Patients with head and neck cancer (HNC) have a high risk of sarcopenia, which is associated with poor prognosis. Skeletal-muscle area and index at the third lumbar (L3) vertebra level (L3MA and L3MI) are recommended for the detection of sarcopenia. However, L3 level is not included in many imaging protocols and there are no data for optimal levels and cutoffs for the diagnosis of sarcopenia in head and neck computed tomography (HNCT) scans. Our aim was to assess the relationship between cervical paravertebral muscle values and L3MI and to investigate optimal level to diagnose sarcopenia on HNCTs.
Patients with HNC (n=159) who underwent positron emission tomography-CT for tumor staging were retrospectively analyzed. On CT images, paravertebral and sternocleidomastoid muscle areas at second (C2), third (C3), and fourth (C4) cervical vertebrae levels (C2MA, C3MA, C4MA, SCMA) and L3MA were measured. Cross-sectional areas were normalized for stature (muscle area/height square) and muscle index (C2MI, C3MI, C4MI, SCMI, L3MI) values were obtained. Spearman correlation and linear regression analyses were used for assessing correlations. To calculate the diagnostic performance of SCMI, C2MI, C3MI, and C4MI for the diagnosis of sarcopenia with respect to the cutoffs of L3MI, receiver operating characteristic (ROC) analysis was used.
Males had significantly higher muscle areas than females. Although C2MI, C3MI, C4MI, and SCMI values all showed very strong and significant correlation with L3MI (
C2MI, C3MI, C4MI, and SCMI values can be used as alternatives for the diagnosis of sarcopenia in routine HNCT examinations.
Cachexia is a common finding in patients with malignancy, especially in head and neck cancers (HNCs) [
For assessing SMM on CT, skeletal muscle cross-sectional area (SMA) and skeletal muscle index (SMI; height2-adjusted SMA) at the level of the third lumbar vertebra (L3) is prevalently used and highly correlates with whole body SMM [
Our aim was to assess the relationship between SMA of the paravertebral muscles at three separate cervical vertebrae levels (C2, C3, and C4) and at the L3 level, and to investigate optimal level to diagnose sarcopenia on head and neck-only CT examinations. We also aimed to define means, and sarcopenia cutoff values for both sexes, on head and neck CT examinations at the cervical vertebrae levels mentioned above.
This study was approved by Ethics Committee of the Pamukkale University Faculty of Medicine (No. 60116787-020/71492) and a waiver of the requirement for informed consent was granted.
The medical records of 197 patients (63.3±11.9 years) with HNC who underwent PET-CT for tumor staging at our tertiary medical center between January 2012 and May 2018 were retrospectively analyzed. Patients with unenhanced CT examination of the entire body were included. Exclusion criteria were missing or incomplete CT images and insufficient CT image quality. In the staging of all patients, the eighth edition of TNM Classification for Head and Neck Cancer by the American Joint Committee on Cancer was used. We used performance status (PS) scaling defined by the Eastern Cooperative Oncology Group (ECOG) to assess the patients’ PS [
All CT examinations of the whole body were performed in a craniocaudal direction, using a multi-detector CT scanner (16-detector row, Brilliance; Philips Healthcare, Best, the Netherlands). CT protocol parameters were tube current of 50–120 mA, tube voltage of 90–140 kV, matrix of 512×512, rotation time of 0.75 second, pitch of 1 and slice thickness of 5 mm. The tube voltage and tube current were adjusted to the patient’s body weight. Contrast medium was not used before or during CT exams.
Before the quantitative tissue assessment, all CT images were fully anonymized and assessed for the presence of artifacts or posterior paravertebral mass or lymphadenopathy which may hinder SMA measurement by a board-certified radiologist (FU) who was unaware of any patient information. Osirix software was used (v9.0; Pixmeo SARL, Bernex, Switzerland) for quantitative SMA measurements. Skeletal muscle was identified and quantified by use of Hounsfield unit thresholds (–29 to +150). All quantitative assessments were made in two different sessions to avoid recall by a single trained radiologist (FU).
In the first session, the second, third, and fourth cervical vertebrae levels (C2, C3, and C4) were chosen to measure paravertebral muscle areas. Paravertebral skeletal muscle areas were measured from the levels of transverse processes of C2, C3, and C4 vertebrae (
A second session was held 1 week after the first session and, at the lower part of the third lumbar vertebrae (L3) level sum of the all skeletal muscle area was measured which was described by Prado et al. (
Cross-sectional areas (cm2) were computed for each image, and these values were normalized for stature (height in m2) (C2, C3, C4, and L3 SMI, SCM muscle index; cm2/m2). Body mass index (BMI) was also computed as weight divided by height squared (kg/m2). For detection of low SMM, existing cutoff values of L3MI of ≤52.4 cm2/m2 in males and ≤38.9 cm2/m2 in females were used, as previously recommended [
IBM SPSS ver. 21.0 (IBM Corp., Armonk, NY, USA) was used for analyses. Continuous variables are represented as mean with standard deviation or median with range, respectively in normally and non-normally distributed data. Categorical variables are represented as percentage. Mann-Whitney
Of the 197 patients with HNC who underwent CT examination, 38 were excluded (27 for missing or incomplete CT examination; 11 for poor image quality due to artifacts). The total study population consisted of 159 subjects (age, 62.2±12.1 years; range, 19 to 86 years), of whom 85 (53.4%) were male. Of the 159 patients with HNC, seven (4.4%) had stage 0 disease and 69 (43.4%) had stage 1 disease (
Males had significantly higher C2MA, C2MI, C3MA, C3MI, C4MA, C4MI, SCMA, L3MA, and L3MI values than females. Although, SCMI values in males were greater than those of females, no statistical significance was observed (
When previously described, L3MI cutoff values [
There was a moderate negative correlation between all of the measured skeletal muscle areas and indexes with PS (
In the multivariable regression analysis using covariables (age, weight, and BMI), C2MI, C3MI, C4MI, and SCMI were the significant predictor of L3MI (
Our study demonstrated that C2MI, C3MI, C4MI, and SCMI values showed very strong and significant correlation with L3MI, and all these values can be used as alternatives for the diagnosis of sarcopenia in routine head and neck CT examinations. Though SCM muscle may be affected in patients with HNC, SCMI measurement can be used as a simple and successful method for the detection of sarcopenia in both males and females when not affected.
Cachexia and sarcopenia are common in patients with HNC [
Patients with sarcopenia are prone to higher toxicity of chemotherapy, longer postoperative recovery, increased infection rate, impaired health related quality of life, higher risk of falling, increased health care costs and increased mortality [
There may also be a negative correlation between SMM and stage of the malignancy (HNC), as larger masses have a higher metabolism [
Assessment of SCMA on CT images in patients with HNC can be impaired by the tumor infiltration or lymphadenopathies. Most of the HNC patients have lymphadenopathy at the initial diagnosis [
In conclusion, assessing C2MI, C3MI, C4MI, and SCMI on head and neck CT appears to be good alternatives to abdominal CT scans, based on the very strong correlation with L3MI. Regardless of gender differences, SCMI is the best and most easily applicable alternative method for L3MI in the detection of sarcopenia. We also established the sex-specific cutoff values for those levels which will allow researchers to investigate the clinical importance and impact of sarcopenia in HNC patients, without additional costs or radiation exposure.
• Sarcopenia is associated with poor prognosis.
• Patients with head and neck cancer have a high risk of sarcopenia.
• Assessing muscle parameters on head and neck computed tomography is a successful tool to detect sarcopenia.
No potential conflict of interest relevant to this article was reported.
Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Visualization & Writing - original draft: FU. Writing - review & editing: DH, DY.
The authors acknowledge Dr. Hande Şenol (Department of Biostatistics, Pamukkale University, Denizli, Turkey) for her help with statistical analyzes for this study.
Paravertebral and sternocleidomastoid muscle areas measurements. The paravertebral muscle area measurements at second (C2, A), third (C3, B), and fourth (C4, C) cervical vertebrae levels. The sternocleidomastoid muscle area measurements at C2 (D), C3 (E), and C4 (F) levels.
Example of skeletal muscle area measurement at the lower part of the third lumbar vertebrae (L3) level.
Receiver operation characteristic curve for prediction of sarcopenia for the assessment of C2MI, C3MI, C4MI, and SCMI in the study total population (males and females). C2MI–C4MI, normalized paravertebral skeletal muscle areas at the level of C2–C4 vertebrae values for stature; SCMI, normalized sum of the bilateral sternocleidomastoid muscle areas at the levels of C2, C3, and C4 vertebrae values for stature.
Receiver operation characteristic curve for prediction of sarcopenia for the assessment of C3MI in males (A) and C4MI in females (B). AUC, area under the curve; CI, confidence interval; C2MI–C4MI, normalized paravertebral skeletal muscle areas at the level of C2–C4 vertebrae values for stature; SCMI, normalized sum of the bilateral sternocleidomastoid muscle areas at the levels of C2, C3, and C4 vertebrae values for stature.
Bland-Altman plots show the agreement of the L3MI and cervical paravertebral muscle indexes (C2MI, C3MI, C4MI, and SCMI) measurements. C2MI–C4MI, normalized paravertebral skeletal muscle areas at the levels of second (C2), third (C3), and fourth (C4) cervical vertebrae values for stature; SCMI, normalized sum of the bilateral sternocleidomastoid muscle areas at the levels of C2, C3, and C4 vertebrae values for stature; L3MI, normalized sum of the all skeletal muscle area at the lower part of L3 level value for stature.
Performance status scale [
Performance status | Definition |
---|---|
0 | Fully active; no performance restrictions |
1 | Strenuous physical activity restricted; fully ambulatory and able to carry out light work. |
2 | Capable of all self-care but unable to carry out any work activities. Up and about >50% of waking hours. |
3 | Capable of only limited self-care; confined to bed or chair >50% of waking hours. |
4 | Completely disabled; cannot carry out any self-care; totally confined to bed or chair. |
Detailed characteristics of the patients
Variable | All subjects | Male | Female | |
---|---|---|---|---|
No. of subjects | 159 | 85 | 74 | - |
Age (yr) | 62.2±12.1 (19–86) | 62.2±10.1 (37–83) | 62.3±14.2 (19–86) | 0.661 |
Weight (kg) | 72±11.1 (39–99) | 71.9±15.8 (45–109) | 67.4±13.3 (38–98) | 0.051 |
Height (cm) | 165.3±8.3 (148–190) | 170.1±6.8 (154–190) | 159.9±7.5 (148–175) | <0.001 |
Body mass index (kg/m2) | 25.6±5.7 (15.9–44.8) | 24.9±5.7 (15.9–44.8) | 26.4±5.5 (16.7–41.4) | 0.091 |
Stage | 0.004 | |||
0 | 7 (4.4) | 1 (1.2) | 6 (8.1) | |
1 | 69 (43.4) | 30 (35.3) | 39 (52.7) | |
2 | 43 (27) | 29 (34.1) | 14 (18.9) | |
3 | 35 (22) | 21 (24.7) | 14 (18.9) | |
4 | 5 (3.2) | 4 (4.7) | 1 (1.4) | |
Performance status | 0.859 | |||
0 | 31 (19.5) | 16 (18.8) | 15 (20.3) | |
1 | 35 (22) | 18 (21.2) | 17 (23) | |
2 | 51 (32.1) | 32 (37.7) | 19 (25.7) | |
3 | 37 (23.3) | 16 (18.8) | 21 (28.3) | |
4 | 5 (3.2) | 3 (3.5) | 2 (2.7) | |
Sarcopenia | 0.021 | |||
Yes | 80 (50.3) | 50 (59.5) | 30 (40) | |
No | 79 (48.7) | 34 (40.5) | 45 (60) | |
Location of HNC | <0.001 | |||
Nasopharynx | 60 (37.7) | 51 (60) | 9 (12.2) | |
Oral cavity | 4 (2.5) | 4 (4.7) | - | |
Hypopharynx | 32 (20.1) | 8 (9.4) | 24 (32.4) | |
Oropharynx | 34 (21.4) | 6 (7.1) | 28 (37.8) | |
Larynx | 29 (18.2) | 16 (18.8) | 13 (17.6) |
Values are presented as mean±standard deviation (range) or number (%).
HNC, head and neck cancer.
Nonparametric test was used.
Prevalence of sarcopenia according to the stage of disease
Stage | Sarcopenia |
||
---|---|---|---|
Negative | Positive | Total | |
0 | 7 (8.9) | 0 | 7 (4.4) |
1 | 50 (63.3) | 19 (23.7) | 69 (43.4) |
2 | 6 (7.6) | 37 (46.3) | 43 (27) |
3 | 16 (20.2) | 19 (23.7) | 35 (22) |
4 | 0 | 5 (6.3) | 5 (3.2) |
Total | 79 (100) | 80 (100) | 159 (100) |
Values are presented as number (%).
Sex-specific measurement results
Variable | Total | Male | Female | |
---|---|---|---|---|
C2MA (cm2) | 23.1±6.6 | 26.2±6.7 | 19.7±4.4 | <0.001 |
C2MI (cm2/m2) | 8.4±2.2 | 9.1±2.4 | 7.7±1.6 | 0.001 |
C3MA (cm2) | 22.7±6.2 | 25.2±6.3 | 19.9±4.8 | <0.001 |
C3MI (cm2/m2) | 8.3±2.1 | 8.8±2.3 | 7.7±1.6 | 0.012 |
C4MA (cm2) | 26.3±7.2 | 28.7±7.3 | 23.5±6 | <0.001 |
C4MI (cm2/m2) | 9.6±2.4 | 9.9±2.6 | 9.2±2.2 | 0.047 |
SCMA (cm2) | 6±1.9 | 6.5±2.1 | 5.4±1.3 | 0.002 |
SCMI (cm2/m2) | 2.2±0.6 | 2.2±0.7 | 2.1±0.5 | 0.548 |
L3MA (cm2) | 122.7±32.7 | 136.2±34.8 | 107.1±21.6 | <0.001 |
L3MI (cm2/m2) | 44.7±10.4 | 47.1±11.9 | 41.8±7.5 | 0.003 |
Values are presented as mean±standard deviation.
C2MA–C4MA, paravertebral skeletal muscle areas at second (C2), third (C3), and fourth (C4) cervical vertebrae levels; C2MI–C4MI, normalized C2MA–C4MA values for stature; SCMA, sum of the bilateral sternocleidomastoid muscle areas at the levels of C2, C3, and C4 vertebrae; SCMI, normalized SCMA values for stature; L3MA, sum of the all skeletal muscle area at the lower part of the third lumbar vertebrae (L3) level; L3MI, normalized L3MA values for stature.
Correlation of variables with L3MI
Variable | ||
---|---|---|
C2MI | 0.810 | <0.001 |
C3MI | 0.877 | <0.001 |
C4MI | 0.827 | <0.001 |
SCMI | 0.801 | <0.001 |
Sex | –0.255 | 0.001 |
Age | –0.191 | 0.016 |
Height | 0.092 | 0.249 |
Weight | 0.383 | <0.001 |
BMI | 0.327 | <0.001 |
Stage of disease | –0.119 | 0.135 |
Location of tumor | –0.109 | 0.172 |
L3MI, normalized sum of the all skeletal muscle areas at the lower part of the third lumbar vertebrae (L3) level value for stature; C2MI–C4MI, normalized paravertebral skeletal muscle areas at the levels of second (C2), third (C3), and fourth (C4) cervical vertebrae values for stature; SCMI, normalized sum of the bilateral sternocleidomastoid muscle areas at the levels of C2, C3, and C4 vertebrae values for stature; BMI, body mass index.
The cutoff values and prediction rule for C2MI, C3MI, C4MI, and SCMI obtained by multivariable analysis and the diagnostic performance of these values in the diagnosis of sarcopenia
Variable | Cutoff value | Prediction rule | Sensitivity (%) | Specificity (%) | |
---|---|---|---|---|---|
Male | |||||
C2MI | 9.3 | L3MI=3.724×C2MI+17.496 | 88.2 | 85.3 | 0.714 |
C3MI | 9.3 | L3MI=4.226×C3MI+12.960 | 94.1 | 91.2 | 0.840 |
C4MI | 10.8 | L3MI=3.416×C4MI+15.623 | 90.2 | 79.4 | 0.718 |
SCMI | 2.8 | L3MI=10.978×SCMI+21.075 | 98.0 | 58.8 | 0.692 |
Female | |||||
C2MI | 8 | L3MI=3.282×C2MI+12.498 | 87.1 | 72.8 | 0.460 |
C3MI | 6.3 | L3MI=3.857×C3MI+14.644 | 48.4 | 90.7 | 0.616 |
C4MI | 9.5 | L3MI=2.755×C4MI+12.730 | 87.0 | 60.5 | 0.649 |
SCMI | 1.5 | L3MI=14.009×SCMI+18.381 | 93.5 | 100.0 | 0.668 |
C2MI–C4MI, normalized paravertebral skeletal muscle areas at the levels of second (C2), third (C3), and fourth (C4) cervical vertebrae values for stature; SCMI, normalized sum of the bilateral sternocleidomastoid muscle areas at the levels of C2, C3, and C4 vertebrae values for stature; L3MI, normalized sum of the all skeletal muscle area at the lower part of the third lumbar vertebrae (L3) level value for stature.