doi: 10.1111/j.1365-2265.2011.04162.x
ORIGINAL ARTICLE
Central cervical lymph node metastases in papillary thyroid cancer: a systematic review of imaging-guided and prophylactic removal of the central compartment
Mubashir Mulla*, and Klaus-Martin Schulte*,, *Department of Endocrine Surgery, Kings College Hospital, Denmark Hill, Kings Health Partners and Kings College London, London, UK
Summary
Background Papillary thyroid cancer (PTC) is a common endocrine cancer and commonly presents with lymph node (LN) metastases. The role of surgical removal of the central cervical LN compartment is poorly dened. There are no prospective randomized controlled trials addressing the relevance to the extent of the initial surgical approach. Design and Methods A systematic review of studies of patients with PTC undergoing either prophylactic or therapeutic lymphadenectomy of the central LNs was carried out. Studies involving imaging modalities in the detection of LNs in PTC were also analysed. Results Twenty-one studies contained data on 4188 patients undergoing prophylactic or imaging-guided removal of the central compartment. Imaging-guided surgery retrieved cancerous central LNs in 346 or 30% of eligible patients, whilst prophylactic central neck dissection yielded histopathological proof of cancer in 898 or 262% of patients. Five imaging studies revealed data on the use of ultrasound (US) and/or computerized tomography (CT). The sensitivity of US and CT was poor, ranging from 50% to 70% when accurately calculated. Conclusion Metastatic central LNs are found in nearly half of all patients with PTC when prophylactic central lymph node dissection (CLND) is performed. With unreliable imaging modalities, prophylactic CLND should be performed on all patients with PTC. (Received 9 March 2011; returned for revision 14 May 2011; nally revised 21 June 2011; accepted 27 June 2011)
decient areas.1 Cervical lymph node (LN) metastases occur commonly in PTC. The incidence of metastases ranges from 20% to 90% with an average of 60%.2 The central compartment is commonly referred to as level VI of the neck. Its boundaries include the hyoid bone superiorly and the carotid arteries laterally. The inferior border has been variably dened as the sternal notch or the innominate (brachiocephalic) artery.3 Central lymph node dissection (CLND) is dened as the removal of the level VI LN that lies in a central position in the neck. Level VI LNs include pretracheal and paratracheal nodes, the precricoid (Delphian) node and the perithyroidal nodes, including nodes along the recurrent laryngeal nerves.3,4 Lymph node metastases are known to be independent risk factors for local recurrence.5,6 Emerging evidence from large studies shows an increase in mortality with regional LN metastases despite earlier reports of no adverse effects on survival.7,8 Whilst it is generally accepted that therapeutic CLND is indicated for PTC with grossly evident metastatic LN in the central compartment, the indications for elective CLND when there is N0 disease on imaging or clinical ground remain highly controversial. There are no randomized controlled trials dealing with the extent of CLND metastases in PTC and its prognostic implications. We here present the results of a systematic review of ndings of CLND in patients with PTC.
Introduction
Papillary thyroid cancer (PTC) is the most common histological type and accounts for about 85% of all thyroid cancers in iodineCorrespondence: Mubashir Mulla, Department of Endocrine Surgery, Kings College Hospital, Denmark Hill, London SE5 9RS, UK. Tel.: 0203 299 1925; Fax: 0203 299 3914; E-mail:mubashirmulla@nhs.net 2011 Blackwell Publishing Ltd
132 M. Mulla and K.-M. Schulte All studies with <100 patients and/or <2 years of follow-up were excluded. The author (MM) searched for articles reported over the last 40 years from 1970 up to October 2009 in PubMed, with the combination of search terms cancer, LNs dissection, central, thyroid, prophylactic, therapeutic, recurrence and survival. The search was restricted to the presence of these key words in the title or abstract of the articles. The preliminary search using these terms yielded 11 924 publications of which 1402 were of potential relevance when reading the title. On detailed scrutiny of the abstracts of these 1402 publications, 95 seemed to contain relevant data. These 95 publications were read in full text and scrutinized for the presence of data. This process identied 23 studies appropriate for analysis based on the criteria set out above. All 23 publications were subjected to forward and backward quotation searches using the Thompson resource ISI Web of Knowledge (http://apps.isiknowledge.com/). This identied three further studies, which are included for analysis. Out of these 26 studies, 21 provided data for CLND929 and ve were imaging studies.3034 All studies included in this publication were reviewed in full by both authors. Statistical analysis For analysis, the standard procedures were not used as there was no study with combined comparative data for both prophylactic and therapeutic LN dissection. To simplify the results, we have used simple percentages.
Results
Data were available for 4188 patients from 21 studies for prophylactic and therapeutic CLND. For imaging, ve studies provided the data for 1060 patients. No randomized controlled trials were found despite an extensive literature search. All studies were cohort studies. Fourteen studies provided results of only prophylactic CLND. Four studies provided results of therapeutic CLND, and in three studies, results were available for both prophylactic and therapeutic CLND. Table 1 shows results of prophylactic and therapeutic CLND in patients with PTC. A total of 4188 patients underwent therapeutic and/or prophylactic CLND, of which 1244 had positive LN on histology (297%).
Table 1. Results of central lymph node dissection (CLND) in papillary thyroid cancer
Prophylactic CLND Study Costa (2009) Zuniga (2009) Koo (2009) Lee (2007) Shindo (2009) Moo (2009) *Choi (2008) Son (2008) Bonnet (2009) Perrino (2009) Mercante (2009) Low (2008) Swyak (2006) Chung (2009) Palestini (2008) Wada (2009) *Wada (2003) R1 Rosenbaum (2009) Roh (2007) Sadowski (2009) Davidson (2008) Total (R2) Total (R) Percentages Total patients (n) 244 266 111 103 100 104 101 114 115 251 445 100 447 245 305 120 259 3430 110 155 310 183 1442 4188 Methods to guide therapeutic CLND n 126 136 111 103 94 104 48 114 115 92 226 40 56 206 93 47 235 1946 p LN+ 59 112 60 39 33 47 18 52 42 39 112 26 21 45 20 30 143 898
64 73 24 22 56 180 19 438
46 72 23 17 51 125 12 346
898/1946 = 461%
346/438 = 79%
US, ultrasonography; CT, computerised tomography; p LN, histologically proven positive LN; R1, total number of patients included in the prophylactic LN dissection group; R2, total number of patients included in the therapeutic LN dissection group; R, grand total number of patients in both groups. Total: 1244 (898 + 346)/4188 = 297%; prophylactic only: 898/3430 = 262%; therapeutic only: 346/ 1442 = 24%; *Papillary Microcarcinoma <1cm. 2011 Blackwell Publishing Ltd, Clinical Endocrinology, 76, 131136
Central lymph node dissection in thyroid cancer 133 Prophylactic CLND alone was performed in 1946 patients, of which 898 had positive lateral LN at surgery (461%). Hence, 262% (898/3430) of patients who were evaluated for prophylactic CLND were found to have positive lateral LN at surgery. A total of 1442 patients were assessed by various pre-operative methods for the presence of central LN. Therapeutic central LN dissection was performed in 438/1442 (304%) patients guided by clinical examination and/or imaging. Of the 438 patients who underwent therapeutic CLND, positive central LN was found in 79% (346/438). Hence, 24% (346/1442) of patients who were evaluated for therapeutic CLND were found to have positive central LN at surgery. We found six studies in literature that had results of primary tumours <2 cm. These include two studies with papillary microcarcinoma (Wada 2003 and Choi 2008). Table 2 shows the results of CLND of these six studies. Out of 819 patients with small primary tumours, LN metastases were found to be positive in 48% of patients (393/819). Table 3 shows further results of CLND according to primary tumour size. The results were available as shown from three studies with patients classied as having tumours less than or greater than 1 cm. In tumours <1 cm, out of a total of 143 patients, 37 were found to have positive central lymph nodes (CLN) (26%). Out of 143 patients with tumours >1 cm, 106 were found to be positive for metastases (74%). Table 4 shows results of CLND according to tumour stages. The stages were classied according to results available from the included studies. In patients classied as T1T2, a total of 50
Table 4. Results of central lymph node dissection according to T stage Study Total patients Total pN+ T12 33 55 88 23 27 50 50/88 = 568% >T3 10 28 38 38/88 = 432%
patients showed metastases out of 88 (57%). In T3T4 patients, 38 were positive out of a total of 88 (43%). Tables 5 and 6 provide the results of the imaging of central LNs pre-operatively. We found ve such studies comparing the sensitivities and specicities of ultrasound scan (US) and or computerized tomography (CT) scan. In three of these studies (Kim 2008, Soler 2008 and Sugitani 2008), central LNs were dissected only if detected pre-operatively by US/CT. It is therefore not possible to calculate the sensitivities accurately as false negatives cannot be known without dissecting all the central LNs. Figures given by the authors are hence shown in brackets. The term sensitivity has been used accurately in two studies (Choi 2009 and Roh 2008), where it is 53% and 61%, respectively, for US and 67% for CT (Choi 2009). The calculation of negative prediction similarly implies knowledge of the false-negative rate. Accordingly, negative predictive value (NPV) calculated by the authors has been put in brackets where false-negative rates were not known.
Discussion
Lymph node metastases are common in PTC ranging from 20% to 90% of cases (2). Generally, central LNs are more commonly involved and usually before lateral LN involvement.35,36 The role of prophylactic CLND remains controversial. The general agreement on guidelines from various associations about CLND is to dissect central nodes only in cases of conrmed node involvement pre-operatively or in T3/T4 tumours.37 Because of the large number of patients with subclinical nodal disease that might resurface at a later stage and the questionable accuracy of imaging modalities, this practice remains controversial. The frequency of positive LN in the central compartment in our meta-analysis was found to be 461% in prophylactic and 79% in directed surgical (therapeutic) approaches. In therapeutic CLND, only those LN detected clinically or by imaging pre-operatively are dissected. This implies that those LN left behind may still be positive for metastases. Our results further show that in tumours <2 cm, which is effectively T1 stage according to AJCC Cancer Staging (37), 48% patients showed metastases in CLN. In other words, in nearly half of patients staged as T1, if the CLN are not dissected electively, positive CLN are left behind. This indicates gross under-staging of these patients, which is clearly unacceptable. It is worth noting that the 5-year relative survival rates between these stages are remarkably different as identied in the AJCC Cancer Staging Manual.37 Whilst stage I and II have a survival of 100%, stage III survival is 93% with a sharp drop to 51% at stage IV. Stage progression from stage I or II to stage III is dened as the presence
Table 2. Results of central lymph node dissection (CLND) for papillary carcinoma <2 cm (T1) Prophylactic CLND Study Mercante (2009) Bonnet (2009) Perrino (2009) Choi (2008) Lee (2007) Wada (2003) Total Percentages Total patients (n) 445 115 251 101 103 259 1274 393/1274 = 308% n p LN+ 112 42 39 18 39 143 393 = 48% Therapeutic CLND n p LN+
24
23
Table 3. Results of central lymph node dissection according to the size of primary tumour Study Total patients Total pN+ <1 cm 45 56 42 143 >1 cm
Moo (2009) 104 Koo (2009) 111 Bonnett (2009) 115 Total 330 Percentages
NPV, negative predictive value; PPV, positive predictive value. *Only LNs detected by US were removed - sensitivity cannot be calculated accurately.
Table 6. Computed tomography for the detection of central lymph node metastases in papillary thyroid cancer Study Total patients (n) Sensitivity, % Specicity, % PPV, % NPV, % 67 (50) (58) 79 91 72 65 79 72 80 (74) (58)
NPV, negative predictive value. *Only LNs detected by US were removed sensitivity cannot be calculated accurately.
of LN metastasis, except for women <45 years of age who remain stage I. Therefore, under-staging nearly half of all those patients who are not young women will clearly have a negative impact on their long-term survival. The MACIS system of staging, which is sometimes used for staging PTC, has also shown 20-year mortality ranging from 09% in group 1 to 100% for group 4.38 There is a growing tendency for prophylactic CLND in many centres but this is still not the universal treatment method. A lot of the time it depends on local practice and the expertise available. Opponents of elective CLND argue that the rate of postoperative complications is higher in these patients. The incidence of transient and permanent recurrent laryngeal nerve palsy has been reported anywhere between 473% and 036%, respectively, and transient hypoparathyroidism is reported between 14% and 44%.39 However, several studies have shown no difference in the complication rates after CLND when performed by experienced surgeons26,40 or as a secondary procedure.41 The other argument is that PTC has low local recurrence rates. A recent meta-analysis did not nd any statistical difference in LN recurrence when comparing the addition of a prophylactic central LN dissection with thyroidectomy vs thyroidectomy alone for patients with PTC. However, the follow-up periods of included studies ranged from 6 to 332 months, which render the conclusions highly debatable.42 As we know PTC is a slow-growing tumour, recurrences can occur as late as decades from initial diagnosis. Long-term studies have reported recurrences up to 30%.4345 However, without a long-term prospective randomized controlled trial, an accurate gure for recurrence is difcult to identify. PTC has good long-term survival compared with many cancers, but this does not justify leaving behind potentially malignant LN exposing the patient to recurrence and further morbidity and maybe even mortality. CLND has the advantage of removing potential sources of recurrence, reducing the morbidity of a re-operation, accurately
staging these tumours and optimizing postoperative treatment regimens and follow-up.9,46,47 Some argue that there is limited benet from prophylactic CLND for either disease recurrence or survival outcomes;48 hence, it should be performed only in patients deemed high risk: larger tumours, extra-thyroidal extension or aggressive histological subtypes.49 Although this may be the case, performing a routine prophylactic bilateral CLND as previously mentioned simplies the logistics of cancer surveillance in the postoperative period. If such a patient does develop a recurrence, it may require central neck re-exploration and CLND. Re-operative CLND has the potential to be more challenging and puts the recurrent laryngeal nerve and parathyroids at increased risk.13,50 The argument to restrain CLND to therapeutic as opposed to prophylactic and staging use is not based on evidence in literature. This view is supported by a recent study by So et al.51 on clinical node negative patients on pre-operative US, which concluded that the incidence of subclinical CLN metastases was high (36%) and recommended prophylactic CLND especially in high risk groups namely men, multifocality of tumour and extra-thyroidal tumour extension. Imaging in PTC for detecting metastatic LN is another area of uncertainty. Many centres use US and CT scans to assess the extent of LN involvement in thyroid cancer. US is operator dependent and identies only a fraction of the LNs found at surgery. Problems arise from the presence of the thyroid gland, the size of the LN involved and in obese patients with short necks because of the presence of the overlying thyroid gland.52 Other modalities such as MRI and Positron Emission Tomography (PET) scans are also used inconsistently; the sensitivities of the detection of cervical LN have been reported to be relatively low at 3040%.53 From the ve imaging studies evaluated in this study, only two (Choi 2009 and Roh 2008) have calculated the sensitivity of US and CT to detect central cervical LNs accurately. This was in the range of 5060% for US and 67% for CT, which is disappointingly low considering the impact this has on staging of thyroid cancer and its potential long-term implications for patient outcome. The remaining imaging studies reporting on sensitivity and NPV of imaging for the detection of central LNs did not evaluate false-negative rates and therefore by principle cannot calculate these values. Imaging assessment has high value to direct therapy decisions in PTC when neck metastases are identied. In the absence of ndings, a false-negative result is possible and becomes more likely with increasing tumour T stage, i.e. size of the tumour. Furthermore, a recent study by Gao Li demonstrated that in patients with two or more positive central LN, the incidence of lateral LN metastases was as high as 7093% (11).
2011 Blackwell Publishing Ltd, Clinical Endocrinology, 76, 131136
Central lymph node dissection in thyroid cancer 135 With low sensitivities of imaging, the argument for dissecting central LN only if detected pre-operatively by imaging is difcult to justify. Some argue that the sensitivity of US is high in detection of other tumours, but this has not been proven in PTC. It is also worth noting that the nature of reported results in our study makes it impossible to differentiate between micrometastases and sizeable metastases.
12 Lee, Y.S., Kim, S.W., Kim, S.W. et al. (2007) Extent of routine central lymph node dissection with small papillary thyroid carcinoma. World Journal of Surgery, 31, 19541959. 13 Shindo, M., Wu, J.C., Park, E.E. et al. (2006) The importance of central compartment elective lymph node excision in the staging and treatment of papillary thyroid cancer. Archives of Otolaryngology Head and Neck Surgery, 132, 650654. 14 Moo, T.A., Umunna, B., Kato, M. et al. (2009) Ipsilateral versus bilateral central neck lymph node dissection in papillary thyroid carcinoma. Annals of Surgery, 250, 403408. 15 Son, Y.I., Jeong, H.S., Baek, C.H. et al. (2008) Extent of prophylactic lymph node dissection in the central neck area of the patients with papillary thyroid carcinoma: comparison of limited versus comprehensive lymph node dissection in a 2-year safety study. Annals of Surgical Oncology, 15, 20202026. 16 Bonnet, S., Hartl, D., Leboulleux, S. et al. (2009) Prophylactic lymph node dissection for papillary thyroid cancer less than 2 cm: implications for radioiodine treatment. Journal of Clinical Endocrinology and Metabolism, 94, 11621167. 17 Perrino, M., Vannucchi, G., Vicentini, L. et al. (2009) Outcome predictors and impact of central node dissection and radiometabolic treatments in papillary thyroid cancers < or =2 cm. Endocrine-Related Cancer, 16, 201210. 18 Mercante, G., Frasoldati, A., Pedroni, C. et al. (2009) Prognostic factors affecting neck lymph node recurrence and distant metastasis in papillary microcarcinoma of the thyroid: results of a study in 445 patients. Thyroid, 19, 707716. 19 Low, T.H., Delbridge, L., Sidhu, S. et al. (2008) Lymph node status inuences follow-up thyroglobulin levels in papillary thyroid cancer. Annals of Surgical Oncology, 15, 28272832. 20 Sywak, M., Cornford, L., Roach, P. et al. (2006) Routine ipsilateral level VI lymphadenectomy reduces postoperative thyroglobulin levels in papillary thyroid cancer. Surgery, 140, 10001005. 21 Chung, Y.S., Kim, J.Y., Bae, J.S. et al. (2009) Lateral lymph node metastasis in papillary thyroid carcinoma: results of therapeutic lymph node dissection. Thyroid, 19, 241246. 22 Palestini, N., Borasi, A., Cestino, L. et al. (2008) Is central neck dissection a safe procedure in the treatment of papillary thyroid cancer? Our experience. Langenbecks Archives of Surgery, 393, 693 698. 23 Wada, N., Sugino, K., Mimura, T. et al. (2009) Treatment strategy of papillary thyroid carcinoma in children and adolescents: clinical signicance of the initial nodal manifestation. Annals of Surgical Oncology, 16, 34423449. 24 Wada, N., Duh, Q.Y., Sugino, K. et al. (2003) Lymph node metastasis from 259 papillary thyroid microcarcinomas: frequency, pattern of occurrence and recurrence, and optimal strategy for neck dissection. Annals of Surgery, 237, 399407. 25 Rosenbaum, M.A. & McHenry, C.R. (2009) Central neck dissection for papillary thyroid cancer. Archives of Otolaryngology Head and Neck Surgery, 135, 10921097. 26 Roh, J.L., Park, J.Y. & Park, C.I. (2007) Total thyroidectomy plus neck dissection in differentiated papillary thyroid carcinoma patients: pattern of nodal metastasis, morbidity, recurrence, and postoperative levels of serum parathyroid hormone. Annals of Surgery, 245, 604610. 27 Sadowski, B.M., Snyder, S.K. & Lairmore, T.C. (2009) Routine bilateral central lymph node clearance for papillary thyroid cancer. Surgery, 146, 696703.
Conclusion
Metastatic central lymph nodes are found in nearly half of all patients with papillary thyroid cancer when prophylactic central lymph node dissection is performed. With unreliable imaging modalities, prophylactic central lymph node dissection should be performed on all patients with papillary thyroid cancer.
References
1 Sugitani, I., Kasai, N., Fujimoto, Y. et al. (2004) A novel classication system for patients with PTC: addition of the new variables of large (3 cm or greater) nodal metastases and reclassication during the follow-up period. Surgery, 135, 139148. 2 Rotstein, L. (2009) The role of lymphadenectomy in the management of papillary carcinoma of the thyroid. Journal of Surgical Oncology, 99, 186188. 3 Carty, S., Cooper, D.S., Doherty, G.M. et al. (2009) Consensus statement on the terminology and classication of central neck dissection for thyroid cancer. Thyroid, 19, 11531158. 4 White, M.L., Gauger, P.G. & Doherty, G.M. (2007) Central lymph node dissection in differentiated thyroid cancer. World Journal of Surgery, 31, 895904. 5 Mazzaferri, E.L. & Jhiang, S.M. (1994) Long-term impact of initial surgical and medical therapy on papillary and follicular thyroid cancer. American Journal of Medicine, 97, 418428. 6 DeGroot, L.J., Kaplan, E.L., McCormick, M. et al. (1990) Natural history, treatment, and course of papillary thyroid carcinoma. Journal of Clinical Endocrinology and Metabolism, 71, 414424. 7 Lundgren, C.I., Hall, P., Dickman, P.W. et al. (2006) Clinically signicant prognostic factors for differentiated thyroid carcinoma: a population-based, nested case-control study. Cancer, 106, 524 531. 8 Tisell, L.E., Nilsson, B., Molne, J. et al. (1996) Improved survival of patients with papillary thyroid cancer after surgical microdissection. World Journal of Surgery, 20, 854859. 9 Costa, S., Giugliano, G., Santoro, L. et al. (2009) Role of prophylactic central neck dissection in cN0 papillary thyroid cancer. Acta otorhinolaryngologica Italica, 29, 6169. 10 Zuniga, S. & Sanabria, A. (2009) Prophylactic central neck dissection in stage N0 papillary thyroid carcinoma. Archives of Otolaryngology Head and Neck Surgery, 135, 10871091. 11 Koo, B.S., Choi, E.C., Yoon, Y.H. et al. (2009) Predictive factors for ipsilateral or contralateral central lymph node metastasis in unilateral papillary thyroid carcinoma. Annals of Surgery, 249, 840844.
Copyright of Clinical Endocrinology is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.