].

In addition to comorbidities and tumor burden, selecting the most effective treatment for advanced HCC has to be carefully for the risk-benefit ratio. Systemic therapies for advanced HCC usually incur a high-cost burden on both patients and healthcare system. It is important to know which patients will benefit from nivolumab versus those in whom it will have lower survival benefit. Despite the recent breakthrough in the treatment of patients with advanced HCC, there remains an unmet need for reliable biomarkers of response to immunotherapy. ctDNA have been reported to be a prognostic biomarker in HCC patients [9, 12, 21].

Our group recently reported the mutational landscape of HCC tumorigenesis for the purpose of selecting patients for targeted therapy trials and the potential clinical utility of ctDNA [22], The total number of alterations was 680 (nonunique); median number of alterations/patient was three (range, 1–13); median mutant allele frequency (% cfDNA), 0.49% (range, 0.06%–55.03%). TP53 was the most commonly altered gene [ $>$ 120 alterations (non-unique)] followed by EGFR, MET, ARID1A, MYC, NF1, BRAF, and ERBB2 [20–38 alterations (nonunique)/gene]. Of the patients with alterations, 56.9% (103/181) had $⩾$ 1 actionable alterations, most commonly in MYC, EGFR, ERBB2, BRAF, CCNE1, MET, PIK3CA, ARID1A, CDK6, and KRAS. In these genes, amplifications occurred more frequently than mutations. Hepatitis B (HBV)-positive patients were more likely to have ERBB2 alterations, 35.7% (5/14) versus 8.8% in HBV-negative patients ( $P=$ 0.04) [22].

In this study, we demonstrated the potential clinical utility of ctDNA in patients treated with Nivolumab and the prognostic value of ctDNA for patients in real-world clinical practice. Our study is the first to report ctDNA association with outcomes among patients with HCC treated with nivolumab. Mutations involving PIK3CA, BRCA1, and CCND1 were associated with shorter OS. Mutations involving KIT and PIK3CA were associated with shorter PFS, while mutation involving CTNNB1 were associated with longer PFS.

The clinical significance of our study lies in its potential to improve personalized treatment strategies for advanced HCC. By identifying ctDNA alterations that correlate with survival outcomes, clinicians can better stratify patients based on their likelihood of responding to nivolumab. This can lead to more tailored treatment plans, optimizing the use of immunotherapy and potentially improving patient outcomes. Additionally, the use of ctDNA as a non-invasive biomarker offers a practical tool for ongoing monitoring of treatment response and disease progression, reducing the need for more invasive procedures.

Our study had some limitation. Firstly, this investigation was conducted at a single center study. Moreover, the small sample size of patients included in the study. Whereas PIK3CA and BRCA1 were associated with a shorter OS, KIT and PIK3CA were also associated with shorter PFS; Hence, these observations need to be further explored and validated in a lager cohort of HCC patients.

Secondly, given the low incidence of molecular alterations in HCC the current study which was exploratory and retrospective, the study was not powered. Lastly, we carried out ctDNA sequencing using commercially available kit (Guardian Health, CA), which only allowed us to study a limited number of genes. Thus, the study is not very inclusive of all genes and alterations. Therefore, herein we focused on any alteration and its relations with the outcome irrespective of the gene being actionable or non-actionable. A larger sample size with enough power is required to validate these observations and further confirm our results.

To the best of our knowledge, this is the first study to assess clinical prognostic value of ctDNA in patients who have received Nivolumab in clinical setting. Our results provide evidence to suggest the clinical utility of ctDNA as a noninvasive biomarker for prediction of OS and PFS in patients with HCC treated with nivolumab and potentially other immunotherapy approaches.

Although this study focused on its baseline prognostic value, serial ctDNA evaluation in future studies is expected have prognostic significance and provide useful data to eventually guide therapy decision in clinical routine practice. This study provides a strong argument for the use of ctDNA as a personalized clinical tool. It has the potential to serve as a prognostic tool for advanced HCC patients and may guide treatment choices. Future studies in larger randomized clinical trial are warranted to further validate these findings.

Funding

National Cancer institute R01 CA260872(AOK, HMA, DGD). The University of Texas MD Anderson Cancer Center, SPORE in HCC, NCI# P50CA217674-01A1(AOK). D.G. Duda’s work is supported through NIH (grant nos. R01CA260872, R01CA260857, and R01CA247441) and by Department of Defense PRCRP (grant nos.W81XWH-19–438 1-0284, W81XWH1910482, and W81XWH-21–1-0738).

Author contributions

Study concept, design, and supervision: Ahmed O. Kaseb.

Acquisition of data: Yehia I. Mohamed, Manal M. Hassan, Tarik Demir, Ahmed O. Kaseb.

Analysis and interpretation of data: Yehia I. Mohamed, Manal M. Hassan, Lianchun Xiao, Ahmed O. Kaseb.

Drafting of the manuscript: Yehia I. Mohamed, Lianchun Xiao, Asif Rashid, Manal Hassan, Hesham Amin, Ahmed Omar Kaseb.

Administrative, technical, or material support: Yehia I. Mohamed, Lianchun Xiao, Rikita Hiatia, Aliya Qayyum, Asif Rashid, James C. Yao, Armeen Mahvash, Manal Hassan, Hesham Amin, Ahmed Omar Kaseb.

Critical revision of the manuscript for important intellectual content: Yehia I. Mohamed, Sunyoung S Lee, Tarik Demir, MD, Zishuo Ian Hu, Shadi Chamseddine, Michael LaPelusa, Lianchun Xiao, Khaled Elsayes, Jeffrey S. Morris, Robert A. Wolff, Rikita Hiatia, Aliya Qayyum, Asif Rashid, Dan G. Duda, James C. Yao, Eugene J Koay, Armeen Mahvash, Ahmed Al Azzam, Ecaterina E Dumbrava, Manal Hassan, Hesham Amin, Ahmed Omar Kaseb.

Administrative, technical, or material support: Yehia I. Mohamed, Lianchun Xiao, Rikita Hiatia, Aliya Qayyum, Asif Rashid, James C. Yao, Armeen Mahvash, Manal Hassan, Hesham Amin, Ahmed Omar Kaseb.

Footnotes

Acknowledgments

None.

Conflict of interest

All authors report no commercial associations (e.g., consultancies, stock ownership, equity interests, or patent-licensing arrangements) that might pose a conflict of interest in connection with the submitted article.