Prognosis Prediction by ¹⁸F-fluorodeoxyglucose Positron Emission Tomography Parameters in Patients Undergoing Chemoradiation Therapy for Esophageal Cancer

Abstract

The aim of this study was to identify ¹⁸F-fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG-PET/CT) parameters that could predict the prognosis of patients with esophageal cancer before and after undergoing chemoradiation therapy. We retrospectively reconstructed images under the same conditions for patients who underwent pre- and post-treatment ¹⁸F-FDG-PET for chemoradiation therapy for esophageal cancer. Correlations between 2-year survival rates and pre-treatment values, differences between pre- and post-treatment quotients, and their ratios were examined for various standardized uptake values (SUV), metabolic tumor volumes (MTV), and each SUVmean (Mean SUV)*MTV (Vol.mean). We enrolled 29 patients who underwent pre-and post-treatment ¹⁸F-FDG-PET. The median overall survival was 21.4 months (range, 3.6–100.9 months). Pre-treatment MTV had the most favorable hazard ratio (HR) for survival. However, the MTV product (Vol.meanQ), SUV corrected for basal metabolic rate using Mifflin-St Jeor estimation (BMR.ms), Vol.mean (SUVmeanQ) using the qPET method, SUVmean, and HR using Vol.meanQ corrected for body weight were nearly equivalent. No significant results were obtained for the pre- and post-treatment quotients. The pre- and post-treatment Vol.meanQ is a useful prognostic parameter that considers the effect of age-related loss of lean body mass. The use of parameters, including metabolism, will facilitate more appropriate use of ¹⁸F-FDG-PET before and after chemoradiation therapy.

Introduction

Most esophageal cancers are regional or distant metastatic cases (locally advanced disease) at the time of initial diagnosis, and multimodal therapy including surgery is the standard treatment.^1,2 We have previously reported favorable long-term outcomes for patients with esophageal cancer.³ Definitive chemoradiotherapy has the merit of esophageal preservation and is one of the standard treatments.^1,2 However, there are known early and late adverse events caused by chemoradiation⁴; therefore, it is useful to predict the therapeutic effect to determine the treatment strategy.

¹⁸F-Fluorodeoxyglucose positron emission tomography/computed tomography (¹⁸F-FDG-PET/CT) has been established as a useful diagnostic modality in oncology because of its excellent tumor detection abilities. It is widely used for detecting distant metastases, unknown primary tumors, and tumor recurrence.⁵ With regard to esophageal cancer, Marzola et al. reported that ¹⁸F-FDG-PET was not suitable for the detection of tumors with very small volumes, although it was superior to CT/endoscopic ultrasound, especially for the detection of distant metastases.⁶

¹⁸F-FDG-PET/CT is widely used for not only tumor detection but also for determination of therapeutic effects because it provides semi-quantitative numerical values.^7,8 In recent years, its utility for determining the prognosis after concurrent chemoradiotherapy (CCRT) in lymph node-positive patients has also been investigated.⁹ Although this modality is widely used to determine therapeutic effects via comparisons before and after treatment,^10,11 the treatment efficacy is based on visual assessment of uptake, and the details of reliable parameters for predicting the patient’s prognosis remain unclear.

Studies on prognosis prediction based on the judgment of therapeutic effects are being conducted, with several studies reporting that the standardized uptake value (SUV) was a useful predictor in oral cancer,^12,13 small cell lung cancer,¹⁴ and lymphoma.¹⁵ However, another study reported that the metabolic tumor volume (MTV) was a more reliable predictor than SUV.¹⁶ Therefore, it is unclear which parameter is more reliable. Although both pre- and post-treatment values appear to be powerful parameters for prognosis prediction, the disadvantage lies in the fact that post-treatment parameters are required and cannot be predicted before treatment.

The purpose of this study was to examine ¹⁸F-FDG-PET/CT parameters before and after treatment and determine the useful parameters for predicting the prognosis before and after chemoradiation for esophageal cancer. In addition, we examined the robustness of SUV and MTV values through detailed examination of changes in their usefulness depending on metabolic correction.

Materials and Methods

Ethical approval

All procedures were in accordance with the ethical standards of the responsible committee at which the studies were conducted (Ethics Committee of Keio University School of Medicine; 31 July 2015; No. 2015/20150137) and were in accordance with the 1964 Declaration of Helsinki and later versions. Informed consent was obtained from all patients.

Patients

We retrospectively reviewed our institutional database of electronic medical records and all available ¹⁸F-FDG-PET/CT imaging studies between 1 August 2012 and 31 December 2019. During this period, 112 patients with esophageal squamous carcinoma received CCRT treatment. We included patients who received CCRT as the first-line treatment and underwent ¹⁸F-FDG-PET/CT before and after CCRT. The first ¹⁸F-FDG-PET/CT (pre-test) examination was performed within 50 days before CCRT initiation, and the second examination was performed 14–150 days after CCRT completion (post-test). Patients with a history of chemotherapy or radiotherapy for other cancers and those with metastatic recurrence were excluded.

¹⁸F-FDG-PET/CT image acquisition

Following confirmation of a fasting blood glucose level below 6 mmol/L (≥6 h of fasting), we intravenously administered ¹⁸F-FDG. Approximately 60–75 min (until 31 August 2016: 60 min; from 1 September 2016: 75 min according to changes in facility policies) after ¹⁸F-FDG uptake, a topogram was obtained in the craniocaudal direction to plan for the scan. This was followed by CT (without contrast enhancement) for anatomical localization, attenuation correction, and diagnostic purposes. The patients were scanned using a Siemens BioGraph 40/64 TruePoint hybrid PET/CT scanner (Siemens Medical Solutions, Malvern PA, USA). Several emission scans were recorded at seven bed positions with the following parameters: duration method, step and shoot (from 1 August 2012 to 31 July 2014, BioGraph 64 used continuous bed motion); scan duration/bed, variable based on body weight (<50 kg: 60 s, 50–70 kg: 90 s, >70 kg: 120 s) until 31 August 2016 and uniformly set to 135 s from 1 September 2016. The collected data were reconstructed using Syngo (Siemens Medical Solutions) with the Iterative+Time-of-Flight method. The processing parameters were as follows: three iterations/21 subsets, Gaussian-Filter, relative accuracy of the scatter correction method, and 256 × 256-pixel image size.

Measurements

We measured various SUV-based parameters such as maximum SUV (SUVmax), mean SUV (SUVmean), peak SUV (SUVpeak),¹⁷ the standard deviation of SUVmean (SUVmean_sd), SUVpeak based on qPET (SUVqpeak),¹⁸ the average of the SUVqpeak (MeanQ), the average of Wahl’s SUVpeak¹⁹ using qPET (MeanW), and MTV on the PET/CT viewer.²⁰ These parameters were based on the distributions of Fiji²¹/ImageJ.²² We manually set volumes of interest (VOIs) and defined targets using the following thresholds:

• 1. Fixed SUV threshold of 2.0 g/mL (SUV2.0, MTV2.0)

• 2. Fixed SUV threshold of 2.5 g/mL (SUV2.5, MTV2.5)

• 3. Fixed SUV threshold of 3.0 g/mL (SUV3.0, MTV3.0)

• 4. Adaptive at 41% of maximum SUV on each lesion (SUV41, MTV41)

We evaluated all SUV thresholds and selected the best representative value. We measured SUVs (maximum, mean, and peak) based on the ¹⁸F-FDG-PET/CT: European Association of Nuclear Medicine procedure guidelines for tumor imaging, version 2.0.¹⁷ SUVqPeak was proposed by Hasenclever et al.¹⁸ for lymphoma cases. Average PET (MeanW) and average Q-PET (MeanQ) were calculated as the average SUVs for each slice including the delimited VOI of SUVpeak¹⁹ and SUVqpeak.¹⁸ Accordingly, we evaluated the following parameters: SUVmax, SUVpeak, SUVqpeak, SUVmean, MeanW, and MeanQ. We tested three patterns of the mean based on SUVpeak in addition to SUVmean for MeanW and MeanQ. The parameters also included MTV and MTV*mean, with three patterns of the mean for MeanW and MeanQ tested based on SUVpeak in addition to SUVmean. MTV*SUVmean represented total lesion glycolysis.

SUV measurement and highest SUVmax detection

We detected the highest SUVmax using a VOI of 1 cm³ for each organ (esophagus, heart, liver, spleen, multifidus muscle, abdominal aorta, colon, and jejunum), because the SUVmax of the contoured organs could not be derived directly on CT. Subsequently, we measured the average and peak values for the organs based on the arranged VOI.

Parameter correction (method)

We corrected SUV values using several parameters,²³ including ideal body weight (IBW),²⁴ predicted normal weight (PNW),²⁵ body surface area (BSA),²⁶ fat-free mass (FFM),²⁷ two types of basal metabolic rate (BMR),^28,29 and two types of lean body mass (LBM).^30,31Table 1 shows the formulae used to calculate each parameter. In addition to Janmahasatian’s estimate (LBM.j)³¹ for SUV corrected by LBM (SUL) in the PET/CT viewer, we calculated LBM using Green and Duffull’s formula (LBM.gd).³⁰

Table 1SUV and metabolic correction functions

Parameter	Method	Sex	Function	Purpose
SUVcorr	-	Both	${\text{SUV}}_{\text{corr}}=\frac{{\text{C}}_{\text{img}}}{{\text{Dose}}_{\text{inj}}÷\text{Correction value (ex}\text{. BW)}}$	For use in each SUV correction formula
BMI	-	Both	$\frac{\text{Bw}\left(\text{kg}\right)}{\text{Ht}{\left(\text{m}\right)}^2}$	For use in each correction formula
BSA	DuBois’ formula	Both	$0.007184\times {\text{Bw}}^{0.425}\left(\text{kg}\right)\times {\text{Ht}}^{0.725}\left(\text{cm}\right)$	Indicator of metabolism used in chemotherapy
IBW	-	Male	$49.9+0.89\times \left(\text{Ht}\left(\text{cm}\right)-152.4\right)$	Index that is not affected by age and weight
		Female	$45.4+0.89\times \left(\text{Ht}\left(\text{cm}\right)-152.4\right)$
PNW	-	Male	$1.57\times \text{Bw}\left(\text{kg}\right)-0.0183\times \text{BMI}\times \text{Bw}\left(\text{kg}\right)-10.5$	LBM related indicators
		Female	$1.75\times \text{Bw}\left(\text{kg}\right)-0.0242\times \text{BMI}\times \text{Bw}\left(\text{kg}\right)-12.6$
FFM	-	Male	$\left(0.285\times \text{Bw}\left(\text{kg}\right)\right)+\left(12.1\times \text{Ht}{\left(\text{m}\right)}^2\right)$	LBM related indicators
		Female	$\left(0.287\times \text{Bw}\left(\text{kg}\right)\right)+\left(9.74\times \text{Ht}{\left(\text{m}\right)}^2\right)$
LBM	Janmahasatian’s estimate (LBM.j)	Male	$\frac{9270\times \text{Bw}\left(\text{kg}\right)}{6680+\left(216\times \text{BMI}\right)}$	As basic parameters
		Female	$\frac{9270\times \text{Bw}\left(\text{kg}\right)}{8780+\left(244\times \text{BMI}\right)}$
	Green and Duffull’s formula (LBM.gd)	Male	$1.10\times \text{Bw}\left(\text{kg}\right)-0.0128\times \text{BMI}\times \text{Bw}\left(\text{kg}\right)$
		Female	$1.07\times \text{Bw}\left(\text{kg}\right)-0.0148\times \text{BMI}\times \text{Bw}\left(\text{kg}\right)$
BMR	Roza’s formula (BMR.r)	Male	$13.397\times \text{Bw}\left(\text{kg}\right)+4.799\times \text{Ht}\left(\text{cm}\right)-5.677\times \text{Age}\left(\text{y}\right)+88.362$	Indicators affected by sex, age, and weight
		Female	$9.247\times \text{Bw}\left(\text{kg}\right)+3.098\times \text{Ht}\left(\text{cm}\right)-4.330\times \text{Age}\left(\text{y}\right)+447.593$
	Mifflin-St Jeor estimate (BMR.ms)	Male	$9.99\times \text{Bw}\left(\text{kg}\right)+6.25\times \text{Ht}\left(\text{cm}\right)-4.92\times \text{Age}\left(\text{y}\right)+5$
		Female	$9.99\times \text{Bw}\left(\text{kg}\right)+6.25\times \text{Ht}\left(\text{cm}\right)-4.92\times \text{Age}\left(\text{y}\right)-161$

SUV_corr, SUV correction; Bw, body weight; Ht, height; BMI, body mass index; BSA, body surface area; IBW, ideal body weight; PNW, predicted normal weight; FFM, fat-free mass; LBM, lean body mass; BMR, basal metabolic rate; C_img, image-derived radioactivity concentration; Dose_inj, injected radioactivity dose.

Verification of PET values and correlation between overall survival and disease-free survival

We verified the correlation between overall survival and disease-free survival (OS-DFS) for the PET values in the following three groups: Pre alone (Pre_alone), difference between Pre and Post (Diff_PrePost), and quotient of Pre and Post (Quot_PrePost). This process is ongoing, and useful parameters and their cutoffs have been identified.

Visual evaluation

A single doctor visually evaluated various organs using the 0–4-point method. However, the entire brain was not imaged or evaluated in multiple cases. For myocardial assessment, we could not reproduce the reconstruction of the vertical long-axis/horizontal long-axis/short-axis images. Therefore, we associated 17 segments of the fusion image with three-way maximum intensity projection images [anterior to posterior: AP; lateral (left to right): LR; left anterior oblique: LAO].

Statistical analysis

We analyzed the data using R software (version 3.1.0, R Foundation)³² and three additional packages: pROC,³³ survival,³⁴ and MASS.³⁵ We computed and evaluated the correlation coefficients between each parameter and event using Cox regression analysis (Cox proportional hazards model). We also evaluated each parameter using stepwise regression³⁶ on Akaike’s information criterion³⁷ using the MASS package. Receiver operating characteristic curve analysis was conducted to derive the optimal cut-off value for each parameter to predict 2-year survival using the Youden index.³⁸ Survival curves were estimated using Kaplan–Meier methods based on the survival package (R software). Generalized logistic regression analysis of binary data was performed using the glm function in the R software package. We also used the logistf package³⁹ with Firth’s penalized likelihood method⁴⁰ to verify the exact logistic regression because of the small number of cases. The history of the program files was managed by Git⁴¹ (a free, open-source system) and GitHub (provides hosting for software development and version control using Git). All P values except those for hazard ratio (HR; obtained from Cox regression model), likelihood ratio (obtained from likelihood test), and log-rank test were calculated using the Wilcoxon rank-sum exact test or Wilcoxon rank-sum test. A two-sided P value of <0.05 was considered statistically significant.

Results

Patient and treatment characteristics

Amongst the 112 patients, 44 received neoadjuvant chemotherapy and 35 did not undergo ¹⁸F-FDG-PET/CT (in most cases, post-CCRT PET was not performed). Moreover, the images of 4 patients could not be technically evaluated because of corrupted or unreadable data. Finally, 29 patients who met the inclusion criteria were examined. The observations for 13 patients were right-censored, and the time of death was recorded for all other patients. At the final follow-up on 26 June 2021, the median overall survival was 21.4 months (range, 3.6–100.9 months), with 8 survivors. Table 2 presents the clinical characteristics of the study population. Following CCRT, the initial evaluation revealed a complete response in 12 patients, a partial response in 8 patients, stable disease in 4 patients, and progressive disease in 5 patients. Surgery (n = 6), endoscopic submucosal dissection (n = 2), chemotherapy (n = 7), and radiotherapy (n = 3) were also used as salvage therapies.

Table 2Clinical characteristics of the study population

Characteristic	n =29
Initial age (years)	66 [61–76]
Height (cm)	168 [160–171]
Diabetes
Yes	2 (6.9%)
No	27 (93%)
Tobacco experience
Yes	26 (90%)
No	2 (6.9%)
Unknown	1 (3.4%)
Tobacco (pack-year)	28 [12–35]
Alcohol experience
Social	28 (97%)
Never	1 (3.4%)
Alcohol face flush
Yes	15 (52%)
Yes (former)	3 (10%)
No	6 (21%)
No (Unknown for previous)	2 (6.9%)
Unknown	3 (10%)
Alcohol (drinks/day)a	4.40 [3.89–6.60]
Initial stage (UICC 8th)
I	5 (17%)
II	4 (14%)
III	7 (24%)
IV	13 (45%)
Pathology
SCC	28 (97%)
ADC	1 (3.4%)
RTx dose (Gy)
<50	3 (10%)
≥50	26 (90%)
CTx regimen
FP1000/75b	23 (79%)
FP700/70c	4 (14%)
5-FU1000d	1 (3.4%)
PTX30e	1 (3.4%)

Data given as median [25%–75%] or number (percentage).

UICC, Union for International Cancer Control; SCC, squamous cell carcinoma; ADC, adenocarcinoma; RTx, radiation therapy; CTx, chemotherapy.

^a One drink =10 g ethanol; ^b Fluorouracil 1000 mg/m² + cisplatin 75 mg/m²; ^c Fluorouracil 700 mg/m² + cisplatin 70 mg/m²; ^d Fluorouracil 1000 mg/m²; ^e Paclitaxel 30 mg/m².

Correlation between weight and each weight correction parameter

Fig. 1 shows plots of each SUV correction parameter for examining the effects. The correction parameters differed only in the IBW derived from height; however, the other values increased with height and weight. LBM.j, LBM.gd, and PNW showed similar trends for men and women and for weight and height; however, LBM.gd and PNW decreased more with short stature and increased weight. In addition, Roza’s BMR (BMR.r) and BMR.ms accounted for the age-related decline. Considering the age distribution of the patients, the weight/height graph for BMRs narrowed between the ages of 40 and 90 years. The height was fixed at 165 cm for all parameters because it represented the mean patient height rounded down to the nearest whole number, and a three-dimensional plot would be too complex to interpret. The correlations of body weight and age with correction parameters are shown in Supplement Figures 1 and 2. BMR.r and BMR.ms decreased with age; however, BMR.ms was fundamentally similar to FFM. BMR.r had different slopes for men and women. The difference in the SUV correction value for BMR.r between men and women was almost constant, regardless of age and weight. However, for BMR.ms, the slope for men was steeper than that for women, and SUV was reversed.

Fig. 1

Correlation between male and female weight, height, and weight correction parameters.

Correlation of weight, height, and weight correction parameters between males and females. The basal metabolic rate values for each sex are displayed with a gradient. BW, body weight; BMI, body mass index; BSA, body surface area; IBW, ideal body weight; PNWT, predicted normal weight; FFM, fat-free mass; LBM.j, lean body mass (Janmahasatian’s estimate); LBM.gd, lean body mass (Green and Duffull’s formula); BMR.r, basal metabolic rate (Roza’s formula); BMR.ms, basal metabolic rate (Mifflin-St’s estimate).

PET/CT scan assessment

¹⁸F-FDG-PET/CT images of all patients allowed evaluation from the neck to the pelvis. Among the patients, one had different bed movement settings and four had different imaging timings for pre-treatment and post-treatment scans, although none had both. Table 3 and Table 4 show the SUV-correlated body parameters based on sex.

Table 3Characteristics of FDG-PET scan

Characteristic	Male		P	Female		P	Overall		P
	Pre (n =24)	Post (n =24)		Pre (n =5)	Post (n =5)		Pre (n =29)	Post (n =29)
FDG dose (MBq)	238 [216–259]	232 [204–244]	0.31	153 [139–180]	165 [144–172]	>0.99	236 [180–256]	227 [191–236]	0.37
BW adapted RA (MBq/kg)a	3.89 [3.69–3.99]	3.93 [3.69–4.06]	0.77	3.98 [3.87–4.30]	4.08 [3.88–4.18]	>0.99	3.90 [3.69–4.05]	3.93 [3.71–4.09]	0.71
Interval (days)	15 [10–20]	34 [27–52]	<0.001	15 [14–21]	29 [27–35]	0.008	15 [11–21]	33 [27–52]	<0.001

Data given as median [25%–75%]; P values from Wilcoxon rank-sum exact test.

^a Radioactivity by body weight.

Table 4Characteristics of patients' body parameters

Group	Characteristic	Male		P	Female		P	Overall		P
		Pre (n =24)	Post (n =24)		Pre (n =5)	Post (n =5)		Pre (n =29)	Post (n =29)
Patients	Age	67 [65–77]	68 [65–78]	0.72	59 [56–61]	60 [57–61]	0.69	66 [61–76]	67 [61–77]	0.75
Correction parameters	BW	61 [54–65]	59 [53–64]	0.35	45 [35–46]	41 [33–44]	0.69	60 [50–63]	57 [47–62]	0.35
	BMI	21.4 [19.9–24.1]	20.8 [18.2–23.0]	0.35	16.8 [14.0–20.4]	14.9 [14.0–20.2]	0.84	21.0 [18.9–24.1]	20.2 [17.6–23.0]	0.38
	BSA	1.67 [1.62–1.82]	1.66 [1.55–1.77]	0.57	1.36 [1.28–1.50]	1.35 [1.25–1.42]	0.84	1.67 [1.50–1.74]	1.62 [1.50–1.71]	0.56
	IBW	65.3 [57.6–67.5]	65.3 [57.6–67.5]	>0.99	50.4 [44.0–51.7]	50.4 [44.0–51.7]	>0.99	64 [56–66]	64 [56–66]	>0.99
	PNW	60 [56–68]	59 [53–65]	0.46	44 [37–50]	44 [35–45]	0.84	59 [50–63]	57 [49–62]	0.45
	FFM	51.4 [48.3–55.5]	51.0 [47.3–54.0]	0.68	34.4 [34.4–40.3]	34.3 [33.9–38.8]	0.69	51 [44–54]	50 [44–53]	0.70
	LBM.j	49 [47–55]	48 [44–53]	0.52	30.3 [26.6–33.5]	30.1 [25.7–30.8]	0.84	49 [41–51]	47 [40–51]	0.59
	LBM.gd	50 [47–55]	48 [44–53]	0.46	34.6 [30.2–38.2]	34.3 [29.1–35.0]	0.84	49 [43–52]	47 [42–51]	0.53
	BMR.r	1291 [1223–1443]	1294 [1161–1407]	0.51	1067 [928–1149]	1061 [943–1098]	0.84	1255 [1149–1432]	1261 [1121–1399]	0.56
	BMR.ms	1308 [1238–1439]	1301 [1195–1411]	0.53	925 [791–1067]	919 [798–1011]	0.84	1281 [1145–1430]	1267 [1137–1407]	0.59

Data given as median [25%–75%]; P values from Wilcoxon rank-sum exact test.

BW, body weight; BMI, body mass index; BSA, body surface area; IBW, ideal body weight; PNW, predicted normal weight; FFM, fat-free mass; LBM.j, lean body mass (Janmahasatian's estimate); LBM.gd, lean body mass (Green and Duffull's formula); BMR.r, basal metabolic rate (Roza's formula); BMR.ms, basal metabolic rate (Mifflin-St Jeor estimate).

SUV evaluation in VOI

Fig. 2 shows the SUVmean values based on the pre-and post-treatment ¹⁸F-FDG accumulation in various organs. Supplement Figure 3 shows the pre-and post-treatment data plots for the corresponding data. A significant difference was observed in the distribution of the myocardium and multifidus before and after treatment (P = 0.008 and P = 0.001, respectively). In multiple cases, the accumulation of ¹⁸F-FDG in the myocardium decreased or recovered significantly before and after treatment. The myocardium was also visually evaluated by applying each segment model based on the 17-segment model. (Supplement Figure 4).

Fig. 2

Paired data plot of standardized uptake values for pre- and post-treatment accumulation.

Paired data plot for pre-and post-treatment fluorodeoxyglucose accumulation in the abdominal aorta, colon, jejunum, myocardium (left ventricle), liver, spleen, and multifidus muscle are shown. P values are based on the Wilcoxon signed-rank test. Left ventricular accumulation was measured as myocardial accumulation. The liver and spleen were assessed with a 1-cm³ VOI, but the total volume accumulation could not be assessed. Pre, pre-treatment; Post, post-treatment.

Changes in ¹⁸F-FDG accumulation before and after CCRT

Fig. 3 displays representative examples of changes in ¹⁸F-FDG accumulation. SUV levels were insignificant and stable in the multifidus muscles; however, they were uneven in the liver and spleen, often showing changes before and after treatment.

Fig. 3

Time course of ¹⁸F-FDG accumulation in three patients.

All three patients underwent imaging once before treatment and twice after treatment. Arrows indicate the primary tumor, and arrowheads indicate the myocardial area.

Fig. 4

Frequency distribution of each weight correction parameter for SUVmean, SUVmeanW, and SUVmeanQ.

SUV before treatment is shown on the left side, and SUV after treatment is shown on the right side (P <0.001). SUVmeanW, average PET; SUVmeanQ, average Q-PET; BW, body weight; IBW, ideal body weight; BSA, body surface area; FFM, fat-free mass; PNW, predicted normal weight;LBM.j, lean body mass (Janmahasatian’s estimate); LBM.gd, lean body mass (Green and Duffull’s formula); BMR.r, basal metabolic rate (Roza’s formula); BMR.ms, basal metabolic rate (Mifflin-St Jeor estimate).

Fig. 5

Histogram of each weight correction parameter for MTV, Vol.mean,Vol.meanQ, and Vol.meanW.

The left side shows each count before treatment, and the right side shows each count after treatment. Vol.mean, MTV*SUVmean; Vol.meanQ, MTV*SUVmeanQ; Vol.meanW, MTV*SUVmeanW; BW, body weight; IBW, ideal body weight; BSA, body surface area; FFM, fat-free mass; PNW, predicted normal weight;LBM.j, lean body mass (Janmahasatian’s estimate); LBM.gd, lean body mass (Green and Duffull’s formula); BMR.r, basal metabolic rate (Roza’s formula); BMR.ms, basal metabolic rate (Mifflin-St Jeor estimate).

Distribution and linear regression

Fig. 4 and Fig. 5 show the distribution of SUVs and SUV*Volumes by weight corrections for SUVmeans (SUVmean, SUVmeanW, and SUVmeanQ). Vol.means (Vol.mean, Vol.meanW, and Vol.meanQ) had higher correlation coefficients for MTV than the SUVmeans (SUVmean, SUVmeanW, or SUVmeanQ) for all body weight corrections before and after treatment, when the threshold was fixed. Table 5 presents the detailed characteristics of the corrected and measured data from the ¹⁸F-FDG-PET scans. In the relative cutoff, the maximum correlation coefficient for MTV to Vol.means was observed when SUVmean was corrected using IBW and the Mifflin-St Jeor estimate (BMR.ms). The detailed parameters are provided in Supplementary tables. Supplementary Table S1 presents the top 20 correlation coefficients between Vol.means (Table S1–1) and SUVmeans (Table S1–2).

Table 5Characteristics of corrected and measured data based on ¹⁸F-FDG-PET scan

Characteristic	Male		P	Female		P	Overall		P
	Pre (n =24)	Post (n =24)		Pre (n =5)	Post (n =5)		Pre (n =29)	Post (n =29)
Vol.ml	22 [12–33]	17 [9–25]	0.21	78 [40–113]	36 [15–44]	0.22	23 [12–40]	18 [10–27]	0.13
Vol.mean	99 [47–185]	41 [20–63]	0.001	641 [191–1,079]	112 [27–117]	0.10	142 [49–191]	41 [21–73]	<0.001
SUVMean	4.53 [2.75–6.97]	2.25 [1.97–2.58]	<0.001	7.68 [4.82–8.26]	2.67 [1.98–3.03]	0.056	4.82 [2.76–7.62]	2.28 [1.98–2.67]	<0.001
SUVPeak	9.9 [5.3–12.7]	3.4 [2.8–4.1]	<0.001	13.5 [10.6–13.8]	4.7 [2.9–4.8]	0.056	10.6 [5.4–12.9]	3.5 [2.9–4.7]	<0.001
MeanWahl	4.76 [3.16–7.14]	2.62 [2.32–3.14]	<0.001	7.37 [5.50–10.32]	3.18 [2.26–3.46]	0.056	5.40 [3.23–7.86]	2.65 [2.26–3.20]	<0.001
SUVqPeak	11.5 [6.4–13.8]	4.0 [3.2–4.9]	<0.001	14.2 [12.8–15.4]	5.4 [3.2–5.8]	0.056	12.8 [6.5–14.3]	4.1 [3.2–5.4]	<0.001
MeanQpet	5.30 [3.49–7.73]	2.83 [2.51–3.45]	<0.001	7.82 [6.10–11.07]	3.48 [2.47–3.75]	0.056	5.89 [3.56–8.52]	2.87 [2.47–3.55]	<0.001
SUVMax	11.8 [6.4–14.0]	4.1 [3.2–5.1]	<0.001	14.3 [12.9–15.6]	5.4 [3.2–5.8]	0.056	12.9 [6.6–14.7]	4.2 [3.2–5.4]	<0.001

Data given as median [25%–75%]; P values from Wilcoxon rank-sum exact test.

Vol.ml, volume only (= MTV); Vol.mean, volume × SUVmean (= total region glycolysis); SUVMean, mean standardized uptake value; SUVPeak, peak standardized uptake value; MeanWahl, average of SUVs in each slice that include the delimited VOI of SUVpeak; SUVqPeak, method for lymphoma as proposed by Hasenclever et al.¹⁸; MeanQpet, average of SUVs in each slice that include the delimited VOI of SUVqpeak; SUVMax, maximum standardized uptake value.

The correlation coefficient between SUVmean and Vol.means decreased as the cutoff increased and reached its maximum at PNW, with a cut-off value of 2.0. With an SUV cut-off value of 3.0, a negative correlation coefficient was observed between survival time and MTV and Vol.means. Furthermore, the correlation coefficient between Vol.MeanW and Vol.meanQ was higher than that between Vol.meanW and Vol.mean. The highest correlation coefficient was observed when Vol.meanW was corrected for body weight. The top 20 results are summarized in Supplementary Table S2.

Cox regression analysis

Table 6 presents the top three groups in the Cox regression analysis. Pre-treatment MTV alone was selected as the best parameter for calculating results based on each parameter using the difference and quotient. This was followed by pre- and post-treatment Vol.meanQ and pre-treatment Vol.mean. No values exceeded the obtained PET values. Supplementary Table S3–1 shows the top 20 parameters by stepwise regression analysis, and Supplementary Table S3–2 provides detailed thresholds and cut-off values for Table 6.

Table 6Characteristics of the top three groups of Cox regression results

Group	Calculation	SUVs	Thresholds	Correction parameters	Hazard ratio			c-statistic	Likelihood ratio
					HR	P	95% CI		LR	P
1	Prealone	Vol.ml	3.0	-	39.4	<0.01	4.6–339.3	0.81	21.5	<0.01
2	DiffPrePost	Vol.meanQ	2.5	BMR.ms	38.5	<0.01	4.4–333.3	0.80	20.9	<0.01
3	Prealone	Vol.mean	2.5, 3.0	-	38.1	<0.01	4.4–330.9	0.80	20.7	<0.01
	DiffPrePost	Vol.meanW	2.5, 3.0	BMR.ms, BMR.r, BSA, BW, FFM, IBW, LBM.gd, PNW	38.1	<0.01	4.4–330.9	0.80	20.7	<0.01
		Vol.meanW	3.0	LBM.j	38.1	<0.01	4.4–330.9	0.80	20.7	<0.01
		Vol.meanQ	2.5, 3.0	BMR.r, BSA, BW, FFM, LBM.gd, PNW	38.1	<0.01	4.4–330.9	0.80	20.7	<0.01
		Vol.meanQ	3.0	BMR.ms, IBW	38.1	<0.01	4.4–330.9	0.80	20.7	<0.01

SUVs, standard uptake values; HR, hazard ratio; CI, confidence interval; LR, likelihood ratio; Pre_alone, only pre-treatment data are used, with no post-treatment data included; Diff_PrePost, difference between pre-treatment and post-treatment tests; Vol.mean, volume × SUV mean (= total region glycolysis); Vol.meanQ, volume × SUV meanQ; Vol.meanW, volume × SUV meanW; BW, body weight; BMI, body mass index; BSA, body surface area; IBW, ideal body weight; PNW, predicted normal weight; FFM, fat-free mass; LBM.j, lean body mass (Janmahasatian's estimate); LBM.gd, lean body mass (Green and Duffull's formula); BMR.r, basal metabolic rate (Roza's formula); BMR.ms, basal metabolic rate (Mifflin-St Jeor estimate).

Survival

Fig. 6 shows the Kaplan–Meier plots for 2-year overall and progression-free survival in the top three groups based on HR. The selected parameters were considered good predictors of overall and progression-free survival.

Fig. 6

Kaplan–Meier plots of 2-year overall survival and 2-year progression-free survival for the top three groups.

Two-year overall and progression-free survival for the top three groups (each group corresponds to the groups in Table 6). Hazard ratios were used in the plots.

Discussion

PET parameters and prognosis prediction

For pre- and post-treatment comparisons, the difference between pre-treatment and post-treatment (Diff_PrePost) Vol*Means (MTV SUV threshold of 2.5) corrected for BMR.ms was selected as the best HR factor. Although the HR value was not superior to the Pre_alone MTV single factor, which is currently widely used, it was considered to be almost equivalent based on the c-statistic.

A report evaluating the parameters of pre-treatment ¹⁸F-FDG-PET found that MTV with an automatic threshold of 20% or 40% was useful (although a fixed value of 3.0 or 2.5 was more useful for the threshold in our study), consistent with our results.^42,43,44,45 Kanoun et al.¹⁵ reported that the baseline MTV in patients with Hodgkin’s lymphoma was a favorable predictor, and that combination with the difference in SUV before and interim treatment provided a significant insight into the prognosis. These results were consistent with those of the present study.

Previous studies of squamous cell carcinoma on the oral side of the diaphragm have examined ratios and rates of change in MTV and SUV and reported a cutoff for their correlation with the prognosis.^12,13,14 However, in our study the quotient did not show statistically significance, which did not support the usefulness in prognosis prediction.

Among the Diff_PrePost, the difference in Vol.*Means was higher than the difference in MTV for both HR and likelihood ratio. Only Vol.*MeanQ, which used BMR.ms for correction, was different from the others. SUVmeanQ (tumor SUV normalized by liver SUVmean) using BMR.ms for correction was superior to MTV probably because of the small number of cases and the effect of body fat correction in men and women. FFM, PNW, and LBM evaluate metabolism by body mass, but BMR also considers age. It is suggested that not only sex differences but also age correction may be important for LBM. LBM.gd, PNW, and FFM also resembled BSA and body weight because the range of 160–170 cm, which was common in this case, was similar, as observed in Fig. 1.

In addition, in this study, the body weight of women was less than 50 kg both before and after treatment. It is possible that women’s metabolism was overestimated because of the following factors. First, PNW and LBM (especially PNW and LBM.gd) have smaller sex differences for patients with low weight. Second, the BMR.r slope tends to flatten for women. In particular, the values for men and women are reversed when the weight is less than 50 kg.

Only SUVmeanQ and SUVmeanW were more useful in the differential analysis. This may reflect insufficient standardization of SUVmean or poor correction of adipose tissue. It has been reported that the estimated muscle mass from the estimation formula differs between Asians and other races.⁴⁶ Regarding the evaluation of LBM, which is related to metabolism, it is considered necessary to at least consider the root genes (e.g., three races), sex, and age.

In addition to reconstructing the estimation formula, we think that it is one of the countermeasures to consider the prediction of body weight that contributes to metabolism according to resting energy expenditure by other methods, including indirect calorimetry.⁴⁷ This includes Weir’s formula⁴⁸ and its simplifications.⁴⁹The Weir equation is as follows:

  

$\begin{array}{cc}EE\left(\text{kcal}/\text{day}\right)& =(\left(Consumed.O_2\left(\text{L}/\text{min}\right)\times 3.941\right)+[Produced.C{O}_2\left(\text{L}/\text{min}\right)\\ & \times 1.106]-\left(urine.N_2\left(\text{g}/\text{min}\right)\times 2.17\right))\times 1440\left(\text{min}/\text{day}\right)\end{array}$

The simplified Weir equation is as follows:

  

$\begin{array}{cc}EE\left(\text{kcal}/\text{day}\right)& \approx (\left(Consumed.O_2\left(\text{L}/\text{min}\right)\times 3.9\right)+[Produced.C{O}_2\left(\text{L}/\text{min}\right)\\ & \times 1.1\left]\right)\times 1440\left(\text{min}/\text{day}\right)\end{array}$

Non-tumor ¹⁸F-FDG accumulation

Although there was a significant difference in the distribution of myocardial uptake and multifidus muscle uptake before and after treatment, there was no significant trend in the course of increase or decrease. Therefore, the effect of the glucose metabolism environment, such as the effect of fasting before the examination, was considered. In addition to the effect of glucose metabolism, the effects of alcohol^50,51,52 and smoking^53,54 on glucose metabolism have been reported.

Reactive oxygen species (ROS) induce glucose metabolic pathways that are distinct from the normal pathway (e.g., α7nAchR-NF-κB-IL-6 → JAK2/Akt crosstalk).⁵⁵ α7-Nicotinic acetylcholine receptor (α7nAchR) is involved in these pathways and is upregulated by the nuclear factor-κB (NF-κB) pathway involving tumor necrosis factor-α and interleukin-6 (IL-6) induced by lipopolysaccharide (LPS) produced from intestinal bacteria,^56,57 by ethanol,⁵⁸ or by nitrosamine ketones derived from tobacco smoke extract and nicotine.⁵⁹

Although the effects of alcohol withdrawal on LPS are unclear, even though LPS is also induced by alcohol, the dual effect observed on cytochrome P450 2E1 (CYP2E1) was short-lived, with no increase in activity after more than 8–10 days of abstinence.^60,61 However, this does not happen after about 3 consecutive days,⁶² and it takes about 2 weeks to reduce the liver damage caused by chronic alcohol consumption⁶³ and 3 months for normalization of the metabolic environment in the body.⁶⁴

CYP2E1-mediated responses in the microsomal ethanol oxidizing system exhibit sex differences⁶⁵ and are upregulated by chronic alcohol consumption.⁶⁶ This fact may be more influential in some people, such as many Japanese and some Southeast Asians who have the ADH1B2 allele (also known as the alcohol dependence gene), which accelerates the metabolism of ethanol to acetaldehyde, and/or ALDH2*2, which slows down the metabolism of acetaldehyde to acetate. On the contrary, the effect of smoking on CYP2E1 regulation is unclear, as there are reports of both upregulation and downregulation, but the effect of nicotine on α7nAch receptors is known.⁶⁷ Therefore, to reduce the effects of metabolic abnormalities on the test, abstinence from alcohol consumption (including smoking cessation if possible) for at least 10 days before each pre- and post-treatment test is desirable.

This study has some limitations. The patients examined and analyzed in this study could have received irradiation for submucosal tumors that had small volumes before and after treatment, and this may have underestimated SUV levels in one test because of the partial volume effect. Therefore, the availability of SUV values related to the amount or rate of change was a secondary goal. Second, because of the retrospective study design, the time period in which the patients underwent PET-CT was varied. Cases with post-treatment PET-CT within 1 month were included, and this included the effect of the post-treatment inflammatory response. Patients who received preoperative chemotherapy were excluded, limiting the number of cases.

Despite these limitations, the strength of this study was that it evaluated the strength and weakness of metabolism by examining changes in the intensity of SUVs rather than evaluating only the amount of SUV that accumulated above the cutoff value. With regard to the indirect evaluation of metabolism, the backgrounds of the original studies were different, and the metabolic formula used for standardization was examined in detail by comprehensive evaluation of the different distribution characteristics.

In conclusion, considering the effect of age-related decreases in LBM, the difference in the product of MTV and SUVmeanQ before and after CCRT may be a useful parameter to predict the prognosis of patients treated with CCRT for esophageal cancer. However, pre-treatment MTV remains a solid prognostic parameter.

Acknowledgments

The authors thank the late Dr. Shinichi Tominaga (who passed away on September 2, 2018) for his guidance and support.

Conflicts of Interest

The authors have declared that no conflict of interest exists.

References

• 1.  National Comprehensive Cancer Network: Esophageal and esophagogastric junction cancers (version 2.2022). National Comprehensive Cancer Network. 2022. https://www.nccn.org/guidelines/guidelines-detail?category=1&id=1433. Accessed 17 Mar 2023.
• 2.  Smyth EC, Verheij M, Allum W, Cunningham D, Cervantes A, Arnold D, ESMO Guidelines Committee: Gastric cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol 2016; 27: v38–v49. PMID:27664260 https://doi.org/10.1093/annonc/mdw350
• 3.  Kumabe A, Fukada J, Kota R, Koike N, Shiraishi Y, Seki S, Yoshida K, Kitagawa Y, Shigematsu N: Long-term results of concurrent chemoradiotherapy with daily-low-dose continuous infusion of 5-fluorouracil and cisplatin (LDFP) for Stage I-II esophageal carcinoma. Dis Esophagus 2018; 31: 1–8. PMID:29228166 https://doi.org/10.1093/dote/dox138
• 4.  Shapiro J, van Lanschot JJ, Hulshof MC, van Hagen P, van Berge Henegouwen MI, Wijnhoven BP, van Laarhoven HW, Nieuwenhuijzen GA, Hospers GA, Bonenkamp JJ, Cuesta MA, Blaisse RJ, Busch OR, ten Kate FJ, Creemers GJ, Punt CJ, Plukker JT, Verheul HM, Bilgen EJ, van Dekken H, van der Sangen MJ, Rozema T, Biermann K, Beukema JC, Piet AH, van Rij CM, Reinders JG, Tilanus HW, Steyerberg EW, van der Gaast A, CROSS Study Group: Neoadjuvant chemoradiotherapy plus surgery versus surgery alone for oesophageal or junctional cancer (CROSS): long-term results of a randomised controlled trial. Lancet Oncol 2015; 16: 1090–1098. PMID:26254683 https://doi.org/10.1016/S1470-2045(15)00040-6
• 5.  Fletcher JW, Djulbegovic B, Soares HP, Siegel BA, Lowe VJ, Lyman GH, Coleman RE, Wahl R, Paschold JC, Avril N, Einhorn LH, Suh WW, Samson D, Delbeke D, Gorman M, Shields AF: Recommendations on the use of 18F-FDG PET in oncology. J Nucl Med 2008; 49: 480–508. PMID:18287273 https://doi.org/10.2967/jnumed.107.047787
• 6.  Marzola MC, De Manzoni G, Grassetto G, Cordiano C, Al-Nahhas A, Alavi A, Rubello D. Extended staging of oesophageal cancer using FDG-PET - a critical appraisal. Eur J Radiol 2012 ; 81: 21-30. PMID: 21055894, https://doi.org/10.1016/j.ejrad.2010.10.018
• 7.  Hoekstra CJ, Stroobants SG, Smit EF, Vansteenkiste J, van Tinteren H, Postmus PE, Golding RP, Biesma B, Schramel FJ, van Zandwijk N, Lammertsma AA, Hoekstra OS: Prognostic relevance of response evaluation using [18F]-2-fluoro-2-deoxy-d-glucose positron emission tomography in patients with locally advanced non-small-cell lung cancer. J Clin Oncol 2005; 23: 8362–8370. PMID:16293866 https://doi.org/10.1200/JCO.2005.01.1189
• 8.  Larson SM, Schwartz LH: 18F-FDG PET as a candidate for “qualified biomarker”: functional assessment of treatment response in oncology. J Nucl Med 2006; 47: 901–903. PMID:16741296
• 9.  Yap WK, Chang YC, Tseng CK, Hsieh CH, Chao YK, Su PJ, Hou MM, Yang CK, Pai PC, Lin CR, Hsieh CE, Wu YY, Hung TM: Predictive value of nodal maximum standardized uptake value of pretreatment [18F]fluorodeoxyglucose positron emission tomography imaging in patients with esophageal cancer. Dis Esophagus 2017; 30: 1–10. PMID:28575243 https://doi.org/10.1093/dote/dox021
• 10.  Cheson BD, Fisher RI, Barrington SF, Cavalli F, Schwartz LH, Zucca E, Lister TA; Alliance, Australasian Leukaemia and Lymphoma Group; Eastern Cooperative Oncology Group; European Mantle Cell Lymphoma Consortium; Italian Lymphoma Foundation; European Organisation for Research; Treatment of Cancer/Dutch Hemato-Oncology Group; Grupo Español de Médula Ósea; German High-Grade Lymphoma Study Group; German Hodgkin’s Study Group; Japanese Lymphorra Study Group; Lymphoma Study Association; NCIC Clinical Trials Group; Nordic Lymphoma Study Group; Southwest Oncology Group; United Kingdom National Cancer Research Institute: Recommendations for initial evaluation, staging, and response assessment of Hodgkin and non-Hodgkin lymphoma: the Lugano classification. J Clin Oncol 2014; 32: 3059–3067. PMID:25113753 https://doi.org/10.1200/JCO.2013.54.8800
• 11.  Barrington SF, Mikhaeel NG, Kostakoglu L, Meignan M, Hutchings M, Müeller SP, Schwartz LH, Zucca E, Fisher RI, Trotman J, Hoekstra OS, Hicks RJ, O’Doherty MJ, Hustinx R, Biggi A, Cheson BD: Role of imaging in the staging and response assessment of lymphoma: consensus of the International Conference on Malignant Lymphomas Imaging Working Group. J Clin Oncol 2014; 32: 3048–3058. PMID:25113771 https://doi.org/10.1200/JCO.2013.53.5229
• 12.  Shimizu M, Mitsudo K, Koike I, Taguri M, Iwai T, Koizumi T, Oguri S, Kioi M, Hirota M, Inoue T, Tohnai I: Prognostic value of 2-[18 F]fluoro-2-deoxy-d-glucose positron emission tomography for patients with oral squamous cell carcinoma treated with retrograde superselective intra-arterial chemotherapy and daily concurrent radiotherapy. Oral Surg Oral Med Oral Pathol Oral Radiol 2016; 121: 239–247. PMID:26768076 https://doi.org/10.1016/j.oooo.2015.10.018
• 13.  Shimomura H, Sasahira T, Yamanaka Y, Kurihara M, Imai Y, Tamaki S, Yamakawa N, Shirone N, Hasegawa M, Kuniyasu H, Kirita T: [18F]fluoro-2-deoxyglucose-positron emission tomography for the assessment of histopathological response after preoperative chemoradiotherapy in advanced oral squamous cell carcinoma. Int J Clin Oncol 2015; 20: 308–316. PMID:24942501 https://doi.org/10.1007/s10147-014-0711-5
• 14.  Kim H, Yoo IR, Boo SH, Park HL, O JH, Kim SH: Prognostic value of pre- and post-treatment FDG PET/CT parameters in small cell lung cancer patients. Nucl Med Mol Imaging 2018; 52: 31–38. PMID:29391910 https://doi.org/10.1007/s13139-017-0490-9
• 15.  Kanoun S, Rossi C, Berriolo-Riedinger A, Dygai-Cochet I, Cochet A, Humbert O, Toubeau M, Ferrant E, Brunotte F, Casasnovas RO: Baseline metabolic tumour volume is an independent prognostic factor in Hodgkin lymphoma. Eur J Nucl Med Mol Imaging 2014; 41: 1735–1743. PMID:24811577 https://doi.org/10.1007/s00259-014-2783-x
• 16.  Jayachandran P, Pai RK, Quon A, Graves E, Krakow TE, La T, Loo BW, Koong AC, Chang DT: Postchemoradiotherapy positron emission tomography predicts pathologic response and survival in patients with esophageal cancer. Int J Radiat Oncol Biol Phys 2012; 84: 471–477. PMID:22381904 https://doi.org/10.1016/j.ijrobp.2011.12.029
• 17.  Boellaard R, Delgado-Bolton R, Oyen WJ, Giammarile F, Tatsch K, Eschner W, Verzijlbergen FJ, Barrington SF, Pike LC, Weber WA, Stroobants S, Delbeke D, Donohoe KJ, Holbrook S, Graham MM, Testanera G, Hoekstra OS, Zijlstra J, Visser E, Hoekstra CJ, Pruim J, Willemsen A, Arends B, Kotzerke J, Bockisch A, Beyer T, Chiti A, Krause BJ, European Association of Nuclear Medicine (EANM): FDG PET/CT: EANM procedure guidelines for tumour imaging: version 2.0. Eur J Nucl Med Mol Imaging 2015; 42: 328–354. PMID:25452219 https://doi.org/10.1007/s00259-014-2961-x
• 18.  Hasenclever D, Kurch L, Mauz-Körholz C, Elsner A, Georgi T, Wallace H, Landman-Parker J, Moryl-Bujakowska A, Cepelová M, Karlén J, Álvarez Fernández-Teijeiro A, Attarbaschi A, Fosså A, Pears J, Hraskova A, Bergsträsser E, Beishuizen A, Uyttebroeck A, Schomerus E, Sabri O, Körholz D, Kluge R: qPET—a quantitative extension of the Deauville scale to assess response in interim FDG-PET scans in lymphoma. Eur J Nucl Med Mol Imaging 2014; 41: 1301–1308. PMID:24604592 https://doi.org/10.1007/s00259-014-2715-9
• 19.  Wahl RL, Jacene H, Kasamon Y, Lodge MA: From RECIST to PERCIST: evolving considerations for PET response criteria in solid tumors. J Nucl Med 2009; 50: 122S–150S. PMID:19403881 https://doi.org/10.2967/jnumed.108.057307
• 20.  Kanoun S, Tal I, Berriolo-Riedinger A, Rossi C, Riedinger JM, Vrigneaud JM, Legrand L, Humbert O, Casasnovas O, Brunotte F, Cochet A: Influence of software tool and methodological aspects of total metabolic tumor volume calculation on baseline [18F]FDG PET to predict survival in Hodgkin lymphoma. PLoS One 2015; 10: e0140830. PMID:26473950 https://doi.org/10.1371/journal.pone.0140830
• 21.  Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, Preibisch S, Rueden C, Saalfeld S, Schmid B, Tinevez JY, White DJ, Hartenstein V, Eliceiri K, Tomancak P, Cardona A: Fiji: an open-source platform for biological-image analysis. Nat Methods 2012; 9: 676–682. PMID:22743772 https://doi.org/10.1038/nmeth.2019
• 22.  Schneider CA, Rasband WS, Eliceiri KW: NIH Image to ImageJ: 25 years of image analysis. Nat Methods 2012; 9: 671–675. PMID:22930834 https://doi.org/10.1038/nmeth.2089
• 23.  Kubota K, Matsuzawa T, Ito M, Ito K, Fujiwara T, Abe Y, Yoshioka S, Fukuda H, Hatazawa J, Iwata R, Watanuki S, Ido T: Lung tumor imaging by positron emission tomography using C-11 l-methionine. J Nucl Med 1985; 26: 37–42. PMID:2981300
• 24.  Devine B: Gentamicin therapy. Drug Intell Clin Pharm 1974; 8: 650–655.
• 25.  Duffull SB, Dooley MJ, Green B, Poole SG, Kirkpatrick CM: A standard weight descriptor for dose adjustment in the obese patient. Clin Pharmacokinet 2004; 43: 1167–1178. PMID:15568893 https://doi.org/10.2165/00003088-200443150-00007
• 26.  Du Bois D, Du Bois EF: A formula to estimate the approximate surface area if height and weight be known. 1916. Nutrition 1989; 5: 303–313. PMID:2520314
• 27.  Garrow JS, Webster J: Quetelet’s index (W/H2) as a measure of fatness. Int J Obes (Lond) 1985; 9: 147–153. PMID:4030199
• 28.  Roza AM, Shizgal HM: The Harris Benedict equation reevaluated: resting energy requirements and the body cell mass. Am J Clin Nutr 1984; 40: 168–182. PMID:6741850 https://doi.org/10.1093/ajcn/40.1.168
• 29.  Mifflin MD, St Jeor ST, Hill LA, Scott BJ, Daugherty SA, Koh YO: A new predictive equation for resting energy expenditure in healthy individuals. Am J Clin Nutr 1990; 51: 241–247. PMID:2305711 https://doi.org/10.1093/ajcn/51.2.241
• 30.Green B, Duffull S: Caution when lean body weight is used as a size descriptor for obese subjects. Clin Pharmacol Ther 2002; 72: 743–744. https://doi.org/10.1067/mcp.2002.129306
• 31.  Janmahasatian S, Duffull SB, Ash S, Ward LC, Byrne NM, Green B: Quantification of lean bodyweight. Clin Pharmacokinet 2005; 44: 1051–1065. PMID:16176118 https://doi.org/10.2165/00003088-200544100-00004
• 32.  R Core Team: R: a language and environment for statistical computing. R Foundation for Statistical Computing. 2022. https://www.R-project.org/. Accessed 17 Mar 2023.
• 33.  Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez JC, Müller M: pROC: display and analyze ROC curves. 2021. http://expasy.org/tools/pROC/. Accessed 17 Mar 2023.
• 34.  Therneau TM: Survival: survival analysis. 2022. https://github.com/therneau/survival. Accessed 17 Mar 2023.
• 35.  Ripley B: MASS: support functions and datasets for Venables and Ripley’s MASS. 2022. http://www.stats.ox.ac.uk/pub/MASS4/. Accessed 17 Mar 2023.
• 36.  Breaux HJ: A modification of Efroymson’s technique for stepwise regression analysis. Commun ACM 1968; 11: 556–558. https://doi.org/10.1145/363567.363580
• 37.  Akaike H: A new look at the statistical model identification. IEEE Trans Automat Contr 1974; 19: 716–723. https://doi.org/10.1109/TAC.1974.1100705
• 38.  Akobeng AK: Understanding diagnostic tests 3: receiver operating characteristic curves. Acta Paediatr 2007; 96: 644–647. PMID:17376185 https://doi.org/10.1111/j.1651-2227.2006.00178.x
• 39.  Heinze G, Ploner M, Jiricka L: Logistf: Firth’s bias-reduced logistic regression. 2022. https://cemsiis.meduniwien.ac.at/en/kb/science-research/software/statistical-software/firth-correction/. Accessed 17 Mar 2023.
• 40.  Firth D: Bias reduction of maximum likelihood estimates. Biometrika 1993; 80: 27–38. https://doi.org/10.1093/biomet/80.1.27
• 41.  Straub B, Chacon S: Pro Git, 2nd edition. USA: Apress; 2014.
• 42.  Chen SW, Hsieh TC, Ding HJ, Yen KY, Lin CY, Liang JA, Kao CH: Pretreatment metabolic tumor volumes to predict the short-term outcome of unresectable locally advanced squamous cell carcinoma of the esophagus treated with definitive chemoradiotherapy. Nucl Med Commun 2014; 35: 291–297. PMID:24276528 https://doi.org/10.1097/MNM.0000000000000046
• 43.  Lemarignier C, Di Fiore F, Marre C, Hapdey S, Modzelewski R, Gouel P, Michel P, Dubray B, Vera P: Pretreatment metabolic tumour volume is predictive of disease-free survival and overall survival in patients with oesophageal squamous cell carcinoma. Eur J Nucl Med Mol Imaging 2014; 41: 2008–2016. PMID:25037871 https://doi.org/10.1007/s00259-014-2839-y
• 44.  Li Y, Lin Q, Wang L, Sun L, Dai M, Luo Z, Zheng H, Zhao L, Wu H: Application of sequential (18)F-FDG PET/CT scans for concurrent chemoradiotherapy of non-surgical squamous cell esophageal carcinoma. J Balkan Union Oncol 2014; 19: 517–523. PMID:24965415
• 45.  Betancourt SL, Marom EM, Palacio D, Correa M, Hofstetter WL: Prognostic value of initial FDG-PET/CT metabolic tumor volume in patients with resectable esophageal adenocarcinoma. JSM Gastroenterol Hepatol 2015; 3: 1037–1044.
• 46.  Janssen I, Heymsfield SB, Baumgartner RN, Ross R: Estimation of skeletal muscle mass by bioelectrical impedance analysis. J Appl Physiol 2000; 89: 465–471. PMID:10926627 https://doi.org/10.1152/jappl.2000.89.2.465
• 47.  Thomas N, Gupta RD, Ramachandran R, Venkatesan P, Anoop S, Joseph M: Indirect calorimetry: from bench to bedside. Indian J Endocrinol Metab 2017; 21: 594–599. PMID:28670546 https://doi.org/10.4103/ijem.IJEM_484_16
• 48.  Weir JB: New methods for calculating metabolic rate with special reference to protein metabolism. J Physiol 1949; 109: 1–9. PMID:15394301 https://doi.org/10.1113/jphysiol.1949.sp004363
• 49.  Mehta NM, Smallwood CD, Joosten KF, Hulst JM, Tasker RC, Duggan CP: Accuracy of a simplified equation for energy expenditure based on bedside volumetric carbon dioxide elimination measurement—a two-center study. Clin Nutr 2015; 34: 151–155. PMID:24636151 https://doi.org/10.1016/j.clnu.2014.02.008
• 50.  Volkow ND, Wang GJ, Shokri Kojori E, Fowler JS, Benveniste H, Tomasi D: Alcohol decreases baseline brain glucose metabolism more in heavy drinkers than controls but has no effect on stimulation-induced metabolic increases. J Neurosci 2015; 35: 3248–3255. PMID:25698759 https://doi.org/10.1523/JNEUROSCI.4877-14.2015
• 51.  Gispert JD, Figueiras FP, Vengeliene V, Herance JR, Rojas S, Spanagel R: Changes in cerebral [18F]-FDG uptake induced by acute alcohol administration in a rat model of alcoholism. Behav Brain Res 2017; 327: 29–33. PMID:28347826 https://doi.org/10.1016/j.bbr.2017.03.038
• 52.  Jacob A, Wang P: Alcohol intoxication and cognition: implications on mechanisms and therapeutic strategies. Front Neurosci 2020; 14: 102. PMID:32116535 https://doi.org/10.3389/fnins.2020.00102
• 53.  Rengier F, Weber TF, Henninger V, Böckler D, Schumacher H, Kauczor HU, von Tengg-Kobligk H: Heartbeat-related distension and displacement of the thoracic aorta in healthy volunteers. Eur J Radiol 2012; 81: 158–164. PMID:20850234 https://doi.org/10.1016/j.ejrad.2010.08.023
• 54.  Tong L, Sui Y, Jiang S, Yin Y: The association between lung fluorodeoxyglucose metabolism and smoking history in 347 healthy adults. J Asthma Allergy 2021; 14: 301–308. PMID:33840997 https://doi.org/10.2147/JAA.S302602
• 55.  Akbari M, Hassan-Zadeh V: IL-6 signalling pathways and the development of type 2 diabetes. Inflammopharmacology 2018; 26: 685–698. PMID:29508109 https://doi.org/10.1007/s10787-018-0458-0
• 56.  Liangpunsakul S, Toh E, Ross RA, Heathers LE, Chandler K, Oshodi A, McGee B, Modlik E, Linton T, Mangiacarne D, Jimenez C, Dong XC, Wang L, Tu W, Nelson DE: Quantity of alcohol drinking positively correlates with serum levels of endotoxin and markers of monocyte activation. Sci Rep 2017; 7: 4462. PMID:28667254 https://doi.org/10.1038/s41598-017-04669-7
• 57.  Bala S, Marcos M, Gattu A, Catalano D, Szabo G: Acute binge drinking increases serum endotoxin and bacterial DNA levels in healthy individuals. PLoS One 2014; 9: e96864. PMID:24828436 https://doi.org/10.1371/journal.pone.0096864
• 58.  Gorbounova O, Svensson AL, Jönsson P, Mousavi M, Miao H, Hellström-Lindahl E, Nordberg A: Chronic ethanol treatment decreases [3H]epibatidine and [3H]nicotine binding and differentially regulates mRNA levels of nicotinic acetylcholine receptor subunits expressed in M10 and SH-SY5Y neuroblastoma cells. J Neurochem 1998; 70: 1134–1142. PMID:9489734 https://doi.org/10.1046/j.1471-4159.1998.70031134.x
• 59.  Lee AT, Xu Z, Pothula SP, Patel MB, Pirola RC, Wilson JS, Apte MV: Alcohol and cigarette smoke components activate human pancreatic stellate cells: implications for the progression of chronic pancreatitis. Alcohol Clin Exp Res 2015; 39: 2123–2133. PMID:26463405 https://doi.org/10.1111/acer.12882
• 60.  Perrot N, Nalpas B, Yang CS, Beaune PH: Modulation of cytochrome P450 isozymes in human liver, by ethanol and drug intake. Eur J Clin Invest 1989; 19: 549–555. PMID:2515975 https://doi.org/10.1111/j.1365-2362.1989.tb00273.x
• 61.  Oneta CM, Lieber CS, Li J, Rüttimann S, Schmid B, Lattmann J, Rosman AS, Seitz HK: Dynamics of cytochrome P4502E1 activity in man: induction by ethanol and disappearance during withdrawal phase. J Hepatol 2002; 36: 47–52. PMID:11804663 https://doi.org/10.1016/S0168-8278(01)00223-9
• 62.  Yin H, Ingelman-Sundberg M, Lindros KO: Induction mechanisms of cytochrome P450 2E1 in liver: interplay between ethanol treatment and starvation. Biochem Pharmacol 1995; 50: 155–161. PMID:7632158 https://doi.org/10.1016/0006-2952(95)00128-M
• 63.  Kirpich IA, McClain CJ, Vatsalya V, Schwandt M, Phillips M, Falkner KC, Zhang L, Harwell C, George DT, Umhau JC: Liver injury and endotoxemia in male and female alcohol-dependent individuals admitted to an alcohol treatment program. Alcohol Clin Exp Res 2017; 41: 747–757. PMID:28166367 https://doi.org/10.1111/acer.13346
• 64.  Addolorato G, Capristo E, Greco AV, Caputo F, Stefanini GF, Gasbarrini G: Three months of abstinence from alcohol normalizes energy expenditure and substrate oxidation in alcoholics: a longitudinal study. Am J Gastroenterol 1998; 93: 2476–2481. PMID:9860412 https://doi.org/10.1111/j.1572-0241.1998.00707.x
• 65.  Penaloza CG, Cruz M, Germain G, Jabeen S, Javdan M, Lockshin RA, Zakeri Z: Higher sensitivity of female cells to ethanol: methylation of DNA lowers Cyp2e1, generating more ROS. Cell Commun Signal 2020; 18: 111. PMID:32653010 https://doi.org/10.1186/s12964-020-00616-8
• 66.  Lieber CS: The discovery of the microsomal ethanol oxidizing system and its physiologic and pathologic role. Drug Metab Rev 2004; 36: 511–529. PMID:15554233 https://doi.org/10.1081/DMR-200033441
• 67.  Wu Y, Song P, Zhang W, Liu J, Dai X, Liu Z, Lu Q, Ouyang C, Xie Z, Zhao Z, Zhuo X, Viollet B, Foretz M, Wu J, Yuan Z, Zou MH: Activation of AMPKα2 in adipocytes is essential for nicotine-induced insulin resistance in vivo. Nat Med 2015; 21: 373–382. PMID:25799226 https://doi.org/10.1038/nm.3826