Prediction of Nodal Metastasis by the AMES Scoring System in Patients with Papillary Thyroid Cancer

When used as an adjunct to total thyroidectomy, prophylactic central lymph node (CLN) dissection can provide accurate staging, more appropriate adjuvant therapy and establish a long term follow-up for patients with papillary thyroid carcinoma (PTC).(1) Also because of the high incidence of CLN metastasis in PTC even in clinically N0 patient group, many authors who claim that CLN metastasis significantly affect recurrence(2) and survival in PTC strenuously insist on performing a prophylactic CLN dissection in order to reduce a locoregional recurrence rate.(3, 4)

However, there may be a significant risk of complications such as transient or permanent vocal cord palsy, due to unintentional recurrent laryngeal nerve palsy and permanent hypoparathyroidism after prophylactic CLN dissection.(5) There are more side effects after total thyroidectomy associated with CLN dissection compared with total thyroidectomy alone.(6, 7)

For the reasons above, the role of prophylactic CLN dissection and the extent of surgery for PTC remain controversial.(8) Thus the standard treatment of PTC has not yet been established. It demonstrates that it would be better to carry out the CLN dissection if CLN metastasis is highly suspected in selective PTC patients.

[¹⁸F] Fluorodeoxyglucose (FDG) positron emission tomography-computed tomography (PET/CT) has been widely used to discriminate between malignant and benign tumors and to detect CLN metastasis in the preoperative PTC patients in Korea until last year.(9) It is an indication of the total thyroidectomy and CLN dissection if there is increased glucose metabolism in [¹⁸F] FDG PET/CT. In addition to showing reduced survival in those with FDG-avid disease, there is a significant correlation with the maximum SUV, suggesting that tumors with the highest metabolic activity might be those with the most rapid growth potential.(10)

However, we often observe that of tracer is not detected in the primary cancer focus in patients with histologically proven PTC on preoperative [¹⁸F] FDG PET/CT. Tumor size and perithyroidal and lymphovascular invasion of PTC can influence [¹⁸F] FDG PET/CT false negatives.

The MACIS and AMES scoring systems are well known prognostic staging systems for thyroid carcinoma.(11) Under MACIS, patients were cartegorized into 4 risk groups. Twenty-year cause-specific survival rates for PTC patients with MACIS less than 6, 6 to 6.99, 7 to 7.99, and above 8 were 99%, 89%, 56%, and 24%, respectively. Meanwhile twenty-year cause-specific survival rates for PTC patients with AMES low, high risk were 99.2%, 46%.(12)

We could figure out the individual prognostic score with the AMES and, MACIS scoring systems. They are available before and during surgery unlike other predictive system such as the TNM system, which is only available after surgery requiring a pathological report.(13, 14)

We formulated a hypothesis that the AMES and, MACIS scoring systems are related to CLN metastasis in PTC patients.

The purpose of this article is to know whether these prognostic scores could determine the extent of surgery in PTC patients. We also assessed their usefulness and accuracy compared with [¹⁸F] FDG PET/CT in PTC patients.

We reviewed the medical records of 341 patients retrospectively with a preoperative diagnosis of PTC from 2001 to 2012. All patients had a total thyroidectomy with a routine prophylactic CLN dissection.

The clinicopathologic characteristics are presented in Table 1. Of the 341 patients, 149 had metastatic CLNs. Univariate analysis suggested younger age (below 45 years old), greater tumor size, presence of capsular invasion or extrathyroidal extension, and lymphovascular invasion to be significant factors in CLN metastasis (Table 1).(17)

Table 1
Clinicopathological data in relation to CLN in PTC

• Click for larger image

Associations of CLN metastasis with AMES scoring, MACIS scoring, and the maximum SUV of the CLN and the tumor is shown at Table 2. The number of patients with AMES low risk is 187 and that of AMES high risk is 154. Sixty six out of 187 patients had histologically positive lymph nodes following CLN dissection in the AMES low risk group and 83 out of 154 patients in the AMES high risk group. Among these patients, CLN metastasis were histologically detected in 43.7% (149/341). In this cross table, AMES high risk was a significant predictor of CLN metastasis in PTC (P=0.001). On the other hand, the relationship between MACIS score and CLN metastasis was not statistically significant (P=0.378) in this study (Table 2). [¹⁸F] FDG PET/CT scanning was performed on all the 341 patients who were included in this study. Patients were divided into two groups according to the SUV of the CLN (SUV=0 group, 279 patients; SUV>0 group, 62 patients). Ninety eight out of 279 patients had histologically positive CLN in SUV of CLN=0 group and fifty one out of 62 patients had histologically positive CLN in SUV of CLN>0 group following CLN dissection. We deduced the cut off value with an ROC curve: >0 for SUV of lymph node and >4.3 for SUV of a tumor using the Medcalc program (Figs. 1 and 2). Patients number with an SUV of tumor≤4.3 is 92 and >4.3 is 249. Twenty four out of 92 patients had histologically positive lymph nodes following CLN dissection in SUV of tumor ≤ 4.3 group and 125 out of 249 patients in SUV of tumor >4.3 group. There was a correlation between nodal status and the maximum SUV of tumor (P<0.001) and between the nodal status and the maximum SUV of the lymph node (P<0.001).

Fig. 1
Receiver operating characteristic curve of tumor SUVmax. With an SUVmax cutoff value of above 4.3, the surgeon can predict central lymph node (CLN) metastasis, and proceed total thyroidectomy and CLN dissection.

• Click for larger image
• Download as PowerPoint slide

Fig. 2
Receiver operating characteristic curve of SUVmax. With an SUVmax cutoff value of above 0, the surgeon can predict central lymph node (CLN) metastasis, and proceed total thyroidectomy and CLN dissection.

• Click for larger image
• Download as PowerPoint slide

Table 2
Associations of CLN metastasis with AMES, MACIS, SUV

• Click for larger image

In multivariate analysis, the rate of CLN metastasis was considerably higher in the cases of PTC with the higher AMES score (odd ratio [OR], 1.718; 95% confidence interval [CI], 1.073~2.752), the higher SUV of the CLN (>0) (OR, 6.525; CI, 3.184~13.371), the higher SUV of the tumor (> 4.3) (OR, 1.855; CI, 1.065~3.231) (Table 2).

In multivariate analysis, the rate of CLN metastasis was considerably higher in the cases of younger patients(<45 years old) (OR, 2.743; CI, 1.535~4.901), large tumors (>1 cm) (OR, 2.882; CI, 1.271~6.532), positive lymphovascular invasion (OR, 6.680; CI, 2.941~15.998), the higher SUV of the CLN (>0) (OR, 4.519; CI, 2.000~10.212), and the higher SUV of the tumor (>4.3) (OR, 1.217; CI, 0.637~2.327) (Table 3).(18)

Table 3
Associations of CLN metastasis with predictive values by multivariate logistic regression analysis

• Click for larger image

The area under the ROC curve of tumor SUV max was 0.62 (CI, 0.567~0.672; P=0.000), while that of the CLN was 0.65 (CI, 0.599~0.699; P=0.000) (Figs. 1 and 2). In the other words, there is an accuracy in predicting lymph node metastasis about 62% with the SUV of the tumor and 65% with the SUV of the lymph node.

PTC is nearly unique among nondermatologic invasive malignancies because it is rarely lethal, regional metastatically-involved lymph nodes do not portend distant metastases and inevitable death, and the surgical treatment options do not differ radically in functional or cosmetic outcome.(19) Nevertheless, heated debate continues to rage regarding the appropriate extent of surgical resection, principally as it relates to low-risk patients. In the vast portion of PTC patients who are at a low risk state, unilateral lobectomy is sufficient surgical treatment. But, for all high-risk patients, near total or total thyroidectomy and CLN dissection should be performed since these patients will uniformly be given radioactive iodine in an attempt to improve their poor prognosis.(20)

For less aggressive treatment in low-risk patients, several scoring systems have been developed to identify low-risk patients. There are several criteria that will predict the prognosis for example, the MACIS and AMES scoring system. However, CLN metastasis which is not accurate without a postoperative pathologic report was not incorporated in either prognostic staging systems, MACIS and AMES. If we can predict whether or not the CLN metastasis is occurring before surgery, it is helpful to determine the extent of the surgery in order to prevent an unnecessary operation and consequent complications. Then, we would like to verify the correlation between the AMES and MACIS scoring systems that can be obtained via preoperative clinical data and CLN metastasis for use in determining the operation range before surgery.

As a result of this study, AMES scoring showed a significant correlation with CLN metastasis; however, MACIS scoring was not able to prove the relevance despite there not being a big difference between the components of the two scoring systems. We had to decide the cut off value of MACIS, because it is represented by the number unlike AMES scoring which is only divided into two categories, high risk or low risk. However, it was difficult to decide the cut off value because of the insufficient number of patients with a MACIS score above 7.

On the other hand, we assessed usefulness and accuracy of [¹⁸F] FDG PET/CT in PTC patients. FDG PET/CT has a role in the management of various thyroid cancers. Incidental focal thyroid uptake on PET/CT has a high risk of malignancy, 24~36%, necessitating further diagnostic workup with ultrasound and fine-needle aspiration biopsy. PET/CT has proven useful for the detection of recurrent or metastatic disease, as well as provision of prognostic information in patients with differentiated thyroid cancers and elevated thyroglobulin levels with negative iodine scintigraphy. FDG PET/CT has also been shown to have high sensitivity and specificity and to provide prognostic information. The higher the SUV value of [¹⁸F] FDG PET/CT leads to the supposition that there is a high possibility of metastasis of CLN when considering the functional aspects of [18F] FDG PET/CT.(10) It is also interesting that the CLN metastasis is significantly related to tumor SUVs, as well as CLN SUVs. According to this study, a tumor SUV is considered to be proportional to malignant cells. And we also assessed the usefulness of MACIS and AMES scoring compared with [¹⁸F] FDG PET/CT in PTC patients. Based on these results, it is helpful to determine the operation range if considering the value of tumor SUVs, CLN SUVs, and AMES scoring.

Some studies claim that the prognostic significance and optimal treatment of nodal metastasis are not known with certainty in thyroid carcinoma unlike other epithelial malignancies. They propose that lymph node metastasis at presentation do not adversely affect survival for PTC, and therefore extensive lymph node dissection at presentation offers no advantage and may cause increased morbidity in papillary carcinoma.(21, 22, 23) However, there are some opinions that the lymph node metastasis status significantly influenced recurrence and survival in PTC (24) and the survival of patients with well-differentiated PTC is adversely affected by lymph node metastasis.(25)

In this study, the AMES preoperative and intraoperative scoring system can be helpful to make more reasonable decisions regarding operation range about whether CLN dissection should occur along with total thyroidectomy or not. However even though the AMES score is low risk, intraoperative pathologic result of suspected lymph node or the maximum SUV of the [¹⁸F] FDG PET/CT should also be considered to plan the extent of surgery.

The limitation of our study is the limited number of cases and that there was no small cartegorized study between the size or number of CLN metastasis and the prognostic scoring systems. Another weakness in this study was ulticenter and retrospective studies, makes the findings not highly generalizable. Nevertheless, it is meaningful to figure out the association between the AMES score and the SUV with CLN metastasis as a result of this study. More categorized prospective randomized controlled trial of prophylactic central lymph node dissection in PTC are required. Also large sample sizes would be required for sufficient statistical power to demonstrate more significant differences in outcomes. It is thought possible to be able to provide more effective treatment strategies to PTC patients with a new scoring system that can determine the CLN metastasis more precisely has been established.

It will be reasonable to perform a total thyroidectomy with CLN dissection if CLN metastasis is highly suspected. According to the above results, it is worth considering performing a total thyroidectomy with CLN dissection rather than a lobectomy in the cases of younger patients, a large tumor, positive lymphovascular invasion, a higher SUV of CLN, and a higher SUV of the tumor.

The AMES score also could be taken into account in the decision whether or not to perform a CLN dissection for a meaningful correlation between the AMES score and CLN metastasis.

Thus, it is also an available predictor of CLN metastasis which is used to determine the extent of PTC surgery especially considering the cost and effort.

The AMES score and [¹⁸F] FDG PET/CT are complementary but do not replace one another. The AMES scoring system along with other clinicopathological factors including [¹⁸F] FDG PET/CT are useful modalities to predict CLN metastasis which could affect the surgical therapeutic plan of PTC.