Abstract
Background
CYLD, a tumor suppressor gene encoding a deubiquitinase enzyme, regulates tumorigenesis via NF-κB signaling. CYLD mutations are linked to immune evasion and PD-L1 upregulation, particularly in virus-associated cancers. We evaluated CYLD alterations across cancers and their correlation with PD-L1 expression. Given the established link between CYLD dysfunction and viral oncogenesis, we also assessed the co-occurrence of CYLD alterations and human papillomavirus (HPV) status.
Methods
We conducted a real-world data analysis of CYLD alterations in 145,991 tumor samples that underwent clinical-grade comprehensive genomic profiling during routine care; PD-L1 expression via immunohistochemistry (IHC) was assessed in 10,053 specimens. The relationship between CYLD alterations and high-risk HPV across cancers was also evaluated.
Results
CYLD mutations occurred in 0.57% (836/145,991) of tumors, with the highest rates in thymic carcinomas (15.4%, 34/220), anal carcinomas (11.4%, 55/479), and head and neck cancers (6.4%, 154/2,423). Among 10,053 samples with PD-L1 IHC evaluated, 62 (0.6%) harbored CYLD alterations. Overall, 61% (38/62) of CYLD-altered tumors were positive for PD-L1 IHC expression (≥ 1%), with 29% (18/62) exhibiting high PD-L1 expression (≥ 50%). In addition, 27.5% (230/836) of CYLD-altered tumors were HPV+, with strong enrichment in head and neck cancers (odds ratio [OR] = 7.4; p < 0.0001).
Conclusions
We highlight the rarity of CYLD mutations across most cancers but underscore their relatively elevated prevalence in thymic carcinomas, anal carcinomas, and head and neck cancers. The observed associations between CYLD alterations, PD-L1 expression, and HPV positivity suggest that CYLD status, combined with immune and viral biomarkers, may merit exploration as a predictor of immunotherapy benefits.
INTRODUCTION
CYLD, a tumor suppressor gene encoding a deubiquitinase (DUB) enzyme, plays a crucial role in tumor suppression by negatively regulating NF-κB signaling through the removal of Lys63-linked ubiquitin chains from signaling proteins. This regulation is essential in controlling inflammation and immune responses, which are key contributors to cancer progression (1-4). Germline alterations in CYLD are associated with CYLD cutaneous syndrome, characterized by multiple benign skin tumors, while somatic mutations have been detected across various human cancers. Recent studies have also explored the relationship between CYLD mutations and immune evasion mechanisms, particularly through PD-L1 upregulation, which is central to immune suppression in cancers such as human papilloma virus (HPV) positive head and neck squamous cell carcinoma (5).
The loss of CYLD function results in increased NF-κB activity and PD-L1 overexpression, which fosters immune suppression and tumor progression. This mechanism is especially relevant in cancers such as HPV-positive head and neck, cervical, and anal cancers, where CYLD mutations, along with NF-κB hyperactivation, contribute to immune evasion and oncogenesis (6-9). Similarly, thymic epithelial tumors have been linked to Epstein-Barr Virus (EBV)-associated oncogenesis, where immune evasion mechanisms, including PD-L1 overexpression, may play a pivotal role (10). The relevance of PD-1/PD-L1 biology in HPV-positive cancers has been further emphasized by Mishra and colleagues, who noted the growing importance of PD-1/PD-L1 checkpoint inhibitors in HPV-associated oral carcinogenesis (11).
In the current study, we analyzed CYLD mutations across a broad spectrum of cancers profiled using a comprehensive genomic profiling assay and explored their association with PD-L1 expression levels via immunohistochemistry (IHC). Our goal was to better understand the landscape of CYLD mutations and their implications for immunotherapy.
METHODS
The study included a retrospective analyses of a cohort comprising 145,991 patients with solid tumors who received comprehensive genomic profiling (CGP) using the laboratory-developed test FoundationOne® assay as part of routine clinical care. Briefly, DNA was extracted from formalin-fixed, paraffin-embedded (FFPE) tissue sections, and CGP was performed on hybridization-captured, adapter ligation-based libraries of ~ 315 cancer-related and select introns of genes commonly rearranged in cancer, as described previously (12). Analysis of sequencing data and identification of different classes of genomic alterations was carried out as described previously (12-14). The identified alteration types include short variants (base substitutions and indels), copy number alterations, and rearrangements. Variants of unknown significance were not included (14). Only one sample per patient was included in this study.
PD-L1 expression was determined using IHC performed on FFPE tissue sections using the Dako 22C3 PD-L1 antibody. PD-L1 expression was binned into three categories based on the percentage of tumor cells staining with ≥ 1 + intensity (TPS): negative (< 1%), low positive (1–49%), or high positive (≥ 50%). Data on PD-L1 expression was available for 10,053 patients.
A two-step approach was used to detect HPV in this cohort. First, a computational algorithm leveraging the sequencing coverage on a list of pathogens, designed in the hybrid capture assay, is used. Second, a de novo assembly of off-target sequencing reads left unmapped to the human reference genome (hg19) as described previously. These assembled contigs were competitively mapped by BLASTn (BLAST+ v2.6.0) to NCBI database of viral nucleotide sequences to detect oncoviruses, including the HPV types. A positive status was determined by contigs ≥ 80 nucleotides in length and with ≥ 97% identity to the BLAST sequence. Samples that were positive for any of the following high-risk HPV types by either approach were classified as HPV+: HPV-16, HPV-18, HPV-31, HPV-33, HPV-35, HPV-39, HPV-45, HPV-51, HPV-52, HPV-56, HPV-58, HPV-59. These annotations are for research use only. In a prior study, HPV status of samples was confirmed using p16 staining. (15-17) Approval for this study, including a waiver of informed consent and Health Insurance Portability and Accountability Act (HIPAA) waiver of authorization was obtained from the Western Institutional Review Board (WCG IRB, Protocol number: 20152817).
Prevalence of PD-L1 expression among CYLD-altered vs CYLD wild type tumors was compared using a Fisher’s exact test in the pan-solid cohort as well as in each tumor type. Co-occurrence analysis for the presence of high-risk HPV among CYLD altered vs unaltered tumors was also done using a Fisher’s exact test in the overall cohort as well as within each tumor type. Comparisons were done between high-risk HPV + vs HPV− cases. Cases with only low-risk HPV were excluded from this analysis; 145,719/145,991 samples were eligible for this analysis. All P-values were adjusted for multiple hypothesis correction using the Benjamini-Hochberg FDR procedure. All reported P-values in this study are after FDR correction, unless stated otherwise. Statistical significance was set at an FDR-corrected P ≤ 0.05. Statistical analyses and plotting were carried out using Python 2.7.18 (Python Software Foundation) and R 4.2.3 (R Foundation for Statistical Computing).
RESULTS
Prevalence of CYLD Alterations Across Cancers
CYLD alterations were identified in 0.57% (n = 836/145,991) of all patients with solid tumors, indicating that these alterations are relatively rare in the broader cancer population (Table 1, Fig. 1A). Among the different tumor types, CYLD was most commonly altered in thymic carcinomas (15.4%; N = 34/220 patients), followed by anal carcinomas (11.4%; N = 55/479 patients), and head and neck cancers (6.4%; N = 154/2,423 patients) (Table 1, Fig. 1A). Among the CYLD altered cases (n = 836), samples exhibited CYLD short variants (76.8%; n = 642/836), copy number homozygous deletions (19%; n = 156/836) and rearrangements (6%; n = 50/836), with some cases showing multiple of these alteration types (Supplementary Table 1). The most common short variants affecting CYLD were nonsense and frameshift alterations with the most common alterations being in codons 371, 722, 42, 605 and 397 (Fig. 1B, Supplementary Table 1). For example, the nonsense alteration S371* was found in 30 patients in the overall cohort, followed by N722fs*13 found in 20 patients.
Table 1.
Prevalence of pathogenic CYLD alterations across tumor types (N = 145,991 cancers)
| Tumor type | Number of total samples |
Number of samples with altered CYLD |
CYLD alteration prevalence |
|---|---|---|---|
| Pan-solid cancers | 145991 | 836 | 0.57% |
| Thymic carcinoma | 220 | 34 | 15.45% |
| Anus | 479 | 55 | 11.48% |
| Head and neck | 2423 | 154 | 6.36% |
| Skin | 753 | 14 | 1.86% |
| Cervix | 1067 | 11 | 1.03% |
| Unknown primary carcinoma (CUP) | 8543 | 80 | 0.94% |
| Non-small cell lung carcinoma (NSCLC) | 29986 | 166 | 0.55% |
| Bladder | 2224 | 11 | 0.49% |
| Colorectal (CRC) | 17467 | 86 | 0.49% |
| Esophagus | 3587 | 16 | 0.45% |
| Melanoma | 4411 | 19 | 0.43% |
| Endometrial (uterine) | 4066 | 17 | 0.42% |
| Cholangiocarcinoma | 3167 | 10 | 0.32% |
| Breast | 15252 | 44 | 0.29% |
| Ovary | 8270 | 16 | 0.19% |
| Glioma | 6752 | 13 | 0.19% |
| Pancreas | 8188 | 11 | 0.13% |
Figure 1. Spectrum of CYLD alterations across tumor types.
A. Proportion of samples with CYLD alterations in the pan-solid cohort and across specific tumor types. Tumor types with at least 10 cases with alterations in CYLD are shown. B. Lollipop plot showing the distribution of CYLD short variants detected pan-tumor.
Abbreviations: CUP: Unknown primary carcinoma; NSCLC: Non-small cell lung carcinoma; CRC: Colorectal cancer.
Association of CYLD Alterations and High PD-L1 Expression
Among the 10,053 samples with PD-L1 expression available from IHC, 62 (0.6%) harbored CYLD alterations. Almost 61% (38/62) of CYLD-altered tumors were positive for PD-L1 expression (≥ 1%) compared to 44% (4,400/9,991) of CYLD-negative tumors, showing an association between CYLD alterations and PD-L1 expression (OR: 2.01; 95% CI: 1.2–3.5; P < 0.01). Notably, among tumors expressing high PD-L1 (≥ 50%), the association with CYLD alterations showed a similar trend but did not reach statistical significance (OR: 2.03, 95% CI: 1.0-3.9; P = 0.069), with 29% (18/62) of these tumors exhibiting high PD-L1 expression (Table 2, Fig. 2). In thymomas, two of the 34 CYLD-mutated cases also had PD-L1 expression data available, and both cases demonstrated high PD-L1 expression (≥ 50%), further supporting a possible link between CYLD dysfunction and immune checkpoint activation. In other tumor types, including anal carcinomas and head and neck cancers, we were limited by cohort size to conduct additional statistical analyses.
Table 2.
Breakdown of PD-L1 immunohistochemistry expression prevalence by CYLD mutation status across top tumor types.*
| Tumor type | Mutation cohort |
PD-L1 available (n) |
PD-L1 negative (n) |
PD-L1+ (1–49%) (n) |
PD-L1+ (≥ 50%) (n) |
|---|---|---|---|---|---|
| Pan-solid | CYLD+ | 62 | 24 | 20 | 18 |
| Pan-solid | CYLD− | 9991 | 5591 | 2342 | 2058 |
| Thymic carcinoma | CYLD+ | 2 | 0 | 0 | 2 |
| Thymic carcinoma | CYLD− | 5 | 1 | 2 | 2 |
| anus | CYLD+ | 3 | 1 | 2 | 0 |
| anus | CYLD− | 20 | 7 | 11 | 2 |
| head and neck | CYLD+ | 4 | 2 | 1 | 1 |
| head and neck | CYLD− | 109 | 49 | 36 | 24 |
| bladder | CYLD+ | 3 | 1 | 1 | 1 |
| bladder | CYLD− | 78 | 49 | 19 | 10 |
| cervix | CYLD+ | 1 | 1 | 0 | 0 |
| cervix | CYLD− | 42 | 25 | 11 | 6 |
| Unknown primary carcinoma (CUP) | CYLD+ | 6 | 4 | 2 | 0 |
| Unknown primary carcinoma (CUP) | CYLD− | 320 | 199 | 56 | 65 |
| esophagus | CYLD+ | 1 | 1 | 0 | 0 |
| esophagus | CYLD− | 147 | 113 | 24 | 10 |
| Non-small cell lung carcinoma (NSCLC) | CYLD+ | 37 | 11 | 12 | 14 |
| Non-small cell lung carcinoma (NSCLC) | CYLD− | 5796 | 2305 | 1702 | 1789 |
| breast | CYLD+ | 2 | 1 | 1 | 0 |
| breast | CYLD− | 495 | 417 | 62 | 16 |
| colorectal (crc) | CYLD+ | 1 | 1 | 0 | 0 |
| colorectal (crc) | CYLD− | 637 | 580 | 45 | 12 |
| glioma | CYLD+ | 0 | 0 | 0 | 0 |
| glioma | CYLD− | 99 | 68 | 23 | 8 |
| ovary | CYLD+ | 0 | 0 | 0 | 0 |
| ovary | CYLD− | 301 | 244 | 53 | 4 |
| pancreas | CYLD+ | 1 | 0 | 1 | 0 |
| pancreas | CYLD− | 297 | 243 | 41 | 13 |
| skin | CYLD+ | 0 | 0 | 0 | 0 |
| skin | CYLD− | 41 | 26 | 10 | 5 |
| cholangiocarcinoma | CYLD+ | 0 | 0 | 0 | 0 |
| cholangiocarcinoma | CYLD− | 112 | 83 | 22 | 7 |
| melanoma | CYLD+ | 0 | 0 | 0 | 0 |
| melanoma | CYLD− | 193 | 127 | 43 | 23 |
| endometrial | CYLD+ | 0 | 0 | 0 | 0 |
| endometrial | CYLD− | 192 | 156 | 34 | 2 |
CYLD showed statistically significant co-occurrence with PD-L1 expression in the pan-solid cohort (OR: 2.01; P < 0.014; 95% CI: 1.2–3.5). The other tumor types did not reach statistical significance or had limited samples for statistical analyses.
Figure 2. Breakdown of PD-L1 immunohistochemistry staining by CYLD alteration status across tumor types.
The proportion of samples stratified by PD-L1 status (negative, PD-L1+ 1-49%, PD-L1+ ≥50%) among cases with (CYLD+) or without (CYLD-) CYLDalterations in the pan-solid tumor cohort and across specific tumor types. Tumor types with at least 10 cases with assessed PD-L1 status are shown. PD-L1 status was determined using IHC using the Dako 22C3 PD-L1 antibody.
Abbreviations: CUP: Unknown primary carcinoma; NSCLC: Non-small cell lung carcinoma; CRC: Colorectal cancer.
Association of CYLD Alterations and high-risk HPV status
This analyses was performed for 145,719/145,991 pan-solid tumor cases which were either positive high-risk HPV or negative for any HPV. Samples that were only positive for low-risk HPV were excluded from these analysis. Of the 836 patients overall with CYLD alterations, 27.5% (n = 230) were also positive for high-risk HPV (Table 3). Strong co-occurrence of CYLD alterations and high-risk HPV + was observed pan-cancer (OR = 10.0, P < 0.0001) and specifically in head and neck cancers (OR = 7.4; P < 0.0001), non-small cell lung cancer (OR = 3.3; P < 0.01), and colorectal cancers (OR = 4.8; P < 0.01) (Table 3). In anal carcinomas, of the 479 cases, 469 were either positive for high-risk HPV or were HPV negative. Of note, almost 81% (n = 380/469) of these were high-risk HPV + and 11% (n = 50/469) had both CYLD alterations and were high-risk HPV+; we observed a trend of co-occurrence between HPV + and the presence of CYLD alterations in anal carcinomas (OR = 2.5) although it did not reach statistical significance after FDR correction (raw P = 0.045) (Table 3). We were limited by cohort size to conduct additional statistical analyses in other tumor types where HPV viral infections are prominent.
Table 3.
Breakdown of HPV+ prevalence by CYLD mutation status across top tumor types
| Tumor type | Mutation cohort |
HPV+ (n) |
HPV− (n) |
OR | Lower 95% CI |
Upper 95% CI |
P (raw) |
P (FDR) |
|---|---|---|---|---|---|---|---|---|
| Pan-solid | CYLD+ | 230 | 606 | 10.0 | 8.6 | 11.7 | < 0.0001 | < 0.0001 |
| Pan-solid | CYLD− | 5282 | 139601 | |||||
| Thymic carcinoma | CYLD+ | 0 | 34 | NA | NA | NA | NA | NA |
| Thymic carcinoma | CYLD− | 4 | 182 | |||||
| anus | CYLD+ | 50 | 5 | 2.5 | 0.9 | 8.4 | 0.045 | 0.11 |
| anus | CYLD− | 330 | 84 | |||||
| head and neck | CYLD+ | 107 | 47 | 7.4 | 5.1 | 10.8 | < 0.0001 | < 0.0001 |
| head and neck | CYLD− | 532 | 1723 | |||||
| bladder | CYLD+ | 0 | 11 | 0 | 0 | 12.2 | 1 | 1 |
| bladder | CYLD− | 71 | 2134 | |||||
| cervix | CYLD+ | 8 | 3 | 1.5 | 0.4 | 8.9 | 0.75 | 1 |
| cervix | CYLD− | 669 | 378 | |||||
| Unknown primary carcinoma (CUP) | CYLD+ | 28 | 52 | 10.1 | 6.1 | 16.5 | < 0.0001 | < 0.0001 |
| Unknown primary carcinoma (CUP) | CYLD− | 427 | 8014 | |||||
| esophagus | CYLD+ | 0 | 16 | 0 | 0 | 12.3 | 1 | 1 |
| esophagus | CYLD− | 75 | 3492 | |||||
| Non-small cell lung carcinoma (NSCLC) | CYLD+ | 13 | 153 | 3.3 | 1.7 | 5.8 | 0.0003 | 0.001 |
| Non-small cell lung carcinoma (NSCLC) | CYLD− | 748 | 29022 | |||||
| breast | CYLD+ | 1 | 43 | 1.0 | 0.03 | 6.2 | 0.62 | 1 |
| breast | CYLD− | 330 | 14853 | |||||
| colorectal (crc) | CYLD+ | 8 | 78 | 4.8 | 1.9 | 9.9 | 0.0005 | 0.0017 |
| colorectal (crc) | CYLD− | 366 | 16992 | |||||
| glioma | CYLD+ | 0 | 13 | 0 | 0 | 17.4 | 1 | 1 |
| glioma | CYLD− | 126 | 6603 | |||||
| ovary | CYLD+ | 0 | 16 | 0 | 0 | 12.3 | 1 | 1 |
| ovary | CYLD− | 172 | 8064 | |||||
| pancreas | CYLD+ | 0 | 11 | 0 | 0 | 16.6 | 1 | 1 |
| pancreas | CYLD− | 193 | 7975 | |||||
| skin | CYLD+ | 3 | 11 | 5.5 | 0.9 | 21.9 | 0.03 | 0.08 |
| skin | CYLD− | 35 | 702 | |||||
| cholangiocarcinoma | CYLD+ | 1 | 9 | 5.3 | 0.1 | 39.1 | 0.19 | 0.40 |
| cholangiocarcinoma | CYLD− | 65 | 3090 | |||||
| melanoma | CYLD+ | 0 | 19 | 0 | 0 | 9.3 | 1 | 1 |
| melanoma | CYLD− | 100 | 4285 | |||||
| endometrial | CYLD+ | 0 | 17 | 0 | 0 | 8.9 | 1 | 1 |
| endometrial | CYLD− | 109 | 3934 |
A. Proportion of samples with CYLD alterations in the pan-solid cohort and across specific tumor types. Tumor types with at least 10 cases with alterations in CYLD are shown. B. Lollipop plot showing the distribution of CYLD short variants detected pan-tumor.
DISCUSSION
Our findings demonstrate that CYLD alterations, while rare in the overall cancer population, are enriched in specific malignancies, particularly thymic carcinomas, anal carcinomas, and head and neck cancers. These results align with data from the TCGA and AACR Genie databases, highlighting a similar pattern of prevalence. Comparisons between our dataset and existing literature consistently emphasize the importance of CYLD in these cancer types (7, 8, 18-21).
Somatic CYLD alterations are notable across various cancers, with TCGA data showing a 1.4% prevalence across ~ 11,000 tumors (20). We similarly showed a prevalence of CYLD mutations in 0.57% (N = 836/145,991) of all patients with solid tumors. In our cohort, CYLD was most commonly altered in thymic carcinomas (15.4%; N = 34/220 patients), followed by anal carcinomas (11.4%; N = 55/479 patients), and head and neck cancers (6.4%; N = = 154/2,423 patients). In TCGA, CYLD alterations were found in 2.4% of thymomas and in 2.9% of head and neck cancers (19), keeping in mind that the types of patients seen in these TCGA and in our dataset may also differ, since patients sent by clinicians for molecular profiling by the FoundationOne® assay often have advanced and/or refractory disease. Notably, an association exists between CYLD mutations and HPV-related cancers, particularly HPV-positive head and neck squamous cell cancers, cervical cancer, and anal carcinoma. Among 515 head and neck squamous cell tumors reported in TCGA, 2.9% harbored CYLD mutations, with a pronounced enrichment in HPV-positive cases (11%) compared to HPV-negative cases (1.7%) (16). Additional studies corroborate this trend. For example, Gillison et al. reported CYLD alterations in 7.4% of HPV-positive HNSCCs versus 0.6% in HPV-negative tumors, while Seiwert et al. identified CYLD mutations in 6% of HPV-positive and 1% of HPV-negative cases (22, 22)
Consistent with prior studies from TCGA and other datasets, our study confirmed a strong association between CYLD alterations and HPV positivity, with 27.5% of CYLD-altered tumors also harboring high-risk HPV. The enrichment was particularly striking in head and neck cancers, NSCLC, and colorectal cancers.
In another study of anal carcinomas, CYLD mutations were observed in 13% of cases, with 97% of these tumors associated with high-risk HPV (8). In our study, 11% of anal cancers (n = 50/469) had both CYLD alterations and were high-risk HPV+. These observations also underscore a potential association between CYLD alterations and HPV positivity in anal carcinoma, albeit limited by cohort size and warranting additional validation. It has been suggested that CYLD alterations, in conjunction with HPV oncoproteins, contribute to tumor progression by activating the NF-κB signaling pathway, upregulating proinflammatory cytokines, and promoting PD-L1 overexpression, thus driving immune evasion (8, 23).
CYLD mutations are also significantly associated with thymic epithelial tumors (TETs), particularly thymic carcinomas (24). In our study, 15.4% (34/220) of thymic carcinomas harbored CYLD mutations.
These alterations, along with NFKBIA mutations, are implicated in constitutive NF-κB activation and elevated PD-L1 expression, positioning CYLD as a molecular driver in thymic carcinoma (1).
Further studies demonstrate that low CYLD expression in TETs is strongly associated with increased PD-L1 expression, particularly via IFN-γ-mediated activation of the STAT1/IRF1 pathway (19). These findings suggest the potential utility of PD-1/PD-L1 inhibitors in this context (25). A study by Giaccone et al. identifies CYLD mutations (p.S331* and p.R850*) exclusively in responders to immunotherapy, with these mutations being associated with significantly higher PD-L1 expression (p = 0.0031) (26). Indeed, in that study, all five patients with a CYLD mutation had high expression of PD-L1 (p = 0.00067). Three of these five patients had a complete or partial response after pembrolizumab immunotherapy. The findings suggest that CYLD mutations may serve as predictors for immunotherapy response in thymic carcinoma (25, 26). Additionally, evidence suggests that EBV, a known oncogenic virus, may contribute to thymic carcinoma pathogenesis, particularly in cases with progressive malignancy. While the role of EBV in thymomas remains unclear, its association with thymic carcinomas warrants further exploration (27). Finally, our own study herein across cancers demonstrates that 61% (38/62) of CYLD-altered tumors were positive for PD-L1 expression (≥ 1% by IHC); indeed, almost 30% (18/62) of these tumors exhibited high PD-L1 expression (≥ 50% by IHC).
Collectively, our analysis underscores the potential importance of CYLD mutations in driving immune evasion mechanisms in cancers associated with viral infections, such as HPV. This study builds on prior research linking CYLD dysfunction to immune evasion through PD-L1 upregulation, particularly in virus-associated malignancies. The observed correlation between CYLD mutations and PD-L1 expression highlights the potential for combining CYLD mutation status with PD-L1 levels as a biomarker for patient selection for immunotherapy.
Our study has important limitations that need to be acknowledged. This study looked at genomic alterations, sequencing-based HPV status and limited cases with available PD-L1 IHC. However, unfortunately treatment information and clinical outcomes were not available in this study cohort. Future work assessing the these patterns with respect to immunotherapy exposure may help better understand the clinical implications of our findings. Importantly, PD-L1 IHC testing is more commonly used in specific tumor types which may contribute to potential biases and warrants additional interrogation. While we already stratify all the analyses by tumor type, a multivariable model in larger cohorts with additional clinicogenomic variables may help overcome some of the limitations in the univariate analysis done in this study. The number of cases with available PD-L1 IHC was limited in this cohort and the PD-L1 IHC testing patterns may show bias towards specific tumor types during routine CGP, thereby limiting our capability to perform more detailed statistical comparisons. Specifically, some subsets were very small; for instance, there were only 2 of 34 CYLD-mutated thymic tumors; while both had PDL IHC > 50%, larger numbers of patients require investigation. Larger studies are warranted to corroborate these findings. Other limitations include the retrospective design, possible selection bias, limited PD-L1 and HPV data, functional uncertainty for some CYLD variants and lack of treatment and outcome data”
In conclusion, our results provide a comprehensive real-world overview of CYLD alterations across solid tumors and their relationship with immune and viral biomarkers. The observed associations with PD-L1 expression and HPV status are consistent with emerging mechanistic evidence. This work is hypothesis-generating and represents a large-scale confirmation of prior and recent findings; future studies should delve into mechanisms underlying the observation. CYLD and its relationship with viral and immune markers have previously been implicated in immunoregulation; our work did not include clinical correlations, but further support the concept that the role of CYLD warrants further exploration vis a vis its clinical impact. Specific next steps include analysis of ICI-treated CYLD cohorts, along with looking at PD-L1 and HPV status and future prospective validation of CYLD as a predictive marker .
Acknowledgments:
UG sincerely acknowledges the financial assistance of the Indian Institute of Technology (IIT) Jammu, India. The authors also acknowledge the Central Instrumentation Facility for providing the necessary research infrastructure.
Funding acknowledgement:
RK is funded in part by 5U01CA180888-08 and 5UG1CA233198-05.
Disclosures:
Dr. Kurzrock has received research funding from Boehringer Ingelheim, Debiopharm, Foundation Medicine, Genentech, Grifols, Guardant, Incyte, Konica Minolta, Medimmune, Merck Serono, Omniseq, Pfizer, Sequenom, Sysmex, Takeda, and TopAlliance and from the NCI; as well as consultant and/or speaker fees and/or advisory board/consultant for Actuate Therapeutics, AstraZeneca, Bicara Therapeutics, Inc., Biological Dynamics, Caris, Daiichi, Datar Cancer Genetics, EISAI, EMD Serono, EOM Pharmaceuticals, Iylon, Jackson Laboratories, LabCorp, Lanauria Therapeutics, Merck, NeoGenomics, Neomed, Pfizer, Precirix, Prosperdtx, Quanta Therapeutics, Recordati, Regeneron, Roche, TD2/Volastra, Turning Point Therapeutics, X-Biotech; has an equity interest in CureMatch Inc.; serves on the Board of CureMatch and CureMetrix and XZOM, and is a co-founder of CureMatch. Smruthy Sivakumar and Saumya Dushyant Sisoudiya are employees at Foundation Medicine, Inc. and have an equity interest in Roche.
Footnotes
Declarations: There are no conflicts to declare.
Contributor Information
Umut Dişel, Acıbadem Adana Hospital.
Saumya Dushyant Sisoudiya, Foundation Medicine Inc.
Smruthy Sivakumar, Foundation Medicine Inc.
Razelle Kurzrock, Medical College of Wisconsin Cancer Center.
Data availability:
All data analyzed during this study are included in this manuscript.
References
- 1.Sun S.C. CYLD: a tumor suppressor deubiquitinase regulating NF-kappaB activation and diverse biological processes. Cell Death Differ 17, 25–34 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Patel SP, Kurzrock R. PD-L1 Expression as a Predictive Biomarker in Cancer Immunotherapy. Mol Cancer Ther. 2015;14(4):847–56. [DOI] [PubMed] [Google Scholar]
- 3.Bevins NJ, Okamura R, Montesion M, Adashek JJ, Goodman AM, Kurzrock R. Tumor Infiltrating Lymphocyte Expression of PD-1 Predicts Response to Anti-PD-1/PD-L1 Immunotherapy. J Immunother Precis Oncol. 2022;5(4):90–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Fountzilas E, Kurzrock R, Vo HH, Tsimberidou AM. Wedding of Molecular Alterations and Immune Checkpoint Blockade: Genomics as a Matchmaker. J Natl Cancer Inst. 2021;113(12):1634–1647. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Schrank T. P., Prince A. C., Sathe T., Wang X., Liu X., Alzhanov D. T., Burtness B., Baldwin A. S., Yarbrough W. G., & Issaeva N. (2022). NF-κB over-activation portends improved outcomes in HPV-associated head and neck cancer. Oncotarget, 13(707–722). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Hajek M., Sewell A., Kaech S., Burtness B., Yarbrough W. G., & Issaeva N. (2017). TRAF3/CYLD mutations identify a distinct subset of human papillomavirus-associated head and neck squamous cell carcinoma. Cancer, 123(10), 1778–1790. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Williams EA, Montesion M, Alexander BM, Ramkissoon SH, Elvin JA, Ross JS, Williams KJ, Glomski K, Bledsoe JR, Tse JY, Mochel MC. CYLD mutation characterizes a subset of HPV-positive head and neck squamous cell carcinomas with distinctive genomics and frequent cylindroma-like histologic features. Mod Pathol. 2021;34(2):358–370. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Williams E.A., Montesion M., Sharaf R. et al. CYLD-mutant cylindroma-like basaloid carcinoma of the anus: a genetically and morphologically distinct class of HPV-related anal carcinoma. Mod Pathol 33, 2614–2625 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Prasad M, Cheng AT, Mayer M, et al. Genomic enrichment and functional impact of TP53 and CYLD alterations in recurrent and metastatic HPV-associated head and neck cancer. Clin Cancer Res. Published online January 28, 2026. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jovanovic D., Markovic J., Ceriman V., Peric J., Pavlovic S., & Soldatovic I. Correlation of genomic alterations and PD-L1 expression in thymoma. J Thorac Dis 12, 7561–7570 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mishra A. PD-1/PD-L1 biology and immunotherapy in HPV-positive oral cancers. Future Oncol 13, 1907–1909 (2017). [DOI] [PubMed] [Google Scholar]
- 12.Frampton GM, Fichtenholtz A, Otto GA, Wang K, Downing SR, He J, Schnall-Levin M, White J, Sanford EM, An P, Sun J, Juhn F, Brennan K, Iwanik K, Maillet A, Buell J, White E, Zhao M, Balasubramanian S, Terzic S, Richards T, Banning V, Garcia L, Mahoney K, Zwirko Z, Donahue A, Beltran H, Mosquera JM, Rubin MA, Dogan S, Hedvat CV, Berger MF, Pusztai L, Lechner M, Boshoff C, Jarosz M, Vietz C, Parker A, Miller VA, Ross JS, Curran J, Cronin MT, Stephens PJ, Lipson D, Yelensky R. Development and validation of a clinical cancer genomic profiling test based on massively parallel DNA sequencing. Nat Biotechnol. 2013;31(11):1023–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Milbury CA, Creeden J, Yip WK, Smith DL, Pattani V, Maxwell K, Sawchyn B, Gjoerup O, Meng W, Skoletsky J, Concepcion AD, Tang Y, Bai X, Dewal N, Ma P, Bailey ST, Thornton J, Pavlick DC, Frampton GM, Lieber D, White J, Burns C, Vietz C. Clinical and analytical validation of FoundationOne®CDx, a comprehensive genomic profiling assay for solid tumors. PLoS One. 2022;17(3):e0264138. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Knepper TC, Montesion M, Russell JS, et al. The Genomic Landscape of Merkel Cell Carcinoma and Clinicogenomic Biomarkers of Response to Immune Checkpoint Inhibitor Therapy. Clin Cancer Res. 2019;25(19):5961–5971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Williams Erik A et al. “Vulvar Squamous Cell Carcinoma: Comprehensive Genomic Profiling of HPV+ Versus HPV− Forms Reveals Distinct Sets of Potentially Actionable Molecular Targets.” JCO precision oncology vol. 4 PO.19.00406. 16 Jun. 2020, [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Knepper TC, Montesion M, Russell JS, et al. The Genomic Landscape of Merkel Cell Carcinoma and Clinicogenomic Biomarkers of Response to Immune Checkpoint Inhibitor Therapy. Clin Cancer Res. 2019;25(19):5961–5971. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lechner M, Frampton GM, Fenton T, et al. Targeted next-generation sequencing of head and neck squamous cell carcinoma identifies novel genetic alterations in HPV + and HPV− tumors. Genome Med. 2013;5(5):49. Published 2013 May 29. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Umemura S., Zhu J., Chahine J. J., Kallakury B., Chen V., Kim I. K., Zhang Y. W., Goto K., He Y., & Giaccone G. (2020). Downregulation of CYLD promotes IFN-γ mediated PD-L1 expression in thymic epithelial tumors. Lung Cancer, 147, 221–228. [DOI] [PubMed] [Google Scholar]
- 19.Zhu X, Meng X, Mochel MC, Wiliams EA, Dresser K, Tomaszewicz K, Hutchinson L, Bledsoe JR. CYLD-mutated anal squamous cell carcinoma: an uncommon entity associated with cylindroma-like morphology and adverse clinical features. Hum Pathol. 2025. Apr [DOI] [PubMed] [Google Scholar]
- 20.Cui Z., Kang H., Grandis J. R., & Johnson D. E. (2021). CYLD Alterations in the Tumorigenesis and Progression of Human Papillomavirus–Associated Head and Neck Cancers. Molecular Cancer Research, 19(1), 14–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Gillison ML, Akagi K, Xiao W, Jiang B, Pickard RKL, Li J, Swanson BJ, Agrawal AD, Zucker M, Stache-Crain B, Emde AK, Geiger HM, Robine N, Coombes KR, Symer DE. Human papillomavirus and the landscape of secondary genetic alterations in oral cancers. Genome Res. 2019;29(1):1–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Seiwert TY, Zuo Z, Keck MK, Khattri A, Pedamallu CS, Stricker T, Brown C, Pugh TJ, Stojanov P, Cho J, Lawrence MS, Getz G, Brägelmann J, DeBoer R, Weichselbaum RR, Langerman A, Portugal L, Blair E, Stenson K, Lingen MW, Cohen EE, Vokes EE, White KP, Hammerman PS. Integrative and comparative genomic analysis of HPV-positive and HPV-negative head and neck squamous cell carcinomas. Clin Cancer Res. 2015;21(3):632–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.An J, Mo D, Liu H, Veena MS, Srivatsan ES, Massoumi R, Rettig MB. Inactivation of the CYLD deubiquitinase by HPV E6 mediates hypoxia-induced NF-kappaB activation. Cancer Cell. 2008;14(5):394–407. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Zhang X, Zhang P, Cong A, Feng Y, Chi H, Xia Z, Tang H. Unraveling molecular networks in thymic epithelial tumors: deciphering the unique signatures. Front Immunol. 2023;14:1264325. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.He Y., Ramesh A., Gusev Y., Bhuvaneshwar K., & Giaccone G. (2021). Molecular predictors of response to pembrolizumab in thymic carcinoma. Cell Reports Medicine, 2(15_suppl), 100392. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Giaccone G., Kim C., Thompson J., McGuire C., Kallakury B., Chahine J. J., et al. (2018). Pembrolizumab in patients with thymic carcinoma: a single-arm, single-centre, phase 2 study. The Lancet Oncology, 19(3), 347–355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Zhao L., Ding J.Y., Tao Y.L., Zhu K., & Chen G. (2023). Detection of Epstein–Barr virus infection in thymic epithelial tumors by nested PCR and Epstein–Barr-encoded RNA ISH. Infectious Agents and Cancer, 18(1), 37. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data analyzed during this study are included in this manuscript.