ORIGINAL RESEARCH
Simultaneous Changes of Magnetic Resonance Diffusion-Weighted Imaging and Pathological Microstructure in Locally Advanced Cervical Cancer Caused by Neoadjuvant Chemotherapy Chun Fu, MD,1* Xiaoyan Feng, MS,1 Dujun Bian, MD,2 Yan Zhao, MS,1 Xiaoling Fang, MD,1 Wanping Du, BS,2 Lan Wang, BS,3 and Xiangquan Wang, MS1 Purpose: To investigate the changes to diffusion-weighted imaging (DWI) correlated with histopathology after neoadjuvant chemotherapy (NACT) in patients with locally advanced cervical cancer (LACC). Materials and Methods: Thirty-three patients with LACC were examined with 3T magnetic resonance imaging (MRI) with DWI and apparent diffusion coefficient (ADC) maps. MRIs were performed for each patient at three timepoints: before the first NACT, 2 weeks after the first NACT, and 2 weeks after the second NACT. Uterine cervical specimens were collected at the same timepoints. Specimens were stained for tumor cell density, proliferating cell nuclear antigen (PCNA), and aquaporin 1 (AQP1). Treatment responses were classified as the effective group (complete and partial response) and the ineffective group (stable and progressive disease). Results: The ADC value of the effective group after the first chemotherapy was higher than that before chemotherapy (P 5 0.002), and expressions of three pathological indicators (tumor cell density, PCNA, and AQP1) significantly decreased after the first NACT compared with those prechemotherapy (P < 0.001). Changes of PCNA expression were negatively correlated with changes of ADC values after the first NACT in the effective group (r 5 –0.56, P 5 0.03). Changes of cellular density were negatively correlated with changes of ADC values from the time of prechemotherapy to after the second NACT in the effective group (r 5 –0.51, P 5 0.04). Conclusion: The ADC change after successful chemotherapy is closely related with cellular characteristics preceding size reduction. ADC may be used as an early imaging biomarker of NACT response in LACC. J. MAGN. RESON. IMAGING 2014;00:000–000
C
ervical cancer is the second most common female cancer.1,2 A worse prognosis has been observed even in early cervical cancer with tumor diameter more than 4 cm or locally advanced cervical cancer (LACC). Neoadjuvant chemotherapy (NACT) followed by radical surgery is a valid alternative therapeutic option to concurrent chemotherapy and radiotherapy for patients with LACC.2–4 The achievement of
an optimal pathological response on surgical specimens is a strong predictor of a better clinical outcome.5 The ineffective treatment may increase operative difficulty and affect a patient’s prognosis. Therefore, evaluation of the early effect of NACT may play an important role in the treatment of LACC. Treatment response is mainly judged by tumor size reduction according to the Response Evaluation Criteria in
View this article online at wileyonlinelibrary.com. DOI: 10.1002/jmri.24779 Received Feb 15, 2014, Accepted for publication Sep 29, 2014. *Address reprint requests to: C.F., Department of Gynecology and Obstetrics, Second Xiangya Hospital, Central South University, 139 Middle Renmin Rd., Changsha, Hunan 410011, People’s Republic of China. E-mail: [email protected] From the 1Department of Gynecology and Obstetrics, Second Xiangya Hospital, Central South University, Changsha, People’s Republic of China; Department of Radiology, Second Xiangya Hospital, Central South University, Changsha, People’s Republic of China; and 3Department of Research, Second Xiangya Hospital, Central South University, Changsha, People’s Republic of China.
2
Contract grant sponsor: National Science Foundation of Hunan Province of the People’s Republic of China; contract grant number: 10JJ6044. Contract grant sponsor: Hunan Science and Technology Department Project of the People’s Republic of China; contract grant numbers: 2009SK3161; 2011FJ3008.Contract grant sponsor: Scientific Research Foundation of Hunan Provincial Health Bureau of the People’s Republic of China; contract grant number: B2009022
C 2014 Wiley Periodicals, Inc. V 1
Journal of Magnetic Resonance Imaging
Solid Tumor (RECIST).6 Conventional magnetic resonance imaging (MRI) and computed tomography are insensitive to early changes after treatment.7 Magnetic resonance diffusion-weighted imaging (MR-DWI) acquisition measures the mobility of water in tissue, and can discriminate cervical cancer from normal cervix by calculating the apparent diffusion coefficient (ADC) value.8 MR-DWI has been used to detect response of cervical cancer to chemoradiation in nine clinical trials,7,9–16 and to assess the response of LACC to NACT in one study.17 The results show that ADC values can be a useful tool to monitor the response to therapy. The effective anticancer treatments (including chemoradiation and NACT) result in tumor cell lysis, loss of cell membrane integrity, increased extracellular space, and, therefore, an increase in water diffusion.9 Changes of water flow at the cellular level can reflect changes of the organizational structure, which may be the theoretical basis of DWI applied in cervical cancer. No studies show the relationship of simultaneous changes between pathological microstructure and DWI caused by NACT. In the study, we chose tumor cell density, proliferating cell nuclear antigen (PCNA), and aquaporin 1 (AQP1) as histological indicators. DWI images and pathological microstructure after NACT was observed at the same time. Our purposes were 1) to assess the relationship between simultaneous changes of ADC and pathological microstructure, and 2) to provide a pathological basis for evaluation of curative effect by using DWI in patients with LACC.
MATERIALS AND METHODS Patient Population The study was approved by the hospital Ethics Committee and written informed consent was obtained from all patients. Only data from patients with International Federation of Gynecology and Obstetrics (FIGO) stage IB2–IIB cervical cancer (confirmed histologically with a tumor diameter of >4 cm) were included in the analysis. Thirtythree patients with cervical squamous cell carcinoma (SCC) participated in this prospective study from May 2010 to May 2012. The mean age of patients was (43.24 6 6.11) years (range, 26–48 years). The mean diameter of tumor was (4.86 6 0.93) cm (range, 4–8 cm). There were 13 patients in stage IB2, 12 in stage IIA2, and 8 in stage IIB. Three cases of tumor grading were well differentiated (G1, low grade), 17 cases were moderately differentiated (G2, intermediate grade), 7 cases poorly differentiated (G3, high grade), and 6 cases moderately to poorly differentiated (G2–G3). The inclusion criteria of patients were: normal renal, hepatic, cardiac, pulmonary, and hematologic function; and no previous surgery for cancer, chemotherapy, or radiation therapy. Exclusion criterion was contraindicated for MRI, such as those with pacemakers, metal objects, or claustrophobia disorder. The checked times were nonmenstrual period.
Standard NACT and Response Evaluation All patients were scheduled for NACT. NACT was consisted of two cycles of intravenous docetaxel 75 mg/m2 and cisplatin 2
50 mg/m2 at 2-week intervals. Two weeks after the second NACT, 31 patients underwent type III radical hysterectomy with systematic pelvic lymphadenectomy, and two patients were treated with concurrent chemoradiotherapy. Postoperatively, additional chemotherapy or radiotherapy alone was administered to patients with one or more recurrence risk factors.3 The factors include lymph node metastasis, positive surgical margin, parametric infiltration, deep cervical stromal invasion, and lymphovascular invasion. All the patients were followed up. The mean time was (19.39 6 7.253) months. Tumor response relative to pretreatment data was classified clinically as one of four groups according to the RECIST.6 Complete response (CR) was concluded if there was no residual tumor on T2-weighted imaging (T2WI); partial response (PR) was concluded if the longest diameter of the tumor was less than 70% of the original size; the disease was determined to be stable if there was neither sufficient shrinkage to qualify for partial response nor sufficient increase to qualify for progressive disease; and progressive disease (PD) was concluded if there was at least a 20% increase and a 5-mm absolute increase in the sum of the longest diameter of the tumor, taking as reference the longest diameter recorded before treatment. Treatment responses were classified as the effective group (CR and PR) and the ineffective group (stable disease and PD).
MRI Scanning Protocol and Imaging Analysis Serial MRIs were performed for each patient at three timepoints: before the first NACT, 2 weeks after the first NACT, and 2 weeks after the second NACT. Each time each patient underwent both conventional MRI and DWI scans. All scans were performed on a 3T MR scanner (Philips Achieva 3.0T X-series, Philips Healthcare, Netherlands), using 16 elements phased array of SENSE XL Torso coil to cover the entire pelvis. The imaging protocol included: 1) SENSE T2-weighted fatsuppressed respiratory triggering MRI was obtained in the coronal, sagittal, and axial planes (T2WI, TR/TE 1462/70 msec; average, 3; flip angle, 90 ; matrix size, 308 3 212; field of view [FOV], 36 3 36 cm); 2) axial planes SENSE T1-weighted respiratory triggering MRI (T1WI, TR/TE 10/2.3 msec; average, 3; flip angle, 15 ; matrix, 252 3 194; FOV, 36 3 36 cm); 3) MR DWI obtained in the axial planes using a single-shot spin-echo echoplanar with chemical-shift-selective fat-suppression technique (TR/TE 952/53 msec; flip angle, 90 ; average number of signals, 4; section thickness, 5 mm; gap, 1.5 mm; matrix, 124 3 100; FOV, 36 3 36 cm); the acquisition time was 1 minute, 32 seconds; 4) contrast-enhanced turbo field echo (TFE) fat-suppressed T1weighted sequences were obtained in identical slice thicknesses and gaps in the axial, coronal, and sagittal planes to cover the entire true pelvis. All MRI data were transmitted to MR Workspace Release 2.6.3 Workstations and analyzed with the Functool package. DWI images were analyzed qualitatively by referring to the signal intensity of uterine cervical cancer, which was classified using visual assessment of hypointensity or hyperintensity in comparison with the adjacent skeletal muscle).7,18 ADC measurement was on the axial plane with mass in maximum diameter and using three regions of interest (ROIs). ROI (area size range 20–50 mm2) was Volume 00, No. 00
Fu et al.: Correlation With ADC in NACT of LACC
arranged in three different positions to avoid necrotic areas and the lesions of peripheral parts. Average value of 3 times was the final ADC value. ADC can be calculated from two measurements with different diffusion attenuations according to the following equation:
ADC 5 In ðS2=S1Þ=ðb12b2Þ Where b1 and b2 are the two applied diffusion-sensitizing factors, S1 and S2 are the DWI signal intensity, respectively, under the conditions of b1 and b2. b 5 r2G2d2 (D-d/3), where r is the gyromagnetic ratio, G, d, D separately represent the amplitude, width, and spacing of pulse gradient. The four b values were chosen to be 0, 300, 600, 900 s/mm2 to optimize the signal-to-noise ratio.17 We selected b-factors of 0 and 900 s/mm2 in the study.
Immunohistochemical Staining and Image Analysis Uterine cervical specimens were collected from every patient at three different times (prechemotherapy, 2 weeks after the first NACT, and postsurgery). Tissue samples were routinely fixed in 10% formalin solution, paraffin-embedded, and cut into 4-lmthick sections. After deparaffinizatiom and rehydration, the sections were heated in three 5-minute periods in a microwave oven at 100 C with sodium citrate buffer (10 mM; pH 6), cooled in the same buffer at room temperature, and subsequently incubated for 20 minutes with 3% hydrogen peroxide. The antibodies for PCNA and AQP1 were used at 1:200 and 1:150. The primary antibodies against PCNA and AQP1 were both purchased from GeneTex (Irvine, CA). The serial sections were incubated with primary antibodies in a humid chamber at 4 C overnight. Sections were washed three times in phosphate-buffered saline (PBS) and further incubated with a biotinylated secondary antibody for 30 minutes at room temperature. Streptavidin-horseradish peroxidase conjugate was added and the peroxidase activity was made visible with diaminobenzidine and counterstained with hematoxylin for 30 seconds. As a control experiment, we performed an identical immunohistochemical procedure with omission of the primary antibody. The overall structure, cell morphology, and nuclear atypia were first observed at low magnification. The representative areas were then selected and observed at high magnification (3200 or 400). Five microscopic fields were freely chosen for each tumor specimen using an Olympus BX51/52 electron microscope collection system and saved into the computer. Images were analyzed by the MIAS medical pathology image analysis system. Tumor cell density was the ratio of total area of tumor cell nucleus and statistical field area (hematoxylin and eosin [H&E]stained specimen under the light microscope, 3200). The mean of the five visual fields was considered the tumor cellular density, expressed in percentage. Tumor cells with brown nucleus were regarded as PCNA-positive. Analysis of PCNA-positive expression was performed using the area density. The area density was the ratio of total area of positive or negative target cells and total area of statistical field (unit/mm2). Cell membrane stained brownish yellow was considered an APQ1-positive cell. Analysis of APQ1positive expression was performed by the average gray value, and the unit was the mean integrated optical density (IOD). Two investigators (C.F. and X.F.) assessed the histological parameters and immunohistochemical staining. Whenever a Month 2014
discrepancy occurred, both investigators reexamined the slides to reach a consensus.
Statistical Analysis Statistical analysis was performed using SPSS v. 17.0 for Windows (Chicago, IL). Patients’ age, tumor size, and follow-up time in the effective chemotherapy group versus those in the ineffective group were compared with independent-samples t-test. The composition ratio of clinical stage and cell grade between the effective and the ineffective group were compared with Fisher’s Exact Test (4-fold table). One-way analysis of variance (ANOVA) was used to compare the ADC value and pathological results (cellular density, PCNA, and AQP1 expressions) at three timepoints (before chemotherapy, 2 weeks after the first NACT, and 2 weeks after the second NACT) in the two chemotherapy groups. The Shapiro–Wilk W test was used to test for data normality. A post-hoc test was performed with least-significant difference (LSD) in a significant variance analysis. Correlation analysis between ADC value and pathological indictors (tumor cell density, PCNA, and AQP1positive expression) were determined with bivariate correlation analysis. Pearson’s correlation was used to test normally distributed data. In the absence of a linear relationship, Spearman’s correlation was used to assess the significance between groups. P < 0.05 was considered statistically significant.
RESULTS Clinical Curative Effect In all, 33 patients finished two chemotherapy cycles and 31 cases had radical surgeries. There were 24 patients in the effective chemotherapy group (2 CR and 22 PR), and 9 cases in the ineffective group (2 PD and 7 stable disease). The effective rate was 72.7%. Two cases with PD received radical radiotherapy and had no surgery. Twenty-five patients received additional treatment after surgery for recurrence risk factors (18 cases in the effective group and 7 cases in the ineffective group). All patients were followed up and one patient in each group died. Clinical data showed no significant differences between the two groups (Table 1). The mean diameter of the tumor was significantly reduced after the second NACT in the effective group (P 5 0.001, after the second NACT vs. prechemotherapy; P 5 0.11, after the first NACT vs. prechemotherapy; Fig. 1). The mean diameter of tumor was not significantly changed in the ineffective group at three timepoints (P 5 0.84, after the second NACT vs. prechemotherapy; P 5 0.74, after the first NACT vs. prechemotherapy; Fig. 2). Simultaneous Changes of MRI Images and Pathological Microstructure DWI Image and ADC Value Before treatment, all cervical tumors were hyperintense to adjacent skeletal muscle on both T2WI and DWI images; on ADC maps, they all manifested as low-signal-intensity 3
Journal of Magnetic Resonance Imaging
TABLE 1. Clinical Data of Patients With LACC in Different Chemotherapy Response Groups
Parameter
Effective group (n 5 24)
Ineffective group (n 5 9)
P value
Age, mean (SD), yrs
45.46 (9.19)
49 (6.44)
0.52
Tumor diameter (SD), cm
4.69 (0.31)
5.58 (2.03)
0.53
IB2
9
4
0.51
IIA2
9
3
0.58
IIB
6
2
0.63
G1
3
0
0.37
G2
12
5
0.59
G2-3
4
2
0.53
G3
5
2
0.64
Deep myometrial invasion
10
3
Lymphovascular space invasion
4
1
Pelvic Lymph node metastasis
4
3
Parametrial invasion
0
0
chemotherapy
10
0
radiotherapy
4
7
Radiotherapy and chemotherapy
4
2
Time(months)
18.46 6 7.59
21.8 6 6.38
Survival status
23
8
FIGO stage
Cell grade
Postoperative pathology
Postoperative complementary therapy
Follow-up 0.09
LACC 5 locally advanced cervical cancer; SD 5 standard deviation; FIGO 5 International Federation of Gynecology and Obstetrics.
regions. After two cycles of effective chemotherapy, cervical lesions showed a significant decrease in size on T2WI and heterogeneously high signal intensity on DWI images; an increase of signal intensity was observed with the decreased range of the low-signal-intensity region in ADC maps. After the ineffective chemotherapy, the signal intensity of DWI and ADC images was unchanged (Fig. 2). ANOVA was calculated on ADC value at three timepoints. The analysis of the effective group was significant, F(2,69) 5 30.88, P < 0.001; the value of the ineffective group was not significant, F(2,21) 5 1.081, P 5 0.36. Multiple comparisons showed the differences were all significant between two different timepoints in the effective group (prechemotherapy vs. after the first NACT, P 5 0.002; prechemotherapy vs. after the second chemotherapy, P < 0.001; after the second chemotherapy vs. after the first NACT, P < 0.001; Table 2, Fig. 1). 4
Tumor Cell Density, PCNA, and AQP1 Protein Expressions Cervical SCC had high cell density and cells arranged in sheets before chemotherapy. Tumor cells usually had a large and deep-dyed nucleus, high nuclear plasma ratio, and atypia. After effective chemotherapy, tumor cell density significantly decreased and nuclear-cytoplasmic staining was vague. More red necrotic material was in the cytoplasm, and part of the tumor cell nucleus pyknosis or disappeared (Fig. 1). Cervical cancer cells had high PCNA expressions. AQP1-positive expressions were seen on the microvascular endothelial cell membrane, interstitial cell membrane, and the membrane of cancer cells (Figs. 3, 4). The ANOVA results of tumor cell density, PCNA, and AQP1 expressions all showed statistical differences between timepoints in the effective group (P < 0.001) and those in the ineffective group were not statistically different Volume 00, No. 00
Fu et al.: Correlation With ADC in NACT of LACC
FIGURE 1: MR images and H&E staining of a 49-year-old patient with cervical SCC in effective chemotherapy group. a: MR images show changes of mass size and signal at three different timepoints (arrows). Left represents DWI image (b 5 0) and right represents ADC image (b 5 900 s/mm2) in each MRI picture. b: Photomicrographs (H&E, magnification, 320) at different timepoints. c: The size of tumor diameter was reduced after the second NACT (䉬 P < 0.01). d: ADC values were increased after treatment (䉬 P < 0.01). e: Cellular density was decreased after treatment (䉬 P < 0.01).
(Pcell density 5 0.17; PPCNA 5 0.18; PAQP1 5 0.26). Multiple comparisons of tumor cell density, PCNA, and AQP1 expressions all showed significant differences between two different timepoints in the effective group (prechemotherapy vs. after the first NACT, P < 0.001; prechemotherapy vs. after the second chemotherapy, P < 0.001; after the second chemotherapy vs. after the first NACT, P < 0.001; Table 2, Fig. 4). Relationship Between Simultaneous Changes of MR Images and Pathological Microstructure All the expression levels of three indictors (tumor cell density, PCNA, and AQP1) showed no significant correlation with the ADC value at three different timepoints in both groups (P > 0.05). Month 2014
Changes of PCNA expression were significantly negative association with changes of ADC value from the time of prechemotherapy to 2 weeks after the first NACT in the effective group (r 5 –0.56, P 5 0.03; Fig. 3), and the same relationship was seen from the time of prechemotherapy to 2 weeks after the second NACT (r 5 –0.64, P 5 0.001; Fig. 3). Changes of cellular density were significantly negative association with changes of ADC value from the time of prechemotherapy to 2 weeks after the second NACT in the effective group (r 5 –0.51, P 5 0.04). There were no relationships between expression changes of three indictors (tumor cell density, PCNA, and AQP1) and ADC value among three different timepoints in the ineffective group (P > 0.05). 5
Journal of Magnetic Resonance Imaging
FIGURE 2: MR images and H&E staining of a 42-year-old patient with cervical SCC in the ineffective group. a: MR images show mass size and signal at three different timepoints (arrows). Left represents DWI image (b 5 0) and right represents ADC image (b 5 900 s/mm2) in each MRI picture. b: Photomicrographs (H&E, magnification, 320) at different timepoints. The size of tumor diameter (c), ADC value (d), and cellular density (e) were all not significant changes after treatment. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
DISCUSSION The main finding of this study is both the changes of PCNA expression and tumor cellular density are negatively correlated with ADC change between pre- and posttherapy in the effective group. Especially the change of decreased PCNA expression is negatively related with the change of increased ADC early after the first NACT. DWI is a functional imaging technique that displays information about the extent and direction of random water motion in tissues. The administration of effective nonsurgical treatment results in necrosis, apoptosis, tumor lysis, loss of cell membrane integrity, increased extracellular space, and therefore an increase in water diffusion. There are nine clinical studies on monitoring the response of cervical cancer to combined chemoradiation and one study on NACT using DWI.7–15,17 All the results showed the mean ADC value of cervical cancer after effective treatment was statistically higher than that before therapy. The results support the 6
conclusion that DWI images and ADC measurement may be useful for early assessment of response to nonsurgical treatment for cervical cancer. However, few studies can provide histological information about the ADC in prediction of tumor response; further study is necessary to investigate pathological changes in cervical tumor corresponding to ADC change over the same period. Chemotherapy is a category of cancer treatment that uses cytotoxic antineoplastic drugs to kill or inhibit the growth of cancer cells. Cervical SCC is considered a sensitive tumor to cisplatin-based chemotherapy. After successful therapy, the increased ADC value can be ascribed to increased displacement of water molecules in the extracellular space and between intra- and extracellular compartments for cell shrinkage, free movements across cell membrane remnants after apoptosis, and in the presence of necrosis. Therefore, multiple pathophysiological factors may affect the movement of water molecules. In the study, we Volume 00, No. 00
Fu et al.: Correlation With ADC in NACT of LACC
TABLE 2. ADC Value and Expressions of Pathological Indicators
Group
Time
ADC value (1023mm2/s)
Cell density (%)
PCNA expression (%)
AQP1 expression (mean IOD)
Effective
Prechemotherapy
0.86 6 0.01
37.57 6 1.11
59.08 6 1.80
123.83 6 1.84
(n 5 24)
After the first NACT
0.94 6 0.02
33.09 6 1.28
49.67 6 1.68
108.33 6 1.92
After the second NACT
1.05 6 0.02
21.13 6 1.00
32.92 6 1.65
95.01 6 2.37
Ineffective
Prechemotherapy
0.86 6 0.02
35.25 6 2.07
55.01 6 2.01
116.78 6 3.17
(n 5 9)
After the first NACT
0.88 6 0.03
33.75 6 1.90
51.89 6 1.72
111.67 6 4.14
After the second NACT
0.90 6 0.02
30.5 6 2.02
50.11 6 1.68
107.78 6 3.90
ADC 5 apparent diffusion coefficient; LACC 5 locally advanced cervical cancer; PCNA 5 proliferating cell nuclear antigen; AQP1 5 aquaporin 1; NACT 5 neoadjuvant chemotherapy; IOD 5 integrated optical density.
investigated the relationship between water movement in the tumor microenvironment and ADC value from three points. The first was cellular density; the second was the state of cell proliferation (PCNA); the third was water channels on the cellular plasma membrane (AQP1). An inverse correlation between cellular density and ADC values has been well described for cervical cancer
10,18
. In our study, tumor cell density significantly decreased after the first NACT; however, the change of cellular density was significantly negative correlation with ADC change only after the second NACT in the effective group. This finding may be attributed to cellular swelling before apoptosis and a decreased ADC value. After initiation of chemotherapy, early apoptotic cell death, loss of membrane integrity, and
FIGURE 3: Immunohistochemical staining of PCNA at three different timepoints (magnification, 320 or 10). a: PCNA expression was decreased after treatment in the effective group (䉬P < 0.01, arrows refer to positive stain). b: PCNA expression was not significantly changed after treatment in the ineffective group. c: Refers to the time between after the first NACT and prechemotherapy. d: Refers to the time between after the second NACT and prechemotherapy, illustrates the relationship of different values of ADC and PCNA (c, r 5 –0.56, P 5 0.03; d, r 5 –0.64, P 5 0.001). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
Month 2014
7
Journal of Magnetic Resonance Imaging
FIGURE 4: Immunohistochemical staining of AQP1 at three different timepoints (magnification, 320). a: AQP1 expression was decreased after treatment in the effective group (P < 0.01). b: AQP1 expression was not significantly changed after treatment in the ineffective group. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
increased extracellular space may lead to alterations in intratumoral water diffusion (ie, more restricted diffusion) and lower tumor ADC.12 Our findings support this conjecture to some extent. After the second effective NACT, therapyinduced increases in ADC coincided with cell lysis and necrosis. The main mechanism of platinum-based combination chemotherapy is interference and suppression of DNA synthesis and mitosis in cervical cancer cells. PCNA is an auxiliary protein of DNA polymerase d, and exists only in the proliferative cells. Tumor cells in the active state of proliferation are more sensitive to cisplatin.20 This study could not show a direct relationship between ADC value and PCNA protein expression in the cervical lesion. Further analysis of the study found the change of PCNA expression had a negative correlation with ADC change after the first course of effective chemotherapy. This indicated more proliferative cells involved in apoptosis and necrosis, which is more likely to promote the diffusion of water molecules and increase the ADC value after effective chemotherapy. AQP1 is a transmembrane protein that controls the water in and out of cells by composing “channels” on the cell membrane.21 Chen et al reported water metabolism through AQP1 and AQP3 was maintained during neoplastic transformation in human cervical carcinoma. AQP overexpression may increase tumor cells permeability to water to alter tumor cell volume and shape.22,23 Our immunohistochemical assay showed that AQP1 protein was expressed mainly in vascular endothelial cell, and also localized in the interstitial cell and cervical cancer tissues. We observed that high AQP1 expression was significantly decreased after 8
effective chemotherapy. We also analyzed the correlation between the change of AQP1 expression and ADC before and after treatment, and found no direct relationship. These results indicate that decreased AQP1 expression may be related to destruction and disintegration of membrane structure induced by apoptosis. There are some limitations to this study. First, the study only included cases of SCC of the cervix and no other histological types. Second, the study used human pathological specimens. It is better to study the relation between ADC value and pathological changes by using tissue in vitro or an animal model studying tumor heterogeneity. Finally, the status of pelvic lymph nodes by DWI was not included. The metastasis of pelvic lymph node is a key factor affecting the prognosis of patients with cervical cancer. DWI is feasible for differentiating metastatic from nonmetastatic lymph nodes.24–26 In conclusion, with the first effective NACT, the change of PCNA expression shows a negative correlation with ADC change, and the change of tumor cellular density shows a negative correlation with ADC change after the second NACT; however, the time of significant change in tumor size is after the second NACT. All the data suggest an ADC change after effective chemotherapy was closely related with cellular characteristics preceding size reduction, and ADC may be used as an early imaging biomarker of therapeutic response in LACC.
REFERENCES 1.
Foley G, Alston R, Geraci M, et al. Increasing rates of cervical cancer in young women in England: an analysis of national data 1982–2006. Br J Cancer 2011;105:177–184.
Volume 00, No. 00
Fu et al.: Correlation With ADC in NACT of LACC 2.
Angioli R, Plotti F, Montera R, et al. Neoadjuvant chemotherapy plus radical surgery followed by chemotherapy in locally advanced cervical cancer. Gynecol Oncol 2012;127:290–296.
3.
Gong L, Lou JY, Wang P, Zhang JW, Liu H, Peng ZL. Clinical evaluation of neoadjuvant chemotherapy followed by radical surgery in the management of stage IB2-IIB cervical cancer. Int J Gynaecol Obstet 2012;117:23–26.
4.
Ferrandina G, Distefano MG, De Vincenzo R, et al. Paclitaxel, epirubicin, and cisplatin (TEP) regimen as neoadjuvant treatment in locally advanced cervical cancer: long-term results. Gynecol Oncol 2013;128: 518–523.
5.
Landoni F, Sartori E, Maggino T, et al. Is there a role for postoperative treatment in patients with stage Ib2–IIb cervical cancer treated with neo-adjuvant chemotherapy and radical surgery? An Italian multicenter retrospective study. Gynecol Oncol 2014;132:611–617.
6.
7.
8.
Nishino M, Jagannathan JP, Ramaiya NH, Van den Abbeele AD. Revised RECIST guideline version 1.1: what oncologists want to know and what radiologists need to know. Am J Roentgenol 2010;195:281– 289. Liu Y, Bai R, Sun H, Liu H, Zhao X, Li Y. Diffusion-weighted imaging in predicting and monitoring the response of uterine cervical cancer to combined chemoradiation. Clin Radiol 2009;64:1067–1074. Haack S, Pedersen EM, Jespersen SN, Kallehauge JF, Lindegaard JC, Tanderup K. Apparent diffusion coefficients in GEC ESTRO target volumes for image guided adaptive brachytherapy of locally advanced cervical cancer. Acta Oncol 2010;49:978–983.
15.
Makino H, Kato H, Furui T, Morishige KI, Kanematsu M. Predictive value of diffusion-weighted magnetic resonance imaging during chemoradiotherapy for uterine cervical cancer. J Obstet Gynaecol Res 2014;40:1098–1104.
16.
Naganawa S, Sato C, Kumada H, Ishigaki T, Miura S, Takizawa O. Apparent diffusion coefficient in cervical cancer of the uterus: comparison with the normal uterine cervix. Eur Radiol 2005;15:71–78.
17.
Fu C, Bian D, Liu F, Feng X, Du W, Wang X. The value of diffusionweighted magnetic resonance imaging in assessing the response of locally advanced cervical cancer to neoadjuvant chemotherapy. Int J Gynecol Cancer 2012;22:1037–1043.
18.
Levy A, Caramella C, Chargari C, et al. Accuracy of diffusionweighted echo-planar MR imaging and ADC mapping in the evaluation of residual cervical carcinoma after radiation therapy. Gynecol Oncol 2011;123:110–115.
19.
Olsen JR, Esthappan J, DeWees T, et al. Tumor volume and subvolume concordance between FDG-PET/CT and diffusion-weighted MRI for squamous cell carcinoma of the cervix. J Magn Reson Imaging 2013;37:431–434.
20.
Albertella MR, Green CM, Lehmann AR, O’Connor MJ. A role for polymerase eta in the cellular tolerance to cisplatin-induced damage. Cancer Res 2005;65:9799–9806.
21.
Zhang L, Li J, Jiang Z, Sun L, Mei X, Yong B, Zhang L. Inhibition of aquaporin-1 expression by RNAi protects against aristolochic acid Iinduced apoptosis in human proximal tubular epithelial (HK-2) cells. Biochem Biophys Res Commun 2011;405:68–73.
9.
Chen J, Zhang Y, Liang B, Yang Z. The utility of diffusion-weighted MR imaging in cervical cancer. Eur J Radiol 2010;74:e101–106.
10.
Harry VN, Semple SI, Gilbert FJ, Parkin DE. Diffusion-weighted magnetic resonance imaging in the early detection of response to chemoradiation in cervical cancer. Gynecol Oncol 2008;111:213–220.
22.
Chen R, Shi Y, Amiduo R, Tuokan T, Suzuk L. Expression and prognostic value of aquaporin 1, 3 in cervical carcinoma in women of Uygur ethnicity from Xinjiang, China. PLoS One 2014;11:e98576.
11.
Kim HS, Kim CK, Park BK, Huh SJ, Kim B. Evaluation of therapeutic response to concurrent chemoradiotherapy in patients with cervical cancer using diffusion-weighted MR imaging. J Magn Reson Imaging 2013;37:187–193.
23.
Shi YH, Chen R, Talafu T, Nijiati R, Lalai S. Significance and expression of aquaporin 1, 3, 8 in cervical carcinoma in Xinjiang Uygur women of China. Asian Pac J Cancer Prev 2012;13:1971–1975.
12.
Chopra S, Verma A, Kundu S, et al. Evaluation of diffusion-weighted imaging as a predictive marker for tumor response in patients undergoing chemoradiation for postoperative recurrences of cervical cancer. J Cancer Res Ther 2012;8:68–73.
24.
Kim JK, Kim KA, Park BW, Kim N, Cho KS. Feasibility of diffusionweighted imaging in the differentiation of metastatic from nonmetastatic lymph nodes: early experience. J Magn Reson Imaging 2008;28: 714–719.
13.
Somoye G, Harry V, Semple S, et al. Early diffusion weighted magnetic resonance imaging can predict survival in women with locally advanced cancer of the cervix treated with combined chemo-radiation. Eur Radiol 2012;22:2319–2327.
25.
Zhang Y, Chen J, Shen J, Zhong J, Ye R, Liang B. Apparent diffusion coefficient values of necrotic and solid portion of lymph nodes: differential diagnostic value in cervical lymphadenopathy. Clin Radiol 2013; 68:224–231.
14.
Rizzo S, Summers P, Raimondi S, et al. Diffusion-weighted MR imaging in assessing cervical tumour response to nonsurgical therapy. Radiol Med 2011;116:766–780.
26.
Lin G, Ho KC, Wang JJ, et al. Detection of lymph node metastasis in cervical and uterine cancers by diffusion-weighted magnetic resonance imaging at 3T. J Magn Reson Imaging 2008;28:128–135.
Month 2014
9
Simultaneous Changes of Magnetic Resonance Diffusion-Weighted Imaging and Pathological Microstructure in Locally Advanced Cervical Cancer Caused by Neoadjuvant Chemotherapy Chun Fu, MD,1* Xiaoyan Feng, MS,1 Dujun Bian, MD,2 Yan Zhao, MS,1 Xiaoling Fang, MD,1 Wanping Du, BS,2 Lan Wang, BS,3 and Xiangquan Wang, MS1 Purpose: To investigate the changes to diffusion-weighted imaging (DWI) correlated with histopathology after neoadjuvant chemotherapy (NACT) in patients with locally advanced cervical cancer (LACC). Materials and Methods: Thirty-three patients with LACC were examined with 3T magnetic resonance imaging (MRI) with DWI and apparent diffusion coefficient (ADC) maps. MRIs were performed for each patient at three timepoints: before the first NACT, 2 weeks after the first NACT, and 2 weeks after the second NACT. Uterine cervical specimens were collected at the same timepoints. Specimens were stained for tumor cell density, proliferating cell nuclear antigen (PCNA), and aquaporin 1 (AQP1). Treatment responses were classified as the effective group (complete and partial response) and the ineffective group (stable and progressive disease). Results: The ADC value of the effective group after the first chemotherapy was higher than that before chemotherapy (P 5 0.002), and expressions of three pathological indicators (tumor cell density, PCNA, and AQP1) significantly decreased after the first NACT compared with those prechemotherapy (P < 0.001). Changes of PCNA expression were negatively correlated with changes of ADC values after the first NACT in the effective group (r 5 –0.56, P 5 0.03). Changes of cellular density were negatively correlated with changes of ADC values from the time of prechemotherapy to after the second NACT in the effective group (r 5 –0.51, P 5 0.04). Conclusion: The ADC change after successful chemotherapy is closely related with cellular characteristics preceding size reduction. ADC may be used as an early imaging biomarker of NACT response in LACC. J. MAGN. RESON. IMAGING 2014;00:000–000
C
ervical cancer is the second most common female cancer.1,2 A worse prognosis has been observed even in early cervical cancer with tumor diameter more than 4 cm or locally advanced cervical cancer (LACC). Neoadjuvant chemotherapy (NACT) followed by radical surgery is a valid alternative therapeutic option to concurrent chemotherapy and radiotherapy for patients with LACC.2–4 The achievement of
an optimal pathological response on surgical specimens is a strong predictor of a better clinical outcome.5 The ineffective treatment may increase operative difficulty and affect a patient’s prognosis. Therefore, evaluation of the early effect of NACT may play an important role in the treatment of LACC. Treatment response is mainly judged by tumor size reduction according to the Response Evaluation Criteria in
View this article online at wileyonlinelibrary.com. DOI: 10.1002/jmri.24779 Received Feb 15, 2014, Accepted for publication Sep 29, 2014. *Address reprint requests to: C.F., Department of Gynecology and Obstetrics, Second Xiangya Hospital, Central South University, 139 Middle Renmin Rd., Changsha, Hunan 410011, People’s Republic of China. E-mail: [email protected] From the 1Department of Gynecology and Obstetrics, Second Xiangya Hospital, Central South University, Changsha, People’s Republic of China; Department of Radiology, Second Xiangya Hospital, Central South University, Changsha, People’s Republic of China; and 3Department of Research, Second Xiangya Hospital, Central South University, Changsha, People’s Republic of China.
2
Contract grant sponsor: National Science Foundation of Hunan Province of the People’s Republic of China; contract grant number: 10JJ6044. Contract grant sponsor: Hunan Science and Technology Department Project of the People’s Republic of China; contract grant numbers: 2009SK3161; 2011FJ3008.Contract grant sponsor: Scientific Research Foundation of Hunan Provincial Health Bureau of the People’s Republic of China; contract grant number: B2009022
C 2014 Wiley Periodicals, Inc. V 1
Journal of Magnetic Resonance Imaging
Solid Tumor (RECIST).6 Conventional magnetic resonance imaging (MRI) and computed tomography are insensitive to early changes after treatment.7 Magnetic resonance diffusion-weighted imaging (MR-DWI) acquisition measures the mobility of water in tissue, and can discriminate cervical cancer from normal cervix by calculating the apparent diffusion coefficient (ADC) value.8 MR-DWI has been used to detect response of cervical cancer to chemoradiation in nine clinical trials,7,9–16 and to assess the response of LACC to NACT in one study.17 The results show that ADC values can be a useful tool to monitor the response to therapy. The effective anticancer treatments (including chemoradiation and NACT) result in tumor cell lysis, loss of cell membrane integrity, increased extracellular space, and, therefore, an increase in water diffusion.9 Changes of water flow at the cellular level can reflect changes of the organizational structure, which may be the theoretical basis of DWI applied in cervical cancer. No studies show the relationship of simultaneous changes between pathological microstructure and DWI caused by NACT. In the study, we chose tumor cell density, proliferating cell nuclear antigen (PCNA), and aquaporin 1 (AQP1) as histological indicators. DWI images and pathological microstructure after NACT was observed at the same time. Our purposes were 1) to assess the relationship between simultaneous changes of ADC and pathological microstructure, and 2) to provide a pathological basis for evaluation of curative effect by using DWI in patients with LACC.
MATERIALS AND METHODS Patient Population The study was approved by the hospital Ethics Committee and written informed consent was obtained from all patients. Only data from patients with International Federation of Gynecology and Obstetrics (FIGO) stage IB2–IIB cervical cancer (confirmed histologically with a tumor diameter of >4 cm) were included in the analysis. Thirtythree patients with cervical squamous cell carcinoma (SCC) participated in this prospective study from May 2010 to May 2012. The mean age of patients was (43.24 6 6.11) years (range, 26–48 years). The mean diameter of tumor was (4.86 6 0.93) cm (range, 4–8 cm). There were 13 patients in stage IB2, 12 in stage IIA2, and 8 in stage IIB. Three cases of tumor grading were well differentiated (G1, low grade), 17 cases were moderately differentiated (G2, intermediate grade), 7 cases poorly differentiated (G3, high grade), and 6 cases moderately to poorly differentiated (G2–G3). The inclusion criteria of patients were: normal renal, hepatic, cardiac, pulmonary, and hematologic function; and no previous surgery for cancer, chemotherapy, or radiation therapy. Exclusion criterion was contraindicated for MRI, such as those with pacemakers, metal objects, or claustrophobia disorder. The checked times were nonmenstrual period.
Standard NACT and Response Evaluation All patients were scheduled for NACT. NACT was consisted of two cycles of intravenous docetaxel 75 mg/m2 and cisplatin 2
50 mg/m2 at 2-week intervals. Two weeks after the second NACT, 31 patients underwent type III radical hysterectomy with systematic pelvic lymphadenectomy, and two patients were treated with concurrent chemoradiotherapy. Postoperatively, additional chemotherapy or radiotherapy alone was administered to patients with one or more recurrence risk factors.3 The factors include lymph node metastasis, positive surgical margin, parametric infiltration, deep cervical stromal invasion, and lymphovascular invasion. All the patients were followed up. The mean time was (19.39 6 7.253) months. Tumor response relative to pretreatment data was classified clinically as one of four groups according to the RECIST.6 Complete response (CR) was concluded if there was no residual tumor on T2-weighted imaging (T2WI); partial response (PR) was concluded if the longest diameter of the tumor was less than 70% of the original size; the disease was determined to be stable if there was neither sufficient shrinkage to qualify for partial response nor sufficient increase to qualify for progressive disease; and progressive disease (PD) was concluded if there was at least a 20% increase and a 5-mm absolute increase in the sum of the longest diameter of the tumor, taking as reference the longest diameter recorded before treatment. Treatment responses were classified as the effective group (CR and PR) and the ineffective group (stable disease and PD).
MRI Scanning Protocol and Imaging Analysis Serial MRIs were performed for each patient at three timepoints: before the first NACT, 2 weeks after the first NACT, and 2 weeks after the second NACT. Each time each patient underwent both conventional MRI and DWI scans. All scans were performed on a 3T MR scanner (Philips Achieva 3.0T X-series, Philips Healthcare, Netherlands), using 16 elements phased array of SENSE XL Torso coil to cover the entire pelvis. The imaging protocol included: 1) SENSE T2-weighted fatsuppressed respiratory triggering MRI was obtained in the coronal, sagittal, and axial planes (T2WI, TR/TE 1462/70 msec; average, 3; flip angle, 90 ; matrix size, 308 3 212; field of view [FOV], 36 3 36 cm); 2) axial planes SENSE T1-weighted respiratory triggering MRI (T1WI, TR/TE 10/2.3 msec; average, 3; flip angle, 15 ; matrix, 252 3 194; FOV, 36 3 36 cm); 3) MR DWI obtained in the axial planes using a single-shot spin-echo echoplanar with chemical-shift-selective fat-suppression technique (TR/TE 952/53 msec; flip angle, 90 ; average number of signals, 4; section thickness, 5 mm; gap, 1.5 mm; matrix, 124 3 100; FOV, 36 3 36 cm); the acquisition time was 1 minute, 32 seconds; 4) contrast-enhanced turbo field echo (TFE) fat-suppressed T1weighted sequences were obtained in identical slice thicknesses and gaps in the axial, coronal, and sagittal planes to cover the entire true pelvis. All MRI data were transmitted to MR Workspace Release 2.6.3 Workstations and analyzed with the Functool package. DWI images were analyzed qualitatively by referring to the signal intensity of uterine cervical cancer, which was classified using visual assessment of hypointensity or hyperintensity in comparison with the adjacent skeletal muscle).7,18 ADC measurement was on the axial plane with mass in maximum diameter and using three regions of interest (ROIs). ROI (area size range 20–50 mm2) was Volume 00, No. 00
Fu et al.: Correlation With ADC in NACT of LACC
arranged in three different positions to avoid necrotic areas and the lesions of peripheral parts. Average value of 3 times was the final ADC value. ADC can be calculated from two measurements with different diffusion attenuations according to the following equation:
ADC 5 In ðS2=S1Þ=ðb12b2Þ Where b1 and b2 are the two applied diffusion-sensitizing factors, S1 and S2 are the DWI signal intensity, respectively, under the conditions of b1 and b2. b 5 r2G2d2 (D-d/3), where r is the gyromagnetic ratio, G, d, D separately represent the amplitude, width, and spacing of pulse gradient. The four b values were chosen to be 0, 300, 600, 900 s/mm2 to optimize the signal-to-noise ratio.17 We selected b-factors of 0 and 900 s/mm2 in the study.
Immunohistochemical Staining and Image Analysis Uterine cervical specimens were collected from every patient at three different times (prechemotherapy, 2 weeks after the first NACT, and postsurgery). Tissue samples were routinely fixed in 10% formalin solution, paraffin-embedded, and cut into 4-lmthick sections. After deparaffinizatiom and rehydration, the sections were heated in three 5-minute periods in a microwave oven at 100 C with sodium citrate buffer (10 mM; pH 6), cooled in the same buffer at room temperature, and subsequently incubated for 20 minutes with 3% hydrogen peroxide. The antibodies for PCNA and AQP1 were used at 1:200 and 1:150. The primary antibodies against PCNA and AQP1 were both purchased from GeneTex (Irvine, CA). The serial sections were incubated with primary antibodies in a humid chamber at 4 C overnight. Sections were washed three times in phosphate-buffered saline (PBS) and further incubated with a biotinylated secondary antibody for 30 minutes at room temperature. Streptavidin-horseradish peroxidase conjugate was added and the peroxidase activity was made visible with diaminobenzidine and counterstained with hematoxylin for 30 seconds. As a control experiment, we performed an identical immunohistochemical procedure with omission of the primary antibody. The overall structure, cell morphology, and nuclear atypia were first observed at low magnification. The representative areas were then selected and observed at high magnification (3200 or 400). Five microscopic fields were freely chosen for each tumor specimen using an Olympus BX51/52 electron microscope collection system and saved into the computer. Images were analyzed by the MIAS medical pathology image analysis system. Tumor cell density was the ratio of total area of tumor cell nucleus and statistical field area (hematoxylin and eosin [H&E]stained specimen under the light microscope, 3200). The mean of the five visual fields was considered the tumor cellular density, expressed in percentage. Tumor cells with brown nucleus were regarded as PCNA-positive. Analysis of PCNA-positive expression was performed using the area density. The area density was the ratio of total area of positive or negative target cells and total area of statistical field (unit/mm2). Cell membrane stained brownish yellow was considered an APQ1-positive cell. Analysis of APQ1positive expression was performed by the average gray value, and the unit was the mean integrated optical density (IOD). Two investigators (C.F. and X.F.) assessed the histological parameters and immunohistochemical staining. Whenever a Month 2014
discrepancy occurred, both investigators reexamined the slides to reach a consensus.
Statistical Analysis Statistical analysis was performed using SPSS v. 17.0 for Windows (Chicago, IL). Patients’ age, tumor size, and follow-up time in the effective chemotherapy group versus those in the ineffective group were compared with independent-samples t-test. The composition ratio of clinical stage and cell grade between the effective and the ineffective group were compared with Fisher’s Exact Test (4-fold table). One-way analysis of variance (ANOVA) was used to compare the ADC value and pathological results (cellular density, PCNA, and AQP1 expressions) at three timepoints (before chemotherapy, 2 weeks after the first NACT, and 2 weeks after the second NACT) in the two chemotherapy groups. The Shapiro–Wilk W test was used to test for data normality. A post-hoc test was performed with least-significant difference (LSD) in a significant variance analysis. Correlation analysis between ADC value and pathological indictors (tumor cell density, PCNA, and AQP1positive expression) were determined with bivariate correlation analysis. Pearson’s correlation was used to test normally distributed data. In the absence of a linear relationship, Spearman’s correlation was used to assess the significance between groups. P < 0.05 was considered statistically significant.
RESULTS Clinical Curative Effect In all, 33 patients finished two chemotherapy cycles and 31 cases had radical surgeries. There were 24 patients in the effective chemotherapy group (2 CR and 22 PR), and 9 cases in the ineffective group (2 PD and 7 stable disease). The effective rate was 72.7%. Two cases with PD received radical radiotherapy and had no surgery. Twenty-five patients received additional treatment after surgery for recurrence risk factors (18 cases in the effective group and 7 cases in the ineffective group). All patients were followed up and one patient in each group died. Clinical data showed no significant differences between the two groups (Table 1). The mean diameter of the tumor was significantly reduced after the second NACT in the effective group (P 5 0.001, after the second NACT vs. prechemotherapy; P 5 0.11, after the first NACT vs. prechemotherapy; Fig. 1). The mean diameter of tumor was not significantly changed in the ineffective group at three timepoints (P 5 0.84, after the second NACT vs. prechemotherapy; P 5 0.74, after the first NACT vs. prechemotherapy; Fig. 2). Simultaneous Changes of MRI Images and Pathological Microstructure DWI Image and ADC Value Before treatment, all cervical tumors were hyperintense to adjacent skeletal muscle on both T2WI and DWI images; on ADC maps, they all manifested as low-signal-intensity 3
Journal of Magnetic Resonance Imaging
TABLE 1. Clinical Data of Patients With LACC in Different Chemotherapy Response Groups
Parameter
Effective group (n 5 24)
Ineffective group (n 5 9)
P value
Age, mean (SD), yrs
45.46 (9.19)
49 (6.44)
0.52
Tumor diameter (SD), cm
4.69 (0.31)
5.58 (2.03)
0.53
IB2
9
4
0.51
IIA2
9
3
0.58
IIB
6
2
0.63
G1
3
0
0.37
G2
12
5
0.59
G2-3
4
2
0.53
G3
5
2
0.64
Deep myometrial invasion
10
3
Lymphovascular space invasion
4
1
Pelvic Lymph node metastasis
4
3
Parametrial invasion
0
0
chemotherapy
10
0
radiotherapy
4
7
Radiotherapy and chemotherapy
4
2
Time(months)
18.46 6 7.59
21.8 6 6.38
Survival status
23
8
FIGO stage
Cell grade
Postoperative pathology
Postoperative complementary therapy
Follow-up 0.09
LACC 5 locally advanced cervical cancer; SD 5 standard deviation; FIGO 5 International Federation of Gynecology and Obstetrics.
regions. After two cycles of effective chemotherapy, cervical lesions showed a significant decrease in size on T2WI and heterogeneously high signal intensity on DWI images; an increase of signal intensity was observed with the decreased range of the low-signal-intensity region in ADC maps. After the ineffective chemotherapy, the signal intensity of DWI and ADC images was unchanged (Fig. 2). ANOVA was calculated on ADC value at three timepoints. The analysis of the effective group was significant, F(2,69) 5 30.88, P < 0.001; the value of the ineffective group was not significant, F(2,21) 5 1.081, P 5 0.36. Multiple comparisons showed the differences were all significant between two different timepoints in the effective group (prechemotherapy vs. after the first NACT, P 5 0.002; prechemotherapy vs. after the second chemotherapy, P < 0.001; after the second chemotherapy vs. after the first NACT, P < 0.001; Table 2, Fig. 1). 4
Tumor Cell Density, PCNA, and AQP1 Protein Expressions Cervical SCC had high cell density and cells arranged in sheets before chemotherapy. Tumor cells usually had a large and deep-dyed nucleus, high nuclear plasma ratio, and atypia. After effective chemotherapy, tumor cell density significantly decreased and nuclear-cytoplasmic staining was vague. More red necrotic material was in the cytoplasm, and part of the tumor cell nucleus pyknosis or disappeared (Fig. 1). Cervical cancer cells had high PCNA expressions. AQP1-positive expressions were seen on the microvascular endothelial cell membrane, interstitial cell membrane, and the membrane of cancer cells (Figs. 3, 4). The ANOVA results of tumor cell density, PCNA, and AQP1 expressions all showed statistical differences between timepoints in the effective group (P < 0.001) and those in the ineffective group were not statistically different Volume 00, No. 00
Fu et al.: Correlation With ADC in NACT of LACC
FIGURE 1: MR images and H&E staining of a 49-year-old patient with cervical SCC in effective chemotherapy group. a: MR images show changes of mass size and signal at three different timepoints (arrows). Left represents DWI image (b 5 0) and right represents ADC image (b 5 900 s/mm2) in each MRI picture. b: Photomicrographs (H&E, magnification, 320) at different timepoints. c: The size of tumor diameter was reduced after the second NACT (䉬 P < 0.01). d: ADC values were increased after treatment (䉬 P < 0.01). e: Cellular density was decreased after treatment (䉬 P < 0.01).
(Pcell density 5 0.17; PPCNA 5 0.18; PAQP1 5 0.26). Multiple comparisons of tumor cell density, PCNA, and AQP1 expressions all showed significant differences between two different timepoints in the effective group (prechemotherapy vs. after the first NACT, P < 0.001; prechemotherapy vs. after the second chemotherapy, P < 0.001; after the second chemotherapy vs. after the first NACT, P < 0.001; Table 2, Fig. 4). Relationship Between Simultaneous Changes of MR Images and Pathological Microstructure All the expression levels of three indictors (tumor cell density, PCNA, and AQP1) showed no significant correlation with the ADC value at three different timepoints in both groups (P > 0.05). Month 2014
Changes of PCNA expression were significantly negative association with changes of ADC value from the time of prechemotherapy to 2 weeks after the first NACT in the effective group (r 5 –0.56, P 5 0.03; Fig. 3), and the same relationship was seen from the time of prechemotherapy to 2 weeks after the second NACT (r 5 –0.64, P 5 0.001; Fig. 3). Changes of cellular density were significantly negative association with changes of ADC value from the time of prechemotherapy to 2 weeks after the second NACT in the effective group (r 5 –0.51, P 5 0.04). There were no relationships between expression changes of three indictors (tumor cell density, PCNA, and AQP1) and ADC value among three different timepoints in the ineffective group (P > 0.05). 5
Journal of Magnetic Resonance Imaging
FIGURE 2: MR images and H&E staining of a 42-year-old patient with cervical SCC in the ineffective group. a: MR images show mass size and signal at three different timepoints (arrows). Left represents DWI image (b 5 0) and right represents ADC image (b 5 900 s/mm2) in each MRI picture. b: Photomicrographs (H&E, magnification, 320) at different timepoints. The size of tumor diameter (c), ADC value (d), and cellular density (e) were all not significant changes after treatment. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
DISCUSSION The main finding of this study is both the changes of PCNA expression and tumor cellular density are negatively correlated with ADC change between pre- and posttherapy in the effective group. Especially the change of decreased PCNA expression is negatively related with the change of increased ADC early after the first NACT. DWI is a functional imaging technique that displays information about the extent and direction of random water motion in tissues. The administration of effective nonsurgical treatment results in necrosis, apoptosis, tumor lysis, loss of cell membrane integrity, increased extracellular space, and therefore an increase in water diffusion. There are nine clinical studies on monitoring the response of cervical cancer to combined chemoradiation and one study on NACT using DWI.7–15,17 All the results showed the mean ADC value of cervical cancer after effective treatment was statistically higher than that before therapy. The results support the 6
conclusion that DWI images and ADC measurement may be useful for early assessment of response to nonsurgical treatment for cervical cancer. However, few studies can provide histological information about the ADC in prediction of tumor response; further study is necessary to investigate pathological changes in cervical tumor corresponding to ADC change over the same period. Chemotherapy is a category of cancer treatment that uses cytotoxic antineoplastic drugs to kill or inhibit the growth of cancer cells. Cervical SCC is considered a sensitive tumor to cisplatin-based chemotherapy. After successful therapy, the increased ADC value can be ascribed to increased displacement of water molecules in the extracellular space and between intra- and extracellular compartments for cell shrinkage, free movements across cell membrane remnants after apoptosis, and in the presence of necrosis. Therefore, multiple pathophysiological factors may affect the movement of water molecules. In the study, we Volume 00, No. 00
Fu et al.: Correlation With ADC in NACT of LACC
TABLE 2. ADC Value and Expressions of Pathological Indicators
Group
Time
ADC value (1023mm2/s)
Cell density (%)
PCNA expression (%)
AQP1 expression (mean IOD)
Effective
Prechemotherapy
0.86 6 0.01
37.57 6 1.11
59.08 6 1.80
123.83 6 1.84
(n 5 24)
After the first NACT
0.94 6 0.02
33.09 6 1.28
49.67 6 1.68
108.33 6 1.92
After the second NACT
1.05 6 0.02
21.13 6 1.00
32.92 6 1.65
95.01 6 2.37
Ineffective
Prechemotherapy
0.86 6 0.02
35.25 6 2.07
55.01 6 2.01
116.78 6 3.17
(n 5 9)
After the first NACT
0.88 6 0.03
33.75 6 1.90
51.89 6 1.72
111.67 6 4.14
After the second NACT
0.90 6 0.02
30.5 6 2.02
50.11 6 1.68
107.78 6 3.90
ADC 5 apparent diffusion coefficient; LACC 5 locally advanced cervical cancer; PCNA 5 proliferating cell nuclear antigen; AQP1 5 aquaporin 1; NACT 5 neoadjuvant chemotherapy; IOD 5 integrated optical density.
investigated the relationship between water movement in the tumor microenvironment and ADC value from three points. The first was cellular density; the second was the state of cell proliferation (PCNA); the third was water channels on the cellular plasma membrane (AQP1). An inverse correlation between cellular density and ADC values has been well described for cervical cancer
10,18
. In our study, tumor cell density significantly decreased after the first NACT; however, the change of cellular density was significantly negative correlation with ADC change only after the second NACT in the effective group. This finding may be attributed to cellular swelling before apoptosis and a decreased ADC value. After initiation of chemotherapy, early apoptotic cell death, loss of membrane integrity, and
FIGURE 3: Immunohistochemical staining of PCNA at three different timepoints (magnification, 320 or 10). a: PCNA expression was decreased after treatment in the effective group (䉬P < 0.01, arrows refer to positive stain). b: PCNA expression was not significantly changed after treatment in the ineffective group. c: Refers to the time between after the first NACT and prechemotherapy. d: Refers to the time between after the second NACT and prechemotherapy, illustrates the relationship of different values of ADC and PCNA (c, r 5 –0.56, P 5 0.03; d, r 5 –0.64, P 5 0.001). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
Month 2014
7
Journal of Magnetic Resonance Imaging
FIGURE 4: Immunohistochemical staining of AQP1 at three different timepoints (magnification, 320). a: AQP1 expression was decreased after treatment in the effective group (P < 0.01). b: AQP1 expression was not significantly changed after treatment in the ineffective group. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]
increased extracellular space may lead to alterations in intratumoral water diffusion (ie, more restricted diffusion) and lower tumor ADC.12 Our findings support this conjecture to some extent. After the second effective NACT, therapyinduced increases in ADC coincided with cell lysis and necrosis. The main mechanism of platinum-based combination chemotherapy is interference and suppression of DNA synthesis and mitosis in cervical cancer cells. PCNA is an auxiliary protein of DNA polymerase d, and exists only in the proliferative cells. Tumor cells in the active state of proliferation are more sensitive to cisplatin.20 This study could not show a direct relationship between ADC value and PCNA protein expression in the cervical lesion. Further analysis of the study found the change of PCNA expression had a negative correlation with ADC change after the first course of effective chemotherapy. This indicated more proliferative cells involved in apoptosis and necrosis, which is more likely to promote the diffusion of water molecules and increase the ADC value after effective chemotherapy. AQP1 is a transmembrane protein that controls the water in and out of cells by composing “channels” on the cell membrane.21 Chen et al reported water metabolism through AQP1 and AQP3 was maintained during neoplastic transformation in human cervical carcinoma. AQP overexpression may increase tumor cells permeability to water to alter tumor cell volume and shape.22,23 Our immunohistochemical assay showed that AQP1 protein was expressed mainly in vascular endothelial cell, and also localized in the interstitial cell and cervical cancer tissues. We observed that high AQP1 expression was significantly decreased after 8
effective chemotherapy. We also analyzed the correlation between the change of AQP1 expression and ADC before and after treatment, and found no direct relationship. These results indicate that decreased AQP1 expression may be related to destruction and disintegration of membrane structure induced by apoptosis. There are some limitations to this study. First, the study only included cases of SCC of the cervix and no other histological types. Second, the study used human pathological specimens. It is better to study the relation between ADC value and pathological changes by using tissue in vitro or an animal model studying tumor heterogeneity. Finally, the status of pelvic lymph nodes by DWI was not included. The metastasis of pelvic lymph node is a key factor affecting the prognosis of patients with cervical cancer. DWI is feasible for differentiating metastatic from nonmetastatic lymph nodes.24–26 In conclusion, with the first effective NACT, the change of PCNA expression shows a negative correlation with ADC change, and the change of tumor cellular density shows a negative correlation with ADC change after the second NACT; however, the time of significant change in tumor size is after the second NACT. All the data suggest an ADC change after effective chemotherapy was closely related with cellular characteristics preceding size reduction, and ADC may be used as an early imaging biomarker of therapeutic response in LACC.
REFERENCES 1.
Foley G, Alston R, Geraci M, et al. Increasing rates of cervical cancer in young women in England: an analysis of national data 1982–2006. Br J Cancer 2011;105:177–184.
Volume 00, No. 00
Fu et al.: Correlation With ADC in NACT of LACC 2.
Angioli R, Plotti F, Montera R, et al. Neoadjuvant chemotherapy plus radical surgery followed by chemotherapy in locally advanced cervical cancer. Gynecol Oncol 2012;127:290–296.
3.
Gong L, Lou JY, Wang P, Zhang JW, Liu H, Peng ZL. Clinical evaluation of neoadjuvant chemotherapy followed by radical surgery in the management of stage IB2-IIB cervical cancer. Int J Gynaecol Obstet 2012;117:23–26.
4.
Ferrandina G, Distefano MG, De Vincenzo R, et al. Paclitaxel, epirubicin, and cisplatin (TEP) regimen as neoadjuvant treatment in locally advanced cervical cancer: long-term results. Gynecol Oncol 2013;128: 518–523.
5.
Landoni F, Sartori E, Maggino T, et al. Is there a role for postoperative treatment in patients with stage Ib2–IIb cervical cancer treated with neo-adjuvant chemotherapy and radical surgery? An Italian multicenter retrospective study. Gynecol Oncol 2014;132:611–617.
6.
7.
8.
Nishino M, Jagannathan JP, Ramaiya NH, Van den Abbeele AD. Revised RECIST guideline version 1.1: what oncologists want to know and what radiologists need to know. Am J Roentgenol 2010;195:281– 289. Liu Y, Bai R, Sun H, Liu H, Zhao X, Li Y. Diffusion-weighted imaging in predicting and monitoring the response of uterine cervical cancer to combined chemoradiation. Clin Radiol 2009;64:1067–1074. Haack S, Pedersen EM, Jespersen SN, Kallehauge JF, Lindegaard JC, Tanderup K. Apparent diffusion coefficients in GEC ESTRO target volumes for image guided adaptive brachytherapy of locally advanced cervical cancer. Acta Oncol 2010;49:978–983.
15.
Makino H, Kato H, Furui T, Morishige KI, Kanematsu M. Predictive value of diffusion-weighted magnetic resonance imaging during chemoradiotherapy for uterine cervical cancer. J Obstet Gynaecol Res 2014;40:1098–1104.
16.
Naganawa S, Sato C, Kumada H, Ishigaki T, Miura S, Takizawa O. Apparent diffusion coefficient in cervical cancer of the uterus: comparison with the normal uterine cervix. Eur Radiol 2005;15:71–78.
17.
Fu C, Bian D, Liu F, Feng X, Du W, Wang X. The value of diffusionweighted magnetic resonance imaging in assessing the response of locally advanced cervical cancer to neoadjuvant chemotherapy. Int J Gynecol Cancer 2012;22:1037–1043.
18.
Levy A, Caramella C, Chargari C, et al. Accuracy of diffusionweighted echo-planar MR imaging and ADC mapping in the evaluation of residual cervical carcinoma after radiation therapy. Gynecol Oncol 2011;123:110–115.
19.
Olsen JR, Esthappan J, DeWees T, et al. Tumor volume and subvolume concordance between FDG-PET/CT and diffusion-weighted MRI for squamous cell carcinoma of the cervix. J Magn Reson Imaging 2013;37:431–434.
20.
Albertella MR, Green CM, Lehmann AR, O’Connor MJ. A role for polymerase eta in the cellular tolerance to cisplatin-induced damage. Cancer Res 2005;65:9799–9806.
21.
Zhang L, Li J, Jiang Z, Sun L, Mei X, Yong B, Zhang L. Inhibition of aquaporin-1 expression by RNAi protects against aristolochic acid Iinduced apoptosis in human proximal tubular epithelial (HK-2) cells. Biochem Biophys Res Commun 2011;405:68–73.
9.
Chen J, Zhang Y, Liang B, Yang Z. The utility of diffusion-weighted MR imaging in cervical cancer. Eur J Radiol 2010;74:e101–106.
10.
Harry VN, Semple SI, Gilbert FJ, Parkin DE. Diffusion-weighted magnetic resonance imaging in the early detection of response to chemoradiation in cervical cancer. Gynecol Oncol 2008;111:213–220.
22.
Chen R, Shi Y, Amiduo R, Tuokan T, Suzuk L. Expression and prognostic value of aquaporin 1, 3 in cervical carcinoma in women of Uygur ethnicity from Xinjiang, China. PLoS One 2014;11:e98576.
11.
Kim HS, Kim CK, Park BK, Huh SJ, Kim B. Evaluation of therapeutic response to concurrent chemoradiotherapy in patients with cervical cancer using diffusion-weighted MR imaging. J Magn Reson Imaging 2013;37:187–193.
23.
Shi YH, Chen R, Talafu T, Nijiati R, Lalai S. Significance and expression of aquaporin 1, 3, 8 in cervical carcinoma in Xinjiang Uygur women of China. Asian Pac J Cancer Prev 2012;13:1971–1975.
12.
Chopra S, Verma A, Kundu S, et al. Evaluation of diffusion-weighted imaging as a predictive marker for tumor response in patients undergoing chemoradiation for postoperative recurrences of cervical cancer. J Cancer Res Ther 2012;8:68–73.
24.
Kim JK, Kim KA, Park BW, Kim N, Cho KS. Feasibility of diffusionweighted imaging in the differentiation of metastatic from nonmetastatic lymph nodes: early experience. J Magn Reson Imaging 2008;28: 714–719.
13.
Somoye G, Harry V, Semple S, et al. Early diffusion weighted magnetic resonance imaging can predict survival in women with locally advanced cancer of the cervix treated with combined chemo-radiation. Eur Radiol 2012;22:2319–2327.
25.
Zhang Y, Chen J, Shen J, Zhong J, Ye R, Liang B. Apparent diffusion coefficient values of necrotic and solid portion of lymph nodes: differential diagnostic value in cervical lymphadenopathy. Clin Radiol 2013; 68:224–231.
14.
Rizzo S, Summers P, Raimondi S, et al. Diffusion-weighted MR imaging in assessing cervical tumour response to nonsurgical therapy. Radiol Med 2011;116:766–780.
26.
Lin G, Ho KC, Wang JJ, et al. Detection of lymph node metastasis in cervical and uterine cancers by diffusion-weighted magnetic resonance imaging at 3T. J Magn Reson Imaging 2008;28:128–135.
Month 2014
9
