Invest Clin 67(2): 260 - 274, 2026 https://doi.org/10.54817/IC.v67n2a08

Corresponding author: Wei Li. Cancer Center, Beijing Tongren Hospital, Capital Medical University. Beijing

100176, China. Telefax: +86-15811929005. Email: fever1988fever@sina.com

Circulating white blood cells and risk

of tonsillar and base of tongue squamous

cell carcinoma: A retrospective

and mendelian randomization study.

Changyu Zhu1, Shizhi He2, Zhixin Li1, Yijun Shi1, Jingyang Zhao1 and Wei Li1

1Cancer Center, Beijing Tongren Hospital, Capital Medical University. Beijing, China.

2Department of Head and Neck Surgery, Beijing Tongren Hospital, Capital Medical

University.Beijing, China.

Keywords: Tonsillar Neoplasms; Leukocytes; Mendelian Randomization Analysis;

Human Papillomavirus; Squamous Cell Carcinoma of Head and Neck.

Abstract. This study aimed to investigate the relationship between circula-

ting white blood cells (cWBC) and the risk of tonsillar and base of tongue squa-

mous cell carcinoma (TSCC/BOT SCC) using retrospective clinical data and

Mendelian randomization (MR) analysis. A retrospective cohort of 239 TSCC/

BOT SCC patients was analyzed for cWBC subtypes and their association with

clinicopathological variables, stratified by human papillomavirus (HPV) status.

Blood tests, tumor staging, and immunological markers were included. For cau-

sal inference, MR analysis was performed using genome-wide association stu-

dy (GWAS) data on cWBC from the Blood Cell Consortium (UK Biobank) and

TSCC/BOT SCC outcome data from the FinnGen consortium. Single-nucleotide

polymorphisms (SNPs) were chosen based on genome-wide significance (p<5 ×

10⁻⁸), low linkage disequilibrium (r² < 0.001), and F-statistic >10. The inverse-

variance weighted (IVW) method was used as the primary MR approach, supple-

mented by MR-Egger, weighted median, and weighted mode analyses. The retros-

pective analysis showed significant differences in cWBC subtypes by gender, age,

lifestyle factors, and HPV status. Notably, neutrophils (cNEU) and monocytes

(cMON) were strongly associated with tumor stage and immune markers. MR

analysis confirmed a causal link between total cWBC count and TSCC/BOT SCC

risk (OR=1.516, p=0.005), with no evidence of heterogeneity or pleiotropy. No

causal relationship was identified for cWBC subtypes or other head and neck

squamous cell carcinoma (HNSCC) sites. This study provides the first compre-

hensive evidence supporting a causal role of elevated cWBC in the development

of TSCC/BOT SCC. These findings indicate that cWBC may serve as a potential

biomarker and therapeutic target in HPV-related or unrelated TSCC/BOT SCC.

White blood cells and human papillomavirus risk in head and neck squamous cell carcinoma 261

Vol. 67(2): 260 - 274, 2026

Leucocitos circulantes y riesgo de carcinoma de células

escamosas de amígdalas y base de la lengua: un estudio

retrospectivo y de aleatorización mendeliana.

Invest Clin 2026; 67 (2): 260 – 274

Palabras clave: Neoplasias Tonsilares; Leucocitos; Análisis de la Aleatorización

Mendeliana; Virus del Papiloma Humano; Carcinoma de Células

Escamosas de Cabeza y Cuello.

Resumen. Este estudio tuvo como objetivo explorar la relación entre los leuco-

citos circulantes (LC) y el riesgo de carcinoma de células escamosas de amígdala y

de base de la lengua (CCEA/CCEB), mediante datos clínicos retrospectivos y análi-

sis de aleatorización mendeliana (AM). Se analizó una cohorte retrospectiva de 239

pacientes con CCEA/CCEB para determinar los subtipos de LC y su asociación con

variables clínico-patológicas, estratificadas según el estado del virus del papiloma

humano (VPH). Se incluyeron análisis de sangre, estadificación tumoral y marcado-

res inmunológicos. Para la inferencia causal, se realizó un análisis de AM utilizando

datos del estudio de asociación del genoma completo (GWAS) sobre LC del Con-

sorcio de Células Sanguíneas (Biobanco del Reino Unido) y datos de resultados de

CCEA/CCEB del consorcio FinnGen. Los polimorfismos de un nucleótido (SNP) se

seleccionaron en función de su significancia a nivel genómico (p<5 × 10⁻⁸), bajo

desequilibrio de ligamiento (r² <0,001) y un estadístico F >10. Se utilizó la ponde-

ración por el inverso de la varianza (IVW: Inverse Variance Weighted) como método

principal de AM, complementado con análisis de regresión MR-Egger, mediana pon-

derada y moda ponderada. El análisis retrospectivo reveló diferencias significativas

en los subtipos de leucocitos totales (LCt) según el sexo, la edad, los factores del

estilo de vida y el estado del VPH. Cabe destacar que los neutrófilos y monocitos

se asociaron fuertemente con el estadio tumoral y los marcadores inmunitarios. El

análisis de AM confirmó una asociación causal entre el recuento total de LC y el ries-

go de carcinoma de células escamosas de la lengua/base de la vejiga (OR=1,516,

p=0,005), sin evidencia de heterogeneidad ni pleiotropía. No se encontró ningún

vínculo causal entre los subtipos de LC y otros sitios de carcinoma de células esca-

mosas de cabeza y cuello. Este estudio proporciona la primera evidencia integrada

que respalda un papel causal de los LC elevados en la patogénesis del CCEA/CCEB.

Estos hallazgos sugieren que el recuento de leucocitos en sangre periférica (LC) po-

dría servir como biomarcador y objetivo terapéutico en el carcinoma de células esca-

mosas de la lengua/base de la lengua (CCEA/CCEB), relacionado o no con el VPH.

Received: 05-11-2025 Accepted: 19-02-2026

INTRODUCTION

The global incidence of oropharyngeal

squamous cell carcinoma (OPSCC) has been

gradually rising, with new cases reaching 98,

412 in 2020 1. Smoking, alcohol consump-

tion, and human papillomavirus (HPV) in-

fection are three independent risk factors

for its development 2. HPV- positive OPSCC

accounts for over 70% of cases in some re-

262 Zhu et al.

Investigación Clínica 67(2): 2026

gions, with TSCC/BOT SCC being the most

common subtypes 1 1,3-5. The survival rate for

HPV- positive OPSCC is roughly twice as high

as that for HPV- negative cases 6. Although

treatments such as surgery, chemotherapy,

targeted therapy, and immune checkpoint

inhibitors (ICI) have improved outcomes to

some extent, these tumors exhibit high het-

erogeneity, increasing incidence, rapid pro-

gression, and high rates of recurrence and

metastasis. This highlights the urgent need

for better stratification tools and immune-

based biomarkers to personalize therapy and

predict prognosis. The tumor immune mi-

croenvironment (TIME) involves continuous

interactions between tumor cells and vari-

ous immune cells, playing a key role in can-

cer development, progression, and response

to therapy 7. These interactions influence

immune response, tumor cell proliferation,

angiogenesis, and tumor recurrence and

spread. Additionally, complex regulation oc-

curs among different immune cells and their

cytokines within the TIME 8. The hetero-

geneity of the tumor itself can also impact

the TIME 9. Currently, immune checkpoint

inhibitor (ICI) therapy is increasingly used,

utilizing immunotherapy agents to boost the

immune system’s ability to recognize and de-

stroy malignant cells more effectively10. ICI

therapy is considered one of the most prom-

ising approaches for Head and Neck Squa-

mous Cell Carcinoma (HNSCC) 11. It has

significantly improved treatment outcomes

for HNSCC and has become the standard

first-line therapy for advanced cases. How-

ever, despite its success, many patients still

experience disease progression, recurrence,

and metastasis after treatment12, 13. Circulat-

ing white blood cells (cWBCs), which con-

sist of various immune cell types, reflect sys-

temic immune status and may offer insights

into the TIME 14. Previous studies suggest

that certain cWBC subsets, such as lympho-

cytes and monocytes, are linked to tumor

immune surveillance, immunotherapy out-

comes, and disease prognosis 15. Nonethe-

less, the relationship between cWBCs and

TSCC/BOT SCC, especially in the context of

HPV infection, remains poorly understood.

In this study, we explore the association be-

tween cWBC subtypes and TSCC/BOT SCC

through a dual approach: retrospective clini-

cal analysis and Mendelian randomization

(MR) to assess potential causality using ge-

nome- wide association study (GWAS) data.

This combined analysis aims to identify reli-

able immune biomarkers and clarify the im-

munogenic mechanisms underlying TSCC/

BOT SCC pathogenesis.

MATERIALS AND METHODS

Retrospective analysis

The clinical and pathological data of all

patients with TSCC/BOT SCC in our hospi-

tal from July 2020 to January 2025 were col-

lected, including gender, age, smoking and

alcohol consumption status, blood routine

tests (counts of circulating white blood cells

(cWBC), neutrophils (cNEU), lymphocytes

(cLYM), monocytes (cMON), eosinophils

(cEOS), basophils (cBAS), and derived ra-

tios such as neutrophil-to-lymphocyte ratio

(NLR) and lymphocyte-to-monocyte ratio

(LMR) before biopsy or radical surgery. Tu-

mor HPV-related status was also recorded,

along with TNM staging: TI-TIV (indicating

increasing size and/or local extent of the

primary tumor), T stage (T1-4: tumor size

and extent), and N stage (N0 to N3: spread

to regional lymph nodes). The KI-67 value

(a protein expressed in dividing cells and

a marker for tumor cell proliferation), im-

aging stage or postoperative pathological

stage, and Combined Positive Score (CPS)

expression were included as well.

Approval from the Ethics Committee

of Beijing Tongren Hospital, Capital Medical

University, was obtained prior to data collec-

tion and analysis (Approval no. TREC2022-

KY018.R1).

The retrospective data of patients aged

≥18 years of both genders were retrieved by

passing the following criteria: pre-biopsy/

radical surgery and routine blood tests con-

White blood cells and human papillomavirus risk in head and neck squamous cell carcinoma 263

Vol. 67(2): 260 - 274, 2026

ducted at the host institute; pathological

confirmation of TSCC/BOT SCC; available

HPV and CPS data; no history of other ma-

lignancies or immunologic diseases, and no

recent infection or anti-infective therapy.

The exclusion criteria included: patients

under 18 years old; TSCC/BOT SCC treated

at other institutions or hospitals; missing

records of pre-biopsy or radical surgery; no

routine blood tests performed at the host

hospital; prior induction therapy before radi-

cal surgery elsewhere; recent infection or on-

going anti-infective therapy; history of other

cancers or immunologic disorders; and use

of hormonal drugs, anti-infective agents, or

traditional medicine before blood collection.

The patients were divided into two

groups based on the HPV status of TSCC/

BOT SCC. Differences in cWBC were com-

pared between the two groups, accounting for

factors such as gender and age. Subsequently,

subgroup analyses were performed within the

HPV-positive and HPV-negative groups to as-

sess differences in these indicators.

Mendelian Randomization

Sources of GWAS data for cWBC

and HNSCC

The cWBC exposure data were obtained

from the Blood Cell Consortium (BCX) me-

ta-analysis (UKBB cohort, N=562,243) 16.

The details of the GWAS data used in this

study were obtained from the FinnGen data-

base (https://www.finngen.fi/en) 17 and are

presented in Table 1.

Fig. 1 shows the MR framework used

to explore the causal link between cWBCs

and specific head and neck cancers. GWAS

summary statistics for cWBCs served as the

exposure data, while outcome data were ob-

tained from GWAS datasets for tonsillar and

base-of-tongue cancers, along with other

head and neck cancer subsites, including

hypopharyngeal, nasal, oral, and nasopha-

ryngeal cancers. This method allows for the

assessment of potential causal effects while

reducing confounding and reverse causality.

Selection of genetic instruments

We identified SNPs strongly associated

with the exposure, applying a genome-wide

significance threshold of p < 5 × 10⁻⁸. To

control for linkage disequilibrium, we ap-

plied stringent clumping parameters (r² <

0.001 within a 10,000-kb window). Palin-

dromic SNPs with ambiguous allele frequen-

cies were excluded to prevent strand mis-

alignment. Additionally, we filtered out weak

instruments by calculating the F-statistic

(β²/SE²) for each SNP-exposure association,

retaining only those with F > 10 to ensure

robust instrument strength 18.

Table 1. Information of summary level genome-wide association study (GWAS) used in this study.

Phenotype GWAS ID / Name Sample size*

Patients Controls

Malignant cancer

of the tonsil and base

of the tongue

Finngen_R12_C3_Malignant cancer of tonsil

and base of tongue

813 378,749

Hypopharyngeal cancer Finngen_R12_C3_Malignant neoplasm of hypopharynx 124 378,749

Nasal cavity and sinus

cancer

Finngen_R12_C3_Malignant neoplasm of the nasal

cavity and sinuses

345 378,749

Oral cancer Finngen_R12_C3_Malignant neoplasm of the oral

cavity

1,614 378,749

Nasopharyngeal cancer Finngen_R12_C3_Malignant neoplasm of nasopharynx 152 378,749

*All other cancers were excluded except the selected type. Controls against each type of cancer in the patients

column.

264 Zhu et al.

Investigación Clínica 67(2): 2026

MR analysis

Four MR methods were used: inverse-

variance weighted (IVW), MR-Egger, weighted

median, and weighted mode. The IVW meth-

od served as the main approach to estimate

the causal effect. By thoroughly considering

the effects and precisions of multiple SNPs,

the IVW method provided a solid estimate

of causality. Additionally, other methods like

MR-Egger, weighted median, and weighted

mode were also employed in the analysis. If

the statistically significant results from IVW

did not agree with those from the other meth-

ods, such as MR-Egger, weighted median, and

weighted mode, we compared the effect esti-

mates (β and OR) across methods to evalu-

ate the size and direction of the differences.

Heterogeneity and horizontal pleiotropy were

checked using the IVW and MR-Egger tests. A

p-value above 0.05 for Cochran’s Q-statistic

(in MR-IVW) and Rucker’s Q-statistic (in MR-

Egger) suggested no heterogeneity in the MR

analysis 19. Additionally, MR-Egger could iden-

tify and assess potential pleiotropy through

the MR-Egger intercept test 20. If pleiotropy

was detected, the MR results were considered

invalid.

Statistical analysis

For the retrospective analysis, R soft-

ware (version 4.3.1; R Foundation for Sta-

tistical Computing, Vienna, Austria) was

used. Non-parametric tests were employed

to compare quantitative variables: the Wil-

coxon rank-sum test for two independent

samples and the Kruskal-Wallis test for mul-

tiple samples. The chi-square test was used

to compare proportions (rates), and analysis

of variance (ANOVA) was used to compare

continuous variables across groups. Statisti-

cal significance was defined as a two-sided

p-value<0.05. MR analyses were performed

using the TwoSampleMR software (version

0.5.11) in R (version 4.3.1). To rigorously

evaluate the causal relationship, false discov-

ery rate (FDR) correction was applied to the

final results. A corrected p-value<0.05 was

considered to indicate a statistically signifi-

cant result.

RESULTS

Patient Baseline Characteristics

A total of 239 patients with TSCC/BOT

SCC were included. Demographic and clini-

cal characteristics are summarized in Table

2. In terms of gender distribution, there

were 203 male patients, accounting for

84.94%. Regarding age, 206 patients were

under 50 years old, making up 86.19%. Con-

cerning lifestyle habits, 149 patients had a

history of smoking or drinking, represent-

ing 62.34%, while 77 patients had no such

habits, accounting for 32.22%. For HPV in-

fection status, 150 patients were HPV-pos-

itive, constituting 62.76%, and 89 patients

were HPV-negative, representing 37.24%.

In tumor staging, T2 was the most com-

mon (40.17%), while N2 (43.10%) and N3

(33.47%) were the predominant N stages.

TNM stage IV accounted for 25.10%. Regard-

ing the immunohistochemical index Ki-67

expression level, the largest group had val-

ues >70, with 116 cases, or 48.54%. The

distribution of CPS scores was as follows: 48

patients scored 1-10 (20.08%), 44 patients

Fig. 1. Mendelian Randomization framework used

to investigate the relationship between cir-

culating white blood cell subtypes and site-

specific head and neck cancers.

White blood cells and human papillomavirus risk in head and neck squamous cell carcinoma 265

Vol. 67(2): 260 - 274, 2026

scored 11-59 (18.41%), 23 patients scored

60-99 (9.62%), and 6 patients scored less

than 1 (2.51%).

Correlation between cWBC and clinico-

pathological characteristics

In the overall cohort, circulating white

blood cell subtypes (cWBC, cNEU, cEOS,

cMON, and cLMR) were significantly asso-

ciated with demographic and clinical vari-

ables, as shown in Fig. 2. Gender differenc-

es were particularly notable in cMON and

cLMR (p<0.001). Smoking and drinking

habits were significantly linked to cWBC,

cMON, and cEOS levels (p<0.001). Addi-

tionally, cNEU and cNLR were associated

with T-stage (p<0.05 and p<0.01, respec-

tively), whereas cMON differed across CPS

score groups (p<0.05). In the HPV-positive

subgroup (n=150), gender had a significant

impact on cMON and cLMR (p<0.001), with

significant differences also observed in cEOS

levels (p<0.01). Smoking and drinking were

associated with higher cWBC (p<0.01) and

cMON (p<0.001), while cNEU varied signifi-

cantly with TNM stage (p<0.05). For the HPV-

negative subgroup (n=89), gender-related

differences were found in cMON (p<0.05),

and smoking/drinking were linked to cMON

(p<0.05) and cEOS (p<0.01). T-stage was

associated with variations in cLYM and cNLR

(p<0.05), and cMON showed differences

across CPS score categories (p<0.05).

MR analysis

IVW analysis revealed a significant posi-

tive association between cWBC and TSCC/

BOT SCC risk (β=0.416, OR=1.516, 95%

CI=1.189-1.935, P/FDR=0.005). Although

cMON showed a positive trend (β=0.254,

OR=1.289, P/FDR=0.018), results were

excluded due to significant horizontal plei-

otropy (p<0.05). No statistically significant

associations were observed between other

cWBC subtypes and TSCC/BOT SCC or

other HNSCC subsites. These findings are

shown in Fig. 3.

Table 2. Baseline characteristics of all included

oropharyngeal squamous cell carcinoma patients.

Characteristics Patients (N=239) Percentage

Gender

Male 203 84.94

Female 36 15.06

Age

≤50 33 13.81

>50 206 86.19

Smoking or Drinking

No 77 32.22

Ye s 149 62.34

NA 13 5.44

HPV status

Positive 150 62.76

Negative 89 37.24

T Stage

1 43 17.99

2 96 40.17

3 54 22.59

4 46 19.25

N Stage

1 54 22.59

2 103 43.1

3 80 33.47

4 2 0.84

TNM Stage

I 84 35.15

II 57 23.85

III 38 15.9

IV 60 25.1

Ki-67

10-39 27 11.3

40-69 71 29.71

>70 116 48.54

NA 25 10.46

CPS

1-10 48 20.08

11-59 44 18.41

60-99 23 9.62

<1 6 2.51

NA 118 49.37

NA: Not available (data missing); HPV: human papillo-

ma virus; TNM system stages cancer; T stage: tumor

stage; N stage: Node stage; Ki-67: proliferation index;

CPS: Combined Positive Score.

266 Zhu et al.

Investigación Clínica 67(2): 2026

Table 3 shows the results of heteroge-

neity and horizontal pleiotropy tests for the

MR estimates. For the association between

WBC and TSCC/BOT SCC, neither the het-

erogeneity nor the horizontal pleiotropy

tests were statistically significant (p>0.05

for both). In contrast, the horizontal pleiot-

ropy test for the association between MON

and TSCC/BOT SCC was statistically signifi-

cant (p<0.05).

DISCUSSION

Globally, the proportion of HPV-positive

OPSCC has increased significantly. The cu-

mulative risk of OPSCC is 0.21% in males

and 0.05% in females, with a significantly

higher proportion of male patients than fe-

male patients 1. In this study, HPV-positive

patients accounted for 62.76%, with males

making up 84.94%, consistent with the

global epidemiological trend in OPSCC. The

biological behavior of HPV-positive OPSCC

differs markedly from that of HPV-negative

OPSCC. Systemic inflammatory markers

have become reliable prognostic tools in

HNSCC, reflecting the dynamic interaction

between tumor biology and host immunity.

HPV-positive OPSCC shows strong immune

cell infiltration, better treatment response,

and improved prognosis 21-22. Notably, circu-

lating leukocyte levels differ significantly be-

tween HPV-positive and HPV-negative cases,

with variations in immune cell types poten-

tially indicating differences in the tumor im-

mune microenvironment (TIME) between

these groups. These distinct immune states

may influence the effectiveness of immuno-

Total

patients

HPV

positive

HPV

negative

Gender

Age

Smoking or Drinking

HPV Relevance

T Stage

N Stage

TNM Stage

ki-67

CPS

Gender

Age

Smoking or Drinking

T Stage

N Stage

TNM Stage

ki-67

CPS

Gender

Age

Smoking or Drinking

T Stage

N Stage

TNM Stage

ki-67

CPS

WBC LYM MON NEU EOS BAS NLR LMR

***

WBC LYM MON NEU EOS BAS NLR LMR

***

WBC LYM MON NEU EOS BAS NLR LMR

* <0.05

** ≤0.01

*** ≤0.001

≥0.05

<0.05

Fig. 2. Association between circulating white blood cell subtypes with demographic and clinical variables.

cWBC: circulating white blood cell; cNEU: neutrophils; cLYM: lymphocytes; cMON: monocytes; cEOS:

eosinophils; cBAS: basophils; NLR: neutrophils-to-lymphocyte ratio; LMR: lymphocyte-to-monocyte

ratio; TNM: system stages cancer; T stage: tumor stage; N stage: Node stage; Ki-67: proliferation in-

dex; CPS: Combined Positive Score; HPV: Human Papilloma Virus.

White blood cells and human papillomavirus risk in head and neck squamous cell carcinoma 267

Vol. 67(2): 260 - 274, 2026

Fig. 3. Associations between circulating white blood cell (cWBC) subtypes and Tonsillar Squamous Cell Car-

cinoma and Base of Tongue Squamous Cell Carcinoma (TSCC/BOT SCC) or other Head and Neck

Squamous Cell Carcinoma (HNSCC) subsites. cNEU: neutrophils; cLYM: lymphocytes; cMON: mono-

cytes; cEOS: eosinophils; cBAS: basophils; LMR: lymphocyte-to-monocyte ratio; HPV: Human Papillo-

ma Virus; IVW: Inverse variance weighted; MR-Egger: Mendelian randomization.

268 Zhu et al.

Investigación Clínica 67(2): 2026

Table 3. Results of Mendelian Randomization heterogeneity and horizontal pleiotropy.

Exposure Outcome

Heterogeneity Pleiotropy

Method Q Q_pval Egger_intercept se pval

cWBC Tonsil and base of

tongue cancer

MR-Egger 491.095 0.177 0.008 0.006 0.187

IVW 492.95 0.17

cMON Tonsil and base of

tongue cancer

MR-Egger 468.766 0.597 0.009 0.005 0.045*

IVW 472.802 0.558

cLYM Tonsil and base of

tongue cancer

MR-Egger 457.389 0.688 0.002 0.006 0.732

IVW 457.506 0.699

cLYM Hypopharyngeal

cancer

MR-Egger 432.64 0.914 0.013 0.011 0.208

IVW 434.231 0.91

cEOS Hypopharyngeal

cancer

MR-Egger 427.628 0.482 0.03 0.011 0.007*

IVW 435.049 0.397

cNEU Hypopharyngeal

cancer

MR-Egger 382.328 0.705 -0.002 0.011 0.893

IVW 382.346 0.717

cBAS Nasal cavity and

paranasal sinuses cancer

MR-Egger 215.195 0.093 0.009 0.013 0.471

IVW 215.789 0.097

cLYM Nasal cavity and

paranasal sinuses cancer

MR-Egger 481.968 0.39 -0.007 0.009 0.425

IVW 482.615 0.395

cNEU Nasal cavity and

paranasal sinuses cancer

MR-Egger 462.155 0.014* 0.002 0.01 0.826

IVW 462.211 0.016*

cBAS Nasopharyngeal cancer MR-Egger 152.792 0.975 0.032 0.019 0.088

IVW 155.731 0.967

cWBC Nasopharyngeal cancer MR-Egger 452.706 0.638 -0.002 0.013 0.886

IVW 452.727 0.65

cMON Nasopharyngeal cancer MR-Egger 409.642 0.988 -0.003 0.01 0.789

IVW 409.713 0.989

cLYM Nasopharyngeal cancer MR-Egger 462.862 0.634 -0.001 0.013 0.925

IVW 462.871 0.646

cNEU Nasopharyngeal cancer MR-Egger 395.543 0.525 0 0.014 0.982

IVW 395.544 0.539

cWBC Oral cavity cancer MR-Egger 438.241 0.79 -0.001 0.004 0.716

IVW 438.373 0.798

cLYM Oral cavity cancer MR-Egger 532.671 0.032* -0.001 0.004 0.815

IVW 532.733 0.034*

cNEU Oral cavity cancer MR-Egger 376.332 0.765 -0.005 0.004 0.267

IVW 377.566 0.762

cWBC: circulating white blood cell; cNEU: neutrophils; cLYM: lymphocytes; cMON: monocytes; cEOS: eosinophils;

cBAS: basophils; IVW: Inverse variance weighted; MR-Egger: Mendelian randomization Egger Regression. *p<0.05.

White blood cells and human papillomavirus risk in head and neck squamous cell carcinoma 269

Vol. 67(2): 260 - 274, 2026

therapy. This study found that cWBC, cNEU,

and cEOS had significant differences across

various genders, ages, smoking/drinking

habits, and tumor stages. In particular,

among HPV-positive patients, cMON and

cLMR exhibited notable differences. Previ-

ous research has demonstrated that HPV can

modulate the immune microenvironment to

promote tumorigenesis and progression23.

Lower lymphocyte counts and reduced LMR

are associated with poorer outcomes. A sys-

tematic review involving 5,234 HNSCC pa-

tients revealed that higher LMR (≥4) was

linked to better overall survival (HR = 1.36)

and disease-free survival (HR = 0.94) 24.

This threshold effect likely reflects a

balance between lymphocyte-driven immune

surveillance and monocyte-derived tumor-

associated macrophage recruitment, which

fosters immune evasion.

Circulating leukocyte subtypes such as

cMON and cNEU are closely linked to im-

mune suppression and angiogenesis in the

TIME 25, and they may work together to sup-

port tumor growth. Notably, cNEU are the

most common circulating leukocytes 26.

However, their levels show significant varia-

tion across different tumor stages (T/TNM

stage), which could indicate different func-

tions during disease progression in TSCC/

BOT SCC. Research shows that cNEU display

high plasticity within the TIME, with their

roles changing based on tumor stage and

microenvironmental factors. Especially in

advanced tumor stages, cNEU may encour-

age tumor growth through mechanisms like

promoting new blood vessel formation and

suppressing immune responses 27. In colorec-

tal cancer studies, increased neutrophil lev-

els are linked to disruptions in the intestinal

microbiota and facilitate peritoneal metas-

tasis of colorectal cancer by interacting with

tumor cells 28. Additionally, higher pretreat-

ment neutrophil counts and NLR have been

repeatedly confirmed as independent indica-

tors of poorer overall survival (OS) and dis-

ease-free survival (DFS) in meta-analyses in-

volving over 10,000 patients 29. The observed

correlation between cNEU and tumor stage

in this study likely reflects their dynamic

contribution to tumor progression.

Whether cWBC is causally associated

with the occurrence of HNSCC, particular-

ly with TSCC/BOT SCC, we conducted an

MR analysis. The IVW method demonstrat-

ed a robust positive association between

cWBC and TSCC/BOT SCC risk (β=0.416,

OR=1.516, P/FDR=0.005), with consistent

results across MR-Egger and weighted me-

dian methods (no evidence of heterogeneity

or pleiotropy, p>0.05). In contrast, cMON

associations were confounded by horizon-

tal pleiotropy (p<0.05), limiting causal

inference. This finding is consistent with

previous studies, and cWBC is captured by

tumour tissues to become infiltrating leu-

kocytes that promote tumour progression

30. The absence of significant associations

between cWBC and other HNSCC subsites

in our MR analysis may reflect the anatomi-

cal and biological heterogeneity of HNSCC,

which could differentially influence im-

mune cell recruitment and function. ICI

therapy involves the binding of anti-pro-

grammed cell death protein 1 (PD-1)/pro-

grammed cell death ligand 1 (PD-L1) anti-

bodies to induce autologous immune cells

to kill tumors 31. It has demonstrated sig-

nificant efficacy across multiple advanced

tumor types, with some patients achieving

durable responses 32. The role of cWBC in

immunotherapy has also been supported by

several studies. The PD-L1 status of cWBC

is associated with PD-L1 expression in im-

mune cells within the TIME 33, 34. In melano-

ma patients, the efficacy of ICI therapy cor-

relates with an adequate number of cLYM 35.

Studies have also found that the efficacy of

ICI therapy is significantly associated with

baseline cWBC and their subtypes (such

as LYM and MON), where higher baseline

cLYM is associated with better treatment

responses and longer progression-free sur-

vival, while higher cNEU is associated with

poorer prognosis 36, 37. Incidental immune-

related adverse events (irAEs) during im-

270 Zhu et al.

Investigación Clínica 67(2): 2026

munotherapy are also associated with cWBC

and their subtypes. Baseline cNEU, cLYM,

cMON, and cEOS, baseline platelet counts,

and increases in cWBC, cLYM, and cEOS

during follow-up are all associated with an

increased risk of irAEs 38-40. Collectively, our

data support a dual role for cWBC in TSCC/

BOT SCC: they may contribute to tumor ini-

tiation and progression while concurrently

modulating responses to immunotherapy.

Clinically, CPS levels are typically used as

a surrogate for PD-1 expression. Therefore,

this study analyzed the differences between

CPS expression and cWBC. Unfortunately,

CPS was not detected in some patients.

To exclude differences in detection results

between our institution and other institu-

tions, only patients with CPS expression

detected in our hospital were included,

resulting in many patients being recorded

as having no CPS expression values. The re-

sults showed that different CPS expression

levels were only associated with cMON in all

patients and HPV-negative patients, but no

differences were observed in cWBC or their

subtypes in HPV-positive patients, further

indicating the heterogeneity of tumor cells

under different HPV infection statuses. Ad-

ditionally, studies have found that retinoic

acid secreted by tumor tissues can induce

MON to differentiate into immunosuppres-

sive tumor-associated macrophages (TAMs),

thereby inhibiting the efficacy of immu-

notherapy 41. TAMs and dendritic cells, a

monocyte subset, may also indirectly pro-

mote tumor progression, which may explain

the different prognoses of HPV-negative and

HPV-positive tumors due to differences in

immune cells 42. This study has some limi-

tations. First, its focus on TSCC/BOT SCC

may limit the generalizability of findings to

other OPSCC subsites. The retrospective

design introduces potential confounding

due to unmeasured clinical variables, such

as treatment history. Additionally, the rela-

tively small sample size in the HPV-negative

subgroup may reduce the statistical power

of subgroup analyses. Furthermore, lifestyle

factors such as diet and environmental ex-

posures, which were not assessed, could in-

dependently influence the observed associ-

ations between circulating white blood cells

and tumor characteristics. Future research

should further expand the sample size and

incorporate more potential confounding

factors to comprehensively evaluate the as-

sociations between cWBC and TSCC/BOT

SCC. In-depth investigations using single-

cell sequencing technology may help reveal

the specific mechanisms of action of differ-

ent WBC subtypes in the TIME. Finally, ex-

ploring the potential of cWBC as biomark-

ers and therapeutic targets for TSCC/BOT

SCC will provide new insights for clinical

diagnosis and treatment.

This study demonstrates a causal link

between elevated cWBC and the risk of

TSCC/BOT SCC through combined ret-

rospective and Mendelian randomization

analyses. Significant associations between

cWBC subtypes and tumor characteristics,

especially in HPV-positive cases, underscore

their potential role in tumor progression

and immune regulation. The findings indi-

cate that cWBC may serve as non-invasive

biomarkers for risk assessment and treat-

ment stratification. Further research is

necessary to validate these results and in-

vestigate underlying mechanisms in larger,

prospective cohorts.

Funding

No sponsor or funder supported this

study.

ORCID ID of the authors

• Changyu Zhu (CZ):

0009-0003-5440-0288

• Shizhi He (SH):

0009-0003-7228-943X

• Zhixin Li (ZL):

0009-0004-8248-1529

• Yijun Shi (YS):

0009-0002-0097-2654

White blood cells and human papillomavirus risk in head and neck squamous cell carcinoma 271

Vol. 67(2): 260 - 274, 2026

• Jingyang Zhao (JZ):

0000-0004-5326-9537

• Wei Li (WL):

0000-0002-9991-7892

Author's contributions

Conceptualization: JZ; Methodology:

CZ, JZ; Data analysis: CZ, SH, ZL, YS; Writ-

ing- original draft: CZ, SH, ZL, YS; Review &

editing: CZ, JZ; Supervision: JZ. All authors

read and approved the final manuscript.

Conflict of interest

The authors state that they have no

conflicts of interest.

Ethical approval and informed consent

The retrospective analysis of patients

at our hospital has been approved by the

Ethics Committee under approval number

TREC2022-KY018.R1, dated 21st April 2022.

This study was performed in line with the

principles of the Declaration of Helsinki.

Data availability statement

All data generated or analyzed during

this study are included in this article. Fur-

ther inquiries can be directed to the corre-

sponding author.

Declaration of generative AI

and AI-assisted technologies

in the writing process

No generative AI or AI-assisted technol-

ogy was used in the writing process.

REFERENCES

1. Sung H, Ferlay J, Siegel RL, Laversanne

M, Soerjomataram I, Jemal A, et al. Glob-

al Cancer Statistics 2020: GLOBOCAN Es-

timates of Incidence and Mortality World-

wide for 36 Cancers in 185 Countries.

CA Cancer J Clin. 2021;71(3):209-249.

https://doi.org/10.3322/caac.21660.

2. Arif RT, Mogaddam MA, Merdad LA, Farsi

NJ. Does human papillomavirus modify

the risk of oropharyngeal cancer related to

smoking and alcohol drinking? A systemat-

ic review and meta-analysis. Laryngoscope

Investig Otolaryngol. 2022;7(5):1391-

1401. https://doi.org/10.1002/lio2.877.

3. Ernster JA, Sciotto CG, O’Brien MM,

Finch JL, Robinson LJ, Willson T, et

al. Rising incidence of oropharyngeal

cancer and the role of oncogenic hu-

man papilloma virus. Laryngoscope.

2007;117(12):2115-2128. https://doi.org/

10.1097/MLG.0b013 e31813e5fbb.

4. Licitra L, Zigon G, Gatta G, Sánchez MJ,

Berrino F, EUROCARE Working Group.

Human papillomavirus in HNSCC: a Euro-

pean epidemiologic perspective. Hematol

Oncol Clin North Am. 2008;22(6):1143-

1153, vii-viii. https://doi.org/10.1016/j.

hoc.2008.10.002.

5. Chaturvedi AK, Engels EA, Anderson WF,

Gillison ML. Incidence trends for human pap-

illomavirus-related and -unrelated oral squa-

mous cell carcinomas in the United States. J

Clin Oncol. 2008;26(4):612-619. https://doi.

org/10.1200/JCO.2007.14. 1713.

6. Zoschke IN, Bennis SL, Wilkerson JM,

Stull CL, Nyitray AG, Khariwala SS, et

al. HPV-related oropharyngeal cancer

early detection in gay and bisexual men

is an “orphan”: A qualitative analysis

among healthcare providers. Front Pub-

lic Health. 2023;11:1165107. https://doi.

org/10.3389/fpubh.2023.1165107.

7. Shi T, Hu Z, Tian L, Yang Y. Pan-cancer

landscape of CENPO and its underly-

ing mechanism in LUAD. Respir Res.

2023;24(1):113. https://doi.org/10.1186/

s12931-023-02408-3.

8. Chechlinska M, Kowalewska M, Nowak R.

Systemic inflammation as a confounding

factor in cancer biomarker discovery and

validation. Nat Rev Cancer. 2010;10(1):2-

3. https://doi.org/10.1038/nrc2782.

9. Thol K, Pawlik P, McGranahan N. Therapy

sculpts the complex interplay between can-

cer and the immune system during tumour

evolution. Genome Med. 2022;14(1):137.

272 Zhu et al.

Investigación Clínica 67(2): 2026

https://doi.org/10.1186/s13073-022-

01138-3.

10. Wang Y, Jiang H, Fu L, Guan L, Yang J,

Ren J, et al. Prognostic value and immu-

nological role of PD-L1 gene in pan-cancer.

BMC Cancer. 2024;24(1):20. https://doi.

org/10.1186/s12885-023-11267-6.

11. Xu J, Liu H, Wang T, Wen Z, Chen H,

Yang Z, et al. CCR7 Mediated Mimetic

Dendritic Cell Vaccine Homing in Lymph

Node for Head and Neck Squamous Cell

Carcinoma Therapy. Adv Sci (Weinh).

2023;10(17):e2207017. https://doi.org/

10.1002/advs.202207017.

12. Cohen EEW, Soulières D, Le Tour-

neau C, Dinis J, Licitra L, Ahn MJ, et

al. Pembrolizumab versus methotrexate,

docetaxel, or cetuximab for recurrent

or metastatic head-and-neck squamous

cell carcinoma (KEYNOTE-040): a ran-

domised, open-label, phase 3 study. Lancet.

2019;393(10167):156-167. https://doi.org

/10.1016/S0140-6736(18)31999-8.

13. Burtness B, Harrington KJ, Greil R,

Soulières D, Tahara M, de Castro G Jr,

et al. Pembrolizumab alone or with che-

motherapy versus cetuximab with che-

motherapy for recurrent or metastatic

squamous cell carcinoma of the head

and neck (KEYNOTE-048): a randomised,

open-label, phase 3 study. Lancet.

2019;394(10212):1915–1928. https://doi.

org/10.1016/S0140-6736(19)32591-7.

14. Roxburgh CS, McMillan DC. Role of sys-

temic inflammatory response in predicting

survival in patients with primary operable

cancer. Future Oncol. 2010;6(1):149-163.

https://doi.org/10.2217/fon.09.136.

15. Aoki H, Ueha S, Shichino S, Ogiwara H,

Hashimoto SI, Kakimi K, et al. TCR Rep-

ertoire Analysis Reveals Mobilization of

Novel CD8+ T Cell Clones into the Can-

cer-Immunity Cycle Following Anti-CD4

Antibody Administration. Front Immunol.

2019;9:3185. https://doi.org/10.3389/fim

mu.2018.03185.

16. Chen MH, Raffield LM, Mousas A,

Sakaue S, Huffman JE, Moscati A, et

al. Trans-ethnic and Ancestry-Specific

Blood-Cell Genetics in 746,667 Indi-

viduals from 5 Global Populations. Cell.

2020;182(5):1198-1213.e14. https://doi.

org/10.1016/j.cell.2020.06.045.

17. Kurki MI, Karjalainen J, Palta P, Sipilä

TP, Kristiansson K, Donner KM, et al.

FinnGen provides genetic insights from a

well-phenotyped isolated population. Na-

ture. 2023;613(7944):508-518. https://

doi.org/10.1038/s41586-022-05473-8.

18. Bowden J, Del Greco M F, Minelli C, Zhao

Q, Lawlor DA, Sheehan NA, et al. Improv-

ing the accuracy of two-sample summary-

data Mendelian randomization: moving

beyond the NOME assumption. Int J Epi-

demiol. 2019;48(3):728-742. https://doi.

org/10.1093/ije/dyy258.

19. Hemani G, Zheng J, Elsworth B, Wade KH,

Haberland V, Baird D, et al. The MR-Base

platform supports systematic causal infer-

ence across the human phenome. Elife.

2018;7:e34408. https://doi.org/10.7554/

eLife.34408.

20. Burgess S, Thompson SG. Interpreting

findings from Mendelian randomization

using the MR-Egger method. Eur J Epi-

demiol. 2017;32(5):377-389. https://doi.

org/10.1007/s10654-017-0276-5.

21. Chaturvedi AK, Engels EA, Pfeiffer RM,

Hernandez BY, Xiao W, Kim E, et al. Hu-

man papillomavirus and rising oropha-

ryngeal cancer incidence in the United

States. J Clin Oncol. 2011;29(32):4294-

4301. https://doi.org/10.1200/JCO.2011.

36.4596.

22. Pytynia KB, Dahlstrom KR, Sturgis

EM. Epidemiology of HPV-associat-

ed oropharyngeal cancer. Oral Oncol.

2014;50(5):380-386. https://doi.org/10.

1016/j.oraloncology.2013.12.019.

23. Gillison ML, Chaturvedi AK, Anderson

WF, Fakhry C. Epidemiology of Human

Papillomavirus-Positive Head and Neck

Squamous Cell Carcinoma. J Clin On-

col. 2015;33(29):3235-3242. https://doi.

org/10.1200/JCO.2015.61.6995.

24. Wei D, Liu J, Ma J. The value of lympho-

cyte to monocyte ratio in the prognosis of

head and neck squamous cell carcinoma:

a meta-analysis. PeerJ. 2023;11:e16014.

https://doi.org/10.7717/peerj.16014.

White blood cells and human papillomavirus risk in head and neck squamous cell carcinoma 273

Vol. 67(2): 260 - 274, 2026

25. Camargo S, Moskowitz O, Giladi A,

Levinson M, Balaban R, Gola S, et al.

Neutrophils physically interact with tu-

mour cells to form a signaling niche pro-

moting breast cancer aggressiveness. Nat

Cancer. 2025;6(3):540-558. https://doi.

org/10.1038/s43018-025-00924-3

26. Hurtado Gutiérrez MJ, Allard FL, Mo-

sha HT, Dubois CM, McDonald PP. Hu-

man Neutrophils Generate Extracellular

Vesicles That Modulate Their Functional

Responses. Cells. 2022;12(1):136. https://

doi.org/10.3390/cells12010136.

27. Ng M, Cerezo-Wallis D, Ng LG, Hidalgo

A. Adaptations of neutrophils in cancer.

Immunity. 2025;58(1):40-58. https://doi.

org/10.1016/j.immuni.2024.12.009.

28. Li Q, Xiao Y, Han L, Luo W, Dai W, Fang

H, et al. Microbiome dysbiosis, neutro-

phil recruitment and mesenchymal transi-

tion of mesothelial cells promotes perito-

neal metastasis of colorectal cancer. Nat

Cancer. 2025;6(3):493-510. https://doi.

org/10.1038/s43018-025-00910-9.

29. Tham T, Bardash Y, Herman SW, Costan-

tino PD. Neutrophil-to-lymphocyte ratio

as a prognostic indicator in head and neck

cancer: A systematic review and meta-anal-

ysis. Head Neck. 2018;40(11):2546-2557.

https://doi.org/10.1002/hed.25324.

30. Tan LY, Cockshell MP, Moore E, Myo Min

KK, Ortiz M, Johan MZ, et al. Vasculogenic

mimicry structures in melanoma support

the recruitment of monocytes. Oncoimmu-

nology. 2022;11(1):2043673. https://doi.

org/10.1080/2162402X.2022.2043673.

31. Wang J, Gai J, Zhang T, Niu N, Qi H,

Thomas DL 2nd, et al. Neoadjuvant ra-

dioimmunotherapy in pancreatic cancer

enhances effector T cell infiltration and

shortens their distances to tumour cells.

Sci Adv. 2024;10(6):eadk1827. https://

doi.org/10.1126/sciadv.adk1827.

32. Kim MJ, Hong SPD, Park Y, Chae YK. In-

cidence of immunotherapy-related hyper-

progressive disease (HPD) across HPD defi-

nitions and cancer types in observational

studies: A systematic review and meta-

analysis. Cancer Med. 2024;13(3):e6970.

https://doi.org/10.1002/cam4.6970.

33. Liu Y, Wu X, Feng Y, Jiang Q, Zhang S,

Wang Q, et al. Insights into the Onco-

genic, Prognostic, and Immunological

Role of BRIP1 in Pan-Cancer: A Compre-

hensive Data-Mining-Based Study. J Oncol.

2023;2023:4104639. https://doi.org/10.

1155/2023/4104639.

34. Ilié M, Szafer-Glusman E, Hofman V,

Chamorey E, Lalvée S, Selva E, et al. De-

tection of PD-L1 in circulating tumour

cells and white blood cells from patients

with advanced non-small-cell lung cancer.

Ann Oncol. 2018;29(1):193-199. https://

doi.org/10.1093/annonc/mdx636.

35. van Rossum PSN, Deng W, Routman

DM, Liu AY, Xu C, Shiraishi Y, et al. Pre-

diction of Severe Lymphopenia During

Chemoradiation Therapy for Esophageal

Cancer: Development and Validation of a

Pretreatment Nomogram. Pract Radiat

Oncol. 2020;10(1):e16-e26. https://doi.

org/10.1016/j.prro.2019.07.010.

36. Gungabeesoon J, Gort-Freitas NA, Kiss

M, Weissleder R, Klein AM, Pittet MJ,

et al. A neutrophil response linked to tu-

mour control in immunotherapy. Cell.

2023;186(7):1448-1464.e20. https://doi.

org/10.1016/j.cell.2023.02.032.

37. Xu F, Xu P, Cui W, Gong W, Wei Y, Liu B, et

al. Neutrophil-to-lymphocyte and platelet-

to-lymphocyte ratios may aid in identifying

patients with non-small cell lung cancer

and predicting Tumour-Node-Metastasis

stages. Oncol Lett. 2018;16(1):483-490.

https://doi.org/10.3892/ol.2018.8644.

38. Diehl A, Yarchoan M, Hopkins A, Jaffee

E, Grossman SA. Relationships between

lymphocyte counts and treatment-related

toxicities and clinical responses in pa-

tients with solid tumours treated with

PD-1 checkpoint inhibitors. Oncotarget.

2017;8(69):114268-114280. https://doi.

org/10.18632/oncotarget.23217.

39. Liu W, Liu Y, Ma F, Sun B, Wang Y, Luo

J, et. al. Peripheral Blood Markers Associ-

ated with Immune-Related Adverse Effects

in Patients Who Had Advanced Non-Small

Cell Lung Cancer Treated with PD-1 Inhibi-

tors. Cancer Manag Res. 2021; 13:765-771.

https://doi.org/10.2147/CMAR.S29 3200.

274 Zhu et al.

Investigación Clínica 67(2): 2026

40. Michailidou D, Khaki AR, Morelli MP,

Diamantopoulos L, Singh N, Grivas P.

Association of blood biomarkers and au-

toimmunity with immune related adverse

events in patients with cancer treated

with immune checkpoint inhibitors.

Sci Rep. 2021;11(1):9029. https://doi.

org/10.1038/s41598-021-88307-3.

41. Devalaraja S, To TKJ, Folkert IW, Na-

tesan R, Alam MZ, Li M, et al. Tumour-

Derived Retinoic Acid Regulates Intra-

tumoural Monocyte Differentiation to

Promote Immune Suppression. Cell.

2020;180(6):1098-1114.e16. https://doi.

org/10.1016/j.cell. 2020.02.042.

42. Olingy CE, Dinh HQ, Hedrick CC. Mono-

cyte heterogeneity and functions in cancer. J

Leukoc Biol. 2019;106(2):309-322. https://

doi.org/10.1002/JLB.4RI0818-311R.