Harlequin syndrome as a complication of epidural anaesthesia in an infant: Do adjunct medications play a role? : Indian Journal of Anaesthesia

Secondary Logo

Journal Logo

Clinical Communication

Harlequin syndrome as a complication of epidural anaesthesia in an infant: Do adjunct medications play a role?

Hylton, Jared R. E.

Author Information
Indian Journal of Anaesthesia 66(9):p 669-672, September 2022. | DOI: 10.4103/ija.ija_426_22
  • Open


Thoracic epidural catheters are commonly utilised for paediatric patients to provide post-operative analgesia.[1] Complications related to epidural catheter placement are uncommon and rarely lead to significant morbidity.[2]

Harlequin syndrome is a rare complication related to blockade of facial sympathetic input that can occur secondary to thoracic epidural anaesthesia.[3] This syndrome causes unilateral flushing, hyperhidrosis and occasional ipsilateral pupillary miosis. This generally resolves after cessation of local anaesthetic infusion through the epidural catheter. Despite the benign nature of this syndrome, the profound clinical appearance can lead to distress for patients and their families.

The risk factors for this phenomenon are unclear. It is not known if the choice of local anaesthetic or additive adjunct medications augments this risk. This case highlights the observation of a medication-related association not previously reported in the literature. While reported in adults and older paediatric patients, Harlequin syndrome as a complication of epidural anaesthesia has not been previously reported in an infant. It is unclear if the risk of this phenomenon is lower in infants relative to older patients.


A 12-month-old, otherwise healthy female, weighing 9 kg, presented for exploratory laparotomy, right nephrectomy and tumour resection for a Wilm’s tumor, via a large transverse abdominal incision above the umbilicus. An epidural catheter was requested for post-operative analgesia.

Following intravenous (IV) induction with 5 mg of 1% lidocaine, 20 mg of propofol, 12.5 μg of fentanyl and 10 mg of rocuronium, the trachea was intubated. An additional 20 gauge (G) IV catheter and 22 G radial arterial line were placed. The patient was then turned into right lateral decubitus for epidural placement.

Utilising external landmarks, the T7-T8 interspace was identified. Using a midline approach with a 5 cm, 19G Tuohy needle, loss of resistance was appreciated at a depth of 2 cm. A 20G epidural catheter was threaded to 6 cm at the skin. After a negative test dose, the epidural catheter was secured and a 2 mL bolus of 0.1% ropivacaine was administered.

General anaesthesia was maintained with inhalational sevoflurane. An infusion of plain 0.1% ropivacaine was infused through the epidural catheter at a rate of 3 mL/h for the duration of the procedure. The patient received two separate 1 mL boluses of 0.1% ropivacaine for haemodynamic responses to surgical stimulation. Total local anaesthetic dose for the intraoperative time period was 8.8 mL of 0.1% ropivacaine (8.8 mg). In addition, the patient received a total of 37.5 μg of fentanyl, 1 mg of dexamethasone, 60 μg of hydromorphone and 135 mg of acetaminophen, all administered via IV route. The surgery was uncomplicated while the surgical time was approximately 3 h. At the conclusion of surgery, the patient was extubated awake without apparent discomfort and was transported to the post-anaesthesia care unit (PACU).

In the PACU, a new epidural solution of 0.1% ropivacaine, 5 μg/mL of hydromorphone and 0.25 μg/mL of clonidine was started at a rate of 3 mL/h which is a routine mixture for epidural infusions for paediatric patients at our institution. Approximately 20 min after initiation of the new epidural medication solution, the PACU nurse noted new marked left unilateral facial flushing with an obvious midline demarcation [Figure 1]. No other clinical changes were noted, and the neurologic examination was otherwise unremarkable. It was determined that the facial flushing was likely Harlequin syndrome secondary to thoracic sympathetic blockade from the epidural catheter. The epidural infusion was paused, and resolution of the facial flushing occurred 30 min later. The epidural infusion, with additives, was restarted at 2 mL/h and no further facial flushing was reported for the remainder of the patient’s hospital course.

Figure 1:
Patient with Harlequin syndrome with left-sided facial flushings


The sympathetic outflow circuit relevant to head, neck and upper extremity anatomy is described as a 3-neuron chain. The first neurons in the hypothalamus send connections that terminate in the intermediolateral cell column of the spinal cord at the upper thoracic level.[45] Projections from the spinal cord follow the ventral spinal root, exit as white rami communicantes and travel with the sympathetic trunk to connect in the superior cervical ganglion. Preganglionic vasomotor and sudomotor connections exit at T3 and T2, respectively. The third order of this chain, stem from postganglionic neurons in the rostral and caudal end of the superior cervical ganglion.

The pathophysiology of Harlequin syndrome is thought to arise from disruption of T2-T3 sympathetic vasodilator and sudomotor facial innervation [Figure 2].[6] The distinguishing clinical feature of unilateral facial flushing is thought to arise from sympathetic dysfunction of thermoregulatory vasodilation contralateral to the disrupted sympathetic chain.[5] The neurologically disrupted side exhibits pallor and anhidrosis while the intact side appears flushed which is thought to be an exaggerated release phenomenon as compensation for loss of sweating and flushing on the contralateral side.

Figure 2:
Schematic diagrams depicting the anatomy and possible lesion sites of the cervico-thoracic sympathetic chain leading to Horner’s and Harlequin syndromes. A lesion below the level of T1 (a) may cause Harlequin syndrome, but is unlikely to cause Horner’s syndrome. A lesion proximal to the stellate ganglion (b) is likely to give upper limb and facial symptoms. Lesions distal to the stellate ganglion and proximal to the superior cervical ganglion (c and d) can cause concurrent Harlequin and Horner’s facial symptoms without symptoms in the upper limb. A lesion around the internal carotid (e) is likely to disrupt the oculosympathetic supply. (Reproduced with permission from Elsevier)[6]

Harlequin syndrome has both idiopathic and iatrogenic aetiologies. Some reports describe a clearly identified unilateral structural disruption in the sympathetic chain, such as secondary to a tumour or surgical trauma. In the absence of a clear unilateral structural abnormality, it is unclear why epidural analgesia would cause unilateral sympathetic blockade that would lead to Harlequin syndrome, albeit it is well-known that medication spread within the epidural space can be variable and dependent upon numerous physiologic and anatomical factors. Harlequin syndrome secondary to regional anaesthesia is benign and resolves with cessation of local anaesthetic action. There are no reports of regional anaesthetic techniques leading to chronic Harlequin syndrome.

Given that this is a rare complication of epidural anaesthesia, the risk factors are unclear. The onset of Harlequin syndrome in our patient was not observed until after the initiation of a local anaesthetic solution containing clonidine and hydromorphone. The patient did not receive any bolus of medication via the epidural route in the immediate time period prior to the onset of symptoms that could possibly explain our observations. Harlequin syndrome has been reported to result after neuraxial anaesthesia with both plain local anaesthetic as well as local anaesthetic solutions containing adjunct medications.[3] While there does not appear to be any evidence that epidural adjunct medications increase the risk of Harlequin syndrome, the temporal relationship between the onset of symptoms in this patient and the addition of clonidine and hydromorphone to the epidural infusion leads to the speculation that these additive adjunct medications played a role in the development of this clinical phenomenon. The mechanism of action of adjunct medications utilised for neuraxial and regional anaesthesia is dependent upon the class of medication utilised.[7] It is reasonable to hypothesise that these adjunct medications could also work with local anaesthetics to exert an additive and complementary action on facial vasodilator and sudomotor neural pathways.

While initially distressed about their child’s appearance, the patient’s parents were provided reassurance regarding the benign aetiology and clinical course of this rare complication. Given that the infant was experiencing excellent analgesia, the parents elected to continue the epidural infusion for advanced post-operative pain control.


The development of Harlequin syndrome can be dramatic and upsetting to patients and family members. It is important to educate patients, family and nursing staff regarding the aetiology and benign course of this syndrome. This case highlights the possible synergistic role that epidural additive medications may play in the development of Harlequin syndrome. Decreasing the infusion rate or temporarily stopping the infusion will lead to the resolution of facial flushing.

Declaration of patient consent

Appropriate written consent was obtained from parents/guardians for publication of photographs of the patient’s face, with appropriate editing to conceal the patient’s identity as much as possible, for publication in a scientific journal.

Financial support and sponsorship


Conflicts of interest

There are no conflicts of interest.


1. Ponde V Recent trends in paediatric regional anaesthesia Indian J Anaesth 2019 63 746 53
2. Long JB, Joselyn AS, Bhalla T, Tobias JD, De Oliveira GS, Suresh S, et al. The use of neuraxial catheters for postoperative analgesia in Neonates:A multicenter safety analysis from the Pediatric Regional Anesthesia Network Anesth Analg 2016 122 1965 70
3. Boling B, Key C, Wainscott J, Rebel A Harlequin syndrome as a complication of epidural anesthesia Crit Care Nurse 2014 34 57 61
4. Drummond PD, Lance JW Facial flushing and sweating mediated by the sympathetic nervous system Brain 1987 110 793 803
5. Lance JW, Drummond PD, Gandevia SC, Morris JG Harlequin syndrome:The sudden onset of unilateral flushing and sweating J Neurol Neurosurg Psychiatry 1988 51 635 42
6. Mohindra A, Herd MK, Roszkowski N, DownieI P Concurrent Horner's and Harlequin syndromes Int J Oral Maxillofac Surg 2015 443 710 2
7. Emelife PI, Eng MR, Menard BL, Myers AS, Cornett EM, Urman RD, et al. Adjunct medications for peripheral and neuraxial anesthesia Best Pract Res Clin Anaesthesiol 2018 32 83 99
Copyright: © 2022 Indian Journal of Anaesthesia