Commonly Drawn Immunologic and Inflammatory Markers as Risk Predictors for Anal Cancer in Veterans Living with HIV

Abstract

Objectives

This study aimed to determine if immune inflammatory markers (neutrophil lymphocyte ratio (NLR), platelet lymphocyte ratio (PLR), and prognostic nutritional index (PNI)) correlate with anal cancer risk in people living with HIV and to compare these markers with the CD4/CD8 ratio.

Materials and Methods

This is a regional retrospective cohort study of veterans living with HIV who were screened for or diagnosed with anal neoplasia or cancer from 2001 to 2019. NLR, PLR, PNI, and CD4/CD8 ratio within one year of anal pathology results were computed. Patients with anal cancer were compared to patients without anal cancer. Regression modeling was used to estimate the odds of developing anal cancer.

Results

334 patients were included (37 with anal cancer, 297 without anal cancer). In patients with anal cancer, NLR and PLR were higher (2.17 vs 1.69, p = 0.04; 140 vs 110, p = 0.02, respectively), while PNI and CD4/CD8 ratio were lower (44.65 vs 50.01, p < 0.001; 0.35 vs 0.80, p < 0.001, respectively). On multivariate logistic regression modeling, only PNI (OR, 0.90; p = 0.001) and CD4/CD8 ratio (OR, 0.05; p < 0.001) were associated with increased anal cancer risk.

Conclusions

Although NLR and PLR independently correlate with anal cancer risk, when controlling for other risk predictors, only PNI and CD4/CD8 ratio were statistically significant biomarkers for anal cancer. The CD4/CD8 ratio is the strongest immune inflammatory marker that predicts risk of anal cancer among veterans living with HIV.

Précis

When controlling for other risk factors, the prognostic nutritional index and CD4/CD8 ratio are predictors of anal cancer risk in people living with HIV.

Introduction

People living with HIV (PLWH) are at a significant increased risk (20-fold) for developing anal squamous cell carcinoma and experience higher rates of progression to invasive cancer from anal precancerous lesions (i.e., high-grade squamous intraepithelial lesions [HSIL]) compared to HIV-negative individuals.^1,2 Due to this increased risk, it is recommended that PLWH undergo routine screening for anal HSIL and anal cancer.^3–6 Screening and surveillance protocols, however, are invasive, resource intensive, and require specialty training to perform. These challenges lead to poor screening compliance in many settings.^7–9

A recent multicenter randomized trial (ANCHOR) demonstrated that identification and treatment of anal high-grade squamous intraepithelial lesions (HSIL) in PLWH can prevent development of cancer.¹⁰ While anal cancer screening guidelines have been developed by agencies like the New York State Department of Health AIDS Institute in the United States, screening exams have been underutilized in many populations and examination frequency following detection and treatment of precancerous lesions are not always aligned with guidelines due to multifactorial barriers.^5,11 Targeting more frequent anal examination in patients at highest risk for anal cancer is critical as treatment of precursor lesions may prevent cancer. If prevention is not possible, early detection of cancer leads to better outcomes.¹² Accordingly, anal cancer risk stratification in PLWH is desperately needed. One health system in which this burden is particularly evident is the Department of Veterans Affairs (VA), which cares for more PLWH than any other health system in the United States.^4,11 This is especially important as the VA population is predominantly male and U.S.-based epidemiologic studies have demonstrated higher rates of association between HIV infection and anal cancer development among men compared to women (approximately 1 in 3 anal cancers in men occur in PLWH versus 1 in 30 in women).¹ Because the burden of screening and surveilling patients for anal cancer is high, additional non-invasive adjuncts to traditional screening measures are necessary. These adjuncts can aid in individualizing protocols based upon patient risk profiles and decrease the burden across a health system for low-risk patients while capturing those at highest risk. One promising adjunct to traditional screening measures is the use of biomarkers. The commonly drawn immunologic marker, CD4/CD8 ratio, has been shown to correlate with risk of HSIL and anal cancer in PLWH.^9,13 Additional immune inflammatory markers, such as neutrophil lymphocyte ratio (NLR), platelet lymphocyte ratio (PLR), and prognostic nutritional index (PNI) have shown value as prognostic indicators in numerous solid tumors, including anal cancer.^14–19 Although these immune inflammatory markers have shown promise in prognostication in some patient populations with anal cancer, it is unknown if these markers can predict anal cancer risk in veterans and how they compare to CD4/CD8 ratio as a risk predictor.

We sought to investigate the relationship of NLR, PLR, and PNI with anal cancer risk in a regional cohort of veterans living with HIV. We also sought to compare these markers with a known risk predictor, the CD4/CD8 ratio. We hypothesized that immunologic inflammatory markers would be associated with anal cancer risk among veterans living with HIV.

Methods

Data Source and Study Population

This regional retrospective cohort study includes all patients with HIV infection from a regional VA health care system encompassing eight VA hospitals. Patients were identified via query of the VA Corporate Data Warehouse (CDW). Veterans with HIV infection from 2001–2019 were identified with a validated method using two individualized diagnostic criteria (ICD-9/10 codes and immunologic laboratory markers) present in a single patient record.²⁰ This study was approved by the appropriate Institutional Review Boards.

We included only patients who were screened for or diagnosed with anal neoplasia or cancer confirmed by anal pathology results. A previously published and validated natural language processing algorithm was utilized to define anal pathology results (cytology and histopathology).⁹ Patients were grouped into one of five pathologic categories: 1) negative for neoplasia, 2) atypical squamous cells of undetermined significance (ASCUS) or atypical squamous cells cannot rule out high-grade (ASC-H), 3) low-grade squamous intraepithelial lesion (LSIL), 4) HSIL, or 5) anal cancer (anal squamous cell carcinoma only).

We excluded patients who did not have a complete blood count (CBC), CD4 count, CD8 count, or albumin drawn within one year of anal pathology results. Laboratory markers were collected including CBC, CD4 count, CD8 count, viral load, and albumin. Laboratory results were utilized to calculate NLR, PLR, PNI, and CD4/CD8 ratio. Laboratory values closest to anal pathology results were used.

Variables and Outcomes

NLR was calculated by dividing absolute neutrophil count by absolute lymphocyte count. PLR was calculated similarly, dividing absolute platelet count by absolute lymphocyte count. PNI was calculated as (10 × albumin) + (0.005 × absolute lymphocyte count). CD4/CD8 ratio was calculated by dividing CD4 count by CD8 count. Patient characteristics including age, sex, and race were collected. Known risk factors for anal cancer including non-HPV related sexually transmitted infection (STI) history (gonorrhea, chlamydia, hepatitis B, hepatitis C, syphilis, genital herpes simplex), anogenital condyloma history, and smoking history were collected. Additionally, cardiovascular disease history and diabetes history, which can impact immune inflammatory markers, were collected.^21,22

Statistical Analysis

Patient characteristics were compared between two cohorts: patients with anal cancer and those without cancer (negative, ASCUS/ASC-H, LSIL, and HSIL). Pearson’s chi square test was used for categorical variables and Welch’s t-test for continuous variables. Both univariate and multivariate logistic regression models were utilized to investigate the relationship between each biomarker and the development of anal cancer. Because NLR, PLR, and PNI are all associated with lymphocytes, they are highly correlated. Accordingly, these biomarkers were fitted in separate models to avoid multicollinearity. Age, race, and history of STI, anogenital condyloma, cardiovascular disease, and diabetes were controlled as covariates in the multivariate logistic regression models. Receiver operating characteristic (ROC) Youden analysis (J = sensitivity + specificity − 1) was utilized to evaluate biomarker’s classification performance to determine optimal cutoff for anal cancer. Statistical analyses were performed at the 0.05 level of significance using software R (version 3.6.0)

Results

Demographics

A total of 334 patients were included in our study, 37 (11%) patients were in the anal cancer cohort and 297 (89%) were in the non-cancer cohort (121 negative, 56 ASCUS/ASC-H, 45 LSIL, 75 HSIL). There was no significant difference among these two groups in age, sex, race, viral suppression, or history of STIs, anogenital condyloma, smoking, cardiovascular disease, or diabetes (Table 1).

Table 1.

Characteristics of patients with and without anal cancer

Characteristics	Cancer (n=37)	Non-cancer (n=297)	p value
Mean age, years (SD)	57.4 (10.1)	54.1 (12.0)	0.117
Sex, n (%)
Female	0 (0.0)	2 (0.7)	1.000
Male	37 (100.0)	295 (99.3)
Race, n (%)
Other	1 (3.0)	12 (4.1)	0.768
Black	20 (60.6)	159 (54.1)
White	12 (36.4)	123 (41.8)
Viral suppression, n (%)	21 (56.8)	207 (69.7)	0.159
STI, n (%)	22 (59.5)	207 (69.7)	0.281
Anogenital condyloma, n (%)	16 (43.2)	116 (39.1)	0.754
Smoking history, n (%)
Current	16 (43.2)	160 (53.9)	0.438
Former	6 (16.2)	34 (11.4)
Never	15 (40.5)	103 (34.7)
CVD history, n (%)	11 (29.7)	90 (30.3)	1.000
DM history, n (%)	11 (29.7)	76 (25.6)	0.732

STI, sexually transmitted infection; CVD, cardiovascular disease; DM, diabetes mellitus; SD, standard deviation.

Immune Inflammatory Markers

Both NLR and PLR were significantly higher among patients with anal cancer than non-cancer patients (2.17 vs 1.69, p = 0.041; 137.13 vs 114.53, p = 0.021, respectively). PNI and CD4/CD8 ratio were significantly lower among patients in the anal cancer cohort than those in the non-cancer cohort (44.65 vs 50.01, p < 0.001; 0.35 vs 0.80, p < 0.001, respectively) (Table 2). On univariate analysis, PLR (OR, 1.01; 95% CI, 1.00−1.01; p = 0.025), PNI (OR, 0.88; 95% CI, 0.83−0.93; p < 0.001), and CD4/CD8 ratio (OR, 0.05; 95% CI, 0.01−0.18; p < 0.001) were all significantly associated with anal cancer risk. The NLR, however, was not statistically associated with anal cancer (OR, 1.23; 95% CI, 1.00−1.52; p = 0.052). Age, race, viral suppression, and history of STI, anogenital condyloma, smoking, diabetes, and cardiovascular disease were not associated with anal cancer risk (Table 3).

Table 2.

Unadjusted comparison of immune inflammatory markers between patients with and without anal cancer

Markers	Cancer (n=37)	Non-cancer (n=297)	p value
Mean NLR (SD)	2.17 (1.6)	1.69 (1.3)	0.041
Mean PLR (SD)	137.13 (83.6)	114.53 (48.4)	0.021
Mean PNI (SD)	44.65 (8.0)	50.01 (5.9)	<0.001
Mean CD4/CD8 (SD)	0.35 (0.3)	0.80 (0.6)	<0.001

Significant results (p ≤ 0.05) are highlighted in bold. NLR, neutrophil lymphocyte ratio; PLR, platelet lymphocyte ratio; PNI, prognostic nutritional index; SD, standard deviation.

Table 3.

Univariate logistic regression modeling for risk of anal cancer

Characteristics	OR (95% CI)	p value
Age	1.02 (0.99–1.06)	0.118
Race
White	Reference	–
Black	1.29 (0.61–2.74)	0.509
Other	0.85 (0.10–7.15)	0.884
Viral suppression	0.57 (0.29–1.16)	0.114
STI	0.64 (0.32–1.29)	0.209
Anogenital condyloma	1.19 (0.60–2.37)	0.624
Smoking history
Current	Reference	–
Former	1.77 (0.64–4.84)	0.270
Never	1.46 (0.69–3.07)	0.324
CVD history	0.97 (0.46–2.05)	0.943
DM history	1.23 (0.58–2.61)	0.589
NLR	1.23 (1.00–1.52)	0.052
PLR	1.01 (1.001–1.012)	0.025
PNI	0.88 (0.83–0.93)	<0.001
CD4/CD8	0.05 (0.01–0.18)	<0.001

Significant results (p ≤ 0.05) are highlighted in bold. STI, sexually transmitted infection; CVD, cardiovascular disease; DM, diabetes mellitus; NLR, neutrophil lymphocyte ratio; PLR, platelet lymphocyte ratio; PNI, prognostic nutritional index; OR, odds ratio; CI, confidence interval.

On multivariate analysis, when controlling for patient factors known to impact anal cancer risk, only PNI and CD4/CD8 ratio remained significantly associated with malignancy: NLR (OR, 1.15; 95% CI, 0.90–1.47; p = 0.251), PLR (OR, 1.00; 95% CI, 1.00–1.01; p = 0.565), PNI (OR, 0.90; 95% CI, 0.85–0.96; p = 0.001), and CD4/CD8 (OR, 0.05; 95% CI, 0.01–0.22; p < 0.001) (Figure 1). The optimal threshold for using PNI as a risk factor for anal cancer was determined to be 47.1 (sensitivity, 0.69; specificity, 0.65; area under the curve, 0.70) (Figure 2).

FIGURE 1.

Multivariate logistic regression modeling for risk of anal cancer with each model examining a unique immune inflammatory marker while controlling for confounders. Significant results (p ≤ .05) are highlighted in bold. STI, sexually transmitted infection; CVD, cardiovascular disease; DM, diabetes mellitus; NLR, neutrophil lymphocyte ratio; PLR, platelet lymphocyte ratio; PNI, prognostic nutritional index; OR, odds ratio.

FIGURE 2.

Receiver operating characteristic curve with Youden analysis to determine optimal PNI cutoff for maximal sensitivity value predictive of anal cancer. Cutoff value = 47.1, sensitivity = 69%, specificity = 65%; PNI, prognostic nutritional index.

Discussion

In this retrospective analysis of a regional population of veterans living with HIV, the PNI and CD4/CD8 ratio were significantly lower in the anal cancer cohort, signifying their potential as aids in risk stratification of PLWH. The CD4/CD8 ratio demonstrated the greater magnitude of effect, thus reinforcing its place as a strong marker of anal cancer risk in veterans with HIV.

Anal cancer is primarily driven by persistent infection with high-risk strains of the human papilloma virus (e.g., HPV-16 and HPV-18) and invasive cancer occurs when HPV remains unchecked by the immune system.⁷ Although most PLWH have undetectable HIV viral loads and often normalized CD4 counts with modern antiretroviral treatment, infections with high-risk strains of HPV are cleared at lower rates in this population.²³ The persistence of HPV infection allows for anal carcinogenesis to occur and is particularly likely when total CD8 counts remain high.¹³ In the present study, some patients were not virally suppressed, however, this was not associated with anal cancer development.

Accordingly, other markers besides CD4 counts and viral loads are needed to assess for degree of immune recovery, a necessity for risk stratification of anal cancer. Immunologic markers are an ideal tool for risk stratification as they are routinely drawn in PLWH and reflect a patient’s underlying immune status. One marker which has been previously reported to be a strong predictor of anal cancer risk in both civilian and veteran PLWH is the CD4/CD8 ratio.^9,13 In our current study, we sought to determine if other immune inflammatory biomarkers would also predict risk in PLWH. Immune inflammatory markers such as NLR, PLR, and PNI have been shown to stratify risk in patients diagnosed with solid organ cancers, and are calculated based on routinely drawn lab values.^14–19 The predictive value of NLR and PNI for anal cancer has been previously demonstrated in a single center civilian cohort.¹⁹ In this current study of veterans, however, when controlling for confounders, PNI was statistically predictive of anal cancer risk but NLR was not. The CD4/CD8 ratio was also predictive of anal cancer risk but with a greater magnitude of effect. This result reinforces the CD4/CD8 ratio as the most predictive known immune marker in this cohort to aid in counseling patients regarding anal cancer risk. Prior work has identified a CD4/CD8 ratio cutoff of < 0.8 that can be used to risk stratify patients.⁹ The PNI is likely to be of value in patients who do not have a recent CD4 and CD8 count available to the clinician and in this scenario a cutoff of 47 could be utilized.

Experts recommend high-risk individuals, such as PLWH, undergo routine anal cancer screenings.^3–6 This recommendation is supported within the VA health system, which recommends yearly anal cancer screening in veterans living with HIV.⁴ Despite the existence of recommendations, screening rates remain low in many populations of PLWH, with reports of initiation in civilians ranging from approximately 30% to 50%.^24,25 In the VA population, the initiation of screening appears to be even lower (approximately 15%) over a two-decade period.⁹ This lower rate is likely more representative of clinical practice overall considering civilian populations studied are cared for within specialized settings and not likely representative of the population of PLWH as a whole. A 2014 survey of surgeon members of the American Society of Colon and Rectal Surgeons revealed that only 1/3 of respondents had ever performed anal cancer screening with anal cytology collection.²⁶ This low rate of screening performed by experts in the field underscores the importance of incorporating other markers of anal cancer risk that can be more easily monitored to tailor risk and target scarce resources to the highest risk individuals. Our study identifies the PNI and CD4/CD8 ratio as immune inflammatory markers that can be utilized for this purpose. A clinician who is not trained in anal cancer screening exams can examine commonly drawn lab values to help risk-stratify their patient population and refer to specialists if needed.

Another area where commonly drawn biomarkers might prove useful is in the risk stratification of patients already participating in anal cancer screening and surveillance protocols. The natural history of progression from anal neoplasia to invasive cancer is poorly understood, as certain patients progress to advanced anal disease or, at times, spontaneously regress to lower grades of neoplasia.²⁷ PLWH have high rates of neoplasia recurrence, and progression to cancer can occur even in the setting of active treatment of precancerous lesions.^2,3,10,28 This uncertainty has led to a wide range of surveillance protocols with variable temporal recommendations for interval screenings.²⁹ Given this lack of standardization, additional adjuncts are necessary to inform surveillance protocols. The information yielded by immunologic and inflammatory markers, such as the PNI and CD4/CD8 ratio, can help identify patients most in need of surveillance at more frequent time points. Conversely, individuals at lesser risk might undergo screening with less frequency, thereby decreasing the burden on patients, care providers, and health systems.

Our study does have several limitations. This is a retrospective study that focused on a limited patient population within the VA health system that consisted almost exclusively of male patients. Additionally, we are unable to accurately identify numbers of sexual partners as a risk predictor due to medical record limitations. For patients who did not have histology performed after cytopathology, severity of their anal disease was determined by cytology result, which is known to have limitations in sensitivity and specificity.³⁰ Also, it remains unclear if the data obtained from this current study is generalizable to a broader and more diverse patient population, as rates of anal cancer screening in the VA patient population have been shown to be low.⁹

Conclusions

Among a regional cohort of veterans living with HIV, the immune inflammatory markers PNI and CD4/CD8 ratio correlate with anal cancer risk after controlling for confounders. In this analysis, the CD4/CD8 ratio is the strongest predictor of anal cancer in PLWH and provides a valuable adjunct in the development of future anal cancer screening and surveillance protocols. If a recent CD4/CD8 ratio is not available to the clinician, the PNI can be helpful to risk predict.

Acronyms

HIV

human immunodeficiency virus

CD

cluster of differentiation

NLR

neutrophil lymphocyte ratio

PLR

platelet lymphocyte ratio

PNI

prognostic nutritional index

OR

odds ratio

PLWH

people living with human immunodeficiency virus

SIL

squamous intraepithelial lesion

HSIL

high-grade squamous intraepithelial lesion

VA

Veterans Affairs

CDW

corporate data warehouse

ICD

international classification of diseases

ASCUS

atypical squamous cells of undetermined significance

ASC-H

atypical squamous cells cannot rule high-grade

LSIL

low-grade squamous intraepithelial lesion

CBC

complete blood count

HPV

human papilloma virus

STI

sexually transmitted infection

CI

confidence interval

CVD

cardiovascular disease

DM

diabetes mellitus

SD

standard deviation