Med Oncol (2015)32:164 DOI 10.1007/s12032-015-0579-x
REVIEW ARTICLE
Prognostic value of circulating tumor cells in patients with squamous cell carcinoma of the head and neck: a systematic review and meta-analysis Zhendan Wang1 • Kai Cui2 • Yan Xue3 • Feng Tong4 • Sheng Li2
Received: 25 November 2014 / Accepted: 14 March 2015 Ó Springer Science+Business Media New York 2015
Abstract The prognostic value of circulating tumor cells (CTCs) in patients with squamous cell carcinoma of the head and neck (SCCHN) is controversial. The objective of this study was to evaluate the prognostic value of CTCs in patients with SCCHN by conducting a meta-analysis. We systematically searched scientific literature published before June 10, 2014, using the PubMed, ScienceDirect, Cochrane Library, and EMBASE databases. Studies evaluating the correlation of CTC status with tumor-nodemetastasis (TNM) disease stage, nodal involvement, and disease progression (recurrence or metastasis) in patients with SCCHN were selected for the analysis. Pooled odds ratio (OR) and 95 % confidence interval (CI) were calculated using the fixed-effects model or the random-effects model in the presence of heterogeneity. Our analysis consisted of eight studies, enrolling a total of 433 patients. The disease progression (recurrence/metastasis) rate in the CTC-positive patients was significantly higher (OR 3.44; 95 % CI 1.87–6.33; p \ 0.01) compared with the patients without disease progression. However, there was no
& Feng Tong [email protected] & Sheng Li [email protected] 1
Shandong Academy of Medical Science, University of Jinan, Jinan 250022, China
2
Department of Hepatology, Shandong Cancer Hospital, Shandong Academy of Medical Science, 440 Jiyan Road, Jinan 250017, China
3
Department of Head and Neck Surgery, Qilu Children’s Hospital of Shandong University, Jinan 250022, China
4
Department of Dean’s Office, Shandong Cancer Hospital, 440 Jiyan Road, Jinan 250017, China
significant difference between TNM disease stage III–IV and stage I–II in the presence of CTCs (OR 1.54; 95 % CI 0.87–2.72; p [ 0.05). CTC status did not correlate with nodal involvement (OR 1.20; 95 % CI 0.67–1.90; p [ 0.05). This meta-analysis indicates that detection of CTCs has a predictive value in patients with SCCHN, particularly those with tumor progression. The presence of CTCs in patients with SCCHN has a poor prognosis compared with the patients without CTCs. Detection of CTCs might be served as a prognosticator in patients with SCCHN. Further studies based on homogeneous populations are warranted to confirm these findings. Keywords Circulating tumor cells Prognostic Head and neck squamous cell carcinoma Meta-analysis
Introduction Malignant tumors of the head and neck account for approximately 5–10 % of all malignant tumors. Nearly 95 % of the head and neck malignant tumors are squamous cell carcinoma originating from the mucosal epithelia, such as oral cavity, nasopharynx, pharynx, and larynx [1]. The incidence rates of squamous cell carcinoma of the head and neck (SCCHN) were about 10–15 per 10 million subjects, with over 500,000 new cases each year [2, 3]. Despite advances in the diagnosis and treatment in recent years, patients with tumor-node-metastasis (TNM) disease stage III–IV still account for about 50 %, and the 5-year survival rate has not significantly improved [4]. Therefore, early identification of metastasis and recurrence has become increasingly important for the tumor progression and prognosis in patients with SCCHN.
123
164
Page 2 of 8
Dissemination of tumor cells into the blood stream is an important step in tumor metastasis [5]. Circulating tumor cells (CTCs) are tumor cells detectable in peripheral blood released by primary tumors or thorough tumor recurrences and metastases. The first report on tumor cells in bloodstream was reported from Ashworth in 1869 [6]. In recent years, various methods have been developed for CTCs identification. Detecting the expression of epithelial and cancer-specific markers is the most widely used method to identify CTCs. Based on this principle, the CellSearch system was approved by the US Food and Drug Administration (FDA) for clinical use in 2004 [7]. The status of CTCs plays an important role in the early diagnosis, prognosis, and instruction of individual therapy [8, 11, 12]. At present, detection of CTCs has been widely applied in various malignant tumors, including breast cancer [8], lung cancer [9], hepatocellular carcinoma [10], prostate cancer [11], and colorectal cancer [12]. The prognostic value of CTCs detected in patients with SCCHN has been investigated in many studies. However, whether the CTC status is associated with the disease stage, lymph node involvement, or prognosis remains controversial. Here, we conducted the first meta-analysis using currently available studies to quantitatively assess the prognostic value of CTCs in patients with SCCHN.
Med Oncol (2015)32:164
tumors, but with no specific results of SCCHN; (2) lacking key information to calculate the risk estimates; and (3) reviews, letters, or case report types of study design. Data extraction and quality assessment
Materials and methods
Two authors (ZD Wang and K Cui) independently identified the eligible articles and collected the following data: first author’s surname, year of publication, country of study conducted, sample size, number of CTC-positive patients, volumes and site of sample collection, sampling time point, tumor stage, nodal involvement, progression (recurrence or metastasis), methods to detect CTCs, type of tumor markers analyzed, and follow-up duration. Any disagreements were resolved by discussion. The methodological quality of the included studies was assessed in accordance with the risk bias guidelines in the Cochrane reviewers’ Handbook 5.1.0. The following study features were assessed: (1) Were adequate eligibility criteria developed and applied? (2) Was the measurement of both exposure and outcome adequate? (3) Was the confounding adequately controlled? (4) Was the follow-up complete and adequate in duration? (5) Were study reports free of suggestion of selective outcome reporting? (6) Was the study free of other high risk of bias? If all quality criteria were met, the study was sorted into low risk of bias, and if one or more criteria not met, the trial was sorted into high risk of bias.
Search strategy
Statistical methods
We systematically searched the PubMed, ScienceDirect, Cochrane Library, and EMBASE databases for studies prior to June 10, 2014. Potentially relevant studies were identified using the following key words: ‘‘circulating tumor cell(s),’’in combination with ‘‘head and neck cancer,’’ ‘‘carcinoma of the head and neck,’’ or ‘‘squamous cell carcinoma of the head and neck.’’ No language restrictions were imposed. Additionally, reference lists of retrieved articles were screened manually to identify additional eligible studies.
Included patients were stratified according to the CTC status and the subsequent comparisons of presence of CTCs versus absence of CTCs. According to clinical characteristics, stage I and stage II were combined and stage III and stage IV were combined. Odds ratio (OR) and 95 % confidence interval (CI) were computed to determine the effect of CTC status on clinical outcomes. OR[1 indicated worse outcome for the positive CTCs group relative to the negative CTCs group. Heterogeneity was tested using the Isquare and Cochrane Q statistic. We defined the Q statistic (p \ 0.10) or I-square [50 % as significant heterogeneity; then, a fixed-effects model was used for analysis; otherwise, a random-effects model was selected. The potential publication bias was evaluated using the funnel plot. To investigate the change of the overall effect size, the sensitivity analysis was performed omitting one study at each turn. All analyses were performed with RevMan software of the Cochrane Collaboration (version 5.2, Oxford, UK). A two-sided p \ 0.05 was considered statistically significant.
Studies selection Studies satisfying following criteria were selected: (1) patients with histological diagnosis of SCCHN; (2) detection of CTCs in peripheral blood or bone marrow; and (3) investigation of association between CTC status and clinical outcome (TNM disease stage, nodal involvement, and disease recurrence/metastasis). Studies were excluded if they met the following criteria: (1) patients with various
123
Med Oncol (2015)32:164
Results Literature search A total of 157 potentially relevant articles were collected from the initial electronic search. After reviewing the title and abstract, 143 relevant studies were excluded, and six studies were additionally excluded after reviewing full-text manuscript. Finally, eight studies [13–20] met our predefined inclusion criteria (Fig. 1). Characteristics of identified studies and risk of bias Characteristics of the included studies are listed in Table 1. A total of 433 patients were identified. The total number of patients ranged from 9 to 176 in the individual studies. The studies were conducted in Italy, Germany, Britain, France, Turkey, China, and USA and published from 2004 to 2013. The CTCs were collected from peripheral blood in seven studies [13, 15, 17–20], and one study [14] collected CTCs from the bone marrow. The sample collecting ranged from 7.5 to 30 ml before the treatment, except for one study [18], which collected sample during treatment. CTCs were detected using CellSearch [13, 15–17], immunocytochemistry alkaline phosphatase and monoclonal anti-alkaline phosphatase (ICC-APAAP) [14], ICC [18], flow cytometry [19], and magnetic activated cell sorting(MACS) [20] analysis. The detected tumor markers mainly included epithelial cell adhesion molecule (EpCAM), CD45,
Page 3 of 8 164
cytokeratins (CK), and 40 ,6-diamidino-2-phenylindole (DAPI). The included studies were classified as low risk of bias, and only one study [19] was considered as high risk of bias due to incomplete follow-up duration (Table 2). CTC status and TNM disease stage Four studies [14, 15, 18, 20] reported CTC status and TNM disease stage, and stage III–IV group included 77 CTC cases among 226 patients, while the stage I–II group included 29 CTC-positive cases among 92 patients. As shown in Fig. 2, a fixed-effects model was selected because there was no obvious sample heterogeneity (I2 = 0 %; p = 0.48). There was no significant difference between TNM disease stage III–IV and stage I–II patients in the presence of CTCs (OR 1.54; 95 % CI 0.87–2.72; p = 0.13). CTC status and nodal involvement Seven studies [13–19] provided the data of CTC status and nodal involvement, and the group with nodal involvement included 77 CTC-positive cases among 241 patients, while the group without nodal involvement included 57 CTCpositive cases among 171 patients. As shown in Fig. 3a, a fixed-effects model was selected because there was no obvious sample heterogeneity (I2 = 0 %; p = 0.44). There was no significant difference between patient with nodal involvement and without nodal involvement in the pres-
Fig. 1 Flow chart of study selection process
123
164
Page 4 of 8
Med Oncol (2015)32:164
Table 1 Baseline characteristics of included studies Study/ Reference
Country
CTC-positive patients (n/N)
Sampling site/volume
Sampling time
Tumor stage (UICC)
Detection method
Type of tumor markers
Follow-up duration (months)
Buglione et al. [13]
Italy
11/73
PB/ C7.5 ml
Before TM
I–IV
CellSearch
EpCAM, CD45,CK,DAPI
13.6
Wollenberg et al. [14]
Germany
54/176
BM/–
Before TM
I–IV
IHCAPAAP
CK19
60
Nichols et al. [15]
Britain
6/15
PB/10 ml
Before TM
III–IV
CellSearch
EpCAM, CD45, CK, DAPI
8
He et al. [16]
China
3/9
PB/7.5 ml
Before TM
III–IV
CellSearch
EpCAM, CD45, CK, DAPI
16
Bozec et al. [17]
France
8/49
PB/7.5 ml
Before TM
III–IV
CellSearch
EpCAM, CD45, CK, DAPI
6
Jatana et al. [18]
USA
34/48
PB/ 10–18 ml
During TM
I–IV
ICC
CK, CD45, DAPI
38
Hristozova et al. [19]
Germany
18/42
PB/7.5 ml
Before TM
I–IV
Flow Cytom
EpCAM, CK
–
Guney et al. [20]
Turkey
7/21
PB/30 ml
Before TM
I–IV
MACS
–
36
PB peripheral blood, BM bone marrow, TM treatment, EpCAM epithelial cell adhesion molecule, CK cytokeratins, DAPI 40 ,6-diamidino-2phenylindole
Table 2 Quality assessment of the included studies Study/ Reference
Were adequate eligibility criteria developed and applied?
Was the measurement of both exposure and outcome adequate?
Was confounding adequately controlled for?
Was the followup complete and adequate in duration?
Are reports of the study free of suggestion of selective outcome reporting?
Was the study free of other problems that put it at a high risk of bias?
Risk of bias
Buglione et al. [13]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Wollenberg et al. [14] Nichols et al. [15]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Yes
Yes
Yes
Yes
Yes
Yes
Low
He et al. [16]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Bozec et al. [17]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Jatana et al. [18]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Hristozova et al. [19]
Yes
Yes
Yes
No
Yes
Yes
High
Guney et al. [20]
Yes
Yes
Yes
Yes
Yes
Yes
Low
ence of CTCs (OR 1.20; 95 % CI 0.76–1.79; p = 0.44). Subgroup analysis using CellSearch in four studies [13, 15–17] indicated that the pooled OR was 1.44 (95 % CI 0.56–3.71; p = 0.45) when the patients with nodal involvement were compared to patients without nodal involvement in the presence of CTCs (Fig. 3b).
123
CTC status and disease progression (recurrence/ metastasis) Four studies [15–17, 20] reported the data of CTC status and progression (recurrence/metastasis), and the CTCpositive group included 35 cases of disease progression
Med Oncol (2015)32:164
Page 5 of 8 164
Fig. 2 Forest plots showing comparison of circulating tumor cells-positive rate between the tumor stage I–II and III–IV groups in a fixed-effects model
Fig. 3 Forest plots showing CTC status and nodal involvement in the eligible studies (a) and detection methods by CellSearch subgroup in a fixed-effects model (b)
among 69 patients, while the CTC-negative group included 35 cases of disease progression among 150 patients. As shown in Fig. 4, a fixed-effects model was selected due to no obvious heterogeneity (I2 = 13 %; p = 0.33). The disease progression (recurrence/metastasis) rate in the CTC-positive group was significantly higher compared with the without disease progression group (OR 3.44; 95 % CI 1.87–6.33; p \ 0.01). Publication bias and sensitivity analysis There were no studies reporting OR of TNM disease stage, nodal involvement, and disease progression (recurrence/
metastasis) by inspecting the funnel plots (Fig. 5). The sensitivity analysis indicated that there was only a slight change in the quantitative summary measure of OR or 95 % CI, but no change in the direction of OR if any study was omitted in the analysis.
Discussion Tumor metastasis is an important contributor to mortality in cancer patients. However, the existing image technology and tumor markers are not sufficient to detect tumor metastasis in early stages and estimate the prognosis. CTCs
123
164
Page 6 of 8
Med Oncol (2015)32:164
Fig. 4 Forest plots showing comparison of circulating tumor cells-positive rate between the disease progression (recurrence or metastasis) and without disease progression groups in a fixed-effects model
Fig. 5 Funnel plots based on TNM disease stage (a), nodal involvement (b), disease progression (c), and nodal involvement in CellSearch subgroup (d)
accompany tumor invasion into the bloodstream. CTC status provides a real-time and minimally invasive approach for the detection of early disease and assessing prognosis, and therapeutic response in established cancers [21]. There is no consensus on whether CTC status has a prognostic value in patients with SCCHN. To the best of our knowledge, this is the first metaanalysis that assesses the prognostic value of CTCs for SCCHN. Overall, the eight included studies were classified as featuring a low risk of study bias. The current metaanalysis reveals that detection of CTCs has a predictive value on the tumor progression in patients with SCCHN. However, the presence of CTCs in patients with SCCHN was not associated with TNM disease stage and nodal
123
involvement. These negative findings might be correlated with the different types of SCCHN included in the analysis. The proportion of each category of tumors varied across the individual studies. These heterogeneous series of patients might be a potential confounding factor that could have impacted the results. Therefore, the prognostic value, mechanisms of CTCs in tumor TNM stage, and lymph node involvement warrant further investigation. We hypothesize that there were several contributors to insignificant heterogeneity in the pooled analysis: (1) Participants were limited to human beings (not experimental animals); (2) clear diagnosis and staging of patients; and (3) location of sampling collection (most sampling is collected from the peripheral blood before treatment). There
Med Oncol (2015)32:164
are many approaches to detect CTCs, including polymerase chain reaction (PCR), flow cytometry, microfluidic chips, immunomagnetic enrichment, and laser scanning cytometry [22]. The sensitivity and specificity of methods for the detection of CTCs vary depending on the marker used. Despite the presence of various CTC detection methods, immunocytochemistry (CellSearch and Flow Cytom) was the only method used in detecting CTCs in all studies. Moreover, selected antigen markers included EpCAM, CK, and CD45. These approaches could reduce the impact of detecting methods on the final results. The CellSearch system is the only assay approved by the US FDA for clinical use. Subgroup analysis using the CellSearch system further demonstrated that the presence of CTCs was not associated with nodal involvement. Several limitations should be noted in the present study. First, although no publication bias was observed through inspecting the funnel plots, potential publication bias cannot be ruled out due to lack of gray literature and conference papers. Second, the association between presence of CTCs and prognosis would have required a larger number of patients. However, the eligible study number was limited, and the sample size (433 cases) was relatively small. These cofounders prevented us to perform additional subgroup analysis based on different baseline characteristics. Most of the studies investigated the clinical significance of CTCs detected before treatment. The predictive value of CTCs detected during or after treatment is mostly unclear. Finally, the effect of CTCs on survival time could not be analyzed due to incomplete data in the included studies.
Conclusions Our results indicate that detection of CTCs has predictive value in patients with SCCHN particularly on the tumor progression. However, CTC status in patients with SCCHN appears to be not associated with TNM disease stage and nodal involvement. Additional, well-designed prospective studies based on homogeneous populations are needed to confirm these findings. Acknowledgments We thank Dr Bo Zhang for the data verification and statistical analysis. Conflict of interest
None.
References 1. Zarbo RJ, Crissman JD. The surgical pathology of head and neck cancer. Semin Oncol. 1988;15:10–9. 2. Jansen JF, Koutcher JA, Shukla-Dave A. Non-invasive imaging of angiogenesis in head and neck squamous cell carcinoma. Angiogenesis. 2010;13:149–60.
Page 7 of 8 164 3. Mockelmann N, Laban S, Pantel K, Knecht R. Circulating tumor cells in head and neck cancer: clinical impact in diagnosis and follow-up. Eur Arch Otorhinolaryngol. 2014;271:15–21. 4. Jatana KR, Lang JC, Chalmers JJ. Identification of circulating tumor cells: a prognostic marker in squamous cell carcinoma of the head and neck? Future Oncol. 2011;7:481–4. 5. Steeg PS. Tumor metastasis: mechanistic insights and clinical challenges. Nat Med. 2006;12:895–904. 6. Ashworth TR. A case of cancer in which cells similar to those in the tumors were seen in the blood after death. Aus Med J. 1869;14:146–9. 7. Allard WJ, Matera J, Miller MC, Repollet M, Connelly MC, Rao C, et al. Tumor cells circulate in the peripheral blood of all major carcinomas but not in healthy subjects or patients with nonmalignant diseases. Clin Cancer Res. 2004;10:6897–904. 8. Bidard FC, Fehm T, Ignatiadis M, Smerage JB, Alix-Panabieres C, Janni W, et al. Clinical application of circulating tumor cells in breast cancer: overview of the current interventional trials. Cancer Metastasis Rev. 2013;32:179–88. 9. Ma XL, Xiao ZL, Liu L, Liu XX, Nie W, Li P, et al. Metaanalysis of circulating tumor cells as a prognostic marker in lung cancer. Asian Pac J Cancer Prev. 2012;13:1137–44. 10. Jin T, Peng H, Wu H. Clinical value of circulating liver cancer cells for the diagnosis of hepatocellular carcinoma: a meta-analysis. Biomed Rep. 2013;1:731–6. 11. Panteleakou Z, Lembessis P, Sourla A, Pissimissis N, Polyzos A, Deliveliotis C, et al. Detection of circulating tumor cells in prostate cancer patients: methodological pitfalls and clinical relevance. Mol Med. 2009;15:101–14. 12. Gazzaniga P, Raimondi C, Gradilone A, Biondi Zoccai G, Nicolazzo C, Gandini O, et al. Circulating tumor cells in metastatic colorectal cancer: do we need an alternative cutoff? J Cancer Res Clin Oncol. 2013;139:1411–6. 13. Buglione M, Grisanti S, Almici C, Mangoni M, Polli C, Consoli F, et al. Circulating tumour cells in locally advanced head and neck cancer: preliminary report about their possible role in predicting response to non-surgical treatment and survival. Eur J Cancer. 2012;48:3019–26. 14. Wollenberg B, Walz A, Kolbow K, Pauli C, Chaubal S, Andratschke M. Clinical relevance of circulating tumour cells in the bone marrow of patients with SCCHN. Onkologie. 2004;27:358– 62. 15. Nichols AC, Lowes LE, Szeto CC, Basmaji J, Dhaliwal S, Chapeskie C, et al. Detection of circulating tumor cells in advanced head and neck cancer using the cell search system. Head Neck. 2012;34:1440–4. 16. He S, Li P, Long T, Zhang N, Fang J, Yu Z. Detection of circulating tumour cells with the Cell Search system in patients with advanced-stage head and neck cancer: preliminary results. J Laryngol Otol. 2013;127:788–93. 17. Bozec A, Ilie M, Dassonville O, Long E, Poissonnet G, Santini J, et al. Significance of circulating tumor cell detection using the Cell Search system in patients with locally advanced head and neck squamous cell carcinoma. Eur Arch Otorhinolaryngol. 2013;270:2745–9. 18. Jatana KR, Balasubramanian P, Lang JC, Yang L, Jatana CA, White E, et al. Significance of circulating tumor cells in patients with squamous cell carcinoma of the head and neck: initial results. Arch Otolaryngol Head Neck Surg. 2010;136:1274–9. 19. Hristozova T, Konschak R, Stromberger C, Fusi A, Liu Z, Weichert W, et al. The presence of circulating tumor cells (CTCs) correlates with lymph node metastasis in nonresectable squamous cell carcinoma of the head and neck region (SCCHN). Ann Oncol. 2011;22:1878–85. 20. Guney K, Yoldas B, Ozbilim G, Derin AT, Sarihan S, Balkan E. Detection of micrometastatic tumor cells in head and neck
123
164
Page 8 of 8
squamous cell carcinoma. A possible predictor of recurrences? Saudi Med J. 2007;28:216–20. 21. Maheswaran S, Haber DA. Circulating tumor cells: a window into cancer biology and metastasis. Curr Opin Genet Dev. 2010;20:96–9.
123
Med Oncol (2015)32:164 22. Allan AL, Keeney M. Circulating tumor cell analysis: technical and statistical considerations for application to the clinic. J Oncol. 2010;2010:426218.
REVIEW ARTICLE
Prognostic value of circulating tumor cells in patients with squamous cell carcinoma of the head and neck: a systematic review and meta-analysis Zhendan Wang1 • Kai Cui2 • Yan Xue3 • Feng Tong4 • Sheng Li2
Received: 25 November 2014 / Accepted: 14 March 2015 Ó Springer Science+Business Media New York 2015
Abstract The prognostic value of circulating tumor cells (CTCs) in patients with squamous cell carcinoma of the head and neck (SCCHN) is controversial. The objective of this study was to evaluate the prognostic value of CTCs in patients with SCCHN by conducting a meta-analysis. We systematically searched scientific literature published before June 10, 2014, using the PubMed, ScienceDirect, Cochrane Library, and EMBASE databases. Studies evaluating the correlation of CTC status with tumor-nodemetastasis (TNM) disease stage, nodal involvement, and disease progression (recurrence or metastasis) in patients with SCCHN were selected for the analysis. Pooled odds ratio (OR) and 95 % confidence interval (CI) were calculated using the fixed-effects model or the random-effects model in the presence of heterogeneity. Our analysis consisted of eight studies, enrolling a total of 433 patients. The disease progression (recurrence/metastasis) rate in the CTC-positive patients was significantly higher (OR 3.44; 95 % CI 1.87–6.33; p \ 0.01) compared with the patients without disease progression. However, there was no
& Feng Tong [email protected] & Sheng Li [email protected] 1
Shandong Academy of Medical Science, University of Jinan, Jinan 250022, China
2
Department of Hepatology, Shandong Cancer Hospital, Shandong Academy of Medical Science, 440 Jiyan Road, Jinan 250017, China
3
Department of Head and Neck Surgery, Qilu Children’s Hospital of Shandong University, Jinan 250022, China
4
Department of Dean’s Office, Shandong Cancer Hospital, 440 Jiyan Road, Jinan 250017, China
significant difference between TNM disease stage III–IV and stage I–II in the presence of CTCs (OR 1.54; 95 % CI 0.87–2.72; p [ 0.05). CTC status did not correlate with nodal involvement (OR 1.20; 95 % CI 0.67–1.90; p [ 0.05). This meta-analysis indicates that detection of CTCs has a predictive value in patients with SCCHN, particularly those with tumor progression. The presence of CTCs in patients with SCCHN has a poor prognosis compared with the patients without CTCs. Detection of CTCs might be served as a prognosticator in patients with SCCHN. Further studies based on homogeneous populations are warranted to confirm these findings. Keywords Circulating tumor cells Prognostic Head and neck squamous cell carcinoma Meta-analysis
Introduction Malignant tumors of the head and neck account for approximately 5–10 % of all malignant tumors. Nearly 95 % of the head and neck malignant tumors are squamous cell carcinoma originating from the mucosal epithelia, such as oral cavity, nasopharynx, pharynx, and larynx [1]. The incidence rates of squamous cell carcinoma of the head and neck (SCCHN) were about 10–15 per 10 million subjects, with over 500,000 new cases each year [2, 3]. Despite advances in the diagnosis and treatment in recent years, patients with tumor-node-metastasis (TNM) disease stage III–IV still account for about 50 %, and the 5-year survival rate has not significantly improved [4]. Therefore, early identification of metastasis and recurrence has become increasingly important for the tumor progression and prognosis in patients with SCCHN.
123
164
Page 2 of 8
Dissemination of tumor cells into the blood stream is an important step in tumor metastasis [5]. Circulating tumor cells (CTCs) are tumor cells detectable in peripheral blood released by primary tumors or thorough tumor recurrences and metastases. The first report on tumor cells in bloodstream was reported from Ashworth in 1869 [6]. In recent years, various methods have been developed for CTCs identification. Detecting the expression of epithelial and cancer-specific markers is the most widely used method to identify CTCs. Based on this principle, the CellSearch system was approved by the US Food and Drug Administration (FDA) for clinical use in 2004 [7]. The status of CTCs plays an important role in the early diagnosis, prognosis, and instruction of individual therapy [8, 11, 12]. At present, detection of CTCs has been widely applied in various malignant tumors, including breast cancer [8], lung cancer [9], hepatocellular carcinoma [10], prostate cancer [11], and colorectal cancer [12]. The prognostic value of CTCs detected in patients with SCCHN has been investigated in many studies. However, whether the CTC status is associated with the disease stage, lymph node involvement, or prognosis remains controversial. Here, we conducted the first meta-analysis using currently available studies to quantitatively assess the prognostic value of CTCs in patients with SCCHN.
Med Oncol (2015)32:164
tumors, but with no specific results of SCCHN; (2) lacking key information to calculate the risk estimates; and (3) reviews, letters, or case report types of study design. Data extraction and quality assessment
Materials and methods
Two authors (ZD Wang and K Cui) independently identified the eligible articles and collected the following data: first author’s surname, year of publication, country of study conducted, sample size, number of CTC-positive patients, volumes and site of sample collection, sampling time point, tumor stage, nodal involvement, progression (recurrence or metastasis), methods to detect CTCs, type of tumor markers analyzed, and follow-up duration. Any disagreements were resolved by discussion. The methodological quality of the included studies was assessed in accordance with the risk bias guidelines in the Cochrane reviewers’ Handbook 5.1.0. The following study features were assessed: (1) Were adequate eligibility criteria developed and applied? (2) Was the measurement of both exposure and outcome adequate? (3) Was the confounding adequately controlled? (4) Was the follow-up complete and adequate in duration? (5) Were study reports free of suggestion of selective outcome reporting? (6) Was the study free of other high risk of bias? If all quality criteria were met, the study was sorted into low risk of bias, and if one or more criteria not met, the trial was sorted into high risk of bias.
Search strategy
Statistical methods
We systematically searched the PubMed, ScienceDirect, Cochrane Library, and EMBASE databases for studies prior to June 10, 2014. Potentially relevant studies were identified using the following key words: ‘‘circulating tumor cell(s),’’in combination with ‘‘head and neck cancer,’’ ‘‘carcinoma of the head and neck,’’ or ‘‘squamous cell carcinoma of the head and neck.’’ No language restrictions were imposed. Additionally, reference lists of retrieved articles were screened manually to identify additional eligible studies.
Included patients were stratified according to the CTC status and the subsequent comparisons of presence of CTCs versus absence of CTCs. According to clinical characteristics, stage I and stage II were combined and stage III and stage IV were combined. Odds ratio (OR) and 95 % confidence interval (CI) were computed to determine the effect of CTC status on clinical outcomes. OR[1 indicated worse outcome for the positive CTCs group relative to the negative CTCs group. Heterogeneity was tested using the Isquare and Cochrane Q statistic. We defined the Q statistic (p \ 0.10) or I-square [50 % as significant heterogeneity; then, a fixed-effects model was used for analysis; otherwise, a random-effects model was selected. The potential publication bias was evaluated using the funnel plot. To investigate the change of the overall effect size, the sensitivity analysis was performed omitting one study at each turn. All analyses were performed with RevMan software of the Cochrane Collaboration (version 5.2, Oxford, UK). A two-sided p \ 0.05 was considered statistically significant.
Studies selection Studies satisfying following criteria were selected: (1) patients with histological diagnosis of SCCHN; (2) detection of CTCs in peripheral blood or bone marrow; and (3) investigation of association between CTC status and clinical outcome (TNM disease stage, nodal involvement, and disease recurrence/metastasis). Studies were excluded if they met the following criteria: (1) patients with various
123
Med Oncol (2015)32:164
Results Literature search A total of 157 potentially relevant articles were collected from the initial electronic search. After reviewing the title and abstract, 143 relevant studies were excluded, and six studies were additionally excluded after reviewing full-text manuscript. Finally, eight studies [13–20] met our predefined inclusion criteria (Fig. 1). Characteristics of identified studies and risk of bias Characteristics of the included studies are listed in Table 1. A total of 433 patients were identified. The total number of patients ranged from 9 to 176 in the individual studies. The studies were conducted in Italy, Germany, Britain, France, Turkey, China, and USA and published from 2004 to 2013. The CTCs were collected from peripheral blood in seven studies [13, 15, 17–20], and one study [14] collected CTCs from the bone marrow. The sample collecting ranged from 7.5 to 30 ml before the treatment, except for one study [18], which collected sample during treatment. CTCs were detected using CellSearch [13, 15–17], immunocytochemistry alkaline phosphatase and monoclonal anti-alkaline phosphatase (ICC-APAAP) [14], ICC [18], flow cytometry [19], and magnetic activated cell sorting(MACS) [20] analysis. The detected tumor markers mainly included epithelial cell adhesion molecule (EpCAM), CD45,
Page 3 of 8 164
cytokeratins (CK), and 40 ,6-diamidino-2-phenylindole (DAPI). The included studies were classified as low risk of bias, and only one study [19] was considered as high risk of bias due to incomplete follow-up duration (Table 2). CTC status and TNM disease stage Four studies [14, 15, 18, 20] reported CTC status and TNM disease stage, and stage III–IV group included 77 CTC cases among 226 patients, while the stage I–II group included 29 CTC-positive cases among 92 patients. As shown in Fig. 2, a fixed-effects model was selected because there was no obvious sample heterogeneity (I2 = 0 %; p = 0.48). There was no significant difference between TNM disease stage III–IV and stage I–II patients in the presence of CTCs (OR 1.54; 95 % CI 0.87–2.72; p = 0.13). CTC status and nodal involvement Seven studies [13–19] provided the data of CTC status and nodal involvement, and the group with nodal involvement included 77 CTC-positive cases among 241 patients, while the group without nodal involvement included 57 CTCpositive cases among 171 patients. As shown in Fig. 3a, a fixed-effects model was selected because there was no obvious sample heterogeneity (I2 = 0 %; p = 0.44). There was no significant difference between patient with nodal involvement and without nodal involvement in the pres-
Fig. 1 Flow chart of study selection process
123
164
Page 4 of 8
Med Oncol (2015)32:164
Table 1 Baseline characteristics of included studies Study/ Reference
Country
CTC-positive patients (n/N)
Sampling site/volume
Sampling time
Tumor stage (UICC)
Detection method
Type of tumor markers
Follow-up duration (months)
Buglione et al. [13]
Italy
11/73
PB/ C7.5 ml
Before TM
I–IV
CellSearch
EpCAM, CD45,CK,DAPI
13.6
Wollenberg et al. [14]
Germany
54/176
BM/–
Before TM
I–IV
IHCAPAAP
CK19
60
Nichols et al. [15]
Britain
6/15
PB/10 ml
Before TM
III–IV
CellSearch
EpCAM, CD45, CK, DAPI
8
He et al. [16]
China
3/9
PB/7.5 ml
Before TM
III–IV
CellSearch
EpCAM, CD45, CK, DAPI
16
Bozec et al. [17]
France
8/49
PB/7.5 ml
Before TM
III–IV
CellSearch
EpCAM, CD45, CK, DAPI
6
Jatana et al. [18]
USA
34/48
PB/ 10–18 ml
During TM
I–IV
ICC
CK, CD45, DAPI
38
Hristozova et al. [19]
Germany
18/42
PB/7.5 ml
Before TM
I–IV
Flow Cytom
EpCAM, CK
–
Guney et al. [20]
Turkey
7/21
PB/30 ml
Before TM
I–IV
MACS
–
36
PB peripheral blood, BM bone marrow, TM treatment, EpCAM epithelial cell adhesion molecule, CK cytokeratins, DAPI 40 ,6-diamidino-2phenylindole
Table 2 Quality assessment of the included studies Study/ Reference
Were adequate eligibility criteria developed and applied?
Was the measurement of both exposure and outcome adequate?
Was confounding adequately controlled for?
Was the followup complete and adequate in duration?
Are reports of the study free of suggestion of selective outcome reporting?
Was the study free of other problems that put it at a high risk of bias?
Risk of bias
Buglione et al. [13]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Wollenberg et al. [14] Nichols et al. [15]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Yes
Yes
Yes
Yes
Yes
Yes
Low
He et al. [16]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Bozec et al. [17]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Jatana et al. [18]
Yes
Yes
Yes
Yes
Yes
Yes
Low
Hristozova et al. [19]
Yes
Yes
Yes
No
Yes
Yes
High
Guney et al. [20]
Yes
Yes
Yes
Yes
Yes
Yes
Low
ence of CTCs (OR 1.20; 95 % CI 0.76–1.79; p = 0.44). Subgroup analysis using CellSearch in four studies [13, 15–17] indicated that the pooled OR was 1.44 (95 % CI 0.56–3.71; p = 0.45) when the patients with nodal involvement were compared to patients without nodal involvement in the presence of CTCs (Fig. 3b).
123
CTC status and disease progression (recurrence/ metastasis) Four studies [15–17, 20] reported the data of CTC status and progression (recurrence/metastasis), and the CTCpositive group included 35 cases of disease progression
Med Oncol (2015)32:164
Page 5 of 8 164
Fig. 2 Forest plots showing comparison of circulating tumor cells-positive rate between the tumor stage I–II and III–IV groups in a fixed-effects model
Fig. 3 Forest plots showing CTC status and nodal involvement in the eligible studies (a) and detection methods by CellSearch subgroup in a fixed-effects model (b)
among 69 patients, while the CTC-negative group included 35 cases of disease progression among 150 patients. As shown in Fig. 4, a fixed-effects model was selected due to no obvious heterogeneity (I2 = 13 %; p = 0.33). The disease progression (recurrence/metastasis) rate in the CTC-positive group was significantly higher compared with the without disease progression group (OR 3.44; 95 % CI 1.87–6.33; p \ 0.01). Publication bias and sensitivity analysis There were no studies reporting OR of TNM disease stage, nodal involvement, and disease progression (recurrence/
metastasis) by inspecting the funnel plots (Fig. 5). The sensitivity analysis indicated that there was only a slight change in the quantitative summary measure of OR or 95 % CI, but no change in the direction of OR if any study was omitted in the analysis.
Discussion Tumor metastasis is an important contributor to mortality in cancer patients. However, the existing image technology and tumor markers are not sufficient to detect tumor metastasis in early stages and estimate the prognosis. CTCs
123
164
Page 6 of 8
Med Oncol (2015)32:164
Fig. 4 Forest plots showing comparison of circulating tumor cells-positive rate between the disease progression (recurrence or metastasis) and without disease progression groups in a fixed-effects model
Fig. 5 Funnel plots based on TNM disease stage (a), nodal involvement (b), disease progression (c), and nodal involvement in CellSearch subgroup (d)
accompany tumor invasion into the bloodstream. CTC status provides a real-time and minimally invasive approach for the detection of early disease and assessing prognosis, and therapeutic response in established cancers [21]. There is no consensus on whether CTC status has a prognostic value in patients with SCCHN. To the best of our knowledge, this is the first metaanalysis that assesses the prognostic value of CTCs for SCCHN. Overall, the eight included studies were classified as featuring a low risk of study bias. The current metaanalysis reveals that detection of CTCs has a predictive value on the tumor progression in patients with SCCHN. However, the presence of CTCs in patients with SCCHN was not associated with TNM disease stage and nodal
123
involvement. These negative findings might be correlated with the different types of SCCHN included in the analysis. The proportion of each category of tumors varied across the individual studies. These heterogeneous series of patients might be a potential confounding factor that could have impacted the results. Therefore, the prognostic value, mechanisms of CTCs in tumor TNM stage, and lymph node involvement warrant further investigation. We hypothesize that there were several contributors to insignificant heterogeneity in the pooled analysis: (1) Participants were limited to human beings (not experimental animals); (2) clear diagnosis and staging of patients; and (3) location of sampling collection (most sampling is collected from the peripheral blood before treatment). There
Med Oncol (2015)32:164
are many approaches to detect CTCs, including polymerase chain reaction (PCR), flow cytometry, microfluidic chips, immunomagnetic enrichment, and laser scanning cytometry [22]. The sensitivity and specificity of methods for the detection of CTCs vary depending on the marker used. Despite the presence of various CTC detection methods, immunocytochemistry (CellSearch and Flow Cytom) was the only method used in detecting CTCs in all studies. Moreover, selected antigen markers included EpCAM, CK, and CD45. These approaches could reduce the impact of detecting methods on the final results. The CellSearch system is the only assay approved by the US FDA for clinical use. Subgroup analysis using the CellSearch system further demonstrated that the presence of CTCs was not associated with nodal involvement. Several limitations should be noted in the present study. First, although no publication bias was observed through inspecting the funnel plots, potential publication bias cannot be ruled out due to lack of gray literature and conference papers. Second, the association between presence of CTCs and prognosis would have required a larger number of patients. However, the eligible study number was limited, and the sample size (433 cases) was relatively small. These cofounders prevented us to perform additional subgroup analysis based on different baseline characteristics. Most of the studies investigated the clinical significance of CTCs detected before treatment. The predictive value of CTCs detected during or after treatment is mostly unclear. Finally, the effect of CTCs on survival time could not be analyzed due to incomplete data in the included studies.
Conclusions Our results indicate that detection of CTCs has predictive value in patients with SCCHN particularly on the tumor progression. However, CTC status in patients with SCCHN appears to be not associated with TNM disease stage and nodal involvement. Additional, well-designed prospective studies based on homogeneous populations are needed to confirm these findings. Acknowledgments We thank Dr Bo Zhang for the data verification and statistical analysis. Conflict of interest
None.
References 1. Zarbo RJ, Crissman JD. The surgical pathology of head and neck cancer. Semin Oncol. 1988;15:10–9. 2. Jansen JF, Koutcher JA, Shukla-Dave A. Non-invasive imaging of angiogenesis in head and neck squamous cell carcinoma. Angiogenesis. 2010;13:149–60.
Page 7 of 8 164 3. Mockelmann N, Laban S, Pantel K, Knecht R. Circulating tumor cells in head and neck cancer: clinical impact in diagnosis and follow-up. Eur Arch Otorhinolaryngol. 2014;271:15–21. 4. Jatana KR, Lang JC, Chalmers JJ. Identification of circulating tumor cells: a prognostic marker in squamous cell carcinoma of the head and neck? Future Oncol. 2011;7:481–4. 5. Steeg PS. Tumor metastasis: mechanistic insights and clinical challenges. Nat Med. 2006;12:895–904. 6. Ashworth TR. A case of cancer in which cells similar to those in the tumors were seen in the blood after death. Aus Med J. 1869;14:146–9. 7. Allard WJ, Matera J, Miller MC, Repollet M, Connelly MC, Rao C, et al. Tumor cells circulate in the peripheral blood of all major carcinomas but not in healthy subjects or patients with nonmalignant diseases. Clin Cancer Res. 2004;10:6897–904. 8. Bidard FC, Fehm T, Ignatiadis M, Smerage JB, Alix-Panabieres C, Janni W, et al. Clinical application of circulating tumor cells in breast cancer: overview of the current interventional trials. Cancer Metastasis Rev. 2013;32:179–88. 9. Ma XL, Xiao ZL, Liu L, Liu XX, Nie W, Li P, et al. Metaanalysis of circulating tumor cells as a prognostic marker in lung cancer. Asian Pac J Cancer Prev. 2012;13:1137–44. 10. Jin T, Peng H, Wu H. Clinical value of circulating liver cancer cells for the diagnosis of hepatocellular carcinoma: a meta-analysis. Biomed Rep. 2013;1:731–6. 11. Panteleakou Z, Lembessis P, Sourla A, Pissimissis N, Polyzos A, Deliveliotis C, et al. Detection of circulating tumor cells in prostate cancer patients: methodological pitfalls and clinical relevance. Mol Med. 2009;15:101–14. 12. Gazzaniga P, Raimondi C, Gradilone A, Biondi Zoccai G, Nicolazzo C, Gandini O, et al. Circulating tumor cells in metastatic colorectal cancer: do we need an alternative cutoff? J Cancer Res Clin Oncol. 2013;139:1411–6. 13. Buglione M, Grisanti S, Almici C, Mangoni M, Polli C, Consoli F, et al. Circulating tumour cells in locally advanced head and neck cancer: preliminary report about their possible role in predicting response to non-surgical treatment and survival. Eur J Cancer. 2012;48:3019–26. 14. Wollenberg B, Walz A, Kolbow K, Pauli C, Chaubal S, Andratschke M. Clinical relevance of circulating tumour cells in the bone marrow of patients with SCCHN. Onkologie. 2004;27:358– 62. 15. Nichols AC, Lowes LE, Szeto CC, Basmaji J, Dhaliwal S, Chapeskie C, et al. Detection of circulating tumor cells in advanced head and neck cancer using the cell search system. Head Neck. 2012;34:1440–4. 16. He S, Li P, Long T, Zhang N, Fang J, Yu Z. Detection of circulating tumour cells with the Cell Search system in patients with advanced-stage head and neck cancer: preliminary results. J Laryngol Otol. 2013;127:788–93. 17. Bozec A, Ilie M, Dassonville O, Long E, Poissonnet G, Santini J, et al. Significance of circulating tumor cell detection using the Cell Search system in patients with locally advanced head and neck squamous cell carcinoma. Eur Arch Otorhinolaryngol. 2013;270:2745–9. 18. Jatana KR, Balasubramanian P, Lang JC, Yang L, Jatana CA, White E, et al. Significance of circulating tumor cells in patients with squamous cell carcinoma of the head and neck: initial results. Arch Otolaryngol Head Neck Surg. 2010;136:1274–9. 19. Hristozova T, Konschak R, Stromberger C, Fusi A, Liu Z, Weichert W, et al. The presence of circulating tumor cells (CTCs) correlates with lymph node metastasis in nonresectable squamous cell carcinoma of the head and neck region (SCCHN). Ann Oncol. 2011;22:1878–85. 20. Guney K, Yoldas B, Ozbilim G, Derin AT, Sarihan S, Balkan E. Detection of micrometastatic tumor cells in head and neck
123
164
Page 8 of 8
squamous cell carcinoma. A possible predictor of recurrences? Saudi Med J. 2007;28:216–20. 21. Maheswaran S, Haber DA. Circulating tumor cells: a window into cancer biology and metastasis. Curr Opin Genet Dev. 2010;20:96–9.
123
Med Oncol (2015)32:164 22. Allan AL, Keeney M. Circulating tumor cell analysis: technical and statistical considerations for application to the clinic. J Oncol. 2010;2010:426218.
