Localization of Recurrent Thyroid Cancer Using Intraoperative Ultrasound-Guided Dye Injection
- 179 Downloads
Small, nonpalpable lymph node recurrences are frequently identified in the follow-up of patients with thyroid cancer, and finding and removing these lesions in a reoperative field can be very challenging. The goal of this study was to evaluate the utility of preincision ultrasound-guided injection of blue dye into the abnormal lymph nodes to facilitate their safe and efficient removal.
We performed a prospective study between January and June 2007 at a single academic institution. Ten patients with isolated, nonpalpable nodal recurrences of papillary thyroid cancer underwent an operation for a neck recurrence (8 central, 2 lateral). A preincision ultrasound was performed in the operating room to localize the lesions, and 0.1 ml of blue dye was injected under ultrasound guidance into each abnormal lymph node. We examined the feasibility of the injection procedure, the accuracy of identifying pathologic lymph nodes, and the complications of injection.
The pathologic lymph nodes averaged 11 mm in size (range = 6–16 mm) and were detectable by ultrasound in all cases. Ultrasound-guided blue dye injection was successful in all cases. There were no complications related to dye injection. The blue node was easily identified and removed in all cases. The mean operative time was 80.4 min (range = 37–157 min).
Blue dye injection was feasible and was very useful for the identification of lymph node recurrences, especially in the reoperative neck. There were no complications related to the injection in this series. Further study is needed to determine the widespread safety and efficacy of this technique.
Patients with differentiated thyroid cancer have a very good overall prognosis. However, up to 40% of patients will have persistent or recurrent disease in their neck lymph nodes . The routine use of thyroglobulin levels and neck ultrasound in the follow-up of patients with thyroid cancer has lead to an increase in the identification of nonpalpable nodal disease in the central and lateral neck, often within previously operated fields. Prior surgery can cause extensive scarring and can distort the anatomy of the neck, making the safe identification of the recurrent laryngeal nerve and the parathyroid glands unreliable. Because of these challenges, neck reoperations can be technically demanding and can be associated with higher morbidity for the patient . Given the small size of many of these recurrences, it can often be very difficult to distinguish tumor from scar tissue and some patients are exposed to unsuccessful operative attempts at excision . Also, even though lymph nodes may be removed, the index node seen preoperatively may be missed.
Preoperative imaging, including ultrasound mapping, is essential for marking the location of the abnormal lymph nodes and for planning the approach to reoperation. However, once the incision is made and tissues are dissected, the clear orientation seen preoperatively is often lost in the scar tissue. We previously described a technique of hook-needle localization which served to focus the dissection to the affected lymph node . While this technique certainly has merit, we have encountered some difficulty in keeping the needle anchored in place, especially in very superficial locations. The needle may also be difficult to keep in position while mobilizing subplatsymal flaps to expose the thyroid bed. In the lateral neck we found the use of hook-needle localization often limited by the proximity to the vessels.
Given these challenges, we sought to develop a technique that could utilize our preoperative localization with ultrasound to better facilitate our operative identification of the relevant lesion. To do this, using ultrasound guidance we injected a small amount of blue dye directly into the lesion seen on preoperative imaging. We believe that this technique can change a difficult reoperation to an easier and safer procedure, especially in patients with extensive scar formation.
Materials and methods
Between January and June 2007, ten patients underwent surgical excision of one or more isolated lymph node metastases of papillary thyroid cancer. All of the lesions were identified on neck ultrasound, and other imaging modalities were used to verify that there was no distant disease. Eight patients underwent a repeat operation in the central neck, while two patients underwent a primary operation in the lateral neck. In the central neck, the blue dye was injected into all pathologic nodes under ultrasound guidance, with the exception of one case with five abnormal lymph nodes where the most superior and inferior lymph nodes were marked. In the lateral neck, the pathologic nodes were under 1 cm in size and these nodes were injected to ensure that they were contained within the specimen of the selective lateral neck dissection.
Three different blue dyes were used during the study, all used at a nondiluted volume of 0.1 ml/injection. Lymphazurin blue was used in three cases, indigo carmine in three cases, and methylene blue in four cases. All three dyes worked effectively to color the lymph node. There was a small amount of extranodal discoloration in all cases.
During the study period, ten patients with papillary thyroid cancer underwent an operation for a nonpalpable lymph node recurrence in the neck. All patients had a detectable thyroglobulin level at baseline (mean = 23.7 ± 18 ng/ml). All patients had a positive preoperative ultrasound showing suspicious neck lymphadenopathy. Preoperative ultrasounds showed a median of 1 suspicious lymph node (range = 1–5), with a mean size of only 11 ± 0.84 mm (range = 6–16 mm). Additional imaging studies were obtained in all but one patient, and they were positive in some: positron emission tomography (3/6), computed tomography (4/5), magnetic resonance imaging (2/4), and radioactive iodine (0/4). Fine-needle aspiration (FNA) was used to confirm a diagnosis of metastatic papillary thyroid cancer in eight of ten cases. In the other two cases, the lesions were felt to be too close to major blood vessels to access safely for FNA, but their ultrasound appearance was highly suspicious for papillary thyroid cancer.
50 ± 4.7
Gender (% female)
Central neck recurrence
80% (n = 8)
Lateral neck recurrence
20% (n = 2)
No. of prior neck operations
2.6 ± 0.5
No. of prior operations in the field
1.7 ± 0.4
Time to first recurrence (months)
61 ± 20
Time since last neck operation (months)
47 ± 14
In all cases, the suspicious node on ultrasound was identified and removed. Additional pathologic nodes, not seen on ultrasound, were removed in 40% of cases. A mean of 2.7 ± 0.58 pathologic nodes were removed from the area of the blue dye in each case. Despite the fact that only one node was injected, we found that in one case a second adjacent pathologic node (not seen on ultrasound) also turned blue demonstrating rapid communication between nodal metastases. In no case did the presence of the dye inhibit the pathologist’s ability to evaluate or interpret the specimen.
There were no complications related to the injections (no nerve injury/dysfunction, no hematoma). In nine of ten cases the node was located in direct proximity to a nerve (either recurrent laryngeal or spinal accessory). There was no postoperatively identified nerve dysfunction related to the injection or surgery and no patients complained of hoarseness or voice changes. However, routine direct laryngoscopy was not performed. Postoperative calcium was greater than 8.0 mg/dl in all patients (mean = 8.6 ± 0.12 mg/dl) and there were no symptoms of hypocalcemia. One patient had a parathyroid gland incidentally removed; it was contained within a central neck specimen of involved lymph nodes. No patients had permanent skin discoloration related to the injection or an allergic reaction to the dye.
Operative times for reoperative central neck dissections
Central neck cases with blue dye
Control central neck cases
Preincision time (min)
Case time (min)
Total time (min)
Surgery, when feasible, is the first-line treatment of thyroid cancer recurrences in the neck and is the treatment most likely to cure the patient. Since surgery remains the most effective treatment option for patients with nodal recurrences, many attempts have been made to develop techniques to facilitate reoperations, making them both safer and more effective. Due to the scar tissue that is frequently present and the small size of many of these lesions, accurate preoperative and intraoperative localization is essential. Neck ultrasound is the most commonly used intraoperative adjunct to localize nodal disease. We have also previously described a technique of ultrasound-guided needle localization . Radio-guided surgery has also been applied to this difficult clinical situation with variable success. It however has not been generally used as only about 25% of metastatic lymph nodes take up radioactive iodine [5, 6]. Agents that have been used include I-131, I-123, Tc-99 m sestamibi, and FDG-PET [5, 7, 8, 9, 10, 11]. A recent report has described an ultrasound-guided injection of radiotracer directly into the affected lymph nodes in order to facilitate intraoperative identification using a gamma probe . Alternatives to surgery for lymph node recurrences include using systemic radioactive iodine and trying to locally ablate individual lymph nodes with either alcohol or radiofrequency ablation [13, 14, 15].
The most common and sensitive technique for intraoperative localization of cervical nodal metastases is ultrasound. Lymph nodes that are affected by thyroid cancer are identified by their suspicious features including microcalcifications, loss of fatty hilum, cystic changes, and an enlarged anterior/posterior diameter. The first report, to our knowledge, of using intraoperative ultrasound to guide the resection of a neck recurrence was by Desai et al. in 2001 . They reported using intraoperative ultrasound to guide their resection of a recurrent lesion after failing to identify the lesion after initial dissection. A second study in 2002 described the use of intraoperative ultrasound for the localization of recurrent thyroid cancer in 13 patients and found it useful in 54% . Since then many surgeons have started incorporating intraoperative ultrasound, either preincision to mark the locations of the suspicious nodes or intraoperatively to guide their dissection.
Despite the usefulness of preoperative ultrasound, it does have its limitations. Marks on the skin, guided by a preoperative ultrasound, can help one determine optimal incision placement. However, once subplatysmal flaps are elevated and the dissection has begun, the skin no longer provides a good reference point to mark the exact location of the involved node. We have used a hook-needle-guided resection since 2005 and have had some success with this technique . While we have had minimal complications from the needle placement, in some patients the needle became dislodged during retraction, raising concern over the accuracy of this localization technique. A report has recently been published in which preoperative ultrasound was used to inject radioactive material directly into the affected lymph nodes, thus facilitating their identification by using a gamma probe intraoperatively . While helpful, localization with a gamma probe is much more subtle than the visualization provided by our blue dye technique.
The goals of our study were to determine the feasibility, utility, and the safety of this new technique which proved to be feasible in all cases, even in lymph nodes as small as 6 mm. The dye was deemed useful in every case and in no case was it misleading. In the reoperative cases it allowed us to focus our dissection to the relevant area, minimizing unnecessary dissection in a scarred field usually near the recurrent laryngeal nerve and parathyroid glands. In the lateral neck dissections, it ensured that our selective dissection was carried high enough and that the pathologic node seen on ultrasound and usually confirmed by cytologic examination was actually included in our excised specimen.
Blue dyes are used for a variety of applications in the operating room, including sentinel lymph node localization in breast cancer and melanoma and to aid in the identification of injuries to structures such as the ureter. Methylene blue and toluidine blue have also been used intravenously for the localization of parathyroid tumors [18, 19]. These dyes are all considered safe for use in humans. In our small study we had no dye-related complications. However, there are potential toxicities that can be related to the use of these dyes. All three of the dyes used in this study have the potential to cause an allergic reaction, and lymphazurin blue has rarely been associated with an anaphylactic reaction. Methylene blue can have some local toxicity which is likely related to the volume and concentration that are used. Local skin necrosis has been seen with soft tissue injection , and nerve toxicity has been described when it is used for epidural injection . However, these complications are typically seen with much larger-volume injections than were used used for this study. While in this study we used the dyes undiluted, we have since found that methylene blue can be safely used when diluted up to tenfold without a significant decline in its efficacy; this may help minimize the risks of local toxicity. Based on our experience with this study, we found that methylene blue was the most useful of the three dyes because it did not wash out of the tissues as rapidly. However, since there are theoretical risks to its use, we feel that its safety must be further validated for this particular use.
A secondary aim of this study was to examine the effect of this intervention on operative time. In our study we found that the addition of this technique added about 20–25 min to the preincision operative time. Because most of the ultrasound-guided injections were performed by a radiologist, much of the preincision time was spent waiting for the radiologist to arrive in the operating room and set up the equipment. When the ultrasound was performed by the operating surgeon, the additional preincision time was reduced to less than 10 min in most cases. The addition of this technique reduces actual operative time, especially in the reoperative field, but to date much of this saved time is offset by the additional time required to perform the preoperative injection.
Ablative procedures such as injection of alcohol or radiofrequency ablation have been proposed as an alternative to surgery, as a “less invasive” option. These techniques have the advantage that they can be performed on an outpatient basis, do not require general anesthesia, and patients may resume normal activities immediately. Unfortunately, experience with these techniques is limited, recurrent laryngeal nerve injuries have occurred, and the outcomes of these few small series may not be applicable to the greater population [13, 15]. One of the challenges of the ablative procedures is the need for absolute precision, especially within the central neck, due to the proximity of the recurrent laryngeal nerve (RLN). Ethanol, if extravasated from the lymph node, is toxic to the surrounding tissues, including the RLN, and can cause vocal cord paralysis. To minimize toxicity to surrounding tissues, ethanol ablation often requires multiple low-dose treatments over time, which makes it impractical for larger lesions . Radiofrequency ablation (RFA) is difficult to perform in small lesions and cannot be used in close proximity to critical structures, which is frequently the case in the neck. The use of RFA in the central neck has also led to RLN injury [13, 14]. Given the limitations of these techniques, we still feel that operative intervention, after tumor localization, is the best treatment option for most patients with recurrent disease within the neck.
A preincision ultrasound-guided blue dye injection is feasible, even in small nonpalpable lymph nodes. Blue dye localization facilitates the identification and successful removal of these lymph nodes, even in a densely scarred reoperative field. There were no complications related to the dye injection. Although this technique decreased actual operative time in the reoperative cases, it added to the preincision operative time leading to no significant effect on overall operative time. This technique shows promise for facilitating the operative treatment of recurrent thyroid cancer in the central and lateral neck, and further evaluation of the efficacy and safety of this technique is needed.
This work was supported in part by Friends of Endocrinology, The Gerald Heller Family Foundation, The Helen and Sanford Diller Foundation, and the Bell Charitable Trust.