Yttrium-90 Induces an Effector Memory Response with Neoantigen Clonotype Expansion: Implications for Immunotherapy

Abstract Yttrium-90 (90Y) transarterial radioembolization can safely and effectively treat hepatocellular carcinoma (HCC). Clinical trials combining 90Y with immunotherapy are aimed at improving treatment response rates. The impact of transient 90Y-induced lymphopenia on T-cell homeostasis and functional dynamics is unknown. Paired blood specimens were collected prior to first-cycle 90Y and at imaging follow-up in patients with HCC Barcelona Clinic Liver Cancer stages A–B. Flow cytometry and T-cell receptor (TCR) sequencing were used to monitor changes in T-cell subsets and TCR repertoire following 90Y. Objective response (OR) rates were determined using modified RECIST and defined as either OR or nonobjective response. Time-to-progression (TTP) was defined as progression to Barcelona Clinic Liver Cancer stage C within 6 months following 90Y. 90Y induced shifts in both CD4+ (P = 0.049) and CD8+ (P < 0.001) toward an effector memory T-cell response independent of treatment response rate. Nonresponders to 90Y were characterized by a sustained elevation in both naïve CD4+ cells (P = 0.019) and programmed cell death protein 1 expression in CD8+ cells (P = 0.003). Paired analysis of the TCR repertoire revealed a variable induction of neoantigen clonotypes and expansion of existing clonotypes independent of 90Y response. In patients with an OR, changes in TCR clonality did not influence TTP. However, polyclonal profiles in patients without an OR were associated with shorter TTP (P = 0.005; HR, 10.8) and 75% disease progression rates 6 months following treatment. 90Y induces a population shift from central to effector memory accompanied by neoantigen T-cell responses independent of treatment response rate. Monoclonal shifts in the post-90Y T-cell repertoire had superior overall TTP and improved TTP in patients with a first-cycle nonobjective response. Significance: 90Y can safely treat HCC; however, it causes transient lymphopenia. In this article, 90Y stimulates a peripheral effector memory response independent of initial treatment response. TCR sequencing revealed that polyclonal profiles in patients without an OR to treatment were associated with rapid progression rates 6 months after 90Y.


Introduction
Hepatocellular carcinoma (HCC) represents the most common type of liver cancer and ranks fourth worldwide for cancer-related deaths (1).HCC incidence rates continue to increase because of improved surveillance driven by the primary risk factors of viral hepatitis and steatotic liver disease (SLD; 2, 3).Liver transplantation remains the only curative treatment option for patients with HCC in the setting of cirrhosis.Liver-directed therapies have been used in early-to intermediate-stage disease [Barcelona Clinic Liver Cancer (BCLC) stages A-B] as a strategy to definitively treat or bridging/ downstage to liver transplantation and include microwave ablation, transarterial chemoembolization, and transarterial radioembolization.
In solitary HCC (BCLC-A), transarterial radioembolization (TARE) using yttrium-90 ( 90 Y) achieves excellent treatment response rates (4,5).Despite excellent first-cycle target response rates, 12% to 16% of patients still experience disease progression through progression of the treated lesion or the development of new lesions within the liver (4,5).Immune checkpoint inhibitors (ICI) are established as the first-line therapy for advanced stage (BCLC-C), but they represent only a modest overall improvement in the response rate compared with tyrosine kinase inhibitors with only ∼20% of patients responding to therapy (6,7).Currently, several clinical trials are underway combining 90 Y with ICI for unresectable HCC (NCT03099564, NCT06040099, NCT03033446, NCT04605731, and NCT05063565) in hopes of improving overall response rates.
Although optimal patient selection for ICI remains challenging, a deeper understanding of how 90 Y affects the immune system is required to fully appreciate the impacts of combining 90 Y with ICI to identify the responder population.Our recent work demonstrated that 90 Y-induced lymphopenia was associated with a lung shunt fraction and did not impact treatment response or disease progression rates (8).Our study also showed that patients with elevations in programmed cell death protein 1 (PD1) on peripheral CD8 + T cells following 90 Y were at increased risk of progression.
Other studies have suggested an immunomodulatory role of 90 Y for both tumor infiltrative and peripheral T cells, but with some notable limitations (9,10).These studies utilized T cells that were polyclonally (antigen-independent) reactivated in vitro, potentially obscuring their functional status at the time of isolation.Additionally, the patient population consisted of 90 Y as a bridge to resection in patients with intermediate-advanced HCC (BCLC-B and -C) in the absence of cirrhosis and therefore negating the immunomodulatory role of underlying cirrhosis.Although the available literature supports a 90 Y-dependent influence on T-cell dynamics, the specific T-cell subsets implicated as well as their responsiveness to tumor-specific antigens and effector potential remain unknown.T-cell receptor (TCR) repertoire dynamics after 90 Y represent a critical first step to characterizing these responses, but they have yet to be explored in the literature.TCR repertoires offer insights into the adaptive immune response and monitor treatmentinduced T-cell dynamics (11).
To address these gaps, we performed paired T-cell immunophenotyping and TCR sequencing on peripheral T cells from patients with BCLC A-B HCC before and after undergoing first-cycle 90 Y.This allowed for an in-depth analysis of 90 Y-induced changes to peripheral T cells and the ability to address whether 90 Y causes an immunologic change within the periphery.
Understanding the immunologic landscape prior to treatment and how 90 Y impacts this landscape could provide insights into which patients may benefit from combinational therapy.

Y-TARE treatment protocol and assessment of treatment response
All 90 Y-TARE procedures were performed at a single interventional oncology site (Ochsner Health System).HCC diagnoses were determined following review by a multidisciplinary HCC board consisting of interventional radiologists, oncologists, hepatologists, and liver transplant surgeons.
The institutional treatment algorithm for first-cycle treatment with 90 Y required unresectable HCC with an index tumor ≥3 or <3 cm and not amendable to microwave ablation.Inclusion criteria for 90 Y as first-cycle treatment for HCC included (i) Child-Pugh A-B, (ii) absence of portal vein thrombus, (iii) no extrahepatic metastasis, (iv) total bilirubin <4 mg/dL, (v) serum creatinine concentration <1.5 mg/dL, and (vi) absence of gross ascites.
The 90 Y-TARE procedure was performed as a two-phase treatment that included a mapping angiogram followed by deployment of the 90 Y microsphere infusion (TheraSphere, Boston Scientific).Vascular evaluation of the celiac, superior mesenteric, proper hepatic, and all feeding hepatic arteries to the tumor was performed during the mapping angiography.All patients received contrast-enhanced cone beam CT to confirm both coverage of tumor angiosome and to calculate the perfused volume for TARE.During mapping angiogram, technetium 99m-labeled macroaggregated albumin was used to calculate the lung shunt fraction.Medical internal radiation dose was used and incorporated perfused volumes and lung shunt fraction to calculate the desired dose to the perfused location.During radioembolization, all vessels feeding to the areas of tumor were treated with target radiation doses >200 Gy and delivered via segmental or subsegmental.
Pre-and post-90 Y dosimetry calculations were performed using Mirada DBx Build 1.2.0 and Simplicit 90 Y using pre-90 Y CT/MRI imaging with either preor post-90 Y single-photon emission computed tomography nuclear medicine and CT scans.Procedures prior to 2020 did not have post-90 Y single-photon emission computed tomography nuclear medicine and CT scans performed.
All procedures after 2019 had tumor-absorbed dose calculated from Simplicit 90 Y.All 90 Y procedures were performed by board-certified radiologists according to institution's standard protocol.
Routine follow-up imaging post-90 Y was performed using either MRI or CT.
First-cycle response to 90 Y was evaluated using the modified RECIST (mRECIST) for HCC (12) by a board-certified interventional radiologist with >8 years of experience.The target objective response rate (tORR) was evaluated as a response to the initial lesion treated during first-cycle 90 Y and defined as either an objective response (OR) corresponding to complete and partial responses of the mRECIST or nonobjective response (NOR) corresponding to stable disease and disease progression of the mRECIST.The overall objective response rate (oORR) was evaluated as a response to the treated lesion, included the development of new lesions in adjacent liver segments, and was defined as either OR or NOR.

Study endpoints
The primary endpoint was rapid time-to-progression (TTP) defined as BCLC-A or B to BCLC-C within 6 months following first-cycle 90 Y as determined by the institution's multidisciplinary HCC board.The secondary endpoints were tORR and oORR.

Blood collections and peripheral blood mononuclear cell isolation
Peripheral blood was collected immediately prior to first-cycle 90 Y, and a second blood collection occurred during routine post-90 Y imaging.All blood was collected in sodium citrate cell preparation tubes (BD Biosciences) and processed within 4 hours of collection.Peripheral blood mononuclear cells https://doi.org/10.1158/2767-9764.CRC-24-0228 | CANCER RESEARCH COMMUNICATIONS Núñez et al.

Flow cytometry
Multiple panels were used to stain for T-cell phenotypes from unstimulated PBMCs and were grouped based on markers for the following: (i) memory, (ii) senescence, (iii) exhaustion, and (iv) PD1.The memory panel consisted of the following antibodies (Life Technologies): anti-CD28 FITC (clone CD28.

TCR sequencing
Genomic DNA was isolated from cryopreserved PBMCs using QIAamp DNA Blood Midi kit (Qiagen).Extracted DNA was quantified using Qubit 4 (Thermo Fisher Scientific).TCR β-chain sequencing was performed using the immunoSEQ platform (Adaptive Biotechnologies).TCR data were analyzed using immunoSEQ Analyzer version 3.0 (Adaptive Biotechnologies).Analysis with clonotypes was based on the amino acid sequence of the TCR.Productive rearrangements refer to the number of unique clonotypes that result in a functional TCR.Out-of-frame rearrangements are the number of unique clonotypes that result in a misaligned reading frame.Stop rearrangements are the number of unique clonotypes that contain a stop codon and do not produce a functional TCR.The number of unique clonotypes within each sample was determined based on the total number of unique clonotypes in one sample compared with another.For example, unique clonotypes found in the baseline sample were absent in the post-90 Y sample.The complementarity-determining region 3 (CDR3) length of the TCRβ receptor was determined for each sample using the "CDR3 Length" algorithm in immunoSEQ, which calculates the CDR3 length for each clonotype and generates the productive frequency of each length in a given sample.Clonotypes unique to either the baseline or post-90 Y samples along with those found in both (shared) were determined using the "Combined Rearrangement" function within immunoSEQ.The percentage of clonotypes that overlapped between the baseline and post-90 Y samples was determined using the differential abundance tool which calculates the "TCR Overlap" between the two samples.TCR Overlap calculation (13)

Statistical analysis
Data analysis was performed using JMP Pro version 17 (SAS Institute Inc.).
All graphical output was generated using GraphPad Prism version 10 (GraphPad Software Inc.).Continuous variables were listed as median with IQR and categorical variables as a percentage of the total cohort.Matched pair analysis was used to determine changes in continuous variables prior to (baseline) and following 90 Y. Bivariate analysis was performed using linear regression.Two-way ANOVA with the Sidak multiple comparisons test was used to determine significance of CDR3 length.Kaplan-Meier survival curves for TTP were generated in GraphPad Prism and compared using logrank tests.

Y-TARE induced shifts in T-cell populations
Complete blood counts were monitored prior to first-cycle 90 Y and again at routine imaging follow-up.Treatment resulted in a posttreatment drop in both white blood cell as well as absolute lymphocyte count, whereas granulocyte, monocyte, and platelet counts remained stable (Supplementary Table S3).Although lymphopenia is an anticipated adverse event of 90 Y, the global decrease in white blood cell seems attributable to the post-90 Y decrease in absolute lymphocyte count.
Our results suggest that after treatment with 90 Y, there is a shift from TCM to TEM cells following 90 Y for both CD4 + and CD8 + T cells independent of cirrhosis etiology.

T-cell subsets and 90 Y-TARE response rate
To answer whether certain T-cell phenotype shifts are found in 90 Y responders versus nonresponders, patients were grouped based on target or overall response rate to first-cycle 90 Y (Table 1; Supplementary Table S5).Patients with target NOR (target nonresponders) had a higher percentage of CD4 + T cells along with elevated PD1 expression on CD8 + T cells both before and after treatment.However, the  S5).Taken together, although 90 Y induces T-cell population dynamics independent of treatment response, increased and sustained expression of the T-cell exhaustion marker PD1 was the only variable prognostic for or associated with treatment response.The derivation of T cells in the sample from the TCR sequencing results confirmed the 90 Y-induced decrease in circulating T cells at imaging follow-up (Supplementary Table S6).Although paired analysis revealed significantly lower total productive, out-of-frame, and stop rearrangements along with unique clonotypes after 90 Y, the fractional percentage remained stable (Supplementary Table S6).This suggests that the treatment-induced changes in the T-cell compartment did not impact the overall rearrangements observed within the repertoire or the fraction of unique clones within the population.
CDR3 length was investigated at baseline and following 90 Y to determine whether there was a treatment effect.Both CDR3 lengths at baseline and post-90 Y for each patient were determined using the sum frequency of clonotypes for each CDR3 length (30-60 amino acids).

Clonotypes within HCC prior to and following 90 Y
The top 10 clonotypes present in baseline samples are shown in Supplementary  Figure 3A characterizes the clonotype distribution based on their detection across both samples, but it does not account for the clonotype frequency in those samples.Although the number of unique clonotypes shared between samples was small (6%), those shared sequences made up 38% (IQR, 29%-49%) of the sample overlapping repertoire (Fig. 3B).These shared clonotype frequencies were then compared with the fractional share of the new clonotypes generated after 90 Y in the follow-up samples (Fig. 3C).The shared clonotypes made up 40% (IQR, 28%-52%) of the post-90 Y TCR repertoire with the new clonotypes representing the majority component of the repertoire.Potential neoantigens could also be present in the shared TCR repertoire and could be reactivated following treatment.Figure 3D characterizes the fractional percentage of shared TCR clonotypes categorized as expanding (>1.25fold increase), contracting (<1.25-fold decrease), or stable (fold change <±1.25).Stable (median, 42%; IQR, 6%-47%) and contracting (median, 39%; IQR, 23%-47%) clonotypes made up the major fractions with expanding clonotypes having the smallest fraction (median, 26%; IQR, 14%-45%), but with significant patient-to-patient variation.Collectively, the longitudinal clonotype analysis supports a neoantigen response that is highly variable and could contain clonotypes reactivated during treatment.
Clonotype dynamics were then examined for concordance with tORR and oORR (Table 2).The percentages of unique clonotypes within each sample (baseline, post-90 Y, or shared) were not associated with tORR or oORR.

TCR repertoire clonality and T-cell phenotypes
Monitoring clonotype frequencies provides evidence of changes within the peripheral TCR repertoire while identifying potential sequences associated with durable antitumoral responses.The TCR repertoire clonality calculation following treatment can provide an overall determination of unique clonotype abundance and characterize the circulating T-cell pool as polyclonal verses monoclonal.
Overall, 90 Y did not induce a change in population clonality (Fig. 4A).However, a subgroup of patients (36%, 26/73) showed marked increases in polyclonal or monoclonal responses characterized by a shift of clonality >5%.

Post-90 Y TCR repertoire and TTP
Both the longitudinal immunophenotyping and TCR repertoire analysis support that 90 Y activates existing and stimulates new clonotypes following treatment but in a manner not directly associated with the initial target or overall response to treatment.An objective 90 Y response, specifically first-cycle OR and durable overall OR, are strongly prognostic of HCC outcomes.Peripheral clonality and tumoral clonality are the most common data outputs from TCR sequencing analysis in studies of human cancer.Peripheral clonality was analyzed for the total cohort and then stratified based on oORR to compare outcomes associated with a posttreatment monoclonal or polyclonal TCR repertoire (Fig. 4B and D).
A polyclonal profile stratified patients with a higher risk of disease progression to BCLC-C (Fig. 4B) within 6 months post-90 Y (P ¼ 0.023; HR, 7.5).In patients with a first-cycle response to 90 Y (Fig. 4C), 6-month outcomes were excellent and not stratified by peripheral clonality (P ¼ 0.423).However, nonresponders with a polyclonal profile were at increased risk of disease progression (P ¼ 0.005; HR, 10.8) with nearly 75% experiencing disease progression within 6 months after treatment (Fig. 4D).

Discussion
90 Y is an emerging strategy for definitive treatment and bridge/downstage to surgery in HCC with BCLC stages A-B and Eastern Cooperative Oncology Group score of 0 to 1 (14).Excellent target response rates (4) and potential to improve response rates in advanced stage HCC in combination with ICI have been reported in which the treatment responder population has been difficult to predict.There is a great need to understand the immunologic impact of 90 Y to understand how this treatment affects antitumoral T-cell populations and potentially uncover which patients would benefit from the addition of ICI. 90 Y treatment is characterized by a transient period of posttreatment lymphopenia (15) that is associated with the lung shunt fraction and not with response to treatment or HCC-specific outcomes (8).
However, T-cell lineages or phenotypes that are most affected by 90 Y remain  Sustained CD8 + PD1 elevation remained the dominant immunologic risk factor for NOR, consistent with our prior results in early-stage HCC treated with liver-directed therapy (8,26).Elevated PD1 expression in these patients is notably sustained beyond the overall peripheral PD1 burst effect described by Surprisingly, only 6% of the TCR repertoire diversity persisted from baseline to post-90 Y.However, these persistent clonotypes represented 40% of the post-90 Y TCR repertoire although they also widely varied from patient to patient.
Although a dynamic T-cell response was confirmed by TCR sequencing, the frequency and abundance of new and expanding clonotypes were highly variable and ultimately not associated with ORR.Currently, there are no strategies to identify potential tumor antigens based on the clonotype sequences, though new technology aimed at antigen recognition from TCR sequences is under development.Recent studies have paired peripheral blood and tumor site TCR analysis to identify potential neoantigen or reactivated tumor antigen sequences (30,31).However, a benefit of the high ORR to 90 Y is that potential treatmentenriched candidate sequences were identified, notably one sequence (CASSL-GETQYF) was present in 88% of post-90 Y samples.These sequences could provide additional targets for chimeric antigen receptor T-cell therapy for HCC, whereas current clinical trials have focused on the protein glypican 3 and have not been completed (32).
TCR clonality offers a way to evaluate the diversity of the TCR repertoire and monitor T-cell dynamics upon interaction with an antigen.Clonality is the most commonly prognostic variable in sequencing analysis, particularly those focused on ICI response in various cancers.TCR clonality in the periphery and tumor showed a strong overlap in melanoma (33)(34)(35) as well as in HCC (30,31).Overall, the prognostic implications of clonal expansion versus population diversity have been inconsistent in the literature (11) and likely depend specifically upon the TSA pool within the repertoire and population sequenced. 90Y did not induce an overall change in repertoire clonality, although a subgroup of patients had notable shifts in clonality or diversity independent of treatment response rate.When the response rate was used as a subgrouping variable, polyclonal expansion emerged as a critical risk for impending rapid disease progression to BCLC-C.In patients with an OR and very low risk of rapid disease progression, posttreatment clonality and diversity were irrelevant and may play a major role over the course of extended follow-up.The monoclonal TCR response may indicate immune activation in a subset of TSA-specific T cells.Patients with melanoma with higher tumoral clonality and more monoclonal intratumoral TCR repertories had higher responses rates to anti-PD1 therapy (36,37).In HCC, one study found more monoclonal intratumoral TCR clonality and T-cell fractions in patients responding to anti-CTLA4 and ablation (17).The result from this study suggests that a polyclonal repertoire coupled with a lack of treatment response portend a

A
prospective, single-center study was conducted and approved by the Ochsner Health System Institutional Review Board (protocol # 2016.131.B) and was in accordance with the ethical guidelines set forth by the 1975 Declaration of Helsinki.Written informed consent was obtained from each participant with enrollment dates from November 2017 to December 2022.Study inclusion criteria were as follows: (i) confirmed HCC diagnosis by biopsy or triple-phase imaging according to the Liver Imaging Reporting and Data System criteria (v2018 American College of Radiology), (ii) unresectable HCC, (iii) BCLC stages A and B, (iv) ages ≥18 years, and (v) scheduled to undergo 90 Y-TARE as first-cycle treatment.Clinical variables were extracted from the electronic medical record following HCC diagnosis prior to treatment and at clinical follow-up (<6 months post-90 Y).Extracted variables were as follows: general demographics, cirrhosis history, model of end-stage liver disease scores, complete blood counts, and complete metabolic liver panels.Decompensation history at HCC diagnosis was defined as the presence of hepatic encephalopathy, ascites, or bleeding esophageal varices that required surgical or pharmacologic intervention.HCC burden was obtained from triple-phase CT/MRI studies following consensus review by the institution's multidisciplinary HCC board.
was performed in immunoSEQ.Expanding clonotypes were clonotypes whose abundance increased by >1.25-fold in the post-90 Y sample compared with baseline.Contracting clonotypes were clonotypes whose abundances decreased >1.25-fold in the post-90 Y sample compared with baseline.Clonotypes whose abundance did not differ between the baseline and post-90 Y samples were termed unchanged.Clonality diversity was measured using productive Simpson clonality within immunoSEQ, which measures the richness and evenness of the T-cell repertoire in each sample.

Fig. 1B
Fig. 1B and C] cells.The percentage of central memory T (TCM) and TEM cells did not change following treatment (Fig. 1D and E).Senescent and exhaustion markers were also investigated for treatment-induced expression changes.The percentage of CD57 + and KLRG1 + CD4 + T cells did not change following treatment (Fig. 1F and G).Elevations of cytotoxic T-lymphocyte-associated protein 4 [CTLA4; 185 (IQR, 178-192)

Figure 2 JFIGURE 1 FIGURE 2
Figure 2 displays shifts in CD8 + T-cell phenotypes at baseline and after 90 Y.As observed with CD4 + T cells, there was no change in the percentage of CD8 + T cells (Fig. 2A) or naïve (Fig. 2B) after treatment.

FIGURE 3
FIGURE 3 90 Y-TARE induces new clonotypes and the expansion and contraction of existing clonotypes.A, Unique clonotypes in the baseline (blue) and post-90 Y (red) and shared clonotypes between baseline and post-90 Y (purple).B, Percentage of the TCR repertoire that overlaps between baseline and post-90 Y. C, Sum frequency of clonotypes post-90 Y that were new (red) or remained from baseline (purple).D, Percentage of clonotypes shared between baseline and post-90 Y that were either expanding (green), contracting (orange), or remained unchanged (gray).

FIGURE 4
FIGURE 4 TCR repertoire clonality and impacts on TTP.A, Changes in TCR repertoire measured as clonality from baseline to post-90 Y. Shifts in the T-cell repertoire toward more monoclonal or polyclonal were determined by comparing clonality changes from baseline to post-90 Y. B, Patients with shifts toward more polyclonal TCR repertoires had shorter rapid TTP.C, Patients with ORs to 90 Y had excellent TTP.D, Patients with NORs to 90 Y and shifts toward a polyclonal TCR repertoire had worse TTP.

90Y
has been shown to activate T cells both within the tumor microenvironment (TME) and in the periphery(9,10), providing evidence of immunomodulation.However, TME evidence is limited to postresected BCLC-0 (very early) tumors evaluated more than 6 months after 90 Y and in the absence of cirrhosisassociated immune dysfunction.Posttreatment tumor sampling is contraindicated in BCLC A-B due to the risk of tumor-related complications including track seeding.Therefore, immune monitoring strategies must ideally use longitudinal, noninvasive sampling to characterize the T-cell responses.Prior evidence confirms that peripheral T cells provide insights to TME response and serve as biomarkers for ICI response(16)(17)(18). Overlap of TCR repertoire and tumor antigen specificities was found between the periphery and the TME in patients with melanoma(19), providing further evidence of the correlation between the periphery and TME immune landscape.Memory T-cell subsets (stem, central, and effector memory) are central to the antitumoral effector response and serve as a critical biomarker for ICI response (see review ref.20).The first cycle of 90 Y induced a CD8 + TEM response in 66% patients at a median follow-up of 75 days.TEM cells recirculate in the blood, traffic to peripheral tissues(21), and produce antitumoral effector molecules upon antigen-specific activation(22).Although TEM cells are more durable and rapidly secrete larger amounts of IFNγ and perforin upon activation compared with TCM cells(23,24), in mouse models, TCM cells have been shown to have better antitumoral activity(25).TCM cells are denoted for expressing lymph node homing properties through CCR7 expression compared with TEM cells(23), which are CCR7 � .Whether the decrease in TCM cells post-90 Y was due to differentiation into TEM cells or whether activation led to relocation to the lymph nodes is unknown.Regardless, the effects of 90 Y on circulating memory T cells were independent of cirrhosis etiology, the degree of lymphopenia, and treatment response.Rivoltini and colleagues(10) similarly reported an increase in the proliferation marker Ki67 and the T-cell effector molecule granzyme B 1 month following 90 Y, supporting a treatment-induced effector response.Their results also captured the post-90 Y exhaustion signature of CTLA4 and LAG3(10) observed at treatment follow-up in this study.Their study identified only T-cell population (CD4 + Ki67 + GranB + ) to be associated with treatment response, with the posttreatment immune profile not ultimately found to be predictive of long-term disease control, in agreement with our results.The difference in treatment prognostic immune populations could be linked to differences in disease staging (BCLC B-C vs. BCLC A-B) and the corresponding ORR (unreported ∼65% vs. 84%).

Rivoltini and colleagues 1
month following 90 Y.The sustained expression of PD1 in these patients could indicate a treatment responder population who may benefit from combination 90 Y-ICI directly targeting T-cell exhaustion (PD1, CTLA4, LAG3, and TGIT).To make matters more challenging, HCC is known to have different immune signatures (27) likely due to underlying etiology and cirrhosis-associated immune dysfunction, which may impact which patients express an exhausted phenotype within the TME.Although overall response rates for anti-PD1 have remained low (<20%; ref. 28), perhaps selection of patients expressing PD1 would improve response rates.Definitive treatment in 90 Y aims to eradicate tumor burden while eliciting an immunogenic response to promote a durable antitumoral response.Although ablation has been shown to release tumor-specific antigens (TSA) in HCC (29), studies of TSA release following 90 Y are lacking.The TCR repertoire holds key information on antigen exposure and could provide insights into TSA release with TCR sequencing serving to identify antigen biomarker signatures associated with durable treatment outcomes.Neither 90 Y treatment nor the lymphodepleting effect of 90 Y impaired the generation of functional TCRs or the percentage of unique clonotypes in the circulating T-cell population.The TCR repertoire was more diverse at treatment baseline with roughly 1/3 of the post-90 Y repertoire representing unique clonotypes with respect to baseline, although the degree of response widely varied across the cohort.The generation of new clonotypes post-90 Y was not impacted by treatment-induced lymphopenia.

AACRJournals.org Cancer Res Commun; 4 ( 8 )
August 2024 2171 high disease progression risk and may identify a potential treatment responder group to combine with ICI.There are some limitations in this study. 90Y follow-up was collected at the time of imaging follow-up and varied among patients based upon the need to expedite sequential therapy or other unforeseeable circumstances.Memory T-cell phenotypes were, therefore, prioritized over TEM cell analysis, but it is recognized that these populations may remain in circulation beyond the prime response window identified by Rivoltini and colleagues.Although TCR sequencing was strengthened by pairing a treatment baseline, sequencing was performed on the isolated peripheral T cells without sorting based on the memory T-cell phenotype.Isolating the TCR repertoire of the TEM cell fraction would increase the relevance of the isolated sequences with an equal importance of evaluating the sequences associated with sustained elevation in PD1 linked to the treatment nonresponse rate.In conclusion, T-cell lineage shifts were observed following first-cycle 90 Y with TCR sequencing analysis confirming dynamic immune fluctuations following treatment.A larger database of peripheral T cells and tumor-infiltrating lymphocyte sequences associated with response rate and durable HCC treatment outcomes may help identify targetable TSA to globally improve response rates.The polyclonal immune signature in patients not responding to first-cycle 90 Y and associated rapid disease progression as well as inferior outcomes associated with sustained PD1 expression suggests a target population in BCLC A-B who may benefit from an immediate switch to ICI.

TABLE 1
90Y induced changes in T cell phenotypes based on target response rate Abbreviations: ALC, absolute lymphocyte count; CTLA4, cytotoxic T-lymphocyte associated protein 4; IQR, interquartile range; KLGR1, killer lectin-like receptor G1; LAG3, lymphocyte activation gene 3; MFI, median fluorescence intensity; NOR, non-objective response rate; ORR, objective response rate; PD-1, programmed cell death-1; T CM , central memory T cell; T EM , effector memory T cells; T SCM , memory stem T cell; 90 Y, Yttrium-90.90Y-induced shifts in naïve, TSCM, TCM, and TEM cells were independent of response to treatment.Similarly, patients with overall NOR (overall nonresponders) to first-cycle 90 Y had elevated PD1 expression on CD8 + T cells before and after treatment as well as a significant increase in CTLA4 expression on CD8 + T cells post-90 Y (Supplementary Table

Table S7
Effect of 90 Y on TCR repertoireAlthough productive rearrangement and unique clonotype frequencies were similar following 90 Y, the posttreatment expansion in effector memory

TABLE 2
Unique T cell receptor clonotypes based on 90 Y response rate Núñez et al. unclear.In this study, T-cell subsets and TCR repertoire were examined before and after 90 Y in which treatment demonstrated distinct evidence of the immunologic response in peripheral T cells.