Research

JAMA Otolaryngology–Head & Neck Surgery | Original Investigation

Host Factors Independently Associated With Prognosis

in Patients With Oral Cavity Cancer
Cristina Valero, MD, PhD; Daniella K. Zanoni, MD; Anjali Pillai, MSc; Ian Ganly, MD, PhD; Luc G. T. Morris, MD, MSc;
Jatin P. Shah, MD; Richard J. Wong, MD; Snehal G. Patel, MD

Supplemental content
IMPORTANCE The association and interaction of host characteristics with prognosis in
patients with oral cavity squamous cell carcinoma (OSCC) are poorly understood. There is
increasing evidence that host characteristics are associated with treatment outcomes of
many cancers.

OBJECTIVES To examine the host factors associated with prognosis in patients with OSCC and
their interactions to create a numerical index that quantifies the prognostic capacity of these
host characteristics.

DESIGN, SETTING, AND PARTICIPANTS This retrospective cohort study included patients with
OSCC treated surgically at a tertiary care center from January 1, 1998, to December 31, 2015.
From a departmental OSCC database of 1377 previously untreated patients, 68 patients with
missing data on any host variable of interest within a month before the start of treatment
were excluded, leaving 1309 patients. Data analysis was performed from October 21, 2019, to
December 10, 2019.

EXPOSURE Primary surgery for OSCC.

MAIN OUTCOMES AND MEASURES Overall survival (OS) was the primary end point, and
disease-specific survival (DSS) was the secondary end point. Optimal cutoffs for each variable
were identified using recursive-partitioning analysis with the classification and regression
tree method using OS as the dependent variable. Body mass index (BMI) and pretreatment
peripheral blood leukocyte count, platelet count, hemoglobin level, and albumin level were
analyzed. A host index (H-index) was developed using independent factors associated with
OS.

RESULTS A total of 1309 patients (731 [55.8%] male; mean [SD] age, 62 [14.3] years)
participated in the study. When including all the host-related factors in a multivariable
analysis, all except BMI (hazard ratio [HR], 1.14; 95% CI, 0.80-1.63) were independently
associated with outcomes. For example, compared with a hemoglobin level of 14.1 g/dL or
greater, the HR for a level of 12.9 to 14.0 g/dL was 1.42 (95% CI, 1.13-1.77) and for a level of 12.8
g/dL or less was 1.51 (95% CI, 1.18-1.94), and compared with an albumin level of 4.3 g/dL or
greater, the HR for a level of 3.7 to 4.2 g/dL was 1.18 (95% CI, 0.95-1.45) and for a level of 3.6
g/dL or less was 3.64 (95% CI, 2.37-5.58). An H-index of 1.4 or less was associated with a 74%
5-year OS, an H-index of 1.5 to 3.5 with a 65% 5-year OS, and an H-index of 3.6 or higher with
a 38% 5-year OS; for DSS, the 5-year survival was 84%, 80%, and 64%, respectively.
Compared with patients with an H-index score of 1.4 or less, patients with H-index scores of
1.5 to 3.5 (hazard ratio, 1.474; 95% CI, 1.208-1.798) and 3.6 or higher (hazard ratio, 3.221; 95%
CI, 2.557-4.058) had a higher risk of death.

CONCLUSIONS AND RELEVANCE The findings suggest that pretreatment values of neutrophils, Author Affiliations: Head and Neck
Service, Department of Surgery,
monocytes, lymphocytes, hemoglobin, and albumin are independently associated with
Memorial Sloan Kettering Cancer
prognosis in patients with OSCC. The interactions between these host factors were Center, New York, New York (Valero,
incorporated into a novel H-index that quantified the prognostic capacity of host Zanoni, Pillai, Ganly, Morris, Shah,
characteristics associated with OSCC. Wong, Patel); Department of
Oncology, Radiotherapy and Plastic
Surgery, Sechenov University,
Moscow, Russia (Shah).
Corresponding Author: Snehal G.
Patel, MD, Head and Neck Service,
Department of Surgery, Memorial
Sloan Kettering Cancer Center, 1275
JAMA Otolaryngol Head Neck Surg. doi:10.1001/jamaoto.2020.1019 York Ave, New York, NY 10065
Published online June 11, 2020. (patels@mskcc.org).

(Reprinted) E1
© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Uppsala University User on 06/13/2020

 Research Original Investigation
T
he American Joint Committee on Cancer (AJCC)/
Union for International Cancer Control (UICC) staging
system generally estimates prognosis of patients based
on only tumor factors.1 However, there is increasing evi-
dence of several host characteristics being associated with on-
cologic outcomes.2-6 To our knowledge, the only example of
a staging system for head and neck cancer that takes into con-
sideration a host factor is the staging system for differenti-
ated thyroid cancers, in which age is incorporated in the prog-
nostic stratification of patients.
Besides well-known host characteristics, such as age, to-
bacco and alcohol consumption, and medical comorbidities,
the most widely investigated host prognostic factors are pre-
treatment peripheral blood leukocytes.7-10 Additional host fac-
tors, such as platelet count, hemoglobin level, albumin level,
and body mass index (BMI) (calculated as weight in kilo-
grams divided by height in meters squared), have also been
studied as potential prognostic factors in several cancer
types.3,11-13
It is logical to assume that these and other host factors are
not associated with the host-tumor interaction in isolation.
Many of these factors are interdependent and interact in a com-
plex manner to influence prognosis not only directly but also
indirectly by affecting the treatment decisions for patients.
Therefore, attempts have been made to combine various sig-
nificant host factors into a comprehensive index for assess-
ing this interaction in association with prognosis.14,15 For ex-
ample, Jafri et al 14 described the advanced lung cancer
inflammation index (ALI), an index that combines host
factors, such as albumin level, BMI, and the neutrophil-
lymphocyte ratio.
A previous study16 reported the association of pretreat-
ment peripheral blood leukocyte levels with outcomes of oral
cavity squamous cell carcinomas (OSCCs). Higher neutrophil
and monocyte counts and lower lymphocyte counts were as-
sociated with poorer outcomes. The aims of this study were
to assess the association of host factors with prognosis and to
examine the interaction between all the analyzed factors to cre-
ate a numerical index to quantify the prognostic capacity of
host characteristics in OSCCs.
Methods
This cohort study used our departmental database of 1377 pa-
tients with a biopsy-proven invasive OSCC treated with pri-
mary surgery at Memorial Sloan Kettering Cancer Center from
January 1, 1998, to December 31, 2015. Data analysis was per-
formed from October 21, 2019, to December 10, 2019. The study
design was approved by the Memorial Sloan Kettering Cancer
Center Institutional Review Board, which determined that no
informed consent was required from participants because of
the retrospective, deidentified nature of the study.
Exclusion criteria were synchronous head and neck squa-
mous cell carcinomas, prior treatment of the reference carci-
noma, distant metastasis at presentation, and history of
nonendocrine head and neck cancer. In addition to tumor vari-
ables and pretreatment peripheral blood leukocyte levels
E2 JAMA Otolaryngology–Head & Neck Surgery Published online June 11, 2020 (Reprinted)
© 2020 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Uppsala University User on 06/13/2020
Host Factors Independently Associated With Prognosis in Patients With Oral Cavity Cancer
Key Points
Questions What host factors are associated with prognosis in
patients with oral cavity cancer, and can these factors be used to
develop an index that quantifies their prognostic capacity?
Findings In this cohort study of 1309 patients with oral cavity
squamous cell carcinomas, an index (H-index) was developed that
combined all the host factors that were independently associated
with overall survival. The index, which included the pretreatment
neutrophil count, monocyte count, lymphocyte count, albumin
level, and hemoglobin level, accurately stratified patients into
groups with differences in overall and disease-specific survival.
Meaning The findings suggest that the H-index can be used to
quantify the prognostic capacity of host characteristics associated
with oral cavity squamous cell carcinomas.
(neutrophil, monocyte, and lymphocyte counts), the host-
related variables of interest were BMI and the pretreatment pe-
ripheral blood absolute counts of eosinophils, basophils, plate-
lets, hemoglobin, and albumin. We used the blood test closest
to surgery, only considering those blood tests performed within
a month before surgery. We excluded 68 patients with miss-
ing data on any variable of interest, leaving 1309 patients.
Optimal cutoffs for each variable were identified using a
recursive-partitioning analysis with the classification and re-
gression tree method using overall survival (OS) as the pri-
mary outcome of interest.17 The established cutoffs were also
used to analyze differences in disease-specific survival (DSS)
as the secondary outcome of interest. In addition to the afore-
mentioned variables, we analyzed other previously pub-
lished ratios and indexes, such as the platelet-lymphocyte ra-
tio and the ALI, calculated as: (BMI × albumin)/neutrophil-
lymphocyte ratio. 14 Parameters for which the recursive-
partitioning analysis did not find a feasible cutoff or for which
the cutoff stratified patients in 2 groups with 1 of them repre-
senting less than 5% of the cohort were excluded from the
analyses.
We analyzed the prognostic capacity of each host vari-
able (for the variables of interest from this study and a previ-
ous study16) using univariable analysis for both OS and DSS and
then we conducted a multivariable analysis that included as
independent variables the clinicopathologic characteristics and
the host variables that were significant in the univariable analy-
ses (eTable 1 in the Supplement). The host variables that were
independently associated with outcomes in the multivari-
able analysis were then combined to create the host index
(H-index), which is calculated using the following formula:
([neutrophils × monocytes]/[lymphocytes × hemoglo-
bin × albumin]) × 100.
Statistical Analysis
We evaluated the association between host variables and pa-
tient characteristics using the t test or 1-way analysis of vari-
ance. Survival curves were calculated according to the Kaplan-
Meier method, and differences in survival were compared using
the log-rank test. Hazard ratios (HRs) were calculated accord-
ing to the Cox proportional hazards regression model, which
jamaotolaryngology.com
 Host Factors Independently Associated With Prognosis in Patients With Oral Cavity Cancer Original Investigation Research

was also used to perform the multivariable analyses. All sta-

Table 1. Clinicopathologic Characteristics of the Study Cohort
tistical analyses were conducted using SPSS software,
version 25.0 (IBM Corp) and Stata software, version 15 Characteristic Finding (N = 1309)a
(StataCorp). Age, mean (SD) [range], y 62.1 (14.3) [18.3-100.4]
Sex
Male 731 (55.8)
Female 578 (44.2)
Results
Tobacco use
A total of 1309 patients (731 [55.8%] male; mean [SD] age, 62 Never 488 (37.3)
[14.3] years) participated in the study. The clinicopathologic Ever 821 (62.7)
characteristics of the study cohort are given in Table 1. His-
Alcohol use
tory of tobacco and alcohol use was reported by 821 (62.7%)
Never 392 (29.9)
and 917 (70.1%), respectively. We considered patients as smok-
Ever 917 (70.1)
ers (ever) if they reported a history of actively smoking or hav-
WUHNCI scoreb
ing smoked tobacco at any point in their life. Comorbidities
were recorded according to the Washington University Head 0 933 (71.3)

and Neck Comorbidity Index (WUHNCI) (scores range from 0 ≥1 376 (28.7)
to 15, with higher scores indicating worse prognosis because Subsite
of comorbidities), with 376 patients (28.7%) having a WUHNCI Oral tongue 696 (53.2)
score of 1 or higher at the time of diagnosis.18 The most com- Lower gum 174 (13.3)
mon primary tumor subsite was the tongue (696 [53.2%]). Floor of the mouth 155 (11.8)
A total of 630 patients (48.2%) had an advanced pathologic Buccal mucosa 104 (7.9)
stage (stages III–IV) according to AJCC TNM classification.1 Me-
Upper gum 88 (6.7)
dian follow-up time from the date of surgery was 39 months
Retromolar trigone 67 (5.1)
(range, 1-221 months). Five-year OS was 64.2%, and 5-year DSS
Hard palate 25 (1.9)
was 79.7%.
pTc
When conducting the recursive-partitioning analysis for
each of the host variables of interest and following the previ- pT1 420 (32.1)

ously mentioned exclusion criteria (no cutoff found or cutoff pT2 327 (25.0)
creating only 2 groups with 1 of them representing <5% of the pT3 246 (18.8)
cohort), we excluded eosinophil count, basophil count, plate- pT4 252 (19.3)
let count, and platelet-lymphocyte ratio from further analy- Not recorded 64 (4.9)
ses. The recursive-partitioning analysis for these variables is pNc
shown in eFigure 1 in the Supplement. The remaining host vari- pN0 891 (68.1)
ables analyzed had a feasible cutoff and were used for further pN1 118 (9.0)
analyses.
pN2 129 (9.9)
We first analyzed the 5-year OS and DSS for each host vari-
pN3 152 (11.6)
able according to the categories defined by the recursive-
Not recorded 19 (1.5)
partitioning analysis for the variables of interest from this study
pStagec
and a previous study (Table 2).16 All variables were associ-
ated with OS and DSS. An ordered decrease in OS was associ- I 381 (29.1)

ated with decreases in hemoglobin level (≥14.1 vs 12.9-14.0 II 230 (17.6)

g/dL: HR, 1.372 [95% CI, 1.121-1.680]; ≥14.1 vs ≤12.8 g/dL: HR, III 201 (15.4)
2.071 [95% CI, 1.682-2.549]), albumin level (≥4.3 vs 3.7-4.2 g/dL: IV 429 (32.8)
HR, 1.726 [95% CI, 1.438-2.071]; ≥4.3 vs ≤3.6 g/dL: HR, 5.754 Not recorded 68 (5.2)
[95% CI, 3.898-8.492]), BMI (>19.3 vs ≤19.3: HR, 2.078 [95% Grade
CI, 1.526-2.831]), and lymphocyte count (>0.8 vs ≤0.8/μL: HR, Differentiated
2.358 [95% CI, 1.731-3.213]). An ordered decrease in DSS was
Well 219 (16.7)
also associated with decreases in these measures. An ordered
Moderately 830 (63.4)
decrease in OS was associated with increases in neutrophil
Poorly 199 (15.2)
count (≤4.8 vs 4.7-9.0/μL: HR, 1.504 [95% CI, 1.264-1.791]; ≤4.8
Not recorded 61 (4.7)
vs ≥9.1/μL: HR, 3.060 [95% CI, 2.157-4.341]) and monocyte
count (≤0.3 vs >0.3: HR, 1.418 [95% CI, 1.191-1.689]). An or- Perineural invasion

dered decrease in DSS was also associated with increases in Absent 819 (62.6)
neutrophil and monocyte counts. The OS and DSS curves ac- Present 402 (30.7)
cording to hemoglobin levels, albumin levels, and BMI are Not recorded 88 (6.7)
shown in Figure 1 and eFigure 2 in the Supplement, respec- (continued)
tively. When analyzing ALI, patients with the lowest score had

jamaotolaryngology.com (Reprinted) JAMA Otolaryngology–Head & Neck Surgery Published online June 11, 2020 E3

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Uppsala University User on 06/13/2020

 Research Original Investigation
Table 1. Clinicopathologic Characteristics of the Study Cohort
(continued)
Characteristic Finding (N = 1309)a
Lymphovascular invasion
Absent 1044 (79.8)
Present 177 (13.5)
Not recorded 88 (6.7)
Margins
Negative 385 (29.4)
Close 781 (59.7)
Positive 137 (10.5)
Not recorded 6 (0.5)
Treatment
Surgery 807 (61.7)
Surgery and adjuvant radiotherapy 379 (29.0)
Surgery and adjuvant chemoradiotherapy 123 (9.4)
Abbreviation: WUHNCI, Washington University Head and Neck Comorbidity
Index.
a
Data are presented as number (percentage) of patients unless otherwise
indicated.
b
The WUHNCI scores range from 0 to 15, with higher scores indicating worse
prognosis because of comorbidities.18
c
Based on the American Joint Committee on Cancer, 8th edition.1
a decrease in OS compared with the rest of the cohort, and pa-
tients with the highest score had the best OS. However, the 2
intermediate groups did not follow the expected order. For DSS,
the categories did not stratify survival in an ordered manner,
with the 2 intermediate categories being the ones with the best
and worst DSS. The OS and DSS curves according to ALI cat-
egories are shown in eFigure 3 in the Supplement.
When including all the host-related factors in a multivari-
able analysis, all except BMI (hazard ratio [HR], 1.14; 95% CI,
0.80-1.63) were independently associated with outcomes. For
example, compared with a hemoglobin level of 14.1 g/dL or
greater, the HR for a level of 12.9 to 14.0 g/dL was 1.42 (95%
CI, 1.13-1.77) and for a level of 12.8 g/dL or less was 1.51 (95%
CI, 1.18-1.94), and compared with an albumin level of 4.3 g/dL
or greater, the HR for a level of 3.7 to 4.2 g/dL was 1.18 (95%
CI, 0.95-1.45) and for a level of 3.6 g/dL or less was 3.64
(95% CI, 2.37-5.58) (eTable 1 in the Supplement). On the basis
of these results, we created the H-index, an index that ac-
counts for the prognostic capacity and interactions of all the
host-related factors that were independently associated with
outcomes. An H-index of 1.4 or less was associated with a 74%
5-year OS, an H-index of 1.5 to 3.5 with a 65% 5-year OS, and
an H-index of 3.6 or higher with a 38% 5-year OS; for DSS, the
5-year survival was 84%, 80%, and 64%, respectively. Com-
pared with patients with an H-index score of 1.4 or less, pa-
tients with an H-index score of 1.5 to 3.5 (HR, 1.474; 95% CI,
1.208-1.798) and those with an H-index of 3.6 or higher (HR,
3.221; 95% CI, 2.557-4.058) had a greater risk of death.
Survival curves according to H-index categories defined
by the recursive-partitioning analysis cutoffs are shown in
Figure 2. In multivariable analysis using the association of H-
index with OS (Table 3), the H-index showed a good balance
E4 JAMA Otolaryngology–Head & Neck Surgery Published online June 11, 2020 (Reprinted)
© 2020 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Uppsala University User on 06/13/2020
Host Factors Independently Associated With Prognosis in Patients With Oral Cavity Cancer
in the number of patients included in each category. After con-
trolling for clinicopathologic characteristics in the multivari-
able analysis, the H-index showed a 2 times greater risk of death
for patients in the highest category (H-index ≥3.6: HR, 1.988;
95% CI, 1.521-2.598) compared with the patients in the low-
est category (H-index ≤1.4 [reference category]).
Discussion
The current AJCC/UICC staging system for most head and neck
cancers, including OSCC, considers only tumor-related fac-
tors to stratify patients into prognostic groups. Despite the cur-
rent system’s ability to estimate prognosis, heterogeneity
within staging groups still exists. We hypothesized that inclu-
sion of both tumor and host-related factors in the staging sys-
tem would increase its prognostic fidelity by decreasing hetero-
geneity within each group.
Screening of the possible significant host prognostic fac-
tors is the natural first step toward including host factors in
the staging system. The most commonly investigated host fac-
tors besides age, tobacco and alcohol consumption, and medi-
cal comorbidities are pretreatment peripheral blood leuko-
cyte counts. A previous study16 found that elevated neutrophil
count, monocyte count, neutrophil-lymphocyte ratio, or sys-
temic inflammatory response index and a decreased lympho-
cyte count are correlated with worse oncologic outcomes. Other
host factors, such as pretreatment peripheral blood eosino-
phil count, basophil count, platelet count, hemoglobin level,
albumin level, C-reactive protein (CRP) level, and a wide range
of ratios and indexes combining these parameters, have been
analyzed.
Only a few studies19-21 have analyzed the prognostic ca-
pacity of eosinophil and basophil counts by evaluating their
pretreatment values, and the results were consistent with our
findings, in which neither eosinophil nor basophil count were
associated with outcomes. However, Wei et al22 found that
lower eosinophil and basophil counts were associated with
worse outcomes in patients with stage I to III colorectal can-
cer, and Holub and Biete4 found the same results for eosino-
phil counts in patients with cervical cancer.
Several studies11,23-28 have analyzed the association of
platelet count with outcomes, and, similarly to neutrophil-
lymphocyte ratio, the platelet-to-lymphocyte ratio has also
been used. Li et al23 found differences in survival associated
with platelet or platelet-lymphocyte ratio counts, revealing that
higher counts were associated with worse survival, with plate-
let count being a surrogate marker of inflammation and lym-
phocyte count with immune status. On the other hand, in sev-
eral studies,11,24-28 including the present study, platelet count
and platelet-lymphocyte ratio were not associated with out-
comes. It is possible that platelet values would act as a prog-
nostic factor only at extreme values; low values are likely as-
sociated with poor performance status, whereas high values
may be associated with a proinflammatory status, both of
which are associated with worse outcomes.
C-reactive protein level and various indexes that include
CRP level, such as the Glasgow prognostic score, have been
jamaotolaryngology.com
 Host Factors Independently Associated With Prognosis in Patients With Oral Cavity Cancer Original Investigation Research

Table 2. Survival Outcomes According to Host Variables Using the Cutoffs Found in a Previous Study16
and in This Study

Overall survival Disease-specific survival

5-y
No. of 5-y Survival,
Variable patients Survival, % HR (95% CI) % HR (95% CI)
Neutrophil count, /μL
≤4.8 735 69.5 1 [Reference] 83.1 1 [Reference]
4.7-9.0 524 60.0 1.504 (1.264-1.791) 76.7 1.517 (1.162-1.982)
≥9.1 50 31.5 3.060 (2.157-4.341) 60.0 3.045 (1.800-5.153)
Monocyte count, /μL
≤0.3 552 70.9 1 [Reference] 83.1 1 [Reference]
>0.3 757 59.1 1.418 (1.191-1.689) 77.0 1.368 (1.048-1.785)
Lymphocyte count, /μL
>0.8 1242 65.3 1 [Reference] 80.0 1 [Reference]
≤0.8 67 42.7 2.358 (1.731-3.213) 74.9 1.453 (0.817-2.600)
Hemoglobin level, g/dL
≥14.1 563 72.8 1 [Reference] 85.2 1 [Reference]
12.9-14.0 429 62.9 1.372 (1.121-1.680) 78.1 1.516 (1.111-2.067)
≤12.8 317 50.8 2.071 (1.682-2.549) 71.6 2.117 (1.534-2.921)
Albumin level, g/dL
≥4.3 912 69.8 1 [Reference] 80.7 1 [Reference]
3.7-4.2 366 54.5 1.726 (1.438-2.071) 79.1 1.255 (0.941-1.673)
≤3.6 31 15.6 5.754 (3.898-8.492) 61.8 3.360 (1.712-6.593)
BMI
>19.3 1240 65.3 1 [Reference] 80.4 1 [Reference]
≤19.3 69 43.8 2.078 (1.526-2.831) 65.0 2.107 (1.302-3.410)
ALI14 Abbreviations: ALI, advanced lung
cancer inflammation index; BMI,
≥30.0 969 70.6 1 [Reference] 82.7 1 [Reference]
body mass index (calculated as
27.5-29.9 60 44.8 2.455 (1.754-3.436) 58.2 2.902 (1.798-4.684) weight in kilograms divided by height
21.8-27.4 112 65.3 1.155 (0.846-1.576) 86.5 0.911 (0.536-1.550) in meters squared); H-index, host
index; HR, hazard ratio.
≤21.9 168 34.8 2.932 (2.379-3.614) 62.7 2.654 (1.916-3.677)
SI conversion factors: To convert
H-index
neutrophils, monocytes, and
≤1.4 505 73.8 1 [Reference] 84.4 1 [Reference] lymphocytes to ×109/L, multiply by
1.5-3.5 614 64.6 1.474 (1.208-1.798) 80.0 1.372 (1.019-1.847) 0.001; to convert hemoglobin and
albumin to grams per liter, multiply
≥3.6 190 37.7 3.221 (2.557-4.058) 64.1 2.704 (1.772-3.873)
by 10.

shown to be associated with outcomes based on CRP level being
a proinflammatory marker.29-31 We did not include this factor
in our analyses because CRP level is not routinely included in
the preoperative assessment of patients with OSCC and there-
fore is not a good candidate to be used on a routine basis in
clinical practice.
Studies23,32-36 that analyzed hemoglobin level as a prog-
nostic factor found that lower levels of hemoglobin are corre-
lated with worse oncologic outcomes. Hemoglobin can be con-
sidered a surrogate marker for a patient’s general nutritional
and performance status, with patients with lower hemoglo-
bin levels having worse outcomes.32 It has also been pro-
posed that decreased levels of hemoglobin lead to reduced oxy-
gen concentrations that will facilitate a hypoxic environment
that promotes tumor progression by increasing tumor cell re-
sistance to therapy and promoting distant metastases.23,33-36
Many studies3,32,33,37,38 that have analyzed hemoglobin level
in different tumor models have found that hemoglobin level
is independently associated with outcomes. Only a few
studies23,27 did not corroborate these results.
Besides inflammation and the immune system status, nu-
tritional status is associated with patient outcomes. Malnu-
trition can influence the intensity of treatment that a patient
is able to receive as well as the host’s immune response to the
tumor. Nutritional status is partially mirrored by albumin level,
and some studies12,39 have found that lower albumin levels,
corresponding to malnutrition, are correlated with worse out-
comes. Moreover, a low BMI can also act as a surrogate marker
of malnutrition. In our study, patients with the lowest BMI had
worse outcomes. The cutoff point in our study (≤19.3) was close
to the World Health Organization BMI underweight range
(<18.5); our analysis revealed that only severe malnourish-
ment was associated with prognosis.
Bi et al37 have shown that underweight patients had the
worst prognosis, and Peter et al3 have shown that low BMI is
not significantly associated with prognosis. Another study13
reported that the other extreme of BMI measure, obesity, is
also correlated with worse outcomes. These contradictory
results may be seen because the analyses using BMI gener-
ally compare obesity with healthy weight only, not includ-

(Reprinted) JAMA Otolaryngology–Head & Neck Surgery Published online June 11, 2020 E5

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Uppsala University User on 06/13/2020

 Research Original Investigation Host Factors Independently Associated With Prognosis in Patients With Oral Cavity Cancer

were even larger when only analyzing T1/T2 tumors. There-

Figure 1. Overall Survival Curves
fore, excluding malnourished patients, BMI analysis showed
A Hemoglobin that patients with obesity had worse outcomes than
100 patients with healthy weight. It is possible that patients
with obesity have a worse disease-specific outcome because
80 a higher amount of fat tissue can act as a reservoir of inflam-
Overall survival, % 60
matory cells, such as macrophages, which when recruited to
the tumor transform in protumoral cells, facilitating tumor
Hemoglobin
40
≥14.1 g/dL
progression, angiogenesis, and metastases.40-42
20
12.9-14.0 g/dL The BMI correlation with survival is not linear; lower BMI
≤12.8 g/dL
is associated with worse outcomes because it is a surrogate of
0 a poor performance status, and higher BMI is also associated
0 12 24 36 48 60
Time, mo with worse outcomes not only because of a higher proinflam-
No. at risk matory status based on a higher volume of fat tissue but also
Hemoglobin ≥14.1 g/dL 563 498 427 359 303 239
Hemoglobin 12.9-14.0 g/dL 429 359 304 251 209 153 because of associated comorbidities in patients with high BMI.
Hemoglobin ≤12.8 g/dL 317 242 193 157 114 86 We believe that both extremes of BMI harbor an adverse host
environment that allows tumors to progress, but this simple
B Albumin measure is probably not strong enough to emerge as an inde-
100 pendent prognostic factor in all patients.
Albumin ≥4.3 g/dL The ALI was one of the first attempts to combine differ-
80
Overall survival, %

ent host factors in 1 index by analyzing the combination of al-

60 bumin level, BMI, and neutrophil-lymphocyte ratio.14 Albu-
Albumin 3.7-4.2 g/dL min level and BMI represent the nutritional status of the
40
patient. The neutrophil-lymphocyte ratio represents the bal-
20 ance between inflammation and immune system response.
Albumin ≤3.6 g/dL
In our study, ALI showed only significant and ordered differ-
0
0 12 24 36 48 60 ences in OS when comparing the highest and lowest catego-
Time, mo ries obtained in this study but showed a contrary to expected
No. at risk
Albumin ≥4.3 g/dL 912 789 675 567 471 369 trend in survival within the 2 intermediate categories. We also
Albumin 3.7-4.2 g/dL 366 293 239 191 149 106 tested the cutoff reported in the article published by Jafri et al14
Albumin ≤3.6 g/dL 31 17 10 9 6 3
and observed that it stratified patients better than the cutoffs
found in our study (eFigure 4 in the Supplement). However,
C BMI
the balance in the number of patients included in each group
100
was limited in the study by Jafri et al,14 with only 8% of pa-
80 tients in the lower category.
Overall survival, %

BMI >19.3
Other attempts have been made to combine host nutri-
60
tional and immune status and inflammation into 1 index. Chen
40 BMI ≤19.3 et al15 created an index combining hemoglobin level, albu-
min level, lymphocyte count, and platelet count (HALP). We
20
believe that this index can be improved by the incorporation
0 of neutrophil and monocyte counts, which had strong asso-
0 12 24 36 48 60
ciations with outcomes in our study, and by excluding plate-
Time, mo
No. at risk let count, which was not associated with outcomes in our study.
BMI >19.3 1240 1049 887 738 603 460 By creating the H-index, we were able to include all the in-
BMI ≤19.3 69 50 37 29 23 18
dependent host-related prognostic factors identified in our
The group stratification was based on the categories obtained with the studies. Because the 3 individual factors analyzed in the pre-
recursive-partitioning analysis. BMI is calculated as weight in kilograms divided vious study (neutrophil, monocyte, and lymphocyte count)
by height in meters squared. To convert hemoglobin and albumin to grams per were independently associated with prognosis, we included
liter, multiply by 10.
all 3 as components of the H-index jointly with the indepen-
dent host factors found in this study (hemoglobin and albu-
ing the subset of patients with cachexia, and are mostly min levels). To create an index that inversely correlates with
restricted to early-stage tumors. We analyzed BMI as a outcomes, we included in the numerator the variables with a
potential variable associated with outcomes in our study negative correlation with survival (neutrophil and monocyte
cohort comparing healthy weight (BMI, 18.5-24.9) with counts) and in the denominator the variables with a positive
overweight (BMI, 25.0-29.9) and obese (BMI, ≥30.0), follow- correlation (lymphocyte count, hemoglobin level, and albu-
ing the universal World Health Organization BMI ranges. No min level). The H-index stratified patients better than the ALI,
differences were found in OS, but when analyzing DSS, with an improved balance in the number of patients included
patients with obesity had worse outcomes. The differences in each category.

E6 JAMA Otolaryngology–Head & Neck Surgery Published online June 11, 2020 (Reprinted) jamaotolaryngology.com

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Uppsala University User on 06/13/2020

 Host Factors Independently Associated With Prognosis in Patients With Oral Cavity Cancer Original Investigation Research

Some variables included in the H-index, such as leuko-

Table 3. Multivariable Analysis Including the H-Index as an Independent
cyte count, may be associated with other clinicopathologic Variable Associated With Overall Survival
characteristics.2,16 To analyze whether the remaining vari-
HR (95% CI)
ables of the H-index are also correlated with patients’ clinico-
Variable Univariable analysis Multivariable analysis
pathologic characteristics, we reported the mean number of
Age, y
each of the host variables included in this study according to
≤60 1 [Reference] 1 [Reference]
patients’ clinicopathologic characteristics (eTable 2 in the
>60 1.955 (1.631-2.343) 1.807 (1.481-2.203)
Supplement). We observed a correlation between different vari-
ables, such as age, smoking status, or cancer stage, and the dif- Tobacco use

ferent host factors analyzed. Nevertheless, these host factors Never 1 [Reference] 1 [Reference]

maintained independent associations with outcomes when Ever 1.305 (1.091-1.561) 1.033 (0.850-1.255)
controlling for clinicopathologic characteristics in the multi- WUHNCIa
variable analyses. 0 1 [Reference] 1 [Reference]
Furthermore, to elucidate whether the H-index goes ≥1 1.624 (1.363-1.936) 1.273 (1.050-1.544)
beyond tracking comorbidities, subanalyses based on the Lymphovascular invasion
presence or absence of comorbidities are shown in eFigure 5 Absent 1 [Reference] 1 [Reference]
in the Supplement. We categorized patients based on the Present 2.100 (1.690-2.611) 1.181 (0.922-1.512)
WUHNCI. Patients with a score of 0 were considered as the Perineural invasion
group without comorbidities, and patients with a score of 1 Absent 1 [Reference] 1 [Reference]
or higher were considered as the group with comorbidities. Present 2.209 (1.855-2.630) 1.374 (1.111-1.699)
The H-index accurately stratified patients in terms of OS Margin status
and DSS both for the subgroup of patients without comor- Negative 1 [Reference] 1 [Reference]
bidities and for the subgroup of patients with comorbidities. Close 1.544 (1.256-1.900) 1.262 (1.000-1.591)
This observation suggests that the H-index accounts not
Positive 3.051 (2.343-3.973) 1.949 (1.426-2.663)
only for comorbidities and performance status but also for
Histologic grade
other interactions of the host against the tumor. We believe
Differentiated
that the H-index is a more comprehensive evaluation of the
Well 1 [Reference] 1 [Reference]
patient status, improving previous classifications that did
Moderately 1.548 (1.204-1.990) 1.084 (0.808-1.454)
not account for all the independent host-related prognostic
Poorly 2.179 (1.619-2.931) 1.035 (0.721-1.486)
factors. Even though statistically the prognostic capacity of
pTb
the H-index might be slightly better than when analyzing
pT1 1 [Reference] 1 [Reference]
leukocyte count alone, by introducing nutritional status
based on albumin levels and hypoxic status based on hemo- pT2 1.657 (1.274-2.153) 1.197 (0.897-1.596)

globin levels, clinically meaningful factors were added in pT3 2.624 (2.024-3.401) 1.309 (0.958-1.789)

the holistic assessment of patients with OSCC. pT4 4.136 (3.239-5.281) 1.582 (1.161-2.157)
Several studies43-46 that supported the inclusion of host fac- pNb
tors in the assessment of patients with cancer have been pub- pN0 1 [Reference] 1 [Reference]
lished. A systematic effort is needed to identify relevant host fac- pN1 1.487 (1.109-1.993) 1.264 (0.921-1.736)
tors that should be included in prediction models for OSCC. In pN2 2.200 (1.699-2.850) 1.594 (1.200-2.118)
addition, there is the issue of geographic heterogeneity because pN3 5.681 (4.565-7.069) 2.966 (2.255-3.901)
it has been previously proposed that host-related variables, such H-index
as CRP level and neutrophil-lymphocyte ratio, differ among ≤1.4 1 [Reference] 1 [Reference]
populations.5,45,47 A previous study16 reported a cohort from Eu- 1.5-3.5 1.474 (1.208-1.798) 1.262 (1.016-1.567)
rope that had different median values in terms of leukocyte ≥3.6 3.221 (2.557-4.058) 1.988 (1.521-2.598)
counts compared with the cohort in the present study. A multi-
Abbreviations: H-index, host index; HR, hazard ratio; WUHNCI, Washington
institutional study that incorporated cohorts from different con- University Head and Neck Comorbidity Index.
tinents would be the next step to account for geographic varia- a
The WUHNCI scores range from 0 to 15, with higher scores indicating worse
tions. This study would allow the development of a more repro- prognosis because of comorbidities.18
ducible and generalizable model for prediction of outcomes in b
Based on the American Joint Committee on Cancer, 8th edition.1
OSCC. Every patient undergoing surgery for OSCC in any setting
around the world will have a presurgical blood test that measures Limitations
leukocyte count, hemoglobin level, and albumin level, and BMI This study has inherent limitations because of its retrospec-
measurement is also universally available. Therefore, we iden- tive nature. Moreover, values were only analyzed at a single
tified a set of powerful and universally available host-related fac- time point before initial treatment. The association of non–
tors associated with outcomes that can be incorporated into a sta- cancer-related conditions (eg, infection and the treatment and
tistical tool, such as a nomogram, to assess risk among individual its complications) with host factors, such as peripheral blood
patients but also allow worldwide use in future iterations of the leukocyte count, hemoglobin level, albumin level, and body
staging system for OSCC.46 weight, could not be analyzed and accounted for in our study.

jamaotolaryngology.com (Reprinted) JAMA Otolaryngology–Head & Neck Surgery Published online June 11, 2020 E7

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Uppsala University User on 06/13/2020

 Research Original Investigation Host Factors Independently Associated With Prognosis in Patients With Oral Cavity Cancer

Figure 2. Survival Curves for the Host Index (H-Index)

A Overall survival B Disease-specific survival

100 100
H-index ≤1.4

Disease-specific survival, %
80 H-index ≤1.4 80
Overall survival, %
H-index 1.5-3.5

60 H-index 1.5-3.5 60 H-index ≥3.6

40 40
H-index ≥3.6
20 20

0 0
0 12 24 36 48 60 0 12 24 36 48 60
Time, mo Time, mo
No. at risk No. at risk
H-index ≤1.4 505 443 390 334 280 226 H-index ≤1.4 505 425 366 309 253 201
H-index 1.5-3.5 614 516 429 352 285 210 H-index 1.5-3.5 614 482 399 318 256 182
H-index ≥3.6 190 140 105 81 61 42 H-index ≥3.6 190 126 90 63 45 32

The group stratification was based on the categories obtained with the recursive-partitioning analysis.

are independently associated with prognosis in patients

Conclusions with OSCC. The interactions between these host factors
were incorporated into a novel H-index that quantified the
The findings suggest that pretreatment values of neutro- prognostic capacity of host characteristics associated
phils, monocytes, lymphocytes, hemoglobin, and albumin with OSCC.

ARTICLE INFORMATION
Accepted for Publication: April 13, 2020.
Published Online: June 11, 2020.
doi:10.1001/jamaoto.2020.1019
Author Contributions: Drs Valero and Patel had full
access to all the data in the study and take
responsibility for the integrity of the data and the
accuracy of the data analysis.
Concept and design: Valero, Shah, Patel.
Acquisition, analysis, or interpretation of data:
Valero, Zanoni, Pillai, Ganly, Morris, Wong, Patel.
Drafting of the manuscript: Valero, Patel.
Critical revision of the manuscript for important
intellectual content: All authors.
Statistical analysis: Valero, Morris, Patel.
Obtained funding: Patel.
Administrative, technical, or material support:
Valero, Zanoni.
Supervision: Ganly, Shah, Patel.
Conflict of Interest Disclosures: Dr Valero
reported receiving a grant from Fundación Alfonso
Martín Escudero during the conduct of the study. Dr
Morris reported receiving grants from Illumina Inc
and AstraZeneca outside the submitted work. Dr
Patel reported having a patent (Methods of Cancer
Detection Using PARPI-FL) pending, holding equity
in Summit Biomedical Imaging, having a patent
(Apparatus, System and Method for Providing Laser
Steering and Focusing for Incision, Excision and
Ablation of Tissue in Minimally-Invasive Surgery),
holding equity in ColdSteel Laser Inc, having a
patent for systems, methods, and apparatus for
multichannel imaging of fluorescent sources in real
time, having a patent for cyclic peptides with
enhanced nerve-binding selectivity, nanoparticles
bound with said cyclic peptides, and use of same
for real-time in vivo nerve tissue imaging, and
having a patent for imaging systems and methods
for tissue differentiation. No other disclosures were
reported.
Funding/Support: This study was funded by
Fundación Alfonso Martín Escudero and Cancer
Center Support Grant P30 CA008748 from the
National Cancer Institute, National Institutes of
Health.
Role of the Funder/Sponsor: The funding sources
had no role in the design and conduct of the study;
collection, management, analysis, and
interpretation of the data; preparation, review, or
approval of the manuscript; and decision to submit
the manuscript for publication.
REFERENCES
1. Amin MB, Edge SB, Greene FL, et al, eds. AJCC
Cancer Staging Manual. 8th ed. Springer; 2017.
doi:10.1007/978-3-319-40618-3
2. Valero C, Pardo L, López M, et al. Pretreatment
count of peripheral neutrophils, monocytes, and
lymphocytes as independent prognostic factor in
patients with head and neck cancer. Head Neck.
2017;39(2):219-226. doi:10.1002/hed.24561
3. Peter F, Wittekindt C, Finkensieper M, Kiehntopf
M, Guntinas-Lichius O. Prognostic impact of
pretherapeutic laboratory values in head and neck
cancer patients. J Cancer Res Clin Oncol. 2013;139
(1):171-178. doi:10.1007/s00432-012-1320-1
4. Holub K, Biete A. Impact of systemic
inflammation biomarkers on the survival outcomes
of cervical cancer patients. Clin Transl Oncol. 2019;
21(7):836-844. doi:10.1007/s12094-018-1991-4
5. Dolan RD, Lim J, McSorley ST, Horgan PG,
McMillan DC. The role of the systemic inflammatory
response in predicting outcomes in patients with
operable cancer: systematic review and
meta-analysis. Sci Rep. 2017;7(1):16717. doi:10.1038/
s41598-017-16955-5
6. Luvián-Morales J, González-Trejo S, Carrillo JF,
et al. Association of the prognostic nutritional index
and overall survival in patients with colorectal
cancer: a STROBE compliant retrospective cohort
study. Cancer Med. 2019;8(7):3379-3388.
doi:10.1002/cam4.2212
7. Yu W, Dou Y, Wang K, et al. Preoperative
neutrophil lymphocyte ratio but not platelet
lymphocyte ratio predicts survival and early relapse
in patients with oral, pharyngeal, and lip cancer.
Head Neck. 2019;41(5):1468-1474. doi:10.1002/
hed.25580
8. Tham T, Bardash Y, Herman SW, Costantino PD.
Neutrophil-to-lymphocyte ratio as a prognostic
indicator in head and neck cancer: a systematic
review and meta-analysis. Head Neck. 2018;40(11):
2546-2557. doi:10.1002/hed.25324
9. Chen Y, Jiang W, Xi D, et al. Development and
validation of nomogram based on SIRI for
predicting the clinical outcome in patients with
nasopharyngeal carcinomas. J Investig Med. 2019;
67(3):691-698. doi:10.1136/jim-2018-000801
10. Templeton AJ, McNamara MG, Šeruga B, et al.
Prognostic role of neutrophil-to-lymphocyte ratio in
solid tumors: a systematic review and
meta-analysis. J Natl Cancer Inst. 2014;106(6):dju124.
doi:10.1093/jnci/dju124
11. Pardo L, Valero C, López M, et al. The prognostic
value of pretreatment platelet count in patients
with head and neck squamous cell carcinoma. Auris
Nasus Larynx. 2017;44(3):313-318. doi:10.1016/j.anl.
2016.06.009
12. Ayhan A, Günakan E, Alyazıcı İ, Haberal N,
Altundağ Ö, Dursun P. The preoperative albumin
level is an independent prognostic factor for
optimally debulked epithelial ovarian cancer. Arch
Gynecol Obstet. 2017;296(5):989-995.
doi:10.1007/s00404-017-4511-9
13. Iyengar NM, Kochhar A, Morris PG, et al. Impact
of obesity on the survival of patients with
early-stage squamous cell carcinoma of the oral

E8 JAMA Otolaryngology–Head & Neck Surgery Published online June 11, 2020 (Reprinted) jamaotolaryngology.com

© 2020 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Uppsala University User on 06/13/2020

 Host Factors Independently Associated With Prognosis in Patients With Oral Cavity Cancer
tongue. Cancer. 2014;120(7):983-991. doi:10.1002/
cncr.28532
14. Jafri SH, Shi R, Mills G. Advance lung cancer
inflammation index (ALI) at diagnosis is a
prognostic marker in patients with metastatic
non-small cell lung cancer (NSCLC): a retrospective
review. BMC Cancer. 2013;13:158. doi:10.1186/1471-
2407-13-158
15. Chen XL, Xue L, Wang W, et al. Prognostic
significance of the combination of preoperative
hemoglobin, albumin, lymphocyte and platelet in
patients with gastric carcinoma: a retrospective
cohort study. Oncotarget. 2015;6(38):41370-41382.
doi:10.18632/oncotarget.5629
16. Valero C, Zanoni DK, McGill MR, et al.
Pretreatment peripheral blood leukocytes are
independent predictors of survival in oral cavity
cancer. Cancer. 2020;126(5):994-1003.
doi:10.1002/cncr.32591
17. van Puten W. CART: Stata module to perform
classification and regression tree analysis. Stat
Softw Components. November 2006;S456776.
18. Piccirillo JF, Lacy PD, Basu A, Spitznagel EL.
Development of a new head and neck
cancer-specific comorbidity index. Arch Otolaryngol
Head Neck Surg. 2002;128(10):1172-1179.
doi:10.1001/archotol.128.10.1172
19. Cihan YB, Arslan A, Cetindag MF, Mutlu H. Lack
of prognostic value of blood parameters in patients
receiving adjuvant radiotherapy for breast cancer.
Asian Pac J Cancer Prev. 2014;15(10):4225-4231.
doi:10.7314/APJCP.2014.15.10.4225
20. Holub K, Biete A. New pre-treatment
eosinophil-related ratios as prognostic biomarkers
for survival outcomes in endometrial cancer. BMC
Cancer. 2018;18(1):1280. doi:10.1186/s12885-018-
5131-x
21. Grimm M, Rieth J, Hoefert S, et al. Standardized
pretreatment inflammatory laboratory markers and
calculated ratios in patients with oral squamous cell
carcinoma. Eur Arch Otorhinolaryngol. 2016;273
(10):3371-3384. doi:10.1007/s00405-016-3950-4
22. Wei Y, Zhang X, Wang G, et al. The impacts of
pretreatment circulating eosinophils and basophils
on prognosis of stage I-III colorectal cancer. Asia Pac
J Clin Oncol. 2018;14(5):e243-e251. doi:10.1111/
ajco.12871
23. Li Z, Xu Z, Huang Y, et al. Prognostic values of
preoperative platelet-to-lymphocyte ratio, albumin
and hemoglobin in patients with non-metastatic
colon cancer. Cancer Manag Res. 2019;11:3265-3274.
doi:10.2147/CMAR.S191432
24. Tangthongkum M, Tiyanuchit S, Kirtsreesakul
V, Supanimitjaroenporn P, Sinkitjaroenchai W.
Platelet to lymphocyte ratio and red cell
distribution width as prognostic factors for survival
and recurrence in patients with oral cancer. Eur Arch
Otorhinolaryngol. 2017;274(11):3985-3992. doi:10.
1007/s00405-017-4734-1
jamaotolaryngology.com
© 2020 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Uppsala University User on 06/13/2020
25. Chen S, Guo J, Feng C, Ke Z, Chen L, Pan Y. The
preoperative platelet-lymphocyte ratio versus
neutrophil-lymphocyte ratio: which is better as a
prognostic factor in oral squamous cell carcinoma?
Ther Adv Med Oncol. 2016;8(3):160-167. doi:10.1177/
1758834016638019
26. Zhang Y, Zheng L, Quan L, Du L. Prognostic role
of platelet-to-lymphocyte ratio in oral cancer:
a meta-analysis. J Oral Pathol Med. Published online
January 25, 2019. doi:10.1111/jop.12832
27. Sakin A, Sahin S, Yasar N, et al. The relation
between hemogram parameters and survival in
extensive-stage small cell lung cancer. Oncol Res
Treat. 2019;42(10):506-515. doi:10.1159/000501595
28. Solak Mekić M, Pedišić I, Šobat H, et al. The role
of complete blood count parameters in patients
with colorectal cancer. Acta Clin Croat. 2018;57(4):
624-629. doi:10.20471/acc.2018.57.04.03
29. Wang Q, Song X, Zhao Y, et al. Preoperative
high c-reactive protein/albumin ratio is a poor
prognostic factor of oral squamous cell carcinoma.
Future Oncol. 2019;15(19):2277-2286.
doi:10.2217/fon-2019-0063
30. DE Paz D, Young CK, Chien HT, et al. Prognostic
roles of SCC antigen, CRP and CYFRA 21-1 in oral
cavity squamous cell carcinoma. Anticancer Res.
2019;39(4):2025-2033. doi:10.21873/anticanres.13313
31. Hanai N, Sawabe M, Kimura T, et al. The
high-sensitivity modified Glasgow prognostic score
is superior to the modified Glasgow prognostic
score as a prognostic predictor for head and neck
cancer. Oncotarget. 2018;9(97):37008-37016. doi:
10.18632/oncotarget.26438
32. Jazieh AR, Hussain M, Howington JA, et al.
Prognostic factors in patients with surgically
resected stages I and II non-small cell lung cancer.
Ann Thorac Surg. 2000;70(4):1168-1171. doi:10.
1016/S0003-4975(00)01529-0
33. Oblak I, Strojan P, Zakotnik B, Budihna M, Šmid
L. Hemoglobin as a factor influencing the outcome
in inoperable oropharyngeal carcinoma treated by
concomitant radiochemotherapy. Neoplasma.
2003;50(6):452-458.
34. Vaupel P, Mayer A, Höckel M. Impact of
hemoglobin levels on tumor oxygenation: the
higher, the better? Strahlenther Onkol. 2006;182
(2):63-71. doi:10.1007/s00066-006-1543-7
35. Franco P, Montagnani F, Arcadipane F, et al. The
prognostic role of hemoglobin levels in patients
undergoing concurrent chemo-radiation for anal
cancer. Radiat Oncol. 2018;13(1):83. doi:10.1186/
s13014-018-1035-9
36. Nordsmark M, Bentzen SM, Rudat V, et al.
Prognostic value of tumor oxygenation in 397 head
and neck tumors after primary radiation therapy: an
international multi-center study. Radiother Oncol.
2005;77(1):18-24. doi:10.1016/j.radonc.2005.06.038
37. Bi H, Huang Y, Wang G, Ma L, Lu M. Impact of
body mass index and pretreatment hemoglobin
(Reprinted) JAMA Otolaryngology–Head & Neck Surgery Published online June 11, 2020
Original Investigation Research
level on prognosis following radical cystectomy for
bladder cancer in males and females. Urol Int.
2020;104(1-2):28-35. doi:10.1159/000500561
38. Schäfer U, Micke O, Müller SB, Schüller P,
Willich N. Hemoglobin as an independent
prognostic factor in the radiotherapy of head and
neck tumors. Strahlenther Onkol. 2003;179(8):527-
534. doi:10.1007/s00066-003-1117-x
39. Toiyama Y, Yasuda H, Ohi M, et al. Clinical
impact of preoperative albumin to globulin ratio in
gastric cancer patients with curative intent. Am J Surg.
2017;213(1):120-126. doi:10.1016/j.amjsurg.
2016.05.012
40. Wu P, Wu D, Zhao L, et al. Inverse role of
distinct subsets and distribution of macrophage in
lung cancer prognosis: a meta-analysis. Oncotarget.
2016;7(26):40451-40460. doi:10.18632/
oncotarget.9625
41. Yin S, Huang J, Li Z, et al. The prognostic and
clinicopathological significance of tumor-associated
macrophages in patients with gastric cancer:
a meta-analysis. PLoS One. 2017;12(1):e0170042.
doi:10.1371/journal.pone.0170042
42. Zhao X, Qu J, Sun Y, et al. Prognostic
significance of tumor-associated macrophages in
breast cancer: a meta-analysis of the literature.
Oncotarget. 2017;8(18):30576-30586. doi:10.18632/
oncotarget.15736
43. Kao HK, Löfstrand J, Loh CYY, et al. Nomogram
based on albumin and neutrophil-to-lymphocyte
ratio for predicting the prognosis of patients with
oral cavity squamous cell carcinoma. Sci Rep. 2018;
8(1):13081. doi:10.1038/s41598-018-31498-z
44. Mattavelli D, Lombardi D, Missale F, et al.
Prognostic nomograms in oral squamous cell
carcinoma: the negative impact of low neutrophil to
lymphocyte ratio. Front Oncol. 2019;9:339.
doi:10.3389/fonc.2019.00339
45. Park JH, Ishizuka M, McSorley ST, et al. Staging
the tumor and staging the host: a two centre, two
country comparison of systemic inflammatory
responses of patients undergoing resection of
primary operable colorectal cancer. Am J Surg.
2018;216(3):458-464. doi:10.1016/j.amjsurg.
2017.08.044
46. Patel SG, Lydiatt WM. Staging of head and neck
cancers: is it time to change the balance between
the ideal and the practical? J Surg Oncol. 2008;97
(8):653-657. doi:10.1002/jso.21021
47. Azab B, Camacho-Rivera M, Taioli E. Average
values and racial differences of neutrophil
lymphocyte ratio among a nationally representative
sample of United States subjects. PLoS One. 2014;9
(11):e112361. doi:10.1371/journal.pone.0112361
E9