Abstract
Objectives:
To investigate the efficacy of pretreatment ADC and Ki-67 index in the prediction of the response to neoadjuvant chemotherapy (NACT) in advanced olfactory neuroblastoma (ONB) patients.
Methods:
A total of 21 advanced ONB patients (mean 43.48 years ± 14.26; range 25–69 years; 13 men and 8 women) with diffusion-weighted imaging (DWI) before NACT between June 2015 and October 2021 were retrospectively analyzed. Patients were categorized into responders and non-responders according to RECIST 1.1 after two cycles of NACT. The clinical data, ADCmean value, and Ki-67 index were analyzed.
Results:
Kadish stage, ADCmean value, and Ki-67 index showed statistical significance between responders and non-responders. Patients with Kadish C stage were more likely to respond to platinum-based NACT (p = 0.035). Patients with the lower ADCmean value showed response to NACT (p = 0.002) and the cutoff point was 1.04 × 10−3 mm2/s. Patients with the higher Ki-67 index showed response to NACT (p = 0.003) and the cutoff point was 17.5%. Predictive performance of Ki-67 index and ADCmean value showed no significance between responders and non-responders (p = 0.865). A significant negative correlation was found between ADCmean value and Ki-67 index (r = −0.539, p = 0.038).
Conclusions:
The pretreatment ADCmean value, Ki-67 index and Kadish stage have the potential to predict the response to NACT in advanced ONB patients.
Advances in knowledge:
This is the first study that investigated the feasibility of DWI in predicting the response to NACT in ONB patients and showed that Kadish stage, pretreatment ADCmean and Ki-67 index may play an important role in the prediction.
Introduction
Olfactory neuroblastoma (ONB) is a rare malignant neuroectodermal neoplasm that originates from the olfactory membrane of the sinonasal tract. 1–3 At present, minimally invasive surgical resection is the routine management for ONB. 1,4 Endoscopic surgery (ESS) followed by post-operative adjuvant radiation (RT) can achieve satisfactory results, 3,4 especially in the patients with ONB limited to nasal cavity or the paranasal (Kadish stage A, B). However, in the advanced tumors with extensive intracranial or intraorbital involvement (Kadish stage C) and those with neck or distant metastasis (Kadish stage D), radical resection cannot be achieved only by ESS. Therefore, neoadjuvant therapy (NACT) has been increasingly used to improve the rates of resectability by ESS, and control early micrometastatic disease. 2,4–7 However, the use of NACT is associated with chemotherapy toxicity. 6 We found that some patients with ONB could not respond to NACT. For this reason, the identification of biomarkers to predict the response to NACT is essential to improving patient management.
Ki-67 index is a biomarker to evaluate proliferation activity, biological behavior, and prognosis of many tumors. 8,9 However, the detection of Ki-67 expression is routinely available through biopsy. Considering the procedure-related complications, it is urgent to develop non-invasive methods to predict the proportion of proliferating cells in a tumor.
Recent advances in magnetic resonance imaging (MRI) make it possible to assess and predict the response to NACT. 10–14 One common MRI technique is diffusion weighted imaging (DWI), which allows quantifying the random motion of water molecules in tissue by means of apparent diffusion coefficient (ADC). 13,14 In particular, ADC values can indicate the alterations in cellularity and show the potential to evaluate cell proliferation before these changes become obvious in conventional imaging. 15–17 In addition, DWI can be acquired by fast sequences, without the administration of contrast media. These properties make ADC a quantitative biomarker for clinical practice. 10,11,13,14 Some studies showed that ADC values were negatively correlated with Ki-67 expression in many tumors. 16,17 However, the correlation between ADC values and Ki-67 expression in ONBs remains unknown.
The aim of this study is to investigate the feasibility of DWI in predicting the response to NACT in ONB patients, and to explore the correlation between ADC values and Ki-67 index.
Methods and materials
Patients
This retrospective study was approved by the institutional review board of our hospital with a waiver of informed consent. We reviewed electronic medical records of ONB patients who received NACT with MRI examinations between June 2015 and October 2021. The inclusion criteria were as follows: Patients with an advanced form of ONB proved by histopathology (Kaddish C or D); received platinum-based NACT without any preceding treatment option, with available initial pre-treatment with DWIs (b = 1000 s/mm2) and post-treatment follow-up MRI scans, and complete clinical and response data in the medical records. Patients who did not meet the above criteria were excluded. Seven patients underwent MR examinations before NACT without DWI sequence were excluded. One patient whose MRI data was missing was excluded. One patient who had intact MRI data before NACT but without medical record on NACT was also excluded. One patient was excluded from the group because she didn’t complete the whole two cycles NACT. One patient was excluded because he received partial resection of tumor before MRI. The last one was excluded because his ONB stage was B. Finally, 21 patients were enrolled in this study (Figure 1).
Figure 1.
Flow diagram of the study cohort selection.
Paranasal sinus MRI protocol
MRI examinations were performed on 3.0 T MRI (GE HDxt, Milwaukee, WI, USA) using an eight-channel head coil. The MRI scanning protocol included axial T 1-weighted fast spin echo image (T1WI-FSE), T 2-weighted fast spin echo image (T2WI-FSE), DWI, contrast-enhanced sagittal, coronal and axial T1WI. Post-contrast T1WI with frequency-selective fat suppression (FS) was used on both axial and coronal plane. The scanning parameters of T1WI (repetition time [TR]/echo time [TE], 400–500 ms/10 ms; matrix 320 × 256; field of view [FOV] 20 × 20 cm; slice thickness 4–5 mm; gap 0.5 mm; number of slices 24) and T2WI (TR/ TE, 3500–4000 ms/90 ms; matrix 512 × 256; FOV 20 × 20 cm; slice thickness 4–5 mm; gap 0.5 mm; number of slices 24). Axial periodically rotated overlapping parallel lines with enhanced reconstruction (PROPELLER) DWI was performed with two b-values (b = 0 and 1000 s/mm2) before injection with parameters as follow: TR/TE 4000 ms/75 ms, slice thickness 4 mm, gap 0.5 mm, FOV 20 × 20 cm, matrix 128 × 128. The scanning time for DWI was 2 min 16 sec.
Data collection and MRI analysis
ONBs were staged according to the modified Kadish staging system. 1,3,18 The response to NACT was assessed according to the Response Evaluation Criteria in Solid Tumors criteria version 1.1 (RECIST 1.1) after two cycles of NACT. 6,19 The following criteria were used: complete response (CR), disappearance of all target lesions; partial response (PR), > 30% decrease in the longest diameter of the lesion; progressive disease (PD), at least 20% increase in the longest diameter of the lesion; and stable disease (SD), neither qualifying the criteria for PD nor for PR. The tumor response to NACT was assessed based on the longest ONB diameter on coronal contrast-enhanced T1WI. Responders achieved CR or PR, while nonresponders achieved SD or PD. The clinical and pathological data of 21 patients, including age, sex, Kadish stage, different chemotherapy regimens, proliferation marker Ki-67 index and response results were recorded. Additionally, the interval between pre-treatment MRI examination and NACT (interval 1) and the interval between the start of NACT and the first follow-up MRI examination after two cycles of NACT (interval 2) were collected.
MR images of the lesions were analyzed by two radiologists (Y.W. and W.L.Y., with 8 and 20 years of diagnosis experience in head and neck MRI, respectively). Both radiologists were blinded to response status to NACT in every patient. The raw DWI data were transferred to a workstation (Advantage Workstation, AW 4.5, GE Medical Systems) and processed using Function tool software. Regions of interest (ROI)s were delineated at the level of the largest axial dimensions, based on the axial DWI (b = 0 s/mm2, 1000 s/mm2) maps and linked to ADC map images. ROIs were delineated comprising the whole lesion, while excluding necrotic or cystic portions of the tumors. ADC values were automatic calculated.
Immunohistochemistry
Formalin and paraffin-embedded tissue samples were stained by hematoxylin-eosin staining and Ki-67 immunohistochemical staining to determine the histologic type and detected Ki-67 expression of the tumor samples after biopsy. The Ki-67 index was evaluated by calculating the frequency of Ki-67 positive cells and expressed as the percentage.
Statistical analysis
Statistical analyses were performed using SPSS software (version 22.0, IBM Corp). Agreements were evaluated by using intraclass correlation coefficient (ICC) analysis. Shapiro Wilk test was performed for the normality of the test. Measurement variables were expressed as means ± standard deviation (SD) or medians with interquartile ranges (IQRs) . Independent sample t-test, Mann-Whitney test and Fisher’s exact test were used to compare responders (CR and PR) and non-responders (SD and PD). Area under the curves (AUCs) of the receiver operating characteristic curve (ROC) and 95% confidence interval (CI) were used to assess diagnostic performance. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy were calculated. Correlation analyses were performed using Pearson’s test or Spearman correlation analysis. All statistical tests were two sided and results were considered significant if p < 0.05. ROCs were calculated using software Medcalc (version 20.010).
Results
Patient characteristics
Twenty-one patients (13 men and 8 women; mean age 43.48 years ± 14.26; range, 25–69 years) were included in this study. Baseline clinical data and pathological characteristics of the final study cohort were listed in Table 1. According to RECIST 1.1, 6,19 ten patients were classified as responders (all PR patients) (Figure 2), and 11 patients were classified as non-responders including 10 SD patients (Figure 3) and one PD patient.
Table 1.
Baseline clinical data and pathological characteristics between Responders and Non-responders
| Characteristics | Responders (n = 10) | Non-responders (n = 11) | P value |
|---|---|---|---|
| Age (y)* | 41.40 ± 14.89 | 45.36 ± 14.09 | 0.538 |
| Sex | 0.183 | ||
| Male | 8 (38.1%) | 5 (23.8%) | |
| Female | 2 (9.5%) | 6 (28.6%) | |
| Interval 1 (d) * | 13.20 ± 10.36 | 12.91 ± 12.97 | 0.956 |
| Interval 2 (d) * | 51.10 ± 9.96 | 47.45 ± 3.21 | 0.293 |
| Kadish classification | 0.035 | ||
| C | 10 (47.6%) | 6 (28.6%) | |
| D | 0 (0.0%) | 5 (23.8%) | |
| Ki-67 index (%)(n = 15)# | 57.50 ± 24.85 | 18.22 ± 17.39 | 0.003 |
| Regimens | 0.442 | ||
| TP | 1 (4.8%) | 3 (14.3%) | |
| TEP | 6 (28.6%) | 4 (19.0%) | |
| TPI | 1 (4.8%) | 3 (14.3%) | |
| TP + VP-16 | 2 (9.5%) | 1 (4.8%) |
TP: paclitaxel combined with cisplatin (or Nedaplatin); TEP: paclitaxel combined with cisplatin (or Nedaplatin) and etoposide triple therapy; TPI: paclitaxel combined with cisplatin (or Nedaplatin) and ifosfamide; TP +VP-16: paclitaxel combined with cisplatin (or Nedaplatin) followed by etoposide sequential therapy.
Note. Unless otherwise noted, values are numbers of patients with percentages in parentheses. * Numbers are means ± standard deviations. # Six cases of Ki-67 index reports are unavailable.
Figure 2.
An ONB (Kadish C) patient presented partial response (PR) after two cycles of NACT. a. coronal enhanced T1WI before NACT. b. coronal enhanced T1WI after NACT. c. pretreatment ADC map (b value = 1000 s/mm2). The tumor size was significantly reduced after NACT (indicated by white arrows). The ADCmean value was 0.89 × 10−3 mm2/s. Ki-67 index was 20%.
Figure 3.
An ONB (Kadish C) patient presented stable disease (SD) after two cycles of NACT. a. coronal enhanced T1WI before NACT. b. coronal enhanced T1WI after NACT. c. pretreatment ADC map (b value = 1000 s/mm2). There were no significant changes in shape, size, or enhancement of the tumor after NACT (indicated by white arrows). The ADCmean value was 1.30 × 10−3 mm2/s. Ki-67 index was less than 3%.
Agreement analysis
There was a high agreement of ADC value between two observers with ICC of 0.968 (95% CI: 0.907, 0.988, p < 0.001). The mean of ADC between two radiologists were calculated and recorded as ADCmean in Table 2.
Table 2.
Agreement of ADC values and average ADC value of two radiologists
| ADC value (× 10−3 mm2/s) | ICC | P value | ADCmean (× 10−3 mm2/s) | ||
|---|---|---|---|---|---|
| Radiologist 1 | Radiologist 2 | ||||
| All | 1.03 ± 0.25 | 1.06 ± 0.28 | 0.968 (0.907, 0.988) | < 0.001 | 1.04 ± 0.27 |
| Responders | 0.86 ± 0.07 | 0.87 ± 0.09 | 0.877 (0.585,0.968) | < 0.001 | 0.87 ± 0.08 |
| Non-responders | 1.18 ± 0.26 | 1.23 ± 0.29 | 0.953 (0.715,0.989) | < 0.001 | 1.21 ± 0.27 |
ICC: intraclass correlation coefficient, ADCmean: mean apparent diffusion coefficient value of two readers.
Factors associated with response status of ONB to NACT and ROC analysis
Sex (p = 0.183) and age (p = 0.538) showed no statistical significance between the responders and non-responders (Table 1). The interval 1 was 13.20 ± 10.36 days in responders and 12.91 ± 12.97 days in non-responders. The interval 2 was 51.10 ± 9.96 days in responders and 47.45 ± 3.21 days in non-responders. Neither the interval 1 (p = 0.956) nor the interval 2 (p = 0.293) showed statistical significance between the two groups (Table 1).
Compared with stage D patients, Kadish C stage patients were more likely to respond to NACT (p = 0.035, Table 1). Patients with higher Ki-67 index showed response to NACT (p = 0.003, Table 1). Pretreatment ADCmean values were lower in responders than in non-responders (p = 0.002) (Figures 2 and 3, Tables 2 and 3). On the basis of the Youden index from the ROC (Figure 4a), the cutoff value of ADCmean value for differentiating two groups were 1.04 × 10−3 mm2/s (AUC = 0.909; 95% CI: 0.702 to 0.990; p < 0.001). The cutoff value of ADCmean value and Ki-67 index of 15 ONB patients with Ki-67 index were 1.04 × 10−3 mm2/s (AUC = 0.889; 95% CI: 0.623 to 0.990; p < 0.001) and 17.5% (AUC = 0.907; 95% CI: 0.646 to 0.994; p < 0.001), respectively (Table 3). No significant differences were observed between ADCmean value and Ki-67 index in distinguishing responders from non-responders (p = 0.865, Figure 4b). The sensitivity, specificity, PPV, NPV and accuracy were listed in Table 3. In 16 patients with Kadish C, we observed the tendency that responders had lower ADCmean value than non-responders (p = 0.001, Table 4).
Table 3.
Diagnostic performance of pretreatment ADC value and Ki-67 index between Responders and Non-responders
| Responders vs
Non-responders |
AUC (95% CI) | Accuracy (%) |
Sensitivity (%) |
Specificity (%) |
PPV (%) |
NPV (%) |
||
|---|---|---|---|---|---|---|---|---|
| T value | P value | |||||||
| Ki-67 (n = 15） |
T = −3.621 | 0.003 | 0.907 (0.646,0.994) | 80.0 (12/15) | 100.0 (6/6) | 66.7 (6/9) | 66.7 (6/9) | 100.0 (6/6) |
| ADCmean* (n = 15） |
T = 3.21 | 0.007 | 0.889 (0.623, 0.990) | 86.7 (13/15) | 100.0 (6/6) | 77.8 (7/9) | 75.0 (6/8) | 100.0 (7/7) |
| ADCmean
(n = 21） |
T = 3.922 | 0.002 | 0.909 (0.702,0.990） | 90.5 (19/21) | 100.0 (10/10) | 81.8 (9/11） | 83.3 (10/12) | 100.0 (9/9) |
Note. 1.04 × 10−3 mm2/s of ADCmean value and 17.5% of Ki-67 index were the cutoff value; PPV: positive predictive value; NPV: negative predictive value; AUC: area under the curve. * mean ADC value of 15 patients with Ki-67 index.
Figure 4.
Receiver operating characteristic curves (ROCs) of ADCmean value and Ki-67 index to differentiate between responders and non-responders. a. The area under the ROC curve (AUC) of ADCmean of all 21 ONB patients was 0.909. b.The comparison of ROCs of Ki-67 index and ADCmean value for 15 patients with Ki-67 index. The AUC of two indices,Ki-67 index and ADCmean, was 0.907 and 0.889 respectively. ADC value had been transformed to a negative form (nADC).
Table 4.
Comparison of ADCmean between responders and non-responders in ONB with Kadish C group or patients received TEP NACT regimen
| ADCmean (× 10−3 mm2/s) | Statistic | P value | |
|---|---|---|---|
| Kadish C (n = 16) | T = 4.003 | 0.001 | |
| Responders (n = 10) | 0.87 ± 0.08 | ||
| non-responders (n = 6) | 1.30 ± 0.33 | ||
| TEP (n = 10) | Z = −2.558 | 0.001 | |
| Responders(n = 6) | 0.87 (0.79–0.98） | ||
| non-responders(n = 4) | 1.11 (1.09–1.36） |
Data from Kadish C group were means ± standard deviation; Data from TEP group were medians, with interquartile ranges in parentheses.
Different platinum-based chemotherapy regimens showed no statistical significance between responders and non-responders including TP [paclitaxel with cisplatin (or Nedaplatin)], TEP [paclitaxel with cisplatin (or Nedaplatin) combined with etoposide triple therapy], TPI [paclitaxel with cisplatin (or Nedaplatin) plus ifosfamide] and TP +VP-16 [paclitaxel with cisplatin (or Nedaplatin) followed by etoposide sequential therapy] (p = 0.442). However, among the 10 patients who received TEP regimens, responders showed lower ADCmean than non-responders (p = 0.001, Table 4).
Correlation analysis
Kadish stage did not show significant correlation with Ki-67 index (r = −0.455, p = 0.088) or ADCmean value (r = 0.277, p = 0.224).
Among 21 ONB patients, 15 patients had Ki-67 index, including six responders and nine non-responders. We found negative correlation between Ki-67 index and ADCmean value (r = −0.539, p = 0.038, Figure 5). Ki-67 index of ≥17.5% and ADCmean value of ≤1.04 ×10−3 mm2/s were combined to predict the response to NACT. Among 15 patients, 10 patients whose ADCmean value showed inverse correlation with Ki-67 index were predicted correctly. The 5 patients whose ADCmean value did not show inverse correlation with Ki-67 index were proved non-responders (Table 5).
Figure 5.
The correlation of ADCmean value with Ki-67 index. A negative correlation was observed between ADCmean values and Ki-67 index for ONB (r = −0.539, p = 0.038).
Table 5.
Inverse correlation between ADCmean value and Ki-67 index and the accuracy for prediction
| Inverse correlation or not (n = 15) | Ki-67 index (≥17.5%) | ADCmean value (≤1.04×10−3 mm2/s) | Number (n) | Prediction | RECIST result | ADC Accuracy |
|---|---|---|---|---|---|---|
| yes | low | high | 4 | Non-response | Non-response | right |
| yes | high | low | 6 | Response | Response | right |
| no | high | high | 3 | —— | Non-response | right |
| no | low | low | 2 | —— | Non-response | wrong |
RECIST: Response Evaluation Criteria in Solid Tumors.
Discussion
In this study we found that Kadish stage, ADCmean value and Ki-67 index were predictive of response to NACT in ONB patients before treatment. Patients with Kadish stage D didn’t respond to NACT. Similar situation was reported in previous study which showed that patients with Kadish D stage who received platinum-based adjuvant chemotherapy did not exhibit significantly improved outcomes. 5 This may indicate that platinum-based agents may not be suitable for patients with Kadish D stage. However, larger studies are needed to confirm this possibility.
The expression of human Ki-67 protein is associated with cell proliferation; therefore, increased Ki-67 levels in cancer tissue have been considered a poor prognostic factor, while also being a positive predictive factor for response to chemotherapy. 19 Turri-Zanoni et al reported that Ki-67 index ≥20.0% for ONBs was associated with an increased risk of recurrence and poorer outcomes. 20 The result of our study with respect to cutoff value of Ki-67 index was 17.5%, close to the former. In this study, ONB patients with higher Ki-67 index also have been shown to respond better to NACT. These results suggested that the more active the proliferation, the more sensitive for cancer to platinum based chemotherapeutic agents. These agents exert anticancer activity via inhibiting cell cycle progression. 21
The present study showed that the responders had lower pretreatment ADCmean values than non-responders, consistent with the results on the prediction of breast cancer, liver cancer and cervical cancer for the response to NACT. 10,13 This trend was also observed in Kadish C group and TEP group in this study. The lower pretreatment ADCmean was often associated with higher cellularity, indicating that tumors with compact cellularity are more likely to respond to NACT. 22
An inverse relationship between ADCmean value and Ki-67 index were found in this study. Taking 17.5% of Ki-67 index and 1.04 × 10−3 mm2/ s of ADCmean values as cutoff points, we found that the majority of patients (10/15,66.7%) showed inverse correlation. This is in concordance with general rule that high cellular proliferation enhances cellular density, reduces extracellular volume and restricts the diffusion of water molecules, thereby decreasing ADC values. 13 The other 5 patients didn’t show the inverse correlation between ADCmean and Ki-67 index. In addition to tumor cellularity, other histological features such as cell size, nuclear atypia, mitotic activity and other factors may influence tissue diffusivity. 23,24 Notably, those 5 patients who didn’t show the inverse correlation between ADCmean and Ki-67 index were proved to be non-responders. Therefore, the correlation of ADCmean value and Ki-67 index may help predict non-response to NACT.
Our study has several limitations. First, this was a retrospective study performed at a single center with a relatively small sample size. This is unavoidable in the setting of ONBs with rare incidence. However, these cases were strictly enrolled according to the inclusion and exclusion criteria. Second, the ROIs for ADC values did not exactly point-to-point correspond to biopsy areas. Third, follow-up imaging ADC value was not included because some of them were not available. A prospective study with strict design, larger sample size, and more quantitative MRI parameters on ONB patients will be carried out in the future.
In conclusion, ONB patients with Kadish C stage responded better to platinum-based NACT compared with those with Kadish D stage. The pretreatment ADCmean value, Ki-67 index and inverse correlation between them showed the potential to predict treatment response to NACT in advanced ONB patients.
Footnotes
Acknowledgment: This work was supported by Professor Junfang Xian in Department of Radiology of Capital Medical University Affiliated Beijing Tongren Hospital.
Funding: This study was supported by Beijing Municipal Administration of Hospitals’ Ascent Plan (Code: DFL20190203). Beijing Municipal Administration of Hospitals Clinical Medicine Development of Special Funding Support (Code: ZYLX201704).
Contributor Information
Yuan Wang, Email: imagedoctor@126.com.
Shurong Zhang, Email: 13501085646@163.com.
Wenling Yu, Email: yuwenlingtr@163.com.
Yongzhe Wang, Email: yzwang1981@163.com.
Fei Yan, Email: yanever@163.com.
BenTao Yang, Email: ybt_108@163.com.
REFERENCES
- 1. Abdelmeguid AS . Olfactory neuroblastoma . Curr Oncol Rep 2018. ; 20 ( 1 ): 7 . doi: 10.1007/s11912-018-0661-6 [DOI] [PubMed] [Google Scholar]
- 2. Zhang S, Sun K, Zheng R, Zeng H, Wang S, Chen R, et al . Cancer incidence and mortality in China, 2015 . Journal of the National Cancer Center 2021. ; 1 : 2 – 11 . doi: 10.1016/j.jncc.2020.12.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Bell D . Sinonasal neuroendocrine neoplasms: current challenges and advances in diagnosis and treatment, with a focus on olfactory neuroblastoma . Head Neck Pathol 2018. ; 12 : 22 – 30 . doi: 10.1007/s12105-018-0887-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Bartel R, Gonzalez-Compta X, Cisa E, Cruellas F, Torres A, Rovira A, et al . Importance of neoadjuvant chemotherapy in olfactory neuroblastoma treatment: series report and literature review . Acta Otorrinolaringol Esp (Engl Ed) 2018. ; 69 : 208 – 13 . doi: 10.1016/j.otorri.2017.07.001 [DOI] [PubMed] [Google Scholar]
- 5. Miller KC, Marinelli JP, Van Gompel JJ, Link MJ, Janus JR, Foote RL, et al . Utility of adjuvant chemotherapy in patients receiving surgery and adjuvant radiotherapy for primary treatment of esthesioneuroblastoma . Head Neck 2019. ; 41 : 1335 – 41 . doi: 10.1002/hed.25558 [DOI] [PubMed] [Google Scholar]
- 6. Patil VM, Joshi A, Noronha V, Sharma V, Zanwar S, Dhumal S, et al . Neoadjuvant chemotherapy in locally advanced and borderline resectable nonsquamous sinonasal tumors (esthesioneuroblastoma and sinonasal tumor with neuroendocrine differentiation) . Int J Surg Oncol 2016. ; 2016 : 6923730 . doi: 10.1155/2016/6923730 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Su SY, Bell D, Ferrarotto R, Phan J, Roberts D, Kupferman ME, et al . Outcomes for olfactory neuroblastoma treated with induction chemotherapy . Head Neck 2017. ; 39 : 1671 – 79 . doi: 10.1002/hed.24822 [DOI] [PubMed] [Google Scholar]
- 8. Singh L, Ranjan R, Madan R, Arava SK, Deepak RK, Singh MK . Microvessel density and ki-67 labeling index in esthesioneuroblastoma: is there a prognostic role? Ann Diagn Pathol 2015. ; 19 : 391 – 96 . doi: 10.1016/j.anndiagpath.2015.06.008 [DOI] [PubMed] [Google Scholar]
- 9. Liu Y, Jia Y, Zhang B, Liu AW . Tug1 indicate unfavorable prognosis of gastric cancer for promoting proliferation, migration and multidrug resistance . Oncologie 2021. ; 23 : 61 – 72 . doi: 10.32604/Oncologie.2021.015906 [DOI] [Google Scholar]
- 10. Pham TT, Liney GP, Wong K, Barton MB . Functional MRI for quantitative treatment response prediction in locally advanced rectal cancer . Br J Radiol 2017. ; 90 : 1072 : 20151078 . doi: 10.1259/bjr.20151078 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Choi EJ, Youk JH, Choi H, Song JS . Dynamic contrast-enhanced and diffusion-weighted MRI of invasive breast cancer for the prediction of sentinel lymph node status . J Magn Reson Imaging 2020. ; 51 : 615 – 26 . doi: 10.1002/jmri.26865 [DOI] [PubMed] [Google Scholar]
- 12. Jiang JX, Tang ZH, Zhong YF, Qiang JW . Diffusion kurtosis imaging for differentiating between the benign and malignant sinonasal lesions . J Magn Reson Imaging 2017. ; 45 : 1446 – 54 . doi: 10.1002/jmri.25500 [DOI] [PubMed] [Google Scholar]
- 13. Thoeny HC, Ross BD . Predicting and monitoring cancer treatment response with diffusion-weighted MRI . J Magn Reson Imaging 2010. ; 32 : 2 – 16 . doi: 10.1002/jmri.22167 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Park SH, Moon WK, Cho N, Song IC, Chang JM, Park I-A, et al . Diffusion-Weighted MR imaging: pretreatment prediction of response to neoadjuvant chemotherapy in patients with breast cancer . Radiology 2010. ; 257 : 56 – 63 . doi: 10.1148/radiol.10092021 [DOI] [PubMed] [Google Scholar]
- 15. Cui Y, Zhang XP, Sun YS, Tang L, Shen L . Apparent diffusion coefficient: potential imaging biomarker for prediction and early detection of response to chemotherapy in hepatic metastases . Radiology 2008. ; 248 : 894 – 900 . doi: 10.1148/radiol.2483071407 [DOI] [PubMed] [Google Scholar]
- 16. Zhang K, Li C, Liu Y, Li L, Meng X, Feng D, et al . Evaluation of astrocytoma cell proliferation using diffusion-weighted imaging: correlation with expression of proliferating cell nuclear antigen . Transl Neurosci 2015. ; 6 : 265 – 70 . doi: 10.1515/tnsci-2015-0029 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Tang Y, Dundamadappa SK, Thangasamy S, Flood T, Moser R, Smith T, et al . Correlation of apparent diffusion coefficient with Ki-67 proliferation index in grading meningioma . AJR Am J Roentgenol 2014. ; 202 : 1303 – 8 . doi: 10.2214/AJR.13.11637 [DOI] [PubMed] [Google Scholar]
- 18. Marinelli JP, Janus JR, Van Gompel JJ, Link MJ, Foote RL, Lohse CM, et al . Esthesioneuroblastoma with distant metastases: systematic review & meta-analysis . Head Neck 2018. ; 40 : 2295 – 2303 . doi: 10.1002/hed.25209 [DOI] [PubMed] [Google Scholar]
- 19. Yang X, Ma K, Chen R, Zhang N, Meng Y, Wen J, et al . Immunocytochemical examination of PTEN and Ki-67 for endometrial carcinoma using thin-layer endometrial preparations . Biocell 2021. ; 45 : 923 – 32 . doi: 10.32604/biocell.2021.013625 [DOI] [Google Scholar]
- 20. Turri-Zanoni M, Maragliano R, Battaglia P, Giovannardi M, Antognoni P, Lombardi D, et al . The clinicopathological spectrum of olfactory neuroblastoma and sinonasal neuroendocrine neoplasms: refinements in diagnostic criteria and impact of multimodal treatments on survival . Oral Oncol 2017. ; 74 : 21 – 29 . doi: 10.1016/j.oraloncology.2017.09.010 [DOI] [PubMed] [Google Scholar]
- 21. Shin HJ, Baek H-M, Ahn J-H, Baek S, Kim H, Cha JH, et al . Prediction of pathologic response to neoadjuvant chemotherapy in patients with breast cancer using diffusion-weighted imaging and MRS . NMR Biomed 2012. ; 25 : 1349 – 59 . doi: 10.1002/nbm.2807 [DOI] [PubMed] [Google Scholar]
- 22. Wu X, Pertovaara H, Dastidar P, Vornanen M, Paavolainen L, Marjomäki V, et al . Adc measurements in diffuse large B-cell lymphoma and follicular lymphoma: a DWI and cellularity study . Eur J Radiol 2013. ; 82 : e158 – 64 . doi: 10.1016/j.ejrad.2012.11.021 [DOI] [PubMed] [Google Scholar]
- 23. Maitre E, Troussard X . Hairy cell leukemia and HCL-like disorders: diagnosis and treatmentleucie tricholeucocytes et autres prolifations cellules chevelues: diagnostic et traitement . Oncologie 2022. ; 24 : 3 – 24 . doi: 10.32604/oncologie.2022.021490 [DOI] [Google Scholar]
- 24. Surov A, Caysa H, Wienke A, Spielmann RP, Fiedler E . Correlation between different ADC fractions, cell count, Ki-67, total nucleic areas and average nucleic areas in meningothelial meningiomas . Anticancer Res 2015. ; 35 : 6841 – 46 . [PubMed] [Google Scholar]