Chyle leakage or fistula is a troublesome complication usually associated with LLND for thyroid cancer [9]. Several measures have been studied for intraoperative lymphatic treatment in neck dissection, especially in the area around the venous angle, where the main trunk of the lymph merges [10]. For chyle leakage prevention, ligation of lymph vessels, including the surrounding tissue, is generally accepted. Monden et al. [11] reported that chyle leakage was safely treated with an energy device such as Harmonic Focus® (ETHICON Endo-Surgery, Inc.). However, Wang et al. [12] stated that lymphatic vessel treatment with energy devices is insufficient to prevent chyle leakage. Furthermore, several studies reported chylothorax following endoscopic lung or esophageal cancer resection using certain energy devices; therefore, such devices are not completely reliable [13, 14]. Zhang et al. [15] recently reported the usefulness of pedicled omohyoid flap coverings in preventing postoperative lymph node or chyle leakage. However, their study enrolled only a few patients, and the procedure was complicated. Thus, the generalizability of the findings is questionable [15]. As mentioned above, no effective chyle prevention method has yet been established.

The primary reason for the difficulty in preventing chyle leakage is the anatomy of lymphatic vessels and their complex networks [1, 16, 17]. Lymphatic vessels with a diameter of ≥ 0.2 mm are composed of smooth muscle and fibrous tissues, whereas thin lymphatic vessels have insufficient smooth muscle and fibrous tissues and form fragile tubular organs [18]. Prolonged fasting before surgery causes lymphatic vessel collapse, decreased lymphatic flow, and colorless and transparent lymphatic fluid. An effective way to avoid chyle leakage is to identify and ligate lymphatic ducts; however, it is difficult to distinguish lymphatic vessels from the surrounding connective tissues following prolonged fasting.

After experiencing complex cases of chyle leakage, we began providing ice cream preoperatively to prevent chyle leakage or fistula in patients scheduled to undergo LLND. Ice cream containing long-chain fatty acids is chylomicronized in the small intestine and increases lymphatic flow 3- to tenfold [6, 18]. In the study by Shackcloth et al. [7], patients were provided ice cream three hours before surgery; this safe method does not require special equipment and identifies the thoracic duct. This approach is especially indicated in patients with severe lymph node metastasis in region IV and those with postoperative chyle.

The frequency of chyle leakage varies according to the state of lateral neck lymph node metastasis and the extent of surgery. In this study, LND was performed only in patients with clinical lymph node metastasis to the lateral neck, and most patients underwent total thyroidectomy. The incidence rate of postoperative chyle leakage in this study was 2.2%, which is not superior to the rates reported in previous studies (2.9–8.3%) [12, 19,20,21,22]. However, in our patients, the degree of chyle leakage was mild, and most did not require reoperation or long-term hospitalization. The mean time to drain removal was 3.2 days, which is approximately 1 day longer than the 2 days specified in our postoperative path. Furthermore, there was no fasting, long local pressure dressing, or use of expensive medications. We want to emphasize that our prospective study is based on a novel treatment strategy aimed at preventing chyle leakage and managing its occurrence, and the number of patients in this study was higher than in other retrospective studies. According to previous reports, conservative treatment was performed in 50–89%, and reoperation was performed in 11–50% of patients [12, 19,20,21,22]. In addition, the median time to drain removal after conservative management was 21.5 (5–60) days, 11.3 days, and 12 days (range 8–22 days) in studies by Saito et al. [20], Roh et al. [21], and Glenn et al. [22], respectively. Our results may be explained by the small size of the leaking lymphatic vessels and the implementation of the DNPR test.

Prior to the present study, we performed the DNPR test and achieved successful outcomes in our patients, including a consultation case from another facility. In the present study, several patients had only a small amount of chyle, as ice cream was consumed before the surgical procedures. Although the DNPR test seemed irrelevant in the present patients, the amount of chyle was significantly reduced after the test in patients with persistent or increased chyle. The physiological internal pressures are 0–15 mmHg for large lymphatic vessels, such as the thoracic duct, and 0–0.5 mmHg for other lymphatic vessels [23]. Under these circumstances, drain suction pressure increases the wound pressure, which causes persistent chyle leakage. It is presumed that once the wound pressure is higher than the intralymphatic pressure, owing to the release or clamping of the drain, the amount of chyle is significantly reduced, thus allowing for early drain removal. We believe that such hypotensive lymphatics are easily healed by wound pressure. Conversely, persistent negative pressure in the drain is expected to form an iatrogenic fistula between the lymphatic vessels and wound. The DNPR test is helpful in patients with chyle leakage or fistulas from the hypotensive lymphatic vessels. In our patients, the timing of starting the DNPR test was left to the discretion of the attending surgeons; however, our results suggest that DNPR testing should have been performed considerably earlier than it actually was in our study. It may be possible to stop drain suction when diagnosing mild chyle in patients undergoing LLND after ice cream consumption. If it does, chyle leakage should not be considered a serious complication.

One patient required reoperation for chyle leakage from the thoracic duct, which was ligated during the initial surgery. The following mechanisms may trigger the loosening of the thoracic duct ligation. The thoracic duct is highly elastic and prone to dilation and collapse. Ice cream consumption enhanced the thoracic duct, as confirmed intraoperatively; however, postoperative fasting severely shrank and loosened the ligature. Finally, chyle leakage occurred in the ruptured thoracic duct after the oral intake. Considering the possibility of lymphatic vessel collapse during postoperative fasting, it is better to perform double ligature of dilated lymphatic vessels in cases with ice cream consumption. Early determination of the effectiveness of conservative treatment is critical to avoid prolonged hospitalization and special treatment for patients with chyle leakage. There may be pros and cons regarding the timing until reoperation, but considering the restrictions on the patient's hospital stay, we believe it is better to perform it earlier than it is generally performed. If the reason for delaying reoperation is chyle curability, we recommend LLND be performed following ice cream consumption to familiarize surgeons with the anatomy of the thoracic duct. However, based on our experience, ice cream consumption likely reduces the number of patients who will require reoperation for chyle leakage.

Some studies have reported the efficacy of using somatostatin analogs for postoperative chyle [24, 25]. However, in Japan, the use of these drugs for managing chyle is not covered by national health insurance and requires ethics committee approval. Furthermore, these drugs are expensive and place a financial burden on patients, even if they can be used. We recommend the DNPR test be performed in patients who are hesitant to undergo reoperation before using somatostatin analogs.

One limitation of our study is its nonrandomized design, as we initially thought that chyle leakage could be prevented in almost all patients following ice cream consumption. Therefore, a randomized study design was not considered appropriate. We believe that groundbreaking treatment can be established in various cases by concentrating on ice cream consumption. Furthermore, our proposed techniques have potential applications in treating other cervical, thoracic, and abdominal surgeries and are not limited to thyroid surgery.

Our large prospective study concluded that ice cream consumption before LLND did not completely prevent chyle leakage in thyroid cancer patients. However, the degree of chyle leakage was mild owing to the avoidance of damage to major lymph vessels, and the drain was able to be removed within a short time with minimal burden on the patients. We recommend this simple and safe preoperative procedure for patients with malignant tumors requiring neck dissection, especially those with clinical lymph node metastases in region IV and those scheduled for reoperation for chyle leakage or fistula. Timely DNPR testing is also recommended to reduce the duration of drain placement.