Japanese clinical practice guidelines for vascular anomalies 2017

Recommendation:

It is necessary to judge the time to begin endovascular or surgical treatment for AVM individually by evaluating the stage of symptoms and lesion extent and in consideration of the risk of complications.

In most reports, treatment for the overgrowth of the lower limbs were aimed to prevent physical disorders caused by leg-length inequality. Some reports particularly mentioned treatment for foot lesions.

If the leg-length inequality is ≤ 2 cm, the management of leg-length difference and accompanying scoliosis is considered possible by the use of shoe lift [277‐281]. If the leg-length inequality is ≥ 2 cm, significant gait disturbance, postural abnormalities, and compensatory changes in the contralateral limb are likely to develop, and before consequent unphysiological gait leads to irreversible impairment, surgical treatment to correct the leg-length inequality should be considered [277‐281]. Long-leg radiography is useful for determining the best time for surgery [282], and the measurement of the leg-length by long-leg radiography or CT is reported to be effective [278]. Surgical treatment reported in the papers are as follows.

Jacob et al. [277] performed epiphysiodesis in 41 patients with a leg-length inequality of ≥ 2 cm among 252 patients with Klippel–Trenaunay syndrome and reported improvement in more than 90% of the patients. The effectiveness of this surgery is also confirmed by other review articles [277‐281]. The effectiveness of shortening of the femur and tibia was reported in the review by Capraro et al. [278] The fixation period is considered to be shortened as a whole by simultaneously performing femoral or tibial shortening in addition to epiphysiodesis. Redondo et al. [280] recommended endoscopic growth control of the epiphyseal plate in the distal end of the femur for patients with a leg-length inequality of ≥ 2 cm. Capraro et al. [278] did not recommend growth control with the epiphyseal stapling because of the unpredictability of the results and high frequency of complications. The appropriate time for surgical intervention on affected limbs is reported to be around the age of 11 years [280].

Tanaka et al. [283] performed bone elongation of the intact limbs using an external fixator in adult patients with mild structural scoliosis and reported that the procedure was effective for correcting the leg-length inequality and scoliosis. Jacob et al. [277] also recommended bone elongation of the intact limb using Ilizarov external fixation apparatus in their review.

Servelle [284] hypothesized that elongation of the affected limbs was due to a high venous pressure and performed ligation of the popliteal vein of the intact limb in 48 children, and they reported significant improvement in leg-length inequality. However, there are also negative views, saying its effectiveness is uncertain [278].

Redondo et al. [280] recommended resection of the toes (ray resection) and debulking for wearing shoes and cosmetic improvement. Gates et al. [285] notably reported that compared with ray resection, wound healing of the stumps was poor after major resection.

Asymmetry due to hypertrophy of the upper limb less frequently causes impairment of ADL than that of the lower limb. In one article, resection in patients with functional impairment due to marked finger deformities is reported [282], but articles reporting treatment for upper limb overgrowth are very few. While debulking has been reported to be advantageous from the cosmetic viewpoint [279], it has also been reported to induce exacerbate of edema of the affected limb [284], causing complications including cicatricial contracture, recurrence of the lesion, and refractory ulcer [278], and sufficient caution is necessary.

Recommendation:

Although surgical resection is an effective treatment, it should be performed after comprehensive evaluation of cosmetic aspects, prognosis, functional prognosis, resectability, and possibility of recurrence/complications.

Generally, the rate of successful surgical resection is high, and ≥ 90% resection is reported to be possible in 60% or more of the patients [286‐288]. This also applies to the head and neck region, which is the frequent site of the lesion [286]. However, the percentage of resectable lesions decreases from the cystic to mixed and to cavernous type [286]. Since many LM are distributed diffusely in the skin and subcutaneous adipose tissue and around structures including muscles, blood vessels, and nerves, resection of the lesion involves resection of normal tissues in varying degrees. In lesions that show complicated distribution in the head and neck region, complications after surgical resection are observed relatively frequently. Serious complications including nerve paralysis, hematoma, local necrosis, sepsis, deformation, salivary fistula, hoarseness, airway obstruction, and malocclusion have been reported [286, 287, 289‐297], and facial nerve paralysis is likely to result from resection, particularly, of LM infiltrating the parotid region [289]. By the site, the incidence of complications increases as the area of involvement widens from unilateral to bilateral, from below to above the lingual bone, both sides, and both above and below the lingual bone [293, 296]. Postoperative death may occur in patients with a severe neck lesion, but the extent of the effect of surgical resection is unclear [287, 288, 298]. Postoperative recurrence is closely related to the resectability of the lesion depending on its distribution, and lesions that are difficult to resect due to a wide area of involvement and a strong tendency of infiltration have been reported to be associated with recurrence [296].

Indications for surgical resection vary among papers, and differences in the patient background must be considered in the evaluation of the effectiveness of resection. While there were many reports that surgical resection was performed in combination with sclerotherapy, and resection is considered to have been performed when more favorable results were expected from resection rather than sclerotherapy, criteria for their selection are unclear. Therefore, there is certainly the large bias of individual variation in the circumstances, and it was clearly impossible to conclude that resection is uniformly effective.

<Summary>

While the effectiveness of surgical resection for soft tissue/superficial LM was evaluated, there was no literature with a high level of evidence. One of the major reasons is the diversity of the lesion type (cystic or cavernous), area of involvement, and history of other treatments. Because of this diversity, the condition of patients is considered to show an extremely wide variation, and their generalization is impossible. However, if conditions, such as the type of the lesion (cystic or cavernous), site of origin, and relationship with other treatments are restricted, tendencies were observed in functional prognosis, recurrence rate, and contents and incidence of complications.

While the resection rate of lesions by surgical treatment was suggested to be generally high, selection criteria for resection were unclear. Therefore, it is speculated that resection was performed for patients clinically judged to be treated more effectively by surgery. However, since there were some serious complications of surgical resection that persist as sequelae, their possibility should be evaluated carefully in performing surgical resection. The risk of resection has been suggested to vary with conditions of the lesion. The functional outcome is poor, and the recurrence rate and incidence of complications after resection are high, in those that occupy a wide area and those that are accompanied by symptoms such as airway obstruction.

From these observations, we propose at present, “While surgical resection is often effective, it must be selected in consideration of cosmetic aspects, prognosis, functional prognosis, resectability, and possibility of recurrence and complications.”, despite limited scientific grounds. If complete resection of the lesion is possible, surgical resection may be performed as the first-line treatment, but the possibility of other treatments including sclerotherapy, in particular, should be evaluated according to the diverse conditions of individual patients, and surgical resection should be performed when other treatments are ineffective or when surgical resection is considered clearly superior.

Recommendation:

It is impossible to recommend optimal timing of surgery, and judgments according to the condition of each case are necessary.

Despite a careful review of the literature by secondary screening, there was no paper that analyzed cases from the viewpoint of optimal timing of surgery. There was information concerning the age at surgery, but its appropriateness was not evaluated. Papers that mentioned the timing of surgery are shown below.

Concerning the timing (age) of surgery, unless the size of the lesion is small or there are symptoms that require urgent treatment, such as respiratory disturbance, it is recommended to delay surgery until the age of 3 years by expecting spontaneous regression or for the ease of identification of surrounding structures during surgery, ease of control of bleeding, and less problems with postoperative management [294]. There was also a paper [295] that suggested the necessity of the determination of the time of surgery in consideration of problems that change with age including the priority of securing the airway and appropriate nutritional management in neonates with head and neck and giant lesions, control of hemorrhage and infection and measures to prevent dysarthria and dental problems in infants, and skeletal and cosmetic problems in school-age children, although it did not mention the optimal timing of surgery.

However, there was no paper that positively recommended resection without considering the time after the diagnosis or grounds for such a recommendation.

<Summary>

As a result of literature search for evaluating the CQ, “What is the optimal timing of surgery for soft tissue/superficial LM”, there were papers that mentioned the timing of surgery, but none of them objectively evaluated its appropriateness. Therefore, no suggestion about the appropriate timing of surgery was obtained from the literature available at present, but there were a few papers suggesting that the decision to perform surgery should be made carefully.

Similar to the previous CQ, soft tissue/superficial LM of which the background vary in individual cases, and it is difficult to uniformly evaluate the effectiveness of resection. In clinical practice, in addition to medical reasons, social reasons including school attendance are considered to largely influence the decision of the time of resection. The results of RCTs are necessary to obtain objective data, but it is practically very difficult to arrange an RCT fulfilling the above conditions.

While this CQ is a very important issue for patients and families as well as clinicians, there has not been objective evaluation of the optimal timing of surgery in the past. Presently, as quick decisions to perform surgery should be avoided, this guideline proposes, “The optimal timing of surgery cannot be decided in general, and judgments according to the condition of each case are necessary.”

Recommendation:

A wide range of drugs are used for sclerotherapy. Although comparison among drugs has not been made, and consensus regarding the methods or frequency of their administration has not been reached, improvements are observed after sclerotherapy in various symptomatic, functional, and cosmetic (esthetic) aspects. However, complications including functional impairment have also been reported.

Regression of lesions of intra-abdominal LM by sclerotherapy was mentioned in 5 papers [288, 316‐319]. According to the report by Chaudry et al. [316], the reduction rate was ≥ 90% in 7 and ≥ 20% in 1 of the 10 patients with LM of the mesentery and retroperitoneum treated with doxycycline, and evaluation using imaging examination was not performed in 2 cases. The patient who showed a low regression rate had a mixed type of cystic and cavernous lymphangiomas, and the other patients had cystic lesions. Oliveira et al. [317] reported that the lesion regressed by 70% in 1 of the 2 patients with cystic lymphangiomas treated with OK-432. Won et al. [318] reported 1 patient who showed complete disappearance of cystic retroperitoneal lesions after sclerotherapy using acetic acid. Shiels et al. [319] reported that cystic lesions responded to sclerotherapy using STS and ethanol in 2 patients, but there was no mention about the reduction rate. However, according to Alqahtani et al. [288], no effect was observed in 10 patients who underwent sclerotherapy using steroid/tetracycline or 50% glucose solution.

There were 3 papers that referred to symptoms of patients treated by sclerotherapy for intra-abdominal LM [316, 317].

According to Chaudry et al. [316], of the 10 patients who underwent sclerotherapy, 3 had chronic abdominal pain, 3 had acute abdominal pain, 1 had fever/chill, 1 had anemia, and 2 had palpable masses, but the symptoms were alleviated by treatment in all patients, and no recurrence was noted.

Oliveira et al. [317] reported that sclerotherapy was performed in a patient with a palpable mass and in 1 with a palpable mass, abdominal compartment syndrome, and a poor general condition. While the condition was alleviated in the patient who only showed a palpable mass after 2 courses of OK-432 sclerotherapy, but the treatment was changed to surgery in the patient who had abdominal compartment syndrome because of enlargement of the mass due to intracystic hemorrhage.

Three papers specifically mentioned complications of sclerotherapy for intra-abdominal LM. There was no report of deaths due to treatment-related complications. Oliveira et al. [317] treated 3 patients by sclerotherapy using OK-432 and reported that one of them developed subbowel obstruction after the treatment and another required emergency surgery due to exacerbation of abdominal compartment syndrome induced by intracystic hemorrhage. Chaudry et al. [316] reported that doxycycline used for sclerotherapy leaked into the retroperitoneal space in 1 of the 10 patients but that the lesion regressed without any particular problem. Won et al. [318] performed sclerotherapy using acetic acid in 1 patient with retroperitoneal cystic lymphangioma. Although pain and hematuria were observed, they concluded that the relationship of hematuria with the therapy was unclear, because it was observed during menstruation.

Sclerotherapy was often performed before, after, or during surgical resection, and papers that reported the results of sclerotherapy alone were few. There was no paper that directly compared observation without treatment, sclerotherapy, and surgical resection. Few papers analyzed intra-abdominal lesions alone, and many papers included lesions in other areas or evaluated lesions in different intra-abdominal regions including the mesentery, retroperitoneum, and viscera collectively.

Moreover, differences in properties of LM, such as cystic, cavernous, and mixed types, their definitions, criteria for the selection of sclerotherapy (combination with surgery, types of sclerosing agents and methods of their use, number of administrations) varied among papers, and few papers evaluated these matters separately.

Such differences in the patient background and contents of treatment must be considered in evaluating the effectiveness of sclerotherapy. In evaluating this CQ, particularly, differences in morphology of LM and sclerosing agents were excluded.

<Summary>

The CQ, “Is sclerotherapy effective for intra-abdominal LM?” was evaluated from the viewpoints of therapeutic effect, symptoms/functions, and complications, but no paper with a high level of evidence was found. While sufficient regression of the lesion and alleviation of symptoms were achieved by sclerotherapy in some patients, the response rate varied among reports, and information was insufficient for general discussion of sclerotherapy. Concerning treatment-related complications, there have been reports of bowel obstruction associated with sclerotherapy, and attention to this condition as well as intracystic hemorrhage is considered necessary. However, there was no report of chylorrhea, which was reportedly caused by surgery.

Based on the above observations, it is presently difficult to determine indications for sclerotherapy in intra-abdominal LM by setting up criteria, but as there was no literature that strongly ruled out intra-abdominal LM as indications of sclerotherapy, these guidelines propose, “Although there are many reports that sclerotherapy is useful, there is the risk of complications, and careful judgments about matters including the resectability of the lesion and selection of the sclerosing agent are necessary.” For the future evaluation of this CQ, validation by a design with a high level of evidence, such as RCT, is considered necessary.

Recommendation:

Since there is risk of treatment-related complications, it is proposed to consider therapeutic intervention when the lesion tends to enlarge or has become symptomatic.

From the literature reviewed, symptoms of intra-abdominal LM (abdominal pain, bowel obstruction, torsion, infection, hemorrhage, vomiting/sucking difficulty, frequent urination and abdominal mass [320‐326]) are considered to be dependent on factors such as site, size and age. It is necessary to determine risk factors by stratification of these factors in the future [320, 322, 325].

Reported complications in treated cases include recurrence that required re-treatment [321], bowel obstruction [317, 323, 324], chylous ascites [324, 326], embolism [317], hemorrhage [317] and wound infection. Embolism of the inferior vena cava after surgery [317] and abdominal compartment syndrome after adhesion therapy [317] were reported as severe complications. It deserves special attention that, if surgical resection is performed for mesenteric LM, the intestine may have to be resected with the lesion [326].

While there have been reports [320, 322] that intra-abdominal LM with few clinical symptoms regressed during follow-up, they may become symptomatic later (as observed in many case reports). For that reason, the opinion that intervention should not been chosen during the follow-up until the lesion enlarges or new symptoms appear was frequently described.

It should be noted that many asymptomatic cases may be left unreported and some asymptomatic lesions that had been detected were treated. There is no study with a high level of evidence indicating explicit criteria concerning the age, site or situation about whether intervention should be made for asymptomatic intra-abdominal LM.

<Summary>

The necessity of treatment of a patient with intra-abdominal LM with few symptoms should be determined after evaluating the balance between the risk of treatment and non-treatment considering its site and size as well as patient age. However, since research on indications for treatments has been insufficient so far and serious complications after treatment have been reported, deliberate evaluation for each patient is mandatory. When observation is selected, periodic imaging studies are recommended to optimize therapeutic intervention by detecting enlargement of the lesion. Also if any symptom has developed during follow-up, intervention should be considered. For these reasons, the recommendation, “Since there is risk of treatment-related complications, it is proposed to consider therapeutic intervention when the lesion tends to enlarge or has become symptomatic.” was adopted.

Recommendation:

Conservative treatments, such as fasting, high-calorie infusion, and medium chain triglyceride (MCT), should be performed first, but, if they are ineffective, drug treatment, sclerotherapy, and surgery may also be considered.

The results of review are presented below according to the treatment.

Whether the amount of ascites changes by fasting should be checked first.

High-calorie infusion is often used with fasting, and since there was no report that ascites increased under the effect of high-calorie infusion according to our review, it is recommended for nutritional support during fasting. In the multicenter case series reported by Bellini et al. [327], high-calorie infusion/total parenteral nutrition was performed in 15 patients without adverse effects.

Medium chain triglyceride was used before, after, and during treatment [327, 328, 330‐335, 337, 339‐341, 343, 345, 346, 348‐352]. In the multicenter case series by Bellini et al. [327], MCT was reportedly performed in 14 patients without adverse effects.

In drug therapy for chylous ascites, primarily octreotide (a long-acting somatostatin analogue) was used, and no report that discussed the effectiveness of other drug therapies was found by the present literature search.

In the multicenter case series by Bellini et al. [327], octreotide was administered to 6 of the 16 patients with chylous ascites for 8–38 days, and a decrease in chylous ascites was reported in all of them. In the single-center case series by Huang et al. [344], 2 of the 4 patients with chylous ascites treated by high-calorie infusion and octreotide administration were reported to have shown a decrease in ascites within 10 days. However, there has been a report [330] that no effect was observed despite the administration of octreotide for 3 weeks. Concerning the dose of octreotide, it was administered at 1 μg/kg/h [327], at 3 μg/kg/h [332], began to be administered at 0.5 μg/kg/h and increased to 10 μg/kg/h by 1 μg/kg/h [329], administered by continuous intravenous infusion at 0.5–2.0 μg/kg/h [333], and began to be administered by subcutaneous injection at 2.5 μg/kg 2 times/day and increased every 2 days to 8 μg/kg 2 times/day [330]. Regarding the time of the beginning of administration, the administration was started as no improvement was observed in chylous ascites after conservative treatments for 2 weeks [330, 334], and as chylous ascites was alleviated by conservative treatments but was exacerbated again [333]. No adverse effects of octreotide administration were noted in the present review of the literature. Thus, no control study that evaluated the effect of octreotide on chylous ascites was found by the present literature search, and the level of evidence concerning the efficacy is low, but as there are case series and many case reports that chylous ascites was reduced by octreotide administration, it appears reasonable to consider drug treatment using octreotide for chylous ascites that does not respond to conservative treatments.

Sclerotherapy was performed in 6 patients in 5 case reports [339, 347, 349, 351, 352]. The sclerosing agent was OK-432 in 5 of the 6 patients and was Beta-Isodona-solution^® (MUNDIPHARMA GmbH, Cambridge, UK) in 1 [349]. OK-432 was locally injected into the lesion in 4 [347, 351, 352], administered intraperitoneally in 1 [352], and administered via the drain in 2 [347, 352]. Concerning sclerotherapy, the number of reported cases that were reviewed was limited, and further accumulation of cases is considered necessary to establish its usefulness.

Abdominal drainage and abdominal puncture are performed when organ compression symptoms (compartment syndrome and respiratory insufficiency) due to abdominal distention are present or possible or when the drain is inserted postoperatively. However, drainage itself cannot improve chylous ascites, and treatments, such as infusion, blood preparations, and blood transfusion, are necessary to supplement the ascites lost due to drainage [327, 330‐333, 337‐340, 343, 345‐347, 349, 351, 352].

Surgical treatment is reported to be frequently performed after conservative or drug treatments. According to the single-center case series by Zeidan et al. [343], surgical treatment was performed in patients who responded poorly to conservative treatments for over a mean of 25.3 days. In other reports, surgical treatment was performed after conservative treatments for 1–3 months [328, 329] and in patients with congenital chylous ascites 1–4 months after birth [330, 334, 350]. Since it is often impossible to identify the leakage site of chylous ascites [330], attempts to identify the leakage site by orally administering a lipophilic dye (Sudan black, Sudan III) before operation. [328, 329, 336, 343] When the leakage site can be identified, ligation, suturing, clipping, and cauterization have been performed [328, 334, 336, 343, 350]. In addition to reports of the usefulness of techniques to stop leakage, such as applying or sprinkling fibrin glue at the leakage site of chylous ascites or over the surrounding retroperitoneum [329, 331, 343, 350] and applying a patch of oxidized cellulose/resorbable local hemostatic agent [331, 343], there have also been reports of peritoneovenous shunting [349, 353] and peritoneoamniotic shunting for fetal cases [338].

There was no large clinical study in the past literature. Therefore, although the level of evidence is low, we consider that surgical treatment is recommendable for chylous ascites that does not respond to conservative or drug treatments, because it has been performed in case series and case reports for chylous ascites that did not respond to conservative or drug treatments for over about 1 month. Although techniques to enhance the response rate of surgical treatment, such as identifying the leakage site by using a lipophilic dye and applying fibrin glue or a patch of oxidized cellulose/resorbable local hemostatic agent, have been attempted, there are only case series and case reports, and none of the papers retrieved by the present literature search evaluated their usefulness.

There was no literature that defined refractory chylous ascites based on the duration of illness or treatment responses. Therefore, we extracted and summarized factors that were considered to contribute to clinical refractoriness, such as the duration of illness and treatment responses, in each paper related to the treatment for chylous ascites. Also, as the cause of chylous ascites varies widely, the therapeutic effect is expected to differ depending on the cause, but no paper that was reviewed evaluated treatments according to the cause. Therefore, in the present evaluation, the statements are limited to treatments and their effects regardless of the cause.

<Summary>

It was difficult to comprehensively discuss treatments, because its cause varied widely, and treatments for various causes were performed. Therefore, treatments were classified into conservative treatments (fasting, high-calorie infusion, MCT), drug treatments (octreotide), sclerotherapy, abdominal drainage, abdominal puncture, and surgical treatment, and the effects of each treatment were evaluated.

Treatments effective for refractory chylous ascites can be summarized as follows with the understanding that they may depend on the cause and that the level of evidence of the available reports concerning treatments and their effects is low. Conservative treatments, such as fasting, high-calorie infusion, and MCT, should be performed first because of the rareness of adverse effects. In patients who respond insufficiently to conservative treatments, drug treatments using octreotide can be considered as there have been case series and many case reports. Concerning sclerotherapy, the number of reported cases is small, and further large clinical studies will be needed to confirm its usefulness. Abdominal paracentesis and surgical treatments may be considered for chylous ascites that does not response to conservative or drug treatments for about 1 month.

Thus, the draft recommendation is “Conservative treatments, such as fasting, high-calorie infusion, and MCT, should be performed first, and, if they are ineffective, drug treatments, sclerotherapy, and surgical treatments may be considered”. However, evaluation of this CQ by a design with a higher level of evidence, such as RCT, is considered necessary for the future.

Recommendation:

Complications associated with sclerotherapy for intra-abdominal LM include bowel obstruction, hemorrhage, pain, hematuria and chylous ascites. Surgical treatment of the disease can be associated with serious complications such as occlusion of the inferior vena cava and massive resection of the intestine as well as more common, wound infection, bowel obstruction, hemorrhage and chylous ascites.

Sclerotherapy using OK-432 was reported to be associated with bowel obstruction and hemorrhage for mesenteric LM [317], and chylous ascites for retroperitoneal LM [326]. Sclerotherapy using acetic acid was reported to be associated with pain and hematuria in patients with retroperitoneal LM [318].

Complete resection of both mesenteric and retroperitoneal LM by laparotomy was reported to be associated with wound infection [324, 354] and bowel obstruction [323, 354, 355] as common complications. There were reports of serious complications such as occlusion of the inferior vena cava [317] and massive resection of the intestine was necessary due to diffuse infiltration of the LM tissue to the intestinal wall [356].

In a report [357] about complications associated with complete laparoscopic resection of intra-abdominal LM by Tran et al., resection was attempted in 47 patients, and conversion to laparotomy was necessary in 3 (6.4%) due to tight adhesion in 2 and intraoperative hemorrhage in 1.

Partial resection by laparotomy was reported to be associated with persistent ascites over a long period that was refractory to the treatment [355].

Patients with intra-abdominal LM are treated with various modalities including sclerotherapy and surgical procedures. Modalities were combined in many cases, and complications are often reported as those of entire treatment without more detail information about those associated with individual treatment.

<Summary>

For answering the CQ, “What kinds of complications are associated with treatments for intra-abdominal LM?”, no literature with a high level of evidence was found, but foreseeable complications were listed from many case reports. Bowel obstruction, hemorrhage, pain, hematuria, and chylous ascites were reported as complications of sclerotherapy. Serious conditions, such as occlusion of the inferior vena cava and massive resection of the intestine, as well as common complications, such as wound infection, bowel obstruction, hemorrhage and chylous ascites were reported as complications after surgical procedures.

Although the incidences and differences in complications in respect of the site and histological type are not shown in the literature, each patient with intra-abdominal LM should be treated with sufficient evaluation of the site, size and symptoms. In addition, treatment must be implemented with sufficient understanding of the possible complications.

Thus, we propose “Complications associated with sclerotherapy for intra-abdominal LM include bowel obstruction, hemorrhage, pain, hematuria, and chylous ascites. Surgical treatment of the disease can be associated with serious complications such as occlusion of the inferior vena cava and massive resection of the intestine as well as more common, wound infection, bowel obstruction, hemorrhage and chylous ascites.” as a recommendation draft.

Recommendation:

Sclerotherapy is effective for macrocystic lesions, and surgical resection is effective for microcystic lesions. However, as the complication rate is relatively high, treatments should be selected according to the condition of each case.

Boardman et al. [358] reported that, of the 97 patients with LM of the head and neck region, surgical treatments were necessary in 6 of the 12 patients with mediastinal lesions, that complications of surgery occurred in 4 of the 6 patients, and that long-term nerve damage was observed in 3 of them. In addition, they reported that management by tracheotomy was necessary in 15% of all patients. Complete or nearly complete remission was observed in 92% of the patients, but they suggested that surgical treatments should be indicated only when there is airway obstruction or there is the risk of it, because surgical treatment of mediastinal lesions frequently causes complications.

Park et al. [359] reported that they surgically resected mediastinal LM in 12 patients. Seven of them had dyspnea, and 3 were asymptomatic, but they were all judged to have indications for surgery due to symptoms or the tendency of the lesions to enlarge. A total of 5 recurrences were observed in 4 patients (33%) during a mean period of 3.6 years after the initial surgery, and all underwent by re-resection. No perioperative death was observed, and, in a total of 25 cases including past cases, the overall survival was not different compared with that in healthy individuals over a follow-up period of 11.5 years.

Smith et al. [302] performed local injection of OK-432 in 16 patients with mediastinal LM and reported ≥ 60% regression of the lesion in 13 (81%). They also mentioned treatment responses according to the histological types and, by reporting responses (complete or nearly complete remission) in 94% of those with macrocystic lesions, 63% of those with mixed lesions, but 0% in those with microcystic lesions, suggested a macrocystic lesion to be a good indication for sclerotherapy using OK-432. Although not from the viewpoint of airway obstruction, they reported that treatment using OK-432 was more effective than surgical resection and less frequently caused serious complications.

There have been no papers that directly analyzed treatments effective for mediastinal lesions expected to cause airway obstruction, and many papers reported cases of mediastinal lesions that responded to treatments. Therefore, we simply extracted matters relevant to this CQ from these reports.

<Summary>

There was no literature with a high level of evidence concerning effective treatments for LM in the mediastinum causing airway obstruction. A few case reports that referred to the effects of surgery and sclerotherapy were observed, but it was difficult to present objective and specific figures concerning their effectiveness or safety. However, according to the available information, it should be noted that favorable responses have been obtained by OK-432 local injection in macrocystic lesions and that complications due to surgical resection are likely to occur relatively frequently.

From these observations, we consider the following to be a therapeutic approach that can be proposed: “Sclerotherapy, such as that by local injection of OK-432, should be considered for macrocystic lesions, and, for lesions that are technically difficult to treat by sclerotherapy or microcystic lesions, surgical resection should be considered with attention to complications. In addition, it is necessary to pay attention to the appearance of respiratory disturbances before and after these treatments and to constantly evaluate indications for securing the airway by intratracheal intubation or tracheostomy.” Therefore, at present, we recommend, “Sclerotherapy is effective for macrocystic lesions, and surgical resection is effective for microcystic lesions. However, as the complication rate is relatively high, treatments should be selected according to the condition of each case.”

Recommendation:

In a patient with LM around the airway, there is risk of respiratory distress in infancy, while airway obstruction is likely to be exacerbated by sclerotherapy. Particularly, when risk of airway obstruction is judged to be high or when the patient has already presented symptoms, it is proposed to perform sclerotherapy with sufficient preparations including airway management.

According to the SR by Adams et al. [360], the mortality was 4.7% in 277 cases with head and neck LM. Since lesions around the airway were not the only target, and since sclerotherapy was not the only treatment modality, the paper has not quite rightly answered to the CQ. However, since patients who died were all before one year of age and their causes of death are considered to have been mostly airway problems, such as airway obstruction and aspiration due to vocal cord paralysis in 8, and as at least 1 patient is judged to have died due to complications of invasive treatment, the paper is considered to indicate the risk of this disorder during infancy.

Many of the papers that referred to the size regression evaluated it by four categories; (1) excellent or complete (≥ 90% regression), (2) good or substantial (≥ 50% and < 90% regression), (3) fair or intermediate (≥ 20% and < 50% regression), (4) and poor or none (≤ 20% regression).

Ravindranathan et al. [306] treated 5 patients (aged 4–19 months) with cervicofacial LM by sclerotherapy using OK-432 (in addition to Fibrovein^® (STD Pharmaceutical Products Ltd, Hereford, UK) in 2) and reported that the responses were good in 1 (20%) (cystic), partial in 1 (20%) (cavernous), and poor in 3 (60%) (2 with cavernous lesions that required tracheotomy and 1 with cystic lesions in whom the condition improved to good after surgical resection). However, they did not mention the evaluation criteria for good, partial, and poor.

According to the report of 8 cases with head and neck LM by Leung et al. [361], all patients underwent sclerotherapy, ≥ 50% regression was observed in all patients with complete regression in 2. However, the patient age varied from 2 months to 11 years, and the types of LM were not mentioned.

Ogawa et al. [362] reported 9 patients (including 5 preschoolers and toddlers, 2 school children, and 2 adults) who underwent OK-432 sclerotherapy for the neck LM and evaluated it to be markedly effective in 8, in whom the lesions mostly disappeared, and effective in 1, who showed a ≥ 50% regression. Eight patients in whom the treatment was markedly effective consisted of 1 with mixed and 7 with cystic lesions, and the one in whom the treatment was effective had a mixed type.

Cahill et al. [315] reported doxycycline sclerotherapy in 17 patients with head and neck LM (cystic in 10, mixed in 7 (3 required tracheotomy)), and its size regression was reported to be > 90% in 7 (41.2%) (cystic in 6, mixed in 1), 75–89% in 4 (23.5%) (cystic in 2, mixed in 2), 51–74% in 4 (23.5%) (cystic in 1, mixed in 3), and 25–50% in 2 (11.8%) (mixed in 2).

Nehra et al. [311] reported doxycycline sclerotherapy in 11 patients with head and neck LM (cystic in 7, mixed in 4; aged 2 days–21 months) (later combined with surgical resection in 3). The treatment results were excellent in 5 (45.5% of all patients) and satisfactory in 2 (18.2% of all patients) among 7 patients with cystic lesions but poor in all 4 patients with mixed type lesions (36.4% of all patients). Particularly, 3 of the 4 patients with mixed type lesions required tracheal intubation shortly after birth and underwent sclerotherapy while intubated, but the effects were poor in all of them. Surgical resection was added in 1 and was under consideration in another.

According to Ravindranathan et al. [306] who reported 5 patients with cervicofacial LM (aged 4–19 months) treated with sclerotherapy using OK-432 (in addition to Fibrovein^® (STD Pharmaceutical Products Ltd, Hereford, UK) in 2), 4 exhibited symptoms of airway obstruction before treatment. Symptoms included dysphagia in 2 and dyspnea (including croup-like symptoms) in 4 (some both). Symptoms were alleviated by sclerotherapy in 2 out of 4 (cystic 1, cavernous 1), but tracheotomy was necessary in the remaining 2 (cavernous in both) without improvement.

In the report of 8 patients with head and neck LM and 5 patients with VM (aged 2 months–11 years) by Leung et al. [361], their symptoms noted before treatment were mass or swelling [10 patients (77%)], pain after hemorrhage [2 patients (15%)], skin discoloration (blue) [1 patient (8%)], obstructive airway symptoms [6 patients (46%)], and swallowing difficulty [1 patient, (8%)]. All symptoms were alleviated by sclerotherapy (doxycycline for LM, STS foam for VM).

Arimoto et al. [363] reported a patient with cystic LM in the neck presented 3 months after birth. The patient presented with respiratory distress at the age of 10 months due to enlargement of the LM following upper respiratory infection. While left vocal cord fixation due to the mass was confirmed by ultrasonography before treatment, aspiration of the cyst and steroid administration resulted in opening of the glottic area and regression of the mass with relief of wheezing and distress. Since they underwent sclerotherapy 2 months after the disappearance of symptoms, aspiration of internal fluid and steroid administration rather than sclerotherapy were directly effective for the alleviation of symptoms.

Kitagawa et al. [364] reported a patient with giant LM of the neck which had been prenatally diagnosed and was treated under ex utero intrapartum treatment (EXIT) by tracheal intubation after aspiration of the cyst. The lesion was reported to be refractive to subsequent sclerotherapy and tracheotomy was eventually needed.

Nehra et al. [311] reported that, among 11 patients with head and neck LM (cystic type in 7 and mixed type of cystic + cavernous in 4; aged 2 days–21 months), 3 out of 4 with mixed LM presented respiratory distress soon after birth and were managed by intubation, but that all were extubated after sclerotherapy using doxycycline (1–3 times, median: 1.6 times).

No paper has reported cosmetic results in detail. Only sporadically they mentioned about surgery for redundant skin after regression of cystic lesions by sclerotherapy.

4. Complications

Complications associated with treatment for LM around the airway have been reported in many papers. They include temporary conditions caused by sclerotherapy such as fever [292, 303, 312, 362, 365‐371], local swelling [303, 312, 365, 367, 368, 370, 371] pain [288, 312, 362, 367, 370‐372], hemorrhage into the cyst [288, 303, 312, 368], and infection [288, 292, 303, 310, 312, 360, 365, 372]. There also reported complications as the effects of treatment for head and neck lesions such as respiratory distress due to airway obstruction [292, 303, 306, 312, 362, 365, 366], as well as nerve palsy [288, 292, 312, 360, 365].

According to a systematic review about head and neck LM by Adams et al. [360], both nerve damage due to sclerotherapy and post-therapeutic infection were reported in 1 (0.8%) out of 123 patients. Since nerve damage and infection after surgery were observed in 12 (10.2%) and 7 (5.9%) out of 118 patients, respectively, the complication rate would be lower by sclerotherapy than by surgery.

Ogawa et al. [362] reported a 1-year- and-5-month-old patient who developed airway edema after OK-432 sclerotherapy for the cystic neck LM and necessitated tracheal intubation for 3 days and cautioned against sclerotherapy for LM around the airway in young children (particularly, those less than 2 years old).

Kudo et al. [369] also reported 2 patients aged 11 months and 1 year and 11 months who were treated with OK-432 sclerotherapy intubated in advance for fear of airway obstruction due to post-therapeutic swelling. Tomemori et al. [373] also cautioned against sclerotherapy for LM in children aged less than 2 years as in the report by Ogawa et al. [362]

On the other hand, Kudo et al. [369] reported 2 cases whose neck LM having been enlarged rapidly after suffering from measles or upper respiratory tract infection (URTI). Arimoto et al. [363] also reported a patient with cystic LM of the neck 3 months after birth, who developed dyspnea due to enlargement of the lesion after URTI at 10 months and was about to be intubated.

Regarding complications due to sclerosing agents, Cahill et al. [315] reported those by doxycycline, STS and absolute ethanol. They reported delayed complications, such as Horner’s syndrome, transient left lip weakness, right facial nerve palsy, and transient left hemidiaphragm paralysis, in addition to peri-procedural complications such as hemolytic anemia after doxycycline injection in 2 patients, hypoglycemic and metabolic acidosis in 3 neonates, transient hypotension during absolute alcohol instillation and self-limiting skin excoriation secondary to peri-catheter leakage of doxycycline. Other reported complications include permanent vocal cord paralysis after local ethanol injection [374], serious complications after OK-432 injection such as death due to pulmonary embolism [375], deaths due to pulmonary complications after treatment using bleomycin [376, 377] and leukocytopenia due to bleomycin [368].

There are few papers that solely analyzed LM around the cervical airway. Most papers included lesions involving not only the neck but also the craniofacial and other parts of the body and reported LM with different properties such as cystic and mixed types. In addition, definition of cavernous lesions and methods of sclerotherapy (injection techniques, number of injections) were not similar among papers, and differences in these backgrounds must be taken into consideration to evaluate the effectiveness of sclerotherapy.

<Summary>

The CQ, “Should sclerotherapy be performed in infancy for a patient with head and neck LM affecting the airway?”, was evaluated from the viewpoints of responses (prognosis (survival rate or mortality), decrease in size, symptoms, cosmetic improvements) and complications. Since there have been some reports on the risk of respiratory distress due to LM around the airway in infants, and therapeutic intervention is necessary even in infants when the risk is high or they have already developed symptoms. Such intervention is made by sclerotherapy or surgery, and as surgical resection is associated with the risk of more serious complications than sclerotherapy, intervention by less invasive sclerotherapy is recommended. Sclerotherapy is considered to be very effective because of high regression rate of the lesion and symptom/function-improving effect. However, its effect varies depending on the disease type, somewhat less effective in the cavernous and mixed types than in the cystic type. Furthermore, when it was performed for the lesions around the airway, it may be associated with the risk of exacerbation of airway obstruction symptoms due to reactive enlargement of the lesion. Thus, we formulated the recommendation, “In LM around the airway, there is the risk of respiratory disturbances from infancy, but airway obstruction is likely to be exacerbated by sclerotherapy. Particularly, when the risk of airway obstruction is judged to be high or when symptoms have appeared, it is proposed to perform sclerotherapy with sufficient preparations including securing the airway.”

Recommendation:

Surgical resection is effective for reducing the size of the lesion and alleviating symptoms and functional impairment. However, total resection is often difficult, and careful decision is required in consideration of the possibility of complications and recurrence.

Twenty-four cases of tongue lesions treated by surgical resection alone were reported in 4 papers. Catalfamo et al. [378] performed surgical resection of localized masses including normal structures with a margin of 1 cm in the horizontal direction and reported that the size of tongue lesions was reduced in 8 (88.9%) of the 9 patients.

Concerning large lesions impossible to resect totally, Boardman et al. [358] reported 13 cases of partial surgical resection, but several operations were often necessary to reduce lesion size. A total of 2 case have been reported [379, 380], and the lesion size was reduced in both. Although differences were observed in re-enlargement after surgery, they are discussed in detail in “(2) Complications”.

In 1 case report, sclerotherapy was performed 15 times, but the lesion size could not be reduced, and surgical resection was performed, eventually resulting in a favorable outcome without recurrence [381].

According to a report of 89 cases of head and neck LM by Lei et al. [291], the outcome was excellent in 73 (82%) and good in 16 (18%) although it was not a report of cases of tongue lesions alone. They included 43 cases of tongue lesions.

In addition, a few papers [382‐385] that suggested the effectiveness of combinations of surgical resection with sclerotherapy and laser therapy were observed. Wiegand et al. [383] classified the disease into 4 stages according to the area of involvement and reported that the stage can be a prognostic factor. Surgery was effective, and complications were rare, when the lesion was localized in the superficial layer and part of the muscle layer. Surgical resection can also be effective, but complete resection is difficult, when the lesion extends over the entire muscle layer or to the tongue base and neck. Therefore, partial resection is often repeated and combined with laser therapy and sclerotherapy, but the recurrence is observed very frequently, and the results did not contradict the reports mentioned below in the section of the recurrence rate [291, 358].

A wide variety of symptoms have been reported depending on the site of the mass, and they include tongue discomfort, bleeding, pain, and difficulty in oral feeding [386]. Roy et al. [387] reported that bleeding from the tongue surface, pain, and eating difficulty were alleviated by cauterization.

In most patients who exhibited functional impairment, the lesions were so extended that they were no longer indications for one-time surgical resection. Large masses located at sites such as the tongue base cause respiratory disturbances, swallowing disorders, and difficulty in speech. According to the report by Azizkhan et al. [385], oral intake of normally cooked food became possible in 14, and normal vocalization became possible in 8, of the 21 patients with tongue base lesions. In addition, 5 of the 17 patients who needed tracheotomy were weaned.

Objective evaluation of cosmetic effects is also difficult.

Azizkhan et al. [385] reported that, of the 20 patients, excluding 1 with severe deformity who died, deformity of structures around the tongue, such as the mandible and maxilla, was mild in 6, moderate in 5, and severe in 9. There have been a few reports that cosmetic improvements were also observed in patients who showed a reduction of the tongue size by surgical resection, but objective evaluation is insufficient.

Although the properties of the lesions are unclear in some papers, facial nerve paralysis, vagus nerve paralysis, infection, hematoma, seroma, salivary leakage, ruptured suture, and skin flap necrosis have been reported as complications of the facial region. There have also been reports of temporary complications such as pain and hemorrhage.

There have been a few postoperative evaluations reporting that no reactivation that clinically required treatment was observed. Lei et al. [291] reported greater details: Recurrence was observed in 21 (23.6%) of 89 patients and was more frequent in those aged less than 1 year, those with lesions in the oral cavity/face, those with lesions at 3 or more sites, and those with microcystic lesions. According to Boardman et al. [358], LM of the tongue recurred in 12 (48%) of 28 patients, more often than other head and neck lesions. As factors related to this more frequent recurrence of lingual LM, more frequent involvement of other regions, such as the floor of mouth, and a high percentage of microcystic lesions (70%) have been suggested. Of the 2 patients treated by surgical resection alone, 1 who underwent resection of the middle part of the tongue showed no re-enlargement for 1 year or longer after surgery [379], but surgery was repeated 3 times in the 1 who underwent marginal resection [380]. This patient who underwent repeated resections also showed no re-enlargement although the time of the last resection is unclear.

In some papers, surgical resection was combined with other treatments [381‐385, 387], lesions in other areas such as the neck were included [291], and the lesion types were unknown. The lack of standardization of subjects and uniformity of the definition or time of recurrence must be considered in the evaluation of the effectiveness of surgical resection.

<Summary>

Many papers suggest that surgical resection is effective for reducing the size of lingual LM. However, in patients with large lesions, lesions extending to structures other than the tongue, and microcystic lesions, several resections or combination of resection with other treatments such as sclerotherapy and laser therapy were necessary, and the recurrence rate tended to be higher. While a few papers referred to symptoms, functional outcome, and cosmetic improvements, none showed a high level of evidence, and the evidence was insufficient for general discussion of the effectiveness of surgical resection.

Therefore, concerning the effectiveness of surgical resection for LM of the tongue, “Surgical resection is effective for reducing the size of the lesion and alleviating symptoms and functional impairment. However, total resection is often difficult depending on the distribution of the lesion, and careful decision is required in consideration of the possibility of complications and recurrence.” was proposed as a draft recommendation.

Recommendation:

For chylous pleural effusion refractory to conservative treatments, surgical procedures, such as pleurodesis, ligation of the thoracic duct, and pleuroperitoneal shunting, may be effective.

Surgical treatment for neonatal chylothorax is performed in patients who respond insufficiently even to thoracic drainage in addition to nutritional therapy using MCT milk or total parenteral nutrition or drug treatment such as octreotide administration.

The methods for surgical intervention found by the present literature review included ligation of the thoracic duct and pleuroperitoneal shunting as well as pleurodesis with OK-432 administration, intrathoracic infusion of fibrin, and povidone-iodine administration, and some patients diagnosed in utero underwent pleuro-amniotic shunting. Cases in which mildly invasive treatments, such as thoracoscopic ligation of the thoracic duct and intrathoracic fibrin application, have been reported in addition to those who underwent thoracic duct ligation by thoracotomy.

Treatments that were performed before surgery and their periods were not uniform. In addition, since there are cases that developed chylous pleural effusion after surgery and those of congenital chylothorax, the diversity of the patient background must be taken into consideration in the efficacy evaluation.

Among the surgically treated cases, those in whom chylous plural effusion disappeared, respiratory symptoms were alleviated, and weaning from the respirator became possible have been reported [388, 389]. In addition, the absence of recurrence or reactivation is considered to be a point. [388‐391] There were reports that chylous pleural effusion after thoracic surgery was resolved by drainage alone. Cleveland et al. [392] considered conservative treatments, such as total parenteral nutrition, octreotide, and diuretic administration, to be the best and, observing that, of the poor responders, the mortality was 80% in 5 who continued to be managed by conservative treatments but 0% in 4 who underwent additional surgery, reported that surgical treatment contributed to the reduction of the mortality. According to the guidelines for the treatment of chylous thoracic effusion by Buttiker et al. [393], conservative treatment should be continuing for about 3 weeks but should be abandoned thereafter because of the risk of nutritional disturbance, increased susceptibility to infection, and liver disorders. However, Kaji et al. [394] reported that it is difficult to set a clear period of conservative therapy, because the effectiveness and success rate of surgical treatment are unclear.

As complications due to sclerosing agents, fever and increased inflammatory reaction due to the administration of OK-432 as well as pulmonary abscess and temporary flaccidity and protrusion of the upper abdominal region considered to have been due to intercostal nerve damage have been reported. While chyle leakage in the abdominal cavity was noted in a patient who underwent pleuroperitoneal shunting, there were no reports of fatal complications.

Surgical treatment was performed in most reported cases when responses to conservative therapy were not obtained. Therefore, it must be assumed that the results of evaluation of this CQ are based on data concerning the effectiveness of surgery performed with conservative therapy.

<Summary>

The literature was reviewed concerning the effectiveness of aggressive surgical intervention for neonatal chylous thoracic effusion from the viewpoints of responses and complications, but no objective study with a high level of evidence was found. In most reported cases, surgical treatment was performed when responses to conservative treatments were poor. Therefore, it is difficult to compare surgery with other therapies, and the evaluation of the period of conservative treatment before surgery remains insufficient. However, there was a paper that proposed surgical intervention after attempting conservative treatments for 3 weeks as a standard.

Thus, surgical intervention for neonatal chylous pleural effusion is characterized at present as an approach that may be effective but should be evaluated when the condition is not improved by other treatments, and “Surgical procedures, such as pleurodesis, ligation of the thoracic duct, and pleuroperitoneal shunting, may be effective for chylous pleural effusion refractory to conservative treatments.” is proposed as a draft recommendation.

Recommendation:

While treatments including surgery, sclerotherapy, radiotherapy, nutritional therapy, and drug therapy are conducted, there is presently no effective treatment with a high level of evidence. Treatments should be selected in consideration of complications and adverse effects according to individual symptoms.

The cause of chylous pleural and pericardial effusion is lymphorrhea from lymphatic vessel tissue lesions that have primarily invaded the mediastinum and pleura, and lymphorrhea from osteolytic lesions of the ribs and vertebrae was also observed. Respiratory disturbances were caused by pleural effusion, chylous pleural effusion, pericardial effusion, and direct invasion of the mediastinum and lungs.

As surgical treatments for chylous pleural effusion, procedures, such as thoracentesis, thoracic drainage, ligation of the thoracic duct, and pleural decortication, have been performed, and local lesions were surgically resected. In most cases, thoracentesis and thoracic drainage were performed, but chyle leakage was not resolved. As for complications, there were patients who developed hypovolemic shock and required blood transfusion and catecholamine administration or supplementation of albumin, immunoglobulin, and clotting factors [395‐397]. While chylous pleural effusion was controlled in some patients who underwent ligation of the thoracic duct [397‐408], the treatment was performed in combination with other surgical procedure or radiotherapy in all cases [400, 402, 408]. There was 1 case that showed improvement in respiratory disturbance [406]. As complications of ligation of the thoracic duct, splenomegaly and lymphorrhea [405] and left-sided pleural effusion [397, 405] have been reported. In the patients who showed marked improvements in chylous pleural effusion [395, 405, 408] after pleural decortication [395, 396, 401, 403‐405, 408, 409], the procedure was performed in combination with other surgical treatments or sclerotherapy, and there was no mention about complications. There were cases that showed marked improvements in chylous pleural effusion [396, 400, 405, 408] among those who underwent surgical resection of local lesions including splenectomy [396, 397, 400, 405, 408, 410‐412], but the procedure was performed in combination with other surgical treatments in most of them. Hemorrhage was reported as a complication [410]. Among other treatments, pleuroperitoneal shunting [403] and lung transplantation [413] were performed, and alleviation of respiratory disturbance was noted in the patient who underwent lung transplantation.

As a surgical treatment for pericardial effusion, pericardiocentesis was performed [396, 414‐416], and pericardial fenestration was performed when pericardial effusion could not be controlled by pericariocentesis [396, 416]. There was no mention about complications.

As sclerotherapy, pleurodesis was performed using OK-432, talc, and minocycline [395, 397‐399, 404, 408, 411, 416‐418]. There were patients who responded markedly to sclerotherapy alone and sclerotherapy combined with surgical procedures such as pleural decortication or local radiotherapy. There was no mention about complications of sclerotherapy.

There have also been reports [399, 400, 402‐404, 410‐412, 415, 416, 418‐420] on local (e.g., lesion area, thoracic duct region) and thoracic radiotherapy for chylous pleural effusion and local lesions, and marked responses of chylous pleural effusion and responses of respiratory symptoms were noted, but other treatments were performed concomitantly in some patients. Radiation pneumonitis has been reported as a complication [416].

Concerning nutritional therapy, fasting, high-calorie infusion, and MCT diet have been performed alone or in combinations, but few cases that showed alleviation of chylous pleural effusion were observed [395, 396, 398‐400, 403, 405, 408, 421].

For drug therapy against chylous pleural effusion, drugs including interferon α, propranolol, anticancer agents (e.g., vincristine), bisphosphonate, octreotide, steroid, sirolimus, and low-molecular-weight heparin were used. Interferon α was used most frequently [395‐398, 400, 401, 403, 415, 421], and marked improvement in chylothorax was reported in 5 cases. Of these cases, interferon α was used with propranolol in 1 [395] and with low-molecular-weight heparin and local radiotherapy (15 Gy) in 1 [400]. As for complications of drug therapy using interferon α, there were reports of fever, nausea, and headache [421] and thrombocytopenia and hepatic toxicity [397]. There was no report of improvement in chylous pleural effusion by the use of steroid [395, 399, 403, 415] or octreotide [395, 397, 398, 400, 403, 405] alone. Concerning other drug therapies, only a few cases have been reported with no improvement in chylous pleural effusion. One case that showed regression of mediastinal invasion of GLA and alleviation of respiratory disturbance by sirolimus treatment has been reported [414], and hypertension was noted as a complication. In drug therapy for pericardial effusion, diuretics were used for conservative therapy [400].

Although cases that responded to various therapies have been reported, treatments are often performed in combinations, and the evaluation of the effectiveness of each treatment alone is difficult at this point.

<Summary>

Treatments effective for GLA and GSD presenting with refractory chylous pleural effusion, pericardial effusion, and respiratory disturbances were evaluated by a review of the literature, which was primarily case reports. Various treatments, such as surgery, sclerotherapy, radiotherapy, nutritional therapy, and drug therapy, have been performed, but, there was no study with a sufficient number of cases and a high level of evidence because of the rareness of the disease and diversity of symptoms. Although cases that responded to various treatments have been reported, treatments are frequently performed in combinations, and the evaluation of the effectiveness of individual therapies is difficult at present. Sirolimus (a mTOR inhibitor) is considered promising as a drug for this disease, and some clinical trials are currently under way in Japan and abroad.

In clinical situations, these diseases are not recognized as indications of various drug therapies by the Japanese health insurance system, and the therapeutic effects of other treatments are also uncertain. Therefore, the above treatments cannot be recommended, but we propose that treatments “should be selected in consideration of complications and adverse effects according to individual symptoms”. It is necessary to evaluate the invasiveness, complications, and adverse effects, and select the treatments judged to be appropriate for each case.