Apparent Diffusion Coefficient and Its Standard Deviation from Diffusion-weighted Imaging to Differentiate Tumor-adjacent and Tumor-distant liver tissues in T3 staged Resectable Gallbladder Carcinoma

doi:10.21203/rs.3.rs-2110899/v1

Download PDF

Research article

Apparent Diffusion Coefficient and Its Standard Deviation from Diffusion-weighted Imaging to Differentiate Tumor-adjacent and Tumor-distant liver tissues in T3 staged Resectable Gallbladder Carcinoma

https://doi.org/10.21203/rs.3.rs-2110899/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

An extended cholecystectomy is generally recommended for patients with gallbladder carcinoma (GBC) at stage T₂ or above, and involves cholecystectomy and a limited hepatic resection. It is important to differentiate tumor-adjacent and tumor-distant liver tissues before surgery for surgical strategies decision making. Diffusion-weighted imaging (DWI) can provide functional information by depicting the diffusion of water in biological tissue and can be quantitated by apparent diffusion coefficient (ADC). Our study aimed to evaluate feasibility of ADC and its standard deviation (ADC_sd) to differentiate between tumor-adjacent and tumor-distant liver tissues in T₃ staged resectable GBC.

Methods

A total of 41 consecutive patients with T₃ staged resectable GBC were included in this retrospective study and divided into two sets with (n = 27) and without (n = 14) liver invasion. All patients underwent DWI at b-values of 0, 20, 50, 80, 100, 200, 400, 600, 800 and 1000 sec/mm² with a 3.0-T magnetic resonance scanner before surgery. Mean ADC and ADC_sd of tumor-adjacent and tumor-distant liver tissues were measured on DWI, and were compared by Mann-Whitney U tests. If there was a significant difference in any derived parameter, area under the receiver operating characteristic curve (AUC) was used to assess performance of this parameter for differentiating tumor-adjacent and tumor-distant liver tissues.

Results

In the set with liver invasion, mean ADC and ADC_sd of tumor-adjacent liver tissue were lower than those of tumor-distant liver tissue when b = 0, 800 sec/mm², and b = 0, 1000 sec/mm², respectively (all P-values < 0.05). In the set without liver invasion, ADC and ADC_sd at all b values had no statistical differences between tumor-adjacent and tumor-distant liver tissues (all P-values > 0.05). To differentiate tumor-adjacent from tumor-distant liver tissues in the set with liver invasion, AUCs of ADC were 0.687 (b = 0, 800 sec/mm²) and 0.680 (b = 0, 1000 sec/mm²), respectively; and those of ADC_sd were 0.673 (b = 0, 800 sec/mm²) and 0.731 (b = 0, 1000 sec/mm²), respectively.

Conclusions

In patients with liver invasion by T₃ staged GBC, ADC and ADC_sd could have potential value in differentiating tumor-adjacent from tumor-distant liver tissues.

Gallbladder carcinoma

Diffusion-weighted imaging

Tumor-adjacent liver tissue

Tumor-distant liver tissue

Gallbladder carcinoma (GBC) is the commonest and most aggressive primary malignant tumor in the biliary system [1–3]. The prognosis of GBC is extremely poor, with a 5-year survival rate of about 5% and a median survival time of 3–6 months [4–6]. Previous research has shown that histological type, vascular invasion, histological stage and grade are highly correlated with the prognosis of gallbladder cancer [4, 6, 7]. Unfortunately, GBC does not exhibit typical clinical signs and symptoms at early stage, and there are no reliable diagnostic tumor markers. Therefore, GBC is often diagnosed as suffering from terminal and unresectable stages, or the presence of GBC is not even detected until postmortem autopsy [1, 8]. Despite the postoperative prognosis of GBC is poor due to high recurrence and mortality, surgical resection is still the only chance for cure [9–11]. Clinically, an extended cholecystectomy is generally recommended for patients with GBC at stage T₂ or above, and involves cholecystectomy and a limited hepatic resection (typically segments IVb and V) [12], or a wedge resection of the gallbladder fossa within 2 cm [13] or of 2–3 cm non-neoplastic liver tissue [14], which encompass the tumor to ensure negative margins and reduce the possible of recurrence. However, the performance of a empiric extended hepatectomy is not independently associated with long-term survival, on the contrast, the overall perioperative morbidity is significantly increased [9].

In addition, previous publications reported that tumor recurrence is closely associated with field cancerization [14–16] and researchers have found field cancerization in the “macroscopically normal” tissues adjacent to tumor and in the resection margins of head and neck cancer [16–19]. If the tissue showing pre-neoplastic state is not resected by surgery, it may develop into invasive cancer [20]. In consideration of the coverage of previous extended hepatectomy and possible field cancerization in GBC, it is of great significance to differentiate tumor-adjacent and tumor-distant liver tissues before surgery for surgical strategies decision making.

Magnetic resonance imaging (MRI) improves the assessment of disease because of the advantages of superior soft tissue resolution and excellent anatomic image quality. Diffusion-weighted imaging (DWI), an established functional imaging technique, can provide functional information by depicting the different diffusion of water in biological tissue and can be quantitated by apparent diffusion coefficient (ADC) [21, 22]. ADC indicates that the degree of limitation of water molecule diffusion in biological tissue is inversely proportional to the cellularity of tissue and the integrity of cell membrane. Marques et al. reported that the quantity and mobility of water molecules increase during the normal cells transition to cancer cells [23], which will produce differences in the extent of water molecules limited diffusion in tissues at different carcinogenic stages. Previous studies have shown the reasonable accuracy of DWI in the detection of GBC and differentiating GBC from benign gallbladder disease [24, 25]. To the best of our knowledge, there are no publications focusing on the utility of DWI to differentiate the tumor-adjacent from tumor-distant liver tissues in GBC. Therefore, the purpose of our study is to evaluate feasibility of ADC and its standard deviation (ADC_sd) to differentiate between tumor-adjacent and tumor-distant liver tissues in resectable GBC.

Patients

The institutional review board of our hospital approved this retrospective study, and the requirements for informed consent were waived.

Through a computerized search of histopathology and MRI reports from July 2016 to April 2021, 44 patients with T₃ staged GBC who underwent cholecystectomy and a limited hepatic resection together with MRI scans including DWI before operation were identified. The inclusion criteria were as follows: (a) all the patients were diagnosed as T₃ staged GBC by pathology, (b) the tumor was considered resectable and there were no contraindications to surgery, and (c) patients desired to undergo surgical resection. The exclusion criteria were as follows: (a) the patient was treated with preoperative neoadjuvant radiation therapy and/or chemotherapy during the interval between MRI and surgery (n = 1), or (b) the MRI were of poor quality (n = 2). Ultimately, 41 patients were enrolled into our study.

According to the operative observations, the patients were divided into two sets: with (n = 27) and without (n = 14) liver invasion. GBC was confirmed by postoperative pathology, and the surgical cut margins of all resected specimens were not involved by the tumor (R₀). The T stage, N stage and differentiation of GBC were decided according to the eighth edition guidelines of American Joint Committee on Cancer Stage criteria [26] as well as the postoperative histopathologic examination.

MRI Technique

MRI scanning was performed on a 3.0-T scanner (Discovery MR 750, GE Medical Systems, Milwaukee, WI, USA) with a 32-channel torso phase-array coil with cardiac gating and breath gating. All patients were fasted at least 6 hours to weaken the peristalsis of the intestinal tract, and received breathing training before scanning to reduce the respiratory artifacts. Patients were examined in the supine position.

The scanning sequences included the following: DWI at multiple b-values of 0, 20, 50, 80, 100, 200, 400, 600, 800 and 1000 sec/mm²; axial liver acquisition with volume acceleration (LAVA)-Flex sequence; axial fast recovery fast spin-echo T₂-weighted imaging (T₂WI) with fat saturation; single-shot fast spin-echo radial series slab magnetic resonance cholangiopancreatography, and dynamic contrast-enhanced three-dimensional LAVA-flex with fat saturation images. The scanning parameters of the above sequences are listed in Table 1. Three-dimensional LAVA dynamic enhancement was performed with 20 ml of gadopentetate dimeglumine (Magnevist; Bayer AG, Leverkusen, Germany) administered intravenously at 2–3 ml/s, followed by a 20 ml saline solution flush. Dynamic enhancement was performed at 20–40 seconds (arterial phase), 45–65 seconds (portal-venous phase), and 3–5 minutes (equilibrium phase) after the injection of contrast-medium, respectively. The MRI data were transferred to our work station for subsequent analysis.

Table 1

MRI sequences and the corresponding scanning parameters
Parameters	DWI	LAVA-Flex	FS T₂WI	MRCP	LAVA-Flex C+
Plane	axial	axial	axial	coronal	axial
TE (msec)	80	2.3	95	900	2.2
TR (msec)	6000	3.9	2609	4134	3.7
FOV (mm)	340×340	320×272	340×340	340×340	320×272
Section thickness (mm)	6	4	5	50	4
Section gap (mm)	1	0	1	0	0
Matrix size	160×192	260×192	384×384	384×256	224×192
Acquisition time	5 min	15 sec	4 min 16 sec	51 sec	31 sec
Flip angle (deg)	90	12	110	12	12
No. of sections	260	38	31	12	38
NEX	1/2/4	1	2	1	1
Notes: MRI, magnetic resonance imaging; DWI, diffusion-weighted imaging; LAVA-Flex, liver acquisition with volume acceleration flexible; FS T₂WI, T₂ weighted fat-suppressed sequence; MRCP, Magnetic Resonance Cholangiopancreatography; C+, contrast enhanced; TE, echo time; TR, repetition time; FOV, field of view; and NEX, number of excitations. DWI has been performed with b-values of 0, 20, 50, 80, 100, 200, 400, 600 ,800, and 1000 sec/mm2, and the corresponding numbers of excitation are 1, 4, 4, 1, 1, 2, 2, 2, 1, and 4, respectively.

Measurement of ADC and ADC_sd

ADC and its standard deviation (ADC_sd) of tumor-adjacent and tumor-distant liver tissues in GBC were measured on DWI at multiple b-values of 0, 20, 50, 80, 100, 200, 400, 600, 800 and 1000 sec/mm². The tumor-adjacent and tumor-distant liver tissues in GBC were defined as follows. Clinically, an extended cholecystectomy is generally recommended for patients with GBC at stage T₂ or above, and involves cholecystectomy and a limited hepatic resection (typically segments IVb and V), or a wedge resection of the gallbladder fossa within 2 cm or of 2–3 cm non-neoplastic liver tissue which encompass the tumor to ensure negative margins and reduce the possible of recurrence [12–14]. Based on the resection range of liver tissue in GBC, tumor-adjacent and tumor-distant liver tissues were defined on MRI as follows in order to accurately measure the ADC and ADC_sd values of the previous two tissues: the liver tissues about 2 cm range of segments IVb and V away from the margin of the tumor was defined as tumor-adjacent liver tissue; and the rest of the segments IVb and V was defined as tumor-distant liver tissue.

The MRI data were reviewed on the work station (GE Advantage Workstation Version 4.4–09, Sun Microsystems, Palo Alto, CA, USA). The ADC values together with the ADC maps of tumor-adjacent and tumor-distant liver tissues in GBC were calculated from DWI with nine sets of b-values (including 0 and 20, 0 and 50, 0 and 80, 0 and 100, 0 and 200, 0 and 400, 0 and 600, 0 and 800, 0 and 1000 sec/mm²) by using a mono-exponential model with the logarithmic equation ADC = ln(S₂/S₁)/(b₁ − b₂), where b is the degree of diffusion sensitization, S₁ and S₂ are the tissue signal intensity on DWI at corresponding b₁ and b₂, respectively [27]. Two radiologists (radiologist 1 with 10 years of experience in radiology, and radiologist 2 with 24 years of experience in digestive radiology) independently evaluated the ADC and ADC_sd of the tumor-adjacent and tumor-distant liver tissues on DWI.

The measurement method of ADC and ADC_sd of the tumor-adjacent and tumor-distant liver tissues were as follows. Firstly, the tumor was presented as intraluminal mass or irregular thickening of gallbladder wall, and significant high signal intensity on DWI due to the remarkable restricted Brownian motion of water molecules within the tumor, and slightly hyperintensity or hyperintensity on contrast–enhanced images. The location of GBC was identified on DWI and axial contrast–enhanced images, and then the axial DWI of tumor-adjacent and tumor-distant liver tissues were also determined. Subsequently, the region of interests (ROIs) of the tumor-adjacent and tumor-distant liver tissues were drawn on the corresponding axial DWI by using the FuncTool software, which automatically copied and pasted ROIs to ADC maps and worked out the corresponding mean ADC and mean ADC_sd values (Fig. 1).

In particular, the ROIs of tumor-adjacent and tumor-distant liver tissues were continuously outlined slice by slice until the whole corresponding liver tissues were covered on DWI, and the ultimate ADC and ADC_sd values were calculated by average ADC and ADC_sd of all slices, respectively. When outlining the ROIs on DWI, carefully avoid the vessels, cystic or necrotic, hemorrhagic and adipose areas that identified by the two observers in consensus on T₂WI. In order to avoid review bias, the two radiologists were blinded to the histopathological results when they were measuring the ADC and ADC_sd values.

The results independently from the previous two radiologists were used to evaluate the inter-observer reliability. In addition, radiologist 1 repeated the previous measurements after a 2-week interval to evaluate the intra-observer reliability.

Statistics Analysis

The data were statistically analyzed by using SPSS software (version 23 for windows; IBM, Armonk, New York, NY, USA) and MedCalc Statistical Software (version 20.111, MedCalc Software Ltd, Ostend, Belgium).

The intra- and inter-observer agreements of measuring ADC values and ADC_sd values by the two observers were evaluated with intraclass correlation coefficient (ICC). ICC greater than 0.80 could be portended of superior reliability [28]. If good agreements were confirmed between the two observers, values of the average measurement by the two measurements from observer 1 were considered as the final results for further analyses.

Before the statistical analysis, Levene’s test and Shapiro–Wilk test were used for equality of variance and normal distribution, respectively. The Mann-Whitney U test was conducted to compare mean ADC and ADC_sd values between tumor-adjacent and tumor-distant liver tissues. Receiver operating characteristic (ROC) analysis was performed to determine whether and how to use the mean ADC and ADC_sd to differentiate between tumor-adjacent and tumor-distant liver tissues. All statistics are considered statistically significant when p-value is less than 0.05.

Characteristics of Our Cohort

In our cohort, all cases were adenocarcinoma as shown on postoperative pathology, and the histologic differentiation appeared as poorly differentiated (n = 7), moderately differentiated (n = 13), and well-differentiated (n = 21). As for the TNM stage, the tumor staging consisted of pT₃ (n = 41). The nodal staging included pN₀ (n = 15), pN₁ (n = 22) and pN₂ (n = 4). The detailed results are shown in Table 2.

Table 2

Clinical characteristics of 41 patients with T3 staged gall-bladder carcinoma
Variable		With liver invasion (n = 27)	Without liver invasion (n = 14)	P
Age (years ± SD)		66.04 ± 11.82	62.29 ± 11.57	0.016
Sex				0.318
	Male	16 (59%)	6 (42%)
	Female	11 (41%)	8 (58%)
N stage				0.067
	N₀	7 (26%)	8 (57%)
	N₁	18 (67%)	4 (29%)
	N₂	2 (7%)	2 (14%)
Differentiation				0.038
	Well	14 (51%)	7 (50%)
	Moderately	8 (30%)	5 (36%)
	Poorly	5 (19%)	2 (14%)
Note: SD, standard deviation.

Intra- and Inter-observer Variability of Mean ADC and ADC_sd Measurements

There were excellent agreements in terms of ADC and ADC_sd values derived from DWI at all b-values between the two independent observers, and between the two measurements by observer 1, with all ICCs exceeded 0.85 (Table 3). The averaged ADC and ADC_sd values of the two measurements by observer 1 were regarded as the final results for the subsequent analysis.

Table 3

Median (95%CI) of ICC of ADC and ADCsd measurements of tumor-adjacent and tumor-distant liver tissues
b-value (sec/mm²)		Tumor-adjacent liver tissue		Tumor-distant liver tissue
b-value (sec/mm²)		ADC	ADC_sd	ADC	ADC_sd
Inter-observer
	b = 0, 20	0.991 (0.914,0.997)	0.960 (0.703,0.987)	0.991 (0.831,0.997)	0.968 (0.618,0.991)
	b = 0, 50	0.992 (0.957,0.997)	0.959 (0.923,0.978)	0.994 (0.948,0.998)	0.980 (0.913,0.993)
	b = 0, 80	0.989 (0.938,0.996)	0.985 (0.894,0.995)	0.991 (0.944,0.997)	0.986 (0.920,0.995)
	b = 0, 100	0.989 (0.820,0.997)	0.983 (0.886,0.994)	0.991 (0.812,0.998)	0.994 (0.941,0.998)
	b = 0, 200	0.983 (0.854,0.995)	0.908 (0.759,0.958)	0.985 (0.877,0.995)	0.974 (0.873,0.991)
	b = 0, 400	0.967 (0.783,0.989)	0.965 (0.850,0.987)	0.955 (0.777,0.984)	0.962 (0.797,0.986)
	b = 0, 600	0.972 (0.681,0.992)	0.975 (0.880,0.991)	0.971 (0.827,0.990)	0.982 (0.945,0.992)
	b = 0, 800	0.956 (0.843,0.982)	0.974 (0.941,0.988)	0.921 (0.844,0.959)	0.944 (0.887,0.971)
	b = 0, 1000	0.961 (0.414,0.986)	0.856 (0.719,0.925)	0.966 (0.872,0.986)	0.981 (0.950,0.992)
Intra-observer
	b = 0, 20	0.998 (0.996,0.999)	0.992 (0.984,0.996)	0.998 (0.997,0.999)	0.996 (0.993,0.998)
	b = 0, 50	0.998 (0.997,0.999)	0.985 (0.973,0.992)	0.998 (0.996,0.999)	0.978 (0.960,0.988)
	b = 0, 80	0.997 (0.994,0.998)	0.995 (0.990,0.997)	0.998 (0.996,0.999)	0.994 (0.989,0.997)
	b = 0, 100	0.998 (0.996,0.999)	0.994 (0.989,0.997)	0.996 (0.992,0.998)	0.992 (0.986,0.996)
	b = 0, 200	0.997 (0.994,0.998)	0.983 (0.968,0.991)	0.997 (0.995,0.999)	0.996 (0.992,0.998)
	b = 0, 400	0.998 (0.996,0.999)	0.993 (0.987,0.996)	0.968 (0.940,0.983)	0.994 (0.989,0.997)
	b = 0, 600	0.994 (0.989,0.997)	0.983 (0.969,0.991)	0.994 (0.989,0.997)	0.995 (0.990,0.997)
	b = 0, 800	0.985 (0.972,0.992)	0.986 (0.973,0.992)	0.992 (0.984,0.996)	0.986 (0.973,0992)
	b = 0, 1000	0.991 (0.983,0.995)	0.937 (0.885,0.966)	0.987 (0.975,0.993)	0.974 (0.953,0.986)
Notes: ICC, interclass correlation coefficient; CI, confidence interval; and ADC, apparent diffusion coefficient.

Mean ADC and ADC_sd Values at Different b-values: Tumor-adjacent vs. Tumor-distant Liver Tissues

Mean ADC and ADC_sd of tumor-adjacent and tumor-distant liver tissues calculated at different b-values are summarized in Table 4. In the set with liver invasion by GBC, the mean ADC and ADC_sd of tumor-adjacent liver tissue were lower than those of tumor-distant liver tissue when b = 0, 800 sec/mm², and when b = 0, 1000 sec/mm² (all P-values < 0.05). The ADC and ADC_sd values at b-values of 0, 20, 50, 80, 100, 200, 400 and 600 sec/mm² did not show a significant difference between tumor-adjacent and tumor-distant liver tissues (all P-values > 0.05). In the set without liver invasion, there were no statistical differences in the ADC and ADC_sd at all b-values between tumor-adjacent and tumor-distant liver tissues (all P-values > 0.05).

Table 4

Median (Quartile) of mean ADC and ADCsd (×10− 3 mm2/sec) of tumor-adjacent and tumor-distant liver tissues
		ADC					ADC_sd
b-value (sec/mm²)		Tumor-adjacent liver tissue		Tumor-distant liver tissue		p	Tumor-adjacent liver tissue	Tumor-distant liver tissue	p
The set with liver invasion
	b = 0, 20		11.453 (8.615,14.425)		11.060 (9.011,15.720)	0.598	8.163 (6.572,9.360)	8.180 (6.335,9.761)	0.684
	b = 0, 50		5.346 (4.068,8.481)		6.100 (4.624,7.764)	0.592	3.989 (3.080,5.738)	4.567 (3.234,5.670)	0.359
	b = 0, 80		4.265 (3.663,6.013)		4.558 (3.715,6.408)	0.378	3.112 (2.428,4.018)	3.940 (2.460,4.869)	0.257
	b = 0, 100		3.707 (2.873,5.167)		3.907 (3.223,5.475)	0.411	2.460 (1.923,3.627)	3.202 (2.156,3.853)	0.312
	b = 0, 200		2.258 (2.058,2.860)		2.409 (2.101,3.093)	0.257	1.335 (1.037,1.933)	1.637 (1.260,1.798)	0.261
	b = 0, 400		1.586 (1.348,1.779)		1.684 (1.576,1.997)	0.099	0.783 (0.645,1.010)	0.954 (0.710,1.084)	0.197
	b = 0, 600		1.358 (1.290,1.607)		1.506 (1.360,1.685)	0.110	0.610 (0.497,0.767)	0.724 (0.599,0.908)	0.050
	b = 0, 800		1.272 (1.122,1.393)		1.408 (1.270,1.614)	0.018^*	0.557 (0.472,0.636)	0.629 (0.548,0.737)	0.029^*
	b = 0, 1000		1.160 (0.987,1.250)		1.250 (1.163,1.380)	0.023^*	0.473 (0.397,0.516)	0.560 (0.465,0.641)	0.004^*
The set without liver invasion
	b = 0, 20		13.200 (11.272,15.846)		12.788 (10.232,16.246)	0.818	8.318 (7.576,9.195)	8.407 (6.749,9.723)	0.818
	b = 0, 50		6.013 (5.186,8.009)		6.315 (5.173,8.038)	0.783	3.971 3.699,4.844)	4.615 (3.618,5.513)	0.646
	b = 0, 80		4.945 (4.250,6.660)		5.447 (3.940,6.593))	0.927	3.554 (2.891,4.251)	3.581 (2.695,4.722)	0.854
	b = 0, 100		4.354 (3.199,5.579)		4.414 (3.228,5.106)	0.963	2.807 (2.312,3.494)	2.787 (2.099,3.364)	0.927
	b = 0, 200		2.334 (1.802,3.216)		2.305 (1.902,3.291)	0.748	1.376 (1.204,1.659)	1.480 (1.041,1.751)	1.000
	b = 0, 400		1.560 (1.273,2.193)		1.603 (1.506,2.053)	0.491	0.816 (0.649,0.960)	0.844 (0.638,0.987)	0.748
	b = 0, 600		1.261 (1.110,1.538)		1.479 (1.215,1.647)	0.370	0.656 (0.504,0.674)	0.648 (0.496,0.743)	0.679
	b = 0, 800		1.145 (1.065,1.403)		1.318 (1.093,1.542)	0.408	0.543 (0.442,0.594)	0.597 (0.454,0.638)	0.250
	b = 0, 1000		1.012 (0.873,1.336)		1.020 (0.938,1.370)	0.800	0.419 (0.372,0.504)	0.485 (0.372,0.552)	0.358
Notes: ADC, apparent diffusion coefficient; ADC_sd, standard deviation of ADC; and *, P-value < 0.05.

ROC Analysis of ADC and ADC _sd values to Differentiate Tumor-adjacent from Tumor-distant Liver Tissue in GBC with Liver Invasion

The cut-off value, area under the ROC curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy of the ADC and ADC_sd for differentiating the tumor-adjacent from tumor-distant liver tissues in the set with liver invasion are summarized in Table 5. When measured on DWI at b-values of 800 sec/mm², the ADC and ADC_sd obtained the AUCs of 0.687 and 0.673, respectively. When measured on DWI at b-values of 1000 sec/mm², the ADC and ADC_sd obtained the AUCs of 0.680 and 0.731, respectively.

Table 5

ROC analysis to differentiate tumor-adjacent from tumor-distant liver tissues when gall-bladder carcinoma invades liver
b-value (sec/mm²)		Cut-off value (×10^− 3 mm²/sec)	AUC	Sensitivity (%)	Specificity (%)	PPV (%)	NPV (%)	Accuracy (%)
ADC
	b = 0, 800	1.397	0.687	81.5	51.9	62.9	73.9	66.7
	b = 0, 1000	1.245	0.680	74.1	59.3	64.5	69.6	66.7
ADC_sd
	b = 0, 800	0.502	0.673	44.4	85.2	75.0	60.5	64.8
	b = 0, 1000	0.504	0.731	70.4	70.4	70.4	70.4	70.4
Notes: ROC, receiver operating characteristic; ADC, apparent diffusion coefficient; ADC_sd, standard deviation of ADC; AUC, area under the receiver operating characteristic curve; PPV, positive predictive value; and NPV, negative predictive value.

In this study, we measured the mean ADC and ADC_sd of tumor-adjacent and tumor-distant liver tissues in T₃ staged GBC on DWI at multiple b values (b-values of 0, 20, 50, 80, 100, 200, 400, 600, 800 and 1000 sec/mm²) to explore whether there could be a difference in the diffusion motion of water molecules between the previous two liver tissues, and to determine feasibility of ADC and ADC_sd to differentiate between tumor-adjacent and tumor-distant liver tissues.

Our results showed that in GBC patients with liver invasion, the mean ADC and ADC_sd of the tumor-adjacent liver tissue were significantly lower than those of the tumor-distant liver tissue when b = 0, 800 sec/mm², and when b = 0, 1000 sec/mm². Our findings may be explained by the fact that the molecular variation during the carcinogenesis process could lead to abnormal changes of the number and the mobility of water molecules [23], and we could presume that the molecular variation might occur in the field cancerization. Moreover, we could also presume that cancerization could be more remarkable in the tumor-adjacent liver tissue than in the tumor-distant liver tissue in GBC, giving rise to the restricted diffusion of water molecules more notable in the tumor-adjacent liver tissue in comparison with the corresponding tumor-distant liver tissue.

In addition, in GBC patients without liver invasion, the results of this study demonstrated that there could be no significant difference in ADC and ADC_sd at all b values between tumor-adjacent and tumor-distant liver tissues, indicating that there could be similar tissue characteristics in the above two tissues, including tissue composition, cell sequencing and water molecular diffusion. It is therefore reasonable to speculate that in patients without liver invasion, tumor-adjacent and tumor-distant liver tissues might belong to normal liver tissue and had less possibility of cancerization.

Our study showed that in case of b-values greater than or equal to 800 sec/mm², the mean ADC and ADC_sd have certain diagnostic efficacy in differentiating tumor-adjacent from tumor-distant liver tissues in GBC patients with the liver invasion. When b = 0, 800 sec/mm², the AUCs were more than 0.67 with the best sensitivity of 81.5% for ADC cutoff of 1.397×10^− 3 mm²/sec, and the best specificity of 85.2% for ADC_sd cutoff of 0.502×10^− 3 mm²/sec. When b = 0, 1000 sec/mm², the AUCs were more than 0.68 with the best sensitivity of 74.1% for ADC cutoff of 1.245×10^− 3 mm²/sec, and the best specificity of 70.4% for ADC_sd cutoff of 0.504×10^− 3 mm²/sec. On the contrary, the mean ADC and ADC_sd could not differentiating tumor-adjacent from tumor-distant liver tissues in GBC with liver invasion when b-values were not more than 600 sec/mm², which could be explained that the low b-values are greatly influenced by micro-perfusion and cannot reflect the weak difference of diffusion limitation between tumor-adjacent and tumor-distant tissues [29]. Furthermore, high number of b-values may give rise to a long scan time (overall 10 b-values took a total of 5 minutes in this study), or increase the thick slices to shorten the scan time, which may results in limited spatial resolution in the z-axis. Therefore, we recommend reducing the number of low b-values in clinical practice to abridge the scan time and reduce the influence of micro-perfusion on the ADC and ADC_sd. We believe that our findings might be clinically important for the extended cholecystectomy decision making. In detail, it may be necessary to expand the scope of resection to some degree to reduce postoperative recurrence in GBC patients with liver invasion, and to reduce the resection range of liver tissue in patients without liver invasion to decrease the incidence of postoperative hepatic insufficiency.

There are several limitations in our study. Firstly, it was a retrospective analysis. Secondly, our MRI data were obtained from a single center. Therefore, our results need to be verified in other medical centers. Thirdly, there was actually no direct histopathologically verified correlation between MR findings and histopathology of tumor-adjacent liver tissues.

Our study revealed that ADC and ADC_sd when b = 0, 800 sec/mm² and b = 0, 1000 sec/mm² could have potential value in differentiating tumor-adjacent from tumor-distant liver tissues in GBC with liver invasion, while there could be no differences in ADC and ADC_sd between tumor-adjacent and tumor-distant liver tissues at all b values in GBC without liver invasion. We hope that our findings could be helpful for the preoperative assessment of surgical procedure decision making.

ADC

Apparent diffusion coefficients

ADC_sd

Standard deviation of apparent diffusion coefficients

GBC

Gallbladder carcinoma

DWI

Diffusion-weighted imaging

AUC

Area under the receiver operating characteristic curve

MRI

Magnetic resonance imaging

ROIs

Region of interests

ICC

Intraclass correlation coefficient

ROC

Receiver operating characteristic

Ethics approval and consent to participate: This retrospective study was approved by the institutional review board of the Affiliated Hospital of North Sichuan Medical College. Due to a retrospective study, the patient consent for the study was waived.

Consent for publication: All authors gave consent for the publication of this paper.

Availability of data and materials: Please contact the corresponding author for data and materials requests.

Competing interests: All authors declare that they have no competing interests in this study.

Funding: Not applicable

Authors' contributions: TWC, XMZ, HYZ and TZ proposed the study. ZT, YPW, BGT, AML, WWG, CQZ, MX and KSW performed research, and collected the data. ZT and TWC analyzed the data and wrote the first draft. All authors contributed to the interpretation of the study and to further drafts. All authors read and approved the final manuscript. TWC is the guarantor. ZT and YPW contribute equally to this work.

Acknowledgements: Not applicable

Reid KM, Ramos-De la Medina A, Donohue JH. Diagnosis and surgical management of gallbladder cancer: a review. J Gastrointest Surg. 2007;11(5):671–81.
Lee NK, Kim S, Moon JI, Shin N, Kim DU, Seo HI, et al. Diffusion-weighted magnetic resonance imaging of gallbladder adenocarcinoma: analysis with emphasis on histologic grade. Clin Imaging. 2016;40(3):345–51.
Kanthan R, Senger JL, Ahmed S, Kanthan SC. Gallbladder Cancer in the 21st Century. J Oncol. 2015; 2015:967472.
Bartlett DL. Gallbladder cancer. Semin Surg Oncol. 2000;19(2):145–55.
Cubertafond P, Gainant A, Cucchiaro G. Surgical treatment of 724 carcinomas of the gallbladder. Results of the French Surgical Association Survey. Ann Surg. 1994;219(3):275–80.
Lim H, Seo DW, Park DH, Lee SS, Lee SK, Kim MH, et al. Prognostic factors in patients with gallbladder cancer after surgical resection: analysis of 279 operated patients. J Clin Gastroenterol. 2013;47(5):443–8.
Henson DE, Albores-Saavedra J, Corle D. Carcinoma of the gallbladder. Histologic types, stage of disease, grade, and survival rates. Cancer. 1992;70(6):1493–7.
Yang XW, Yang J, Li L, Man XB, Zhang BH, Shen F, et al. Analysis of the relationships between clinicopathologic factors and survival in gallbladder cancer following surgical resection with curative intent. PLoS ONE. 2012;7(12):e51513.
D'Angelica M, Dalal KM, DeMatteo RP, Fong Y, Blumgart LH, Jarnagin WR. Analysis of the extent of resection for adenocarcinoma of the gallbladder. Ann Surg Oncol. 2009;16(4):806–16.
Hueman MT, Vollmer CM Jr, Pawlik TM. Evolving treatment strategies for gallbladder cancer. Ann Surg Oncol. 2009;16(8):2101–15.
Reddy SK, Marroquin CE, Kuo PC, Pappas TN, Clary BM. Extended hepatic resection for gallbladder cancer. Am J Surg. 2007;194(3):355–61.
Benson AB, D'Angelica MI, Abbott DE, Anaya DA, Anders R, Are C, et al. Hepatobiliary Cancers, Version 2.2021, NCCN Clinical Practice Guidelines in Oncology. J Natl Compr Canc Netw. 2021;19(5):541–65.
He XD, Li JJ, Liu W, Qu Q, Hong T, Xu XQ, et al. Surgical procedure determination based on tumor-node-metastasis staging of gallbladder cancer. World J Gastroenterol. 2015;21(15):4620–6.
Lee SE, Kim KS, Kim WB, Kim IG, Nah YW, Ryu DH, et al. Practical guidelines for the surgical treatment of gallbladder cancer. J Korean Med Sci. 2014;29(10):1333–40.
Braakhuis BJ, Tabor MP, Kummer JA, Leemans CR, Brakenhoff RH. A genetic explanation of Slaughter's concept of field cancerization: evidence and clinical implications. Cancer Res. 2003;63(8):1727–30.
Tabor MP, Brakenhoff RH, Ruijter-Schippers HJ, Kummer JA, Leemans CR, Braakhuis BJ. Genetically altered fields as origin of locally recurrent head and neck cancer: a retrospective study. Clin Cancer Res. 2004;10(11):3607–13.
Chen LC, Hao CY, Chiu YS, Wong P, Melnick JS, Brotman M, et al. Alteration of gene expression in normal-appearing colon mucosa of APC(min) mice and human cancer patients. Cancer Res. 2004;64(10):3694–700.
Guo H, Zeng W, Feng L, Yu X, Li P, Zhang K, et al. Integrated transcriptomic analysis of distance-related field cancerization in rectal cancer patients. Oncotarget. 2017;8(37):61107–17.
Hawthorn L, Lan L, Mojica W. Evidence for field effect cancerization in colorectal cancer. Genomics. 2014;103(2–3):211–21.
Spitzwieser M, Entfellner E, Werner B, Pulverer W, Pfeiler G, Hacker S, et al. Hypermethylation of CDKN2A exon 2 in tumor, tumor-adjacent and tumor-distant tissues from breast cancer patients. BMC Cancer. 2017;17(1):260.
Wang H, Cheng L, Zhang X, Wang D, Guo A, Gao Y, et al. Renal cell carcinoma: diffusion-weighted MR imaging for subtype differentiation at 3.0 T. Radiology. 2010;257(1):135–43.
Paudyal B, Paudyal P, Tsushima Y, Oriuchi N, Amanuma M, Miyazaki M, et al. The role of the ADC value in the characterisation of renal carcinoma by diffusion-weighted MRI. Br J Radiol. 2010;83(988):336–43.
Marques MPM, Batista de Carvalho ALM, Mamede AP, Dopplapudi A, García Sakai V, Batista de Carvalho LAE. Role of intracellular water in the normal-to-cancer transition in human cells-insights from quasi-elastic neutron scattering. Struct Dyn. 2020;7(5):054701.
Kang TW, Kim SH, Park HJ, Lim S, Jang KM, Choi D, et al. Differentiating xanthogranulomatous cholecystitis from wall-thickening type of gallbladder cancer: added value of diffusion-weighted MRI. Clin Radiol. 2013;68(10):992–1001.
Kaza RK, Gulati M, Wig JD, Chawla YK. Evaluation of gall bladder carcinoma with dynamic magnetic resonance imaging and magnetic resonance cholangiopancreatography. Australas Radiol. 2006;50(3):212–7.
Amin MB, Edge S, Greene F, Byrd DR, Brookland RK, Washington MK et al, editors. AJCC Cancer Staging Manual (8th edition). Springer International Publishing: American Joint Commission on Cancer; 2017.
Barral M, Taouli B, Guiu B, Koh DM, Luciani A, Manfredi R, et al. Diffusion-weighted MR imaging of the pancreas: current status and recommendations. Radiology. 2015;274(1):45–63.
Hemminki K, Li X. Familial liver and gall bladder cancer: a nationwide epidemiological study from Sweden. Gut. 2003;52(4):592–6.
Thoeny HC, De Keyzer F. Diffusion-weighted MR imaging of native and transplanted kidneys. Radiology. 2011;259(1):25–38.

Download PDF

Version 1

posted

You are reading this latest preprint version

Apparent Diffusion Coefficient and Its Standard Deviation from Diffusion-weighted Imaging to Differentiate Tumor-adjacent and Tumor-distant liver tissues in T3 staged Resectable Gallbladder Carcinoma

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Background

Methods

Patients

MRI Technique

Measurement of ADC and ADC_sd

Statistics Analysis

Results

Characteristics of Our Cohort

Intra- and Inter-observer Variability of Mean ADC and ADC_sd Measurements

Mean ADC and ADC_sd Values at Different b-values: Tumor-adjacent vs. Tumor-distant Liver Tissues

Discussion

Conclusions

List Of Abbreviations

Declarations

References

Status:

Version 1

Apparent Diffusion Coefficient and Its Standard Deviation from Diffusion-weighted Imaging to Differentiate Tumor-adjacent and Tumor-distant liver tissues in T3 staged Resectable Gallbladder Carcinoma

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Background

Methods

Patients

MRI Technique

Measurement of ADC and ADCsd

Statistics Analysis

Results

Characteristics of Our Cohort

Intra- and Inter-observer Variability of Mean ADC and ADCsd Measurements

Mean ADC and ADCsd Values at Different b-values: Tumor-adjacent vs. Tumor-distant Liver Tissues

Discussion

Conclusions

List Of Abbreviations

Declarations

References

Status:

Version 1

Measurement of ADC and ADC_sd

Intra- and Inter-observer Variability of Mean ADC and ADC_sd Measurements

Mean ADC and ADC_sd Values at Different b-values: Tumor-adjacent vs. Tumor-distant Liver Tissues