Lay summary
Tumour burden score is a useful tool to gauge the entity of hepatic tumour infiltration. It is associated with an increased risk of immune-related hepatotoxicity in patients with hepatocellular carcinoma treated with immune checkpoint inhibitors.
BACKGROUND
Immune checkpoint inhibition is the standard of care for a multiplicity of cancer types. After the first pivotal results in melanoma,1 a number of new approvals followed, changing the natural history of a number of tumours, including hepatocellular carcinoma (HCC).2 Immune checkpoint inhibitors (ICIs) have become an established treatment strategy for HCC, as it has traditionally been associated with a background of liver inflammation. Currently, the combination of atezolizumab, an anti-programmed death ligand-1 (PD-L1) monoclonal antibody (mAb), plus bevacizumab, an anti-vascular endothelial growth factor (VEGF) mAb, has been approved by both the United States (US) Food and Drug Administration (FDA) and the European Medicines Agency (EMA) as first-line treatment for patients with unresectable or metastatic HCC.3,4 Other immunotherapy options approved by the FDA in the second-line setting are pembrolizumab, an anti-PD-1 mAb, as monotherapy, based on the positive results of the phase II KEYNOTE-224 trial,5 and the combination of nivolumab, another anti-PD-1 mAb, plus ipilimumab, an anticytotoxic T lymphocyte antigen 4 (CTLA-4) mAb, based on one of the cohorts of the phase I/II CheckMate 040 trial.6 Further combinations have recently shown positive results in the first-line setting and will expand the treatment landscape in the near future.7,8
While immunotherapy constitutes an expanding treatment option in the management of HCC, an important issue is the wide range of immune-related adverse events (irAEs) that these therapeutic agents can cause. Among patients with HCC on ICI therapy, the incidence of immune-related acute hepatitis of all grades is 1–4% with anti-PD-1/L1 monotherapy, 4–9% with anti-CTLA-4 monotherapy, and 18% with the combination of anti-PD-1 and anti-CTLA-4 mAbs.9 Immune-related AEs have been linked to improved prognosis in patients with HCC,10 and it is of capital importance to avoid that hepatic irAEs (HIRAEs) lead to premature treatment interruption or even permanent discontinuation. For this reason, predictive biomarkers of treatment toxicity are needed to identify patients at higher risk of HIRAEs in order to keep them under strict surveillance in collaboration with hepatologists in the frame of an even more important multidisciplinary approach to the patient.
Immune-related hepatotoxicity while on ICIs is more common in patients with HCC than in patients with other primary cancer, possibly because of the background of liver cirrhosis and the presence of the tumour in the liver itself. Therefore, primary hepatic tumour infiltration seems to be a predisposing condition for the development of immune-mediated hepatotoxicity.9 On the other hand, little is known about the role of secondary hepatic tumour infiltration from liver metastases on hepatotoxicity and there is lack of data regarding how primary and secondary tumour infiltration could have a different impact on hepatotoxicity during treatment with ICIs.
Based on these premises and given the lack of biomarkers of treatment toxicity, in this study we aimed to evaluate the role of baseline hepatic tumour burden, measured according to the tumour burden score (TBS), in the development of immune-related hepatotoxicity both in patients with HCC and in patients with liver metastases from other cancer types treated with ICIs. The TBS is defined as the distance from the origin on a Cartesian plate that has two variables incorporated: the maximum tumour size on the x-axis and the number of liver lesions on the y-axis. The Pythagorean theorem is then applied in the following way to calculate the distance of any point from the origin of the plane: [TBS2 = (maximum tumour diameter)2 + (number of liver lesions)2] (Figure 1).11 In addition, we assessed the impact that both the baseline hepatic tumour burden and the development of HIRAEs have on overall survival (OS) in patients treated with ICIs for both HCC and other primary tumours with liver metastases.
METHODS Study design
We conducted a retrospective observational cohort study, including all consecutive patients with HCC and with liver metastases from other primary tumours treated with ICIs at IRCCS Humanitas Research Hospital between January 2014 and March 2021. We divided them into an HCC cohort and a non-HCC cohort, according to their tumour type. All patients were at least 18 years old, with a radiological or histological diagnosis of malignancy. Patients in the HCC cohort had a histological or radiological diagnosis of HCC according to the American Association for this study of Liver Diseases (AASLD) criteria,12 and they were diagnosed with advanced disease, defined according to the Barcelona Clinic Liver Cancer (BCLC) criteria as BCLC C or BCLC B not deemed amenable to locoregional treatments. Clinical and laboratory data were collected retrospectively from electronic records.
Treatment administration and outcome measuresWe included every patient receiving at least one dose of ICI. ICIs were administered as monotherapy or in combination according to the different indications, and toxicity management, including dose reduction or treatment discontinuation, followed the summary of product characteristics. Causality to ICIs of AEs was inferred by the treating physician according to the clinical assessment. For each patient, we assessed the hepatic tumour burden at baseline using the TBS, and we explored its correlation with development of HIRAEs and with survival.
Radiological response was assessed with a CT scan every 6–12 weeks using the Response Evaluation Criteria in Solid Tumours (RECIST) criteria version 1.1.13 Adverse events were graded according to the Common Terminology Criteria for Adverse Events version 5.0 (CTCAE 5.0). In particular, HIRAEs were defined as any increase in alanine aminotransferase (ALT) and/or aspartate aminotransferase (AST) deemed related to ICI by the investigator. Only AEs deemed to be treatment-related were collected, and attribution of causality to immune checkpoint inhibitors was based on the published toxicity profile of the agents, the assessment of treating physicians, and the judgement of the investigators according to the summary of product characteristics. We defined OS as the time from the date of the first dose of the treatment to the date of death.
Statistical methods and data analysisData have been summarized as frequencies and proportions or as medians and ranges, and differences have been estimated by the Chi square test (the Fisher exact test when appropriate) or the t test (the Wilcoxon test when appropriate).
A logistic regression model has been used to estimate odds ratios (OR and their corresponding 95% confidence intervals [CIs]). The area under the curve receiver operating characteristic (AUC ROC) curve's value was evaluated in order to find a possible cut-off for continuous factors. Survival curves have been estimated with the Kaplan-Meier Method and differences between groups have been evaluated by the Log-rank test. Hazard ratios and their corresponding 95% CI have been calculated by the Cox regression hazard model. All evaluations are explorative in nature. All analyses have been performed using SAS 9.4.
Ethical considerationsWe conducted this study to the ethics guidelines in the Declaration of Helsinki. Ethical approval was granted by the local ethical committee. Because of its retrospective nature, a written consent was not deemed necessary.
RESULTS Patient characteristicsA total of 93 patients treated with ICIs at IRCCS Humanitas Research Hospital from January 2014 to March 2021 were analysed in this study and divided into two cohorts: cohort 1 was made up by 42 patients with advanced HCC (Table 1), while cohort 2 was made up by 51 patients with non-HCC cancers of various types who developed liver metastases prior to immunotherapy initiation (Table 2). Of these 51 patients, 26 were affected by lung cancer, 11 by melanoma (of which three by choroidal melanoma), seven by renal cancer, four by neuroendocrine tumours, one each by colorectal cancer (CRC), leiomyosarcoma and ovarian cancer.
TABLE 1 Cohort 1, Baseline patient characteristics
| Variable—no. (%) | Patients with HCC (Cohort 1) N = 42 |
| Median age, years (range) | 70 (40–81) |
| Male sex | 28 (66.67) |
| Liver cirrhosis | 23 (54.76) |
| Aetiology | |
| HCV | 15 (35.71) |
| Alcohol | 9 (21.43) |
| NASH | 3 (7.14) |
| HBV | 2 (4.76) |
| NAFLD | 1 (2.38) |
| Idiopathic | 12 (28.57) |
| ECOG-PS | |
| 0 | 22 (52.38) |
| 1 | 19 (45.24) |
| 2 | 1 (2.38) |
| Child-Pugh score | |
| A5 | 31 (73.81) |
| A6 | 9 (21.43) |
| B | 2 (4.76) |
| BCLC stage | |
| B | 9 (21.43) |
| C | 33 (78.57) |
| AFP | |
| ≥400 ng/mL | 19 (45.24) |
| Portal vein thrombosis | 10 (23.81) |
| Extra-hepatic spread | 28 (66.67) |
| Prior locoregional treatment | 34 (80.95) |
| Surgery | 21 (50) |
| TACE | 17 (40.48) |
| RFA | 9 (21.43) |
| TARE | 1 (2.38) |
| Other | 12 (28.57) |
| Tumour burden | |
| Maximum nodule diameter, median (range) (cm) | 5 (0–15) |
| Number of nodules, median (range) | 2 (0–30) |
| TBS value, median (range) (cm) | 9 (0–31.18) |
Abbreviations: AFP, alpha-fetoprotein; BCLC, Barcelona Clinic Liver Cancer; ECOG-PS, Eastern Cooperative Oncology Group-Performance Status; HBV, hepatitis B virus; HCC, hepatocellular carcinoma; HCV, hepatitis C virus; NAFLD, non-alcoholic fatty liver disease; NASH, non-alcoholic steatohepatitis; RFA, radiofrequency ablation; TACE, trans-arterial chemoembolization; TARE, trans-arterial radioembolization; TBS, tumour burden score.
TABLE 2 Cohort 2, Baseline patient characteristics
| Variable—no. (%) | Patients with liver metastases (Cohort 2) N = 51 |
| Median age, years (range) | 66 (41–82) |
| Male sex | 29 (56.86) |
| ECOG-PS | |
| 0 | 26 (50.98) |
| 1 | 18 (35.29) |
| 2 | 7 (13.73) |
| Primary tumour | |
| Lung cancer | 26 (50.98) |
| Melanoma | 11 (21.57) |
| Renal cancer | 7 (13.73) |
| NET | 4 (7.84) |
| CRC | 1 (1.96) |
| Ovarian cancer | 1 (1.96) |
| Leiomyosarcoma | 1 (1.96) |
| Hepatic dysfunction | 4 (7.84) |
| Steatosis | 2 (3.92) |
| HBV infection | 2 (3.92) |
| Extra-hepatic metastases | 40 (78.43) |
| Tumour Burden | |
| Maximum nodule diameter, median (range) (cm) | 2.8 (0.6–76) |
| Number of nodules, median (range) | 3 (1–30) |
| TBS value, median (range) (cm) | 7.43 (1–76.01) |
Abbreviations: CRC, colorectal cancer; ECOG-PS, Eastern Cooperative Oncology Group-Performance Status; HBV, hepatitis B virus; NET, neuroendocrine tumour; TBS, Tumour Burden Score.
For cohort 1, we found that 15 patients (35.71%) had underlying hepatitis C virus (HCV)-related cirrhosis, nine patients (21.43%) had alcoholic cirrhosis, three patients had non-alcoholic steatohepatitis (NASH) (7.14%), two patients (4.76%) had hepatitis B virus (HBV)-related cirrhosis and one patient had non-alcoholic fatty liver disease (NAFLD) (2.38%). The remaining 12 patients (28.57%) had idiopathic HCC. Performance status (PS) according to the Eastern Cooperative Oncology Group (ECOG) scale was graded as 0 for 22 patients (52.38%), as 1 for 19 patients (45.24%), and as 2 for 1 patient (2.38%). Lastly, 31 patients (73.81%) had Child-Pugh score A5, nine patients (21.43%) had Child-Pugh score A6 and two patients (4.76%) had Child-Pugh score B (Table 1).
Twenty-nine patients (69.05%) were treated with immunotherapy as monotherapy while 13 patients (30.95%) were treated with a combination therapy (Table 3). For 19 patients (45.24%) it was the first-line treatment, for 16 patients (38.10%) the second-line treatment and for seven patients (16.67%) the third-line treatment. Thirty-nine patients (92.86%) discontinued immunotherapy treatment, of which 32 (76.19%) because of progressive disease, one because of toxicity (2.38%) and six (14.29%) because of other reasons.
TABLE 3 Cohort 1, Immunotherapy characteristics
| Variable—no. (%) | Patients with HCC (Cohort 1) N = 42 |
| Line of treatment | |
| 1 | 19 (45.24) |
| 2 | 16 (38.10) |
| 3 | 7 (16.67) |
| Median duration, months (range) | 4.1 (0.5–32.7) |
| Monotherapy | 29 (69.05) |
| Combination therapy | 13 (30.95) |
| Best response | |
| Complete response | 2 (4.76) |
| Partial response | 3 (7.14) |
| Stable disease | 24 (57.14) |
| Progressive disease | 13 (30.95) |
| Discontinuation | 39 (92.86) |
| Reason for discontinuation | |
| Progressive disease | 32 (76.19) |
| Toxicity | 1 (2.38) |
| Other | 6 (14.29) |
| Post-immunotherapy treatment | 33 (78.57) |
Abbreviation: HCC, hepatocellular carcinoma.
In cohort 2, extra-hepatic metastases were present in 40 patients (78.43%), and ECOG-PS: 26 patients (50.98%) had ECOG-PS 0, 18 patients (35.29%) had ECOG-PS 1, and seven patients (13.73%) had ECOG-PS 2 (Table 2).
Fifty patients (98.04%) were treated with immunotherapy as monotherapy, while one patient (1.96%) was treated with immunotherapy combination (Table 4). For 16 patients (31.37%) it was the first-line treatment, for 26 patients (50.98%) the second-line treatment, for five patients (9.80%) the third-line treatment, for three patients (5.88%) the fourth-line treatment and for one patient (1.96%) the fifth-line treatment. Lastly, 47 patients (92.16%) discontinued immunotherapy, of which 20 (39.22%) because of progressive disease, 11 (21.57%) because of toxicity and four (7.84%) because of liver failure.
TABLE 4 Cohort 2, Immunotherapy characteristics
| Variable—no. (%) | Patients with liver metastases (Cohort 2) N = 51 |
| Line of treatment | |
| 1 | 16 (31.37) |
| 2 | 26 (50.98) |
| 3 | 5 (9.80) |
| 4 | 3 (5.88) |
| 5 | 1 (1.96) |
| Median duration, months (range) | 3.27 (0.5–37.33) |
| Monotherapy | 50 (98.04) |
| Combination therapy | 1 (1.96) |
| Best response | |
| Complete response | 1 (1.96) |
| Partial response | 13 (25.49) |
| Stable disease | 12 (23.53) |
| Progressive disease | 20 (39.22) |
| Unknown | 5 (9.80) |
| Discontinuation | 47 (92.16) |
| Ongoing | 4 (7.84) |
| Reason for discontinuation | |
| Progressive disease | 20 (39.22) |
| Toxicity | 11 (21.57) |
| Liver failure or PS worsening | 4 (7.84) |
| Completion of treatment plan | 3 (5.88) |
| Other | 9 (17.65) |
Abbreviation: PS, performance status.
Association between tumour burden score and any-gradeMedian value of TBS at baseline was 9 (range: 0–31.18) (Table 1). Twenty-four patients (57.14%) (Table 5) suffered from any grade immune-related toxicity, of which eight patients (19.05%) had grade ≥3 toxicity. In particular, 18 patients (42.86%) reported any grade immune-related hepatotoxicity, of whom five (11.90%) had grade ≥3 hepatic toxicity. Of these five patients, four patients (9.52%) had to temporarily interrupt the treatment, while one patient (2.38%) had to permanently discontinue immunotherapy because of hepatotoxicity.
TABLE 5 Safety summary
| Variable—no. (%) | Patients with HCC (Cohort 1) N = 42 | Patients with liver metastases (Cohort 2) N = 51 |
| Any grade irAEs | 24 (57.14) | NA |
| Any irAEs ≥G3 | 8 (19.05) | NA |
| Any grade hepatic irAEs | 18 (42.86) | 18 (35.29) |
| Hepatic irAEs ≥G3 | 5 (11.90) | 3 (5.88) |
| Discontinuation because of toxicity | 5 (11.90) | 7 (13.73) |
| Temporarily interruption | 4 (9.52) | 4 (7.84) |
| Permanent discontinuation | 1 (2.38) | 3 (5.88) |
Abbreviation: G, grade; HCC, hepatocellular carcinoma; irAEs, immune-related adverse events; NA, not available.
In patients who did not report HIRAE median TBS value was 5.85 (interquartile range [IQR] = 9.29), while in patients who developed any-grade HIRAE median TBS value was 10.95 (IQR = 16.32), P = .11 (Figure 2A). Considering TBS as a continuous variable with values ranging from 0 to 31.18, for every increase of 1 point of TBS on a continuous scale, the risk of developing any-grade HIRAEs increased by 7% (OR 1.07, 95% CI 0.99–1.14, P = .08, AUC 0.65). On the other hand, considering progressive 5-point increase in TBS, we found an OR of 1.37 (95% CI 0.96–1.94, P = .08, AUC 0.65) with a 37% increase in HIRAE risk for every 1 unit (5 points)-increase in TBS.
Enlarge this image.FIGURE 2. Distribution of Tumour burden score values in Cohort 1 (A) and Cohort 2 (B). Legend: 0 = patients who did not develop hepatotoxicity; 1 = patients who developed hepatotoxicity. Abbreviations: TBS, tumour burden score
When calculating TBS at baseline, we found a median value of 7.43 (range 1.00–76.01) (Table 2).
We then analysed the incidence and grade of immune-related hepatotoxicity (Table 5): 18 patients (35.29%) had any grade immune-related hepatotoxicity, of which three patients (5.88%) had grade ≥3 hepatic toxicity. Four patients (7.84%) had to temporarily interrupt immunotherapy because of toxicity, and three patients (5.88%) had to permanently discontinue it.
Median TBS was not different when comparing patients who did not develop HIRAEs and those with any-grade HIRAEs (5 [IQR 7.28] vs 7.82 [IQR 14.44], P = .32; Figure 2B). Considering TBS as a continuous variable with values ranging from 1 to 76.01, for every increase of 1 point of TBS, the risk of developing any-grade HIRAEs increased by 1% (OR 1.01, 95% CI 0.97–1.06, P = .63, AUC 0.59). On the other hand, considering progressive 5-point increase in TBS, for every 1 unit-increase in TBS the risk of developing any-grade HIRAEs increased by 6% (OR 1.06, 95% CI 0.84–1.33, P = .63, AUC 0.59).
The incidence of any-grade HIRAEs in the two cohorts of patients was not statistically significant (P = .46).
Survival Cohort 1: patients withIn patients who did not develop HIRAE median OS was 9.98 (95% CI 5.58–22.27), 1-year OS was 44.87%, and 2-year OS was 22.44%. On the other hand, in patients who developed any-grade HIRAE median OS was 9.87 (95% CI 6.63–11.89), 1-year OS was 27.78% and 2-year OS was 16.67%, without any significant differences between the two groups (P = .67) (Figure 3A). Also, TBS was found to be not associated with OS improvement (HR 1.03, 95% CI 0.99–1.06, P = .10).
Enlarge this image.FIGURE 3. Kaplan–Meier curved showing overall survival in Cohort 1 (A) and Cohort 2 (B). Legend: No = patients who did not develop hepatotoxicity; Yes = patients who developed hepatotoxicity
In patients who did not develop HIRAEs median OS was 13.10 (95% CI 3.65–20.59), and 1-year OS was 51.95%. On the other hand, in patients who developed HIRAEs median OS was not reached, and 1-year OS was 53.57%, with no significant difference in survival between the two groups (P = .13) (Figure 3B). Also, TBS did not correlate with OS (HR 1.02, 95% CI 1.00–1.04, P = .09).
DISCUSSIONTo date, lack of biomarkers limits the ability to predict treatment toxicity from ICIs across all oncological fields.
Our study is the first to investigate the possible predictive role of baseline hepatic tumour burden, measured according to TBS, in the development of immune-related hepatotoxicity in patients treated with ICIs for both advanced HCC and other primary tumours featuring liver metastases. We showed that patients experiencing any-grade immune-related hepatotoxicity had a numerically higher baseline median TBS compared to those without hepatitis, especially in the HCC cohort. Furthermore, when considering TBS as a continuous variable, we found that for every 5-point increase in TBS in the HCC cohort, the risk of HIRAEs increased by 37%, pointing in the direction of a possible correlation between TBS and HIRAEs, even if not statistically significant.
TBS is a simple, straightforward tool which summarizes in one single variable both the number of liver nodules and the maximum lesion diameter, providing a more accurate estimate of the hepatic tumour burden than the single variables alone.11 The great advantage of TBS as a potential predictor of HIRAEs is its easy and economic application in clinical practice. Analysing other applications of the TBS across various oncological fields, tumour size and number (together with AFP levels) have been used to predict post-transplantation risk of death from HCC-related factors14 and long-term outcomes after hepatic resection for patients with colorectal liver metastases.11 In addition, the impact of both tumour volume and number of nodules on tumour response and OS has been studied across a wide range of malignancies, as melanoma15 and non-small-cell lung cancer.16–18 The different results observed in the two cohorts of this study might be related to the different background of liver impairment and the different role that the burden of primary and secondary hepatic infiltration have in the development of immune-related hepatitis and survival.
Early diagnosis and management of irAEs are important to limit the severity of irAEs, and they can avoid premature treatment interruption or even permanent discontinuation. Furthermore, patients with HCC experiencing a clinically relevant (grade ≥2) irAE during ICI treatment showed an improved prognosis, in terms of OS and PFS.10 For this reason, predictive biomarkers of treatment toxicity are needed to identify patients at higher risk of irAEs, in order to implement a strict surveillance in the frame of an essential multidisciplinary approach and to reduce the risk of treatment discontinuation. At present, few clinical factors and biomarkers have been associated with the development of irAEs during treatment with ICIs. For instance, patients with prior autoimmune diseases were found to be more likely to have irAEs from ipilimumab.19 Also, patients with prior irAEs from ICIs are more likely to re-experience irAEs even from a different class of ICIs.20 For patients with HCC treated with ICIs, viral aetiology and an increased baseline ALT level were shown to be predictive of HIRAEs.21 On the other hand, genome-wide association studies may be able to identify genetic predisposition for developing irAEs, providing a decisive step forward to personalized medicine.22 Recent studies revealed that changes in circulating T-cell repertoire and B-cell repertoire preceded the development of irAEs in specific subsets of patients treated with ICIs.23–26 In clinical practice, prevention of HIRAEs can be partially achieved through a multidisciplinary approach to the patient, a comprehensive history and physical examination, blood tests and patient education, and the TBS could integrate the multidimensional personalized assessment.
As of today, there is a lack of data regarding how primary and secondary tumour infiltration could have a different impact on HIRAEs during treatment with ICIs. For this reason, we decided to analyse the incidence and severity of HIRAEs in both patients with HCC and non-HCC patients with liver metastases; however, we did not find any significant difference between the groups. Reassuringly, despite the different background of liver impairment, patients with HCC were not found to be at higher risk of developing HIRAEs compared to patients with other cancer types. Of note, the distribution of TBS values in the two cohorts was similar, showing an adequate balance in terms of hepatic tumour burden. Our findings point in the direction of a safe use of ICIs in patients with an underlying liver disease: if adequately managed within a multidisciplinary tumour board, these patients can receive ICIs without any major risk of hepatic decompensation or reactivation of a viral hepatitis.
In our study, we assessed the impact that both baseline hepatic tumour burden and HIRAEs might have on OS. Despite a higher liver tumour burden could intuitively impact the hepatic function, our study found that baseline TBS did not significantly predict survival in none of the two cohorts. These data show that an adequate patients' selection can achieve optimal survival outcomes regardless of the underlying liver invasion, further highlighting the importance of a multidisciplinary management within a tertiary referral centre and a close monitoring of the liver function during the treatment.
Looking at the impact of hepatotoxicity on OS, we did not observe any significant difference between patients who developed HIRAEs and those who did not in terms of median OS in none of the two cohorts. Although the survival rates at 1 and 2 years were higher for patients with HCC developing HIRAEs, this difference could be because of the small sample size. However, considering the evidence in literature of an improved survival for patients with HCC developing any irAEs,10 these findings warrant further evaluation in larger, prospective studies, and an increased enrolment could better characterize the possible relationship between HIRAEs, irAEs of any type and survival for patients with HCC. Regarding patients in the non-HCC cohort, the value of the conclusions that can be drawn in terms of survival is reduced by the heterogeneity of the sample. Patients enrolled in this cohort were affected by many different primary tumours, which are per se characterized by a different natural course and a different OS, and this bias might have had an important role in the lack of correlation between HIRAEs and OS.
Several limitations should be acknowledged for our study. In first instance, despite the dataset was prospectively maintained, the retrospective nature of this study negatively affects the conclusions which can be drawn. Potential biases or imbalances between groups cannot be excluded, and this study should be only considered as hypothesis generating. Also, the small samples analysed could have negatively impacted the statistical significance. For instance, the different impact on HIRAEs of ICIs used as monotherapy or in combination was not investigated because of the small sample size. Indeed, given the results achieved, further recruitment of patients is already ongoing, in order to increase the statistical power of the correlation between TBS and the development of immune-mediated hepatotoxicity, in particular in the cohort of patients with HCC treated with ICIs.
In summary, our study did not provide statistically significant results in the primary and secondary endpoints. However, our data identified a possible association between baseline TBS and the development of immune-related hepatotoxicity. Further evaluations of larger cohorts of patients are warranted to further corroborate these findings. This is especially true for patients treated with ICIs for HCC, where the achieved results, although not statistically significant, are clinically meaningful, and suggest a potential association.
CONCLUSIONSAs immunotherapy is taking over the landscape of cancer treatment, it is of vital importance to conduct studies on treatment toxicity biomarkers. The identification of predictive biomarkers would spare a subset of patients from toxicity and would allow early diagnosis and management of irAEs, decreasing their severity and the need of treatment discontinuation, thus improving patient prognosis.
Our results, even if not statistically significant, are promising and encourage us to continue our study with a larger number of patients, in order to identify simple predictive biomarkers with a potential application in clinical practice.
ACKNOWLEDGEMENTSAD is supported by the NIHR Imperial BRC and by grant funding from the European Association for this study of the Liver (Andrew Burroughs Fellowship) and from Cancer Research UK (RCCPDB-Nov21/100008).
COMPETING INTERESTSAD received educational support for congress attendance from Roche. TP received consulting fees from Bayer; and institutional research funding from Bayer, Lilly and Roche. NP received consulting fees from Amgen, Merck Serono and Servier; lectures fees from AbbVie, Gilead, Lilly and Sanofi; travel expenses from Amgen and ArQule; and institutional research funding from Basilea, Merck Serono and Servier. LR received consulting fees from Amgen, ArQule, AstraZeneca, Basilea, Bayer, BMS, Celgene, Eisai, Exelixis, Genenta, Hengrui, Incyte, Ipsen, IQVIA, Lilly, MSD, Nerviano Medical Sciences, Roche, Sanofi, Servier and Zymeworks; lecture fees from AbbVie, Amgen, Bayer, Eisai, Gilead, Incyte, Ipsen, Lilly, Merck Serono, Roche and Sanofi; travel expenses from Ipsen; and institutional research funding from Agios, ARMO BioSciences, AstraZeneca, BeiGene, Eisai, Exelixis, Fibrogen, Incyte, Ipsen, Lilly, MSD, Nerviano Medical Sciences, Roche and Zymeworks. All remaining authors have declared no conflict of interest. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. No writing assistance was utilized in the production of this manuscript.
DATA AVAILABILITY STATEMENTThe data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
© 2022. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Details
; Cammarota, Antonella 3 ; Zanuso, Valentina 3 ; Pressiani, Tiziana 4 ; Bozzarelli, Silvia 4 ; Ferrillo, Giuseppe 5 ; Vatteroni, Giulia 5 ; Pedicini, Vittorio 6 ; Giordano, Laura 7 ; Personeni, Nicola 3
; Rimassa, Lorenza 3
1 Department of Biomedical Sciences, Humanitas University, Milan, Italy
2 Department of Biomedical Sciences, Humanitas University, Milan, Italy; Department of Surgery & Cancer, Imperial College London, Hammersmith Hospital, London, UK
3 Department of Biomedical Sciences, Humanitas University, Milan, Italy; Medical Oncology and Hematology Unit, Humanitas Cancer Center, IRCCS Humanitas Research Hospital, Milan, Italy
4 Medical Oncology and Hematology Unit, Humanitas Cancer Center, IRCCS Humanitas Research Hospital, Milan, Italy
5 Department of Biomedical Sciences, Humanitas University, Milan, Italy; Department of Radiology, IRCCS Humanitas Research Hospital, Milan, Italy
6 Department of Radiology, IRCCS Humanitas Research Hospital, Milan, Italy
7 Biostatistics Unit, Humanitas Cancer Center, IRCCS Humanitas Research Hospital, Milan, Italy