Article Text
PDF
Abstract
Background Comprehensive profiling of autoantibodies (AAbs) in metastatic urothelial cancer (mUC) has not been performed to date. This may aid in diagnosis of UC, uncover novel therapeutic targets in this disease as well as identify associations between AAbs and response and toxicity to systemic therapies.
Methods We used serum from patients with mUC collected prior to and after systemic therapy (immune checkpoint inhibitor (ICI) or platinum-based chemotherapy (PBC)) at Dana-Farber Cancer Institute. 38 age-matched and sex-matched healthy controls (HCs) from healthy blood donors were also evaluated. The SeroTag immuno-oncology discovery array (Oncimmune) was used, with quantification of the AAb reactivity toward 1132 antigens. Bound AAbs were detected using an anti-immunoglobulin G-specific detection antibody conjugated to the fluorescent reporter dye phycoerythrin. The AAb reactivity was reported as the median fluorescence intensity for each color and sample using a Luminex FlexMAP3D analyzer. Clinical outcomes of interest included radiographic response and development of immune-related adverse events (irAEs). Significance analysis of microarray was used to compare mUC versus HC and radiographic response. Associations with irAE were evaluated using a logistic regression model. P<0.05 was considered statistically significant.
Results 66 patients were included with a median age of 68 years; 54 patients (82%) received ICI and 12 patients (18%) received PBC. Compared with HCs, AAbs against the cancer/testis antigens (CTAG1B, CTAG2, MAGEB18), HSPA1A, TP53, KRAS, and FGFR3 were significantly elevated in patients with mUC. AAbs against BRCA2, TP53, and CTNBB1 were associated with response, and those against BICD2 and UACA were associated with resistance to ICI therapy. AAbs against MITF, CDH3, and KDM4A were associated with development of irAEs in patient who received an ICI. A higher variance in pre-to-post treatment fold change in AAb levels was seen in patients treated with ICI versus PBC and was associated with response to ICI.
Conclusions This is the first report of comprehensive AAb profiling of patients with mUC and identified key AAbs that were elevated in patients with mUC versus HCs as well as AAbs associated with therapeutic response to ICI. These findings are hypothesis generating and further mechanistic studies evaluating humoral immunity in UC are required.
- Immunotherapy
- Urinary Bladder Neoplasms
- Immunity, Humoral
- Autoimmunity
Data availability statement
Data are available upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Introduction
Urothelial cancer (UC) is the seventh most common cancer in the USA, with more than 80 000 cases predicted in 2023.1 Metastatic UC (mUC) is an incurable condition, with the mainstay of treatment being systemic therapy, including platinum-based chemotherapy (PBC), immune checkpoint inhibitors (ICIs), and antibody–drug conjugates. Despite these therapies, median survival for mUC remains in the order of 1–2 years.2
The role of the adaptive immune system in cancer has been explored in depth and led to the development of ICIs, which stimulate T-cell responses against tumors. There is much less known about the role of the humoral immune system in cancer, though distinct autoantibody (AAb) signatures have been shown to be detectable prior to the first clinically detectable signs of the cancer in patients with prostate,3 breast,4 and lung cancers5 and potentially have roles in tumor detection as well as prognosis. In UC, AAbs against specific tumor-associated antigens such as matrix metalloproteinase-26 and protein phosphate 1, catalytic subunit, alpha isoform7 have been shown to be associated with outcomes in localized disease, while the presence of antinuclear antibodies at high levels was associated with development of immune-related adverse events (irAEs) in a cohort of patients receiving pembrolizumab for mUC.8
We therefore sought to comprehensively evaluate the AAb profile using a multiplexed approach in patients with mUC, with the aims being to uncover AAbs present at significantly higher levels in patients versus healthy controls (HCs), evaluating the associations between AAbs and response to systemic therapy, progression-free survival (PFS), development of irAEs, tumor mutational burden (TMB), and exploring the changes in the AA profile during therapy. These findings may provide novel insights into B cell and humoral AAb response as a biomarker for predicting clinical response, TMB, and irAEs in mUC and potentially uncover novel therapeutic targets.
Methods
The study cohort comprises patients with mUC who were treated at Dana-Farber Cancer Institute, Boston, Massachusetts, USA, had banked serum available before and after the receipt of ICI and/or PBC, and had provided consent for their clinical data and biospecimens to be used for research purposes. In total, 38 age-matched and sex-matched samples form healthy blood donors (HCs) without any known malignancy were supplied by the Bavarian Red Cross Biobank.
AAb profiling was performed using serum dilutions of 1:100 and employing a previously described methodology.9 The SeroTag multiplex-based immunoassay platform was interrogated to detect humoral immunoglobulin G (IgG) response against 1132 antigens using IgG-specific antibody conjugated to phycoerythrin.
The association of AAb profiling with objective radiographic response, PFS, TMB, and irAEs was evaluated. Significance analysis of microarray10 (SAM, p<0.05, |D|>2.0 and D>2.0, respectively) was used to compare mUC versus HC and radiographic response (complete response (CR) and partial response (PR)) vs stable disease (SD) and progressive disease (PD)). A log rank test based on Kaplan-Meier estimator was used to evaluate associations with PFS (p<0.05; |delta|>20; group split: 25% ‘high’ median fluorescence intensity (MFI), 75% ‘low’ MFI). Associations with irAE was evaluated using a logistic regression model (irAE~log2 MFI(AAb); p<0.05).
Tumor profiling employing next-generation sequencing of a panel of up to 454 genes11 was available in a subset of 48 patients and the association of TMB with AAb profile was investigated using a Harrell-Davis quantile estimator (Q75 and Q85 p<pcrit, |log2 fold change (FC)|≥1).
Longitudinal AAb changes was evaluated by calculating pre–post treatment FCs. One post-treatment sample was identified as an outlier and removed from the analysis. For each patient, the variance of pre–post FCs was calculated and the variance distributions for ICI versus PBC and response were compared using the Kolmogorov–Smirnov test.
Results
A total of 66 patients formed the study cohort (table 1). Median age was 68 years, and 51 patients (77%) were men. In total, 54 patients (82%) received ICI and 12 patients (18%) received PBC. The median duration between the baseline and post-therapy serum samples was 5 months. Overall, among 61 patients in whom response was evaluable, 24 (39%) had a CR or PR, 19 (31%) had SD and 18 (30%) had PD as best response to therapy.
Clinical characteristics of the cohort
SAM analysis revealed that 18 AAbs were observed at significantly higher levels in patients with mUC compared with HC (online supplemental table 1). Key AAbs that were found at significantly higher levels in patients with mUC included those against the cancer/testis antigens (CTAG1B, CTAG2, MAGEB18), HSPA1A, TP53, KRAS, and FGFR3.
Supplemental material
Figure 1 shows a heatmap of AAbs present at baseline that were associated with radiographic response to ICI. The top AAbs associated with lack of response to ICI were those directed against BICD2 and UACA, whereas key AAbs associated with response included those against SPTBN1, FOXO1, CTNBB1, BRCA2, and TP53. Baseline AAbs associated with PFS among ICI-treated patients are shown in online supplemental table 2; AAbs directed against SPTBN1, TP53, and HSPA1B were associated with improved PFS, while those against GNAI2 were associated with poorer PFS. Kaplan-Meier curves showing PFS stratified by ‘high’ versus ‘low’ AAb levels of anti-TP53 anti-SPTBN1 are shown in online supplemental figure 1.
Heatmap of AAb reactivities associated with response to therapy in ICI-treated patients. AAb, autoantibody; CR, complete response; ICI, immune checkpoint inhibitor; PD, progressive disease; PR, partial response; SD, stable disease.
The mutational landscape among 48 patients with available data is shown in online supplemental figure 2. The most frequently mutated genes were TP53 (62%), ARID1A (35%) KMT2D (29%), RB1 (27%), and KDM6A (25%). The association of tumor genomics with AAb profiling was studied among all 39 frequently mutated genes. Only TP53 alterations were significantly associated with detection of anti-TP53 AAb in mutated patients, and anti-TP53 AAb was associated with response (CR/PR) in this subcohort (online supplemental figure 3). AAbs associated with higher TMB included those against CTNNB1, PGR, MLANA, and SERPINB3, while those against PRL and ADRB2 were associated with lower TMB (online supplemental figure 4).
In total, 24 patients treated with ICI (45%) developed an irAE during therapy, and key AAbs present at baseline associated with development of any-grade irAE included those against MITF, CDH3, NPM3, KDM4A, and SNRPF (figure 2). AAbs against IGF2BP1, AXIN1A, AGGF1, and RALY were associated with a decreased risk of developing irAEs.
Heatmap of AAb reactivities associated with development of irAEs in ICI-treated patients. AAb, autoantibody; ICI, immune checkpoint inhibitor; irAEs, immune-related adverse events.
The log-FC in AAb levels between baseline and post-therapy across all patients is shown in online supplemental figure 5. The comparison of pretreatment and post-treatment AAb profiles revealed interindividual differences in the magnitude, trend, and targets of the treatment-induced AAbs among patients, reflecting individual patterns in loss of tolerance to self-antigen and tumor antigen during therapy. A higher variance in pre–post FC in AAb levels was observed in those treated with ICI compared with PBC (figure 3A), and this was also associated with response to ICI therapy (figure 3B).
Variance in fold change (FC) of pre–post AAb profiles in patients treated with ICI and PBC (A), and association in FC variance with response in patients treated with ICI (B). AAb, autoantibody; ICI, immune checkpoint inhibitor; PBC, platinum-based chemotherapy.
Discussion
In this study, we used a high-throughput multiplexed immunoassay platform with over 1100 antigens available for interrogation to comprehensively evaluate the AAb landscape in patients with mUC. This is the first such study in mUC and our findings have several avenues for potential clinical and translational relevance.
First, we noted key AAbs that were present at significantly higher levels in patients with mUC compared with HCs. These included AAbs against the cancer/testis antigens which are known to be expressed in several cancer types, including UC.12 These have been shown to be particularly expressed in higher grade/stage UC and associated with poorer outcomes.13 Humoral responses against these has been detected in a variety of solid tumors,14 but ours is the first study to detect AAbs to these in UC. Further studies are needed to determine whether these could be used to aid with diagnosis of UC or potentially in monitoring for disease recurrence after definitive therapy. Moreover, our findings provide a rationale for evaluating cancer/testis antigen-based immunotherapeutic strategies in UC, clinical studies of which are currently ongoing.15
Second, we noted that AAs against key tumor suppressor genes such as BRCA2, TP53, SPTBN1, FOXO1, and CTNNB1 were associated with response to therapy. Many of these genes are involved in the DNA damage repair pathway, including BRCA2, TP53, and SPTBN1, defects in which have previously been shown to be associated with response to PBC16 and ICI17 in UC. This provides a rationale for the clinical study of combination strategies involving Poly (ADP-ribose) polymerase (PARP) inhibition with ICI, in particular. If validated, our data also suggest that use of a simple blood-based test may provide a suitable biomarker to inform therapy selection.
The relationship between the cellular and humoral immune systems in mediating development of irAEs in patients receiving ICI is complex. While studies have generally reported a positive correlation between the presence of baseline AAbs and irAEs,18 including a prior study in mUC,8 recent work has suggested that it may be an early change (or increase) in AAbs that are associated with irAEs.19 Here, we noted that baseline AAbs against specific antigens, notably CDH3 and MITF, were particularly associated with development of irAEs. CDH3 upregulation has been observed in patients who developed immunotherapy-related colitis on dual ICI therapy,20 while MITF is a transcription factor involved in the regulation of the immune response and antigen presentation by the immune system.21 These observations therefore have biologic plausibility and further work to elucidate potential mechanisms underlying development of irAE based on the AAbs is needed.
Finally, our observation that the FC in AAb levels was greater in patients receiving ICI compared with PBC is likely indicative of the activation of the humoral immune system by ICI. Moreover, a higher FC was associated with response to ICI, which may potentially reflect a greater degree of humoral activation in patients with greater T-cell activation from ICI. A possible explanation may be that host tissue damage by cytotoxic CD8 T cells releases antigens to induce de novo humoral autoimmunity or may trigger expansion of memory B cells that previously circulated only at low levels. These findings support the hypothesis that the humoral immune system may mediate some of the effects of checkpoint blockade, since it is known that subsets of B cells express PD-1, PD-L1, and CTLA-4 which can therefore be targeted by ICIs.22 23
Despite this being the first study to evaluate the AAb landscape in mUC and explore its longitudinal associations with outcomes and toxicity on ICI as well as comparisons to PBC, there are key limitations to our work. This was a small observational cohort and the number of patients who received PBC was low, while there was heterogeneity in terms of the ICI used in the study (including monotherapy and dual therapy). The array capacity was limited to ~1100 antigens, and we could have overlooked other relevant antigens. This work is also purely descriptive in providing associations between AAbs and outcomes, and further studies are needed to validate these findings but also to provide greater translational and mechanistic insights.
In summary, we used a multiplexed bead-based immunoassay to evaluate the AAb landscape in patients with mUC receiving ICI and PBC using minimal amounts of serum. We noted key AAbs elevated in patients compared with HCs, while also seeing associations between AAbs against known oncogenic drivers and response to ICI. The overall FC in AAb during therapy was greater in patients receiving ICI compared with PBC and was associated with response to ICI. These findings are hypothesis generating and further validation and mechanistic studies are required to better understand the role of the humoral immune response in mediating responses and toxicity to ICI therapy in mUC, as well as permitting development of non-invasive biomarkers using AAb profiles as prognostic and/or predictive biomarkers to inform therapy.
Data availability statement
Data are available upon reasonable request.
Ethics approval
This study involves human participants and was approved by DF/HCC IRB 02-021 and 19-583. Participants gave informed consent to participate in the study before taking part.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Twitter @prafulravi1, @PetraBudde, @sonpavde
Contributors Design/conception: PR, GS, BAA, and MB. Acquisition of data: PR, DF, IE, and JT. Provision of patients: PR, AR, CM, BAM, JEB, and GS. Analysis of data: PB, BAA, ER, JG, AS-S, ALK, MB, and H-DZ. Drafting of manuscript: PR, BAA, and MB. Revision of manuscript/approval to be published: all authors.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests PR: research funding (to institution) from Lilly, Bayer, and Telix and speaker’s fees from OncLive. BAM: consulting fees from BMS, Eisai, Exelixis, Gilead, Pfizer, and Seagen and research funding (to institution) from BMS, Exelixis, Pfizer, and Seagen. JEB: speaker honoraria from Guardant Health; consulting fees from Guardant Health, Genome Medical, Oncotect, Precede, TracerDx, Musculo, and JucaBio; equity in Cityblock Health, Genome Medical, Oncotect, Precede, TracerDx, and Musculo; and filed an institutional patent on methods to detect neuroendocrine prostate cancer through tissue-informed cell-free DNA methylation analysis. GS: in advisory board of BMS, Genentech, EMD Serono, Merck, AstraZeneca, Sanofi, Seattle Genetics/Astellas, AstraZeneca, Exelixis, Janssen, Bicycle Therapeutics, Pfizer, Gilead, Scholar Rock, G1 Therapeutics, Eli Lilly/Loxo Oncology, Infinity Pharmaceuticals, Lucence Health, IMV, Vial, Syapse, Tempus, Ellipses Pharma, PrecisCa, and Primum; in consultant/scientific advisory board of Suba Therapeutics, Syapse, Servier, Merck, and Syncorp; received research support (to institution) from Sanofi, AstraZeneca, Gilead, Helsinn, Lucence, BMS, EMD Serono, and Jazz Therapeutics; received speaker's fees from Seagen, Gilead, Natera, Exelixis, Janssen, Bayer, and Aveo; received data safety monitoring committee honorarium from Mereo; received writing/editor fees from UpToDate, Practice Update, and Onviv; and spouse employed in Myriad. PB, BAA, ER, JG, AS-S, ALK, MB, and H-DZ are employees of Oncimmune.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.