Long-Term Efficacy of Ethanol Ablation as Treatment of Metastatic Lymph Nodes From Papillary Thyroid Carcinoma.

CONTEXT: Ethanol ablation (EA) is considered an alternative to surgery for metastatic lymph nodes from papillary thyroid carcinoma (PTC) in selected patients.
OBJECTIVE: The aim of this study was to evaluate the long-term efficacy and safety of this treatment. DESIGN AND
SETTING: Adult patients with PTC who had received EA in lymph node metastasis at a tertiary referral center, and were included in a published study from 2011, were invited to participate in this follow-up study.
METHODS: Radiologic and medical history were reviewed. Ultrasound examination of the neck was performed by radiologists, and clinical examination was performed by an endocrine surgeon. Response was reported according to predefined criteria for satisfactory EA treatment. Adverse events associated with EA were evaluated. Cause of death was reported for deceased patients.
RESULTS: From the 2011 study, 51 of 63 patients were included. Forty-four patients were reexamined (67/109 lesions) and 7 patients were deceased. Median follow-up time from primary surgery was 14.5 years. Median follow-up from the latest performed EA in the 2011 study was 11.3 years. Local control was permanently achieved in most patients (80%). Recurrence within an ablated node was registered in 13 metastases in 10 patients. Seven of these patients also had recurrent disease elsewhere in the neck. No major side effects were reported.
CONCLUSION: EA is a minimally invasive procedure with a low risk of complications. Our data suggest that EA is a safe and efficient treatment, providing excellent results for a large group of patients in the long term.

Treatment of small neck lymph node metastases from papillary thyroid carcinoma (PTC) is challenging. Our institution introduced, among the first institutions worldwide, ultrasound (US)-guided ethanol ablation (EA) as an optional treatment for neck lymph node metastases from PTC in 2004 after promising results from ethanol ablations were published in 2002 by Lewis et al (1). Our results from treatment with EA in a 5-year period (2004-2009) were presented by Heilo et al in 2011 (2).

Neck lymph node metastases are common and found in 20% to 50% of patients with PTC at diagnosis (3). With the use of more sensitive diagnostic methods, studies have reported a frequency of micrometastasis (< 2 mm) approaching 90% (4, 5). The presence of lymph node metastases is considered an independent risk factor for increased morbidity and locoregional recurrent disease, but the effect on survival is still controversial (6). Surgical extirpation is recommended when lymph node metastases are present before or during primary surgery. The benefits of prophylactic lymph node dissection are still controversial, but it is usually recommended in cases with advanced stage primary tumor (T3-4) (7, 8).

Studies have demonstrated a risk of recurrence in neck lymph nodes ranging from 5% to 21%, with most of the cases occurring within 10 years after primary treatment (9-11). High-resolution US, thorough morphologic evaluation and sensitive assays for detection of thyroglobulin (Tg) in fine-needle aspiration biopsy washouts (FNAB-Tg) can detect metastases as small as 2 to 3 mm. The availability of such tools contributes to an increase in the number of lymph node metastases detected early in the clinical course, leading to a clinical dilemma when the appropriate therapeutic approach should be determined (12).

Patients with persistent or recurrent disease in the thyroid bed or neck lymph nodes after thyroidectomy and adjuvant radioiodine therapy may have limited benefit of a second radioactive therapy (13). The standard treatment for recurrent disease is still considered to be surgery, although it is well known that repeated neck surgery can be technically challenging because of adhesions, fibrosis, and scar tissue with disruption of the normal anatomy and tissue planes (14, 15). Therefore, localized treatments like thermal ablation and ethanol ablation have been proposed as alternatives to surgery in selected patients with a limited number of metastases or a high risk of local or general complications related to surgery (3).

In 2011, our institution published the results from treatment of 109 neck lymph node metastases in 63 patients with PTC (2). According to predefined criteria, a total of 92 (84%) of the lymph nodes were successfully treated in the study. The mean follow-up was 38 months (range, 3-72). No major complications related to the EA treatment were recognized. Additional studies with promising results from EA of nodal neck metastasis have later been published, but the data for long-term follow-up are still sparse (16-20). In this article, we present the results from a long-term follow-up survey of the patients from the 2011 study (2).

Materials and Methods

The study was approved by the Regional Committee for Medical and Health Research Ethics. The cause of death was obtained from the Norwegian Cause of Death Registry for deceased patients.

Patients

Patients previously treated with EA in our initial report (2) were identified and invited to participate in the follow-up study. Of the 63 participating patients, 7 were deceased. One patient had surgery following the first study and was lost to follow-up. Forty-four of 55 eligible patients were successfully included and reexamined in this study (Fig. 1). The remaining 11 patients did not give their active consent for participation or did not respond to the invitation. Of the reexamined patients, 59% were female. Median age at follow-up was 63 years (range, 35-86). Median interval since primary surgery was 14.5 years (range, 10-36).

Figure 1.

Flow chart of included patients.

Methods

All the included patients had undergone total or subtotal thyroidectomy in 1 or 2 steps, followed by 131I-ablation. All the patients had 3 or fewer metastatic lymph nodes at the time of inclusion in the first study (2). The patients in the current study were reexamined on an outpatient basis. Last known status concerning the patient’s PTC was obtained from the patient’s medical record and the radiological (RIS/PACS) system. The patients were examined with US using a 12.5 MHz broadband linear US transducer on GE Logiq E9 (GE Healthcare, Milwaukee, WI, USA). The US examinations were performed by radiologists (K.H.F., L.I.F.) with more than 5 years of experience of neck US examinations. We evaluated all previously confirmed neck lymph node metastasis and registered any new lymph nodes being suspicious of metastasis. The anteroposterior, transverse, and longitudinal diameters were recorded, as was the efficacy of previous EA and time to recurrence. Any adverse events associated with EA were evaluated.

US-guided FNAB smear and FNAB Tg needle-wash specimen, as described by Sigstad et al, was undertaken whenever lymph nodes suspicious for metastases were detected (12). EA procedure was performed when considered clinically indicated, usually after a multidisciplinary team meeting, as described by Heilo et al (2). Patients were informed about possible adverse events before the procedure. The procedures were performed without premedication and coagulation tests. Total volume of injected ethanol and side effects were recorded. The patients were discharged a short time after the procedure.

The patients were examined and interviewed by experienced endocrine surgeons. Details from the medical history, experiences with prior treatment, and side effects from EA were reviewed. The included and deceased patients were staged according to the American Joint Committee on Cancer Tumor Node Metastasis, 8th edition (Table 1) (21). Blood samples were obtained for biochemical analysis (S-TSH, S-freeT4, S-Tg, and S-Tg antibodies). The outcomes of EA were considered successful when 1 or more of previously published criteria were fulfilled (Table 2) (2). Recurrent disease was confirmed by morphological confirmation of FNAB smear and/or Tg-FNAB values greater than S-Tg values (not corrected for dilution) in FNAB needle-wash sample. Recurrent disease in a location where EA had not been performed was defined as recurrence outside of previously EA-treated lesions. Lesion volumes were calculated with the formula for the volume of an ellipsoid (). Mann-Whitney U test was used to compare median volumes of lesions with durable response to recurrent lesions. Disease-specific death was calculated for the deceased patients.

Table 1.

Distribution of TNM status and disease stage at initial diagnosis

	Included patients		Deceased patients		Test of significanceb
	No.	%	No.	%
Primary tumor
T1	9	20.5	2	28.6	P = 0.811
T2	8	18.2	2	28.6	U = 122
T3	10	22.7	2	28.6
T4	6	13.6	1	14.3
Tx	11	25.0	0	0.0
Nodal status
N0	6	13.6	1	28.6	P = 0.936
N1	30	68.2	6	71.4	U = 123
Nx	8	18.2	0	0.0
Stage
I	34	81.0	2	28.6	P = 0.042*
II	4	9.5	4	57.1	U = 76
III	4	9.5	1	14.3
IV	0	0.0	0	0.0
Missing	2		-

Abbreviation: TNM, tumor, node, metastasis.

*Statistical significance level P < 0.05.

American Joint Committee on Cancer, 8th edition. All patients were staged as M0 (no distant metastases) at initial diagnosis.

Mann-Whitney U test, 2-sided.

Table 2.

Criteria for satisfactory ethanol ablation

1	Complete lymph node disappearance
2	Reduction of the anteroposterior lymph node diameter to ≤ 4 mm without visible vascularization
3	Normalization of lymph node size and appearance, and in some nodes scar formation
4	No detectable malignant cells in the FNAB specimens and no detectable Tg in FNAB needle wash

Abbreviations: FNAB, fine-needle aspiration biopsy; Tg, thyroglobulin.

Tg (both in serum samples and in FNAB needle-wash samples) was measured with an in-house Tg assay and human TgAb (in serum samples) was analyzed using a competitive assay (Brahms Kryptor, Henningsdorf, Germany) as described by Broughton et al (22). In the in-house Tg-assay 4 monoclonal antibodies (E44 RRID:AB_2894939, E45 RRID:AB_2894941, I24 RRID:AB_2894937, and E40 RRID:AB_2894934) were used.

Results

Out of the 63 eligible patients, 7 had died since the end of the study by Heilo et al (2). The deceased patients represented 15 of the 109 EA-treated lesions in our initial report, with a median of 1 EA-treated lesion per patient (range, 1-4). Six of the 7 deceased patients (86%) were females. The median age at death was 68 years (range, 55-82). Median time from primary surgery to death was 11 years (range, 5-35). The distributions of disease stage (I-IV) in the reexamined and deceased patients differed significantly (median stage I vs II, U = 76, P = 0.042). There was no significant difference in the distribution of tumor stage (T) or nodal status (N) in the 2 groups (Table 1). Thyroid cancer was identified as cause of death in 42.9% (3/7) of the deceased patients, giving a cause-specific death in 4.8% (43/63) of the patients from the 2011 study. Two of the patients with cause-specific death had stage II disease and the last patient had stage III disease. All 3 patients were > 55 years at the time of diagnosis (median 59; range, 58-76).

The 44 reexamined patients in the current study represented 67 of the 109 lesions (62%) in our initial report. The number of EA-treated lesions in each of the reexamined patients ranged from 1 to 4, with most of the patients having 1 (64%) or 2 (23%) lesions treated with ethanol ablations. The remaining 27 lesions were distributed in the 12 nonincluded patients in this follow-up study (median 2; range, 1-6) with 75% of these patients having 1 or 2 lesions treated by EA in the initial report. Most reexamined lesions (40/67) were in the lateral compartments, whereas 27 lesions were located in the central compartment. The median number of EA in each of the reexamined lesions in the initial study was 2 (range, 1-6) with the majority (88%) having 1 to 3 EA procedure(s). The median follow-up time since the latest performed EA in the initial study was 124 months (range, 90-160). Most of the patients (80%) were treated with EA in a single neck field, 18% were treated in 2 fields, and 1 patient was treated in the central field as well as both lateral neck fields. Almost all lesions reexamined in this study (97%) had fulfilled 1 or more of the defined response criteria at the end of the study by Heilo et al. At follow-up, a durable response was demonstrated in 54 metastases (81%), of which a great majority (49) were nondetectable, whereas 2 and 3 lesions fulfilled response criteria 2 and 3, respectively (Table 2). Consequently, no FNAB was performed in these lesions.

We registered a total of 13 recurrent lesions (in 10 patients) at the ablation site of earlier EA-treated lymph nodes, giving a recurrence rate at the ablation site of 19%. Most of these lesions were in the lateral neck (10), whereas the remaining 3 lesions were in the central compartment. Seven of these residual lesions had been detected in the time before the follow-up examination in the current study. Of these 7 lesions, 1 had response from additional EA (Fig. 2A-C), 4 lesions were still receiving EA, and 1 lesion was followed by active surveillance at the time of inclusion. One lesion had been surgically removed, and this patient had no sign of recurrent disease at the time of the follow-up examination. The remaining 6 lesions were detected at the follow-up examination in the current study. Three of these lesions, in 2 patients, were referred to surgery after consensus meetings in our multidisciplinary team. Two recurrent lesions detected at follow-up, 1 in each of 2 patients, were considered unsuitable for surgery because of the candidates’ age and comorbidity, and a new EA was consequently performed. The remaining lesion will be followed by active surveillance. The time frame for EAs, time of recurrence, and actual status of the recurrences in EA-treated lesions are shown in Fig. 3. In the 13 recurrent lesions, about one-half (6) had a diameter ≥ 10 mm (median 13.5 mm; range, 10-18). There was no significant difference in the initial volume of the lesions having durable response (median 125.7 mL; range, 6.3-1809.6), compared with recurring lesions (median 205.3 mL; range, 25.1-622.0) (U = 406, P = 0.38). The median number of ethanol ablations was similar (2) for lesions having durable response (range, 1-5) and recurring lesions (range, 1-6). The rate of recurrence was 19.1% (9/47) in the lesions that had been nondetectable (fulfilling diagnosis criteria 1 at the termination of the study by Heilo et al, and 23.5% (4/17) in the lesions fulfilling any other of the given response criteria.

Figure 2.

A 72-year-old male patient had received EA in a total of 6 metastatic neck lymph nodes. Five had fulfilled the response criteria, but 1 lesion in the right lateral compartment was still under active EA treatment at follow-up. B-mode ultrasound images are showing a successfully treated residual lesion in the central compartment. The lesion (white open arrow) was initially treated with a single injection of 0.3 mL ethanol (A). The lesion was considered successfully treated with a remaining residual lesion of 4 × 4 × 4 mm with no visible vascularity after 4 months (B). Recurrence with growth of the lesion (white solid arrow) and reappearance of vascularization occurred 7 years after the initial ablation (C). The lesion was then treated with a total of 4 additional ethanol injections and has been persistently undetectable on later follow-up examinations.

Figure 3.

The figure displays the time course of 13 recurrences found at the ablation site in 10 patients. The top line displays the number of years (Y) of follow-up. The horizontal lines to the left represent time periods where the initial ethanol injections were performed. The horizontal lines to the right display the time of recurrence and actual status of the recurrent lesions. Closed lines represent completed series of ethanol ablations, whereas open lines are indicating ongoing treatment with ethanol ablations.

We mapped a total of 22 neck lymph node metastasis (in 13 patients) in locations outside of previously EA-treated lesions. Sixteen of the lesions were in the lateral compartments and 6 in the central compartment. Most of these new lesions (14/22) were found in 7 of the 10 patients who also experienced recurrent disease at the ablation site. Response after EA was achieved in 42.9% (6/14) of the lesions in this patient group. The remaining 8 lesions were found in 6 patients who also had durable response in all their EA-treated lesions from the initial report. All the new lesions in these patients responded to EA after a median number of injections of 1.5 (1-4); thus, none of the patients in this group had detectable disease at the follow-up examination. All patients in the latter group, with durable response in all lesions, had stage I disease. This was the fact for only one-half of the patients in the group with recurrences in ablated nodes and inferior efficacy of EA in new lesions (median 1.5; range, 1-3, 1 missing value). The difference in distribution of disease stage between the 2 groups was not significant (U = 9.0, P = 0.180), however. There was no significant difference in the distribution of disease stage between the group of reexamined patients experiencing recurrent disease in the follow-up time (n = 15, 1 missing value) and those who did not have any recurrences (n = 27, 1 missing value) (U = 226.50, P = 0.357).

There was no sign of recurrent disease in the neck in 80% of the patients (35/44) at the time of follow-up. Even though all patients had undergone radioiodine ablation following initial surgery, 6 patients in the group without detectable disease (n = 35) had elevated S-Tg values and negative S-Tg antibodies at follow-up. The measured S-Tg was low (≤ 1 µg/L) for 4 of these patients. For the remaining 2 patients, the values were 3.7 and 4.3 µg/L, respectively. Two patients had nondetectable S-Tg-values because of positive S-Tg antibodies.

In the group in which recurrent disease was detected (n = 9), elevated S-Tg values (range, 0.60-37 µg/L) were seen in 6 patients. Two patients had undetectable S-Tg values and negative S-Tg antibodies and 1 patient had undetectable S-Tg value and positive S-Tg antibodies.

Discussion

We present the long-term results from the so-far largest published patient group treated with EA for metastatic neck lymph nodes in patients with PTC (15). Local control was achieved in most patients (80%), with a median follow up of 11.3 years. However, recurrence within an ablated node was found in 13 EA-treated metastases in 10 patients. More than one-half of these patients (7/10) also had recurrent disease outside of previously ablated lesions. The observation time in our material is markedly longer (11.3 years) than prior studies evaluating this treatment.

Since Lewis et al presented their results from EA in recurrent neck lymph node disease in 2002 (1), several studies have been published, suggesting a favorable outcome of this treatment (2, 16-20, 23-26). Size reduction is seen in > 90% of the treated lesions and a significant number of lesions are not detectable on US after successful EA treatment. The first published studies, up until 2011, reported low rates of local progression or recurrences within ablated nodes (1, 2, 23-26). These findings could be due to the relatively shorter follow-up times compared with later case series. In a study from 2013, local progression was seen in 23.8% (5/21) of the lesions after a mean follow-up of 38.5 months (range, 0-94) (20). This contrasts with the results published by Hay et al the same year, in which no local progression within an ablated node was reported after a mean follow-up-time of 65 months (range, 5-157) (19). The latter study found, however, increasing size of 2/37 lesions in 1 patient during the follow-up time, but these lesions were still considered successfully treated because of the lack of vascularization and undetectable S-Tg-level at 110 months. Strajina et al reported progression within an ablated node in 16% of the EA-treated lateral neck lymph nodes (7/43) after a median follow-up of 54 months (range, 6-102) (17). A recurrence rate of 19% of the treated lesions was found in the current study. The rate is within the range of the previously reported suggesting that a prolonged observation time would not necessarily lead to an increasing number of recurrences (17, 20). Unfortunately, several of the patients in our study were not summoned for regular follow-up. Consequently, the time from the last preceding EA to occurrence of recurrence was not assessable.

The number of complete disappearances in our long-term follow-up was slightly higher (73%), but comparable to the numbers previously reported. There has been some concern about the risk of residual tumor deposits in the lesions that show a decrease in size, but still are visible on US. Lim et al performed FNAB on all detectable lesions after EA and found that tumor cells were present in 75% (15/20) of the cases, but no further increase in the size of these lesions were seen during the follow-up period (25). We found a slightly higher, but comparable recurrence rate in these lesions compared with the lesions with complete disappearance after EA treatment.

It has been proposed that EA is more effective in treating tiny metastatic lesions. Strajina et al found significant differences between the initial size of the lesions and the differences in treatment response from EA (17). The initial diameter of the recurring EA-treated lesions in our material ranged from 4 to 18 mm, with about one-half the lesions (6/13) having a diameter ≥ 10 mm. The recurring lesions in our material did not have a significantly larger initial volume than the successfully treated lesions; however, the numbers are small and encumbered with uncertainty.

Because EA represents a selective treatment of confirmed lymph node metastases, resembling a “berry-picking” procedure, the risk of subsequent recurrences in locations outside of the ablation sites are of concern. We detected “new” neck lymph node metastases in 27% (12/44) of the patients in our material. One-half of these also had recurrent disease in 1 or more of the EA ablation sites. The other one-half had persistent response in their EA-treated lesions while experiencing new lesions elsewhere in the neck. Two studies reported the proportion of patients with “new” lesions to be 9% and 24%, respectively (17, 19). In light of the longer follow-up time in this case series, our results suggest that the proportion of patients with new lesions outside the ablation sites would not increase with time after diagnosis. The results are, however, not conclusive. Most of the new lesions in our material were found in patients who also had recurrences at the ablation site of lesions treated in the initial report. These patients had a poorer efficacy of EA in the new lesions than the patients without recurrent EA-treated lesions, with durable response in less than one-half of the lesions. This contrasts with Hay et al reporting successful EA treatment of 15 lesions outside of the ablation site (19), which is consistent with our results from EA in the new lesions in the group without recurrences within ablated lesions. The reason for the difference in efficacy between the 2 groups is not clear. All patients in the group with durable response in all their lesions had stage I disease, whereas one-half of the patients in the group with recurrences in ablated nodes and inferior efficacy of EA in new lesions had disease stage II-III. This could imply that more aggressive disease in these patients could be a contributing factor to this effect, but the difference in distribution of disease stage between the 2 groups was not significant, however.

Surgery is still considered to be the treatment of choice for lymph node metastases from PTC, but there are still no randomized trials comparing EA and surgery. In a systematic review and pooled analysis by Fontenot et al in 2015, the success rate of surgery was slightly higher than the success rate for EA treatment (94.8% vs 87.5%) (27). They found no significant difference in risk of recurrence in the treated lesion or elsewhere in the neck. The comparison between the methods was complicated by an uneven distribution with more lesions in the central compartment treated with surgery than EA. Despite surgery being the treatment of choice, the indolent nature of the disease is making the goal of managing recurrent or persistent lymph node metastasis a persistent matter of controversy. From the patient’s perspective, it seems like a reasonable strategy to get rid of any sign of disease and prevent recurrences. The choice of treatment must be balanced against the risk for adverse effects and complications from the treatment that might impair the patients function and/or quality of life, however. Since the study by Heilo et al, there has been a change in the approach to locoregional recurrences of PTC. Studies have confirmed that a considerable number of tiny metastatic neck lesions do not progress, or possibly progress slowly and without affecting the patient’s function or compromising neighboring structures (28-30). Thus, it has become more acceptable to follow small and nonprogressive lesions by active surveillance.

Our results support the present consensus that EA should be considered an alternative to repeated neck surgery in selected patients with increased risk associated with surgery and general anesthesia, as well as for patients refusing to undergo additional surgery. The treatment is shown to make most of the lesions fulfill the response criteria. Furthermore, locoregional control in the neck region is achievable in most patients, also in the long term. We achieved local control after a limited number of injections for a limited period for most of the patients. The efficacy of EA in recurrences in previously ablated nodes, as well as in new recurrent lesions in the same patients, was inferior to the overall efficacy. Based on our experiences, a lower threshold for the consideration of other treatment alternatives than EA should be obtained in these cases. We did not detect any major or persisting side effects in this long-term follow-up. The result is in accordance with several previous reports, concluding that EA could be considered a safe and well-tolerated treatment option (1, 2, 16-20, 23-26). In a recent overview article, the authors favor the use of EA for locally recurrent differentiated thyroid carcinoma in nonsurgical candidates (31). However, they emphasize the need for randomized controlled trials to clarify the true value and risks of the treatment.

In conclusion, EA is a safe and efficient treatment, providing excellent results for a large group of patients also in the long run.