Collection and publication of patient data were approved by the IRB of Dar Al Fouad Hospital, Giza, Egypt (DAF2023-19). Written informed consent was obtained from the father for both the intervention and the publication.

An 18-month-old baby girl was referred to the pain clinic with a history of severe lower limb pain that started 2 months earlier. Her parents were unable to recall any trauma before the onset of the pain syndrome. Before the incidental onset of pain, the girl had her normal growth in all aspects. She was able to take her first few steps while walking without assistance, with no known previous diseases or allergies.

The mother reported that the girl gradually started crying and pointing to her right knee. Over the next 2 weeks, the pain increased to cover the entire lower limb, and spread to the gluteus area by the third week. The mother informed us that the girl experienced continuous stinging pain for 24 hours, and diaper changes became problematic due to the severe pain. Due to a lack of sleep and poor appetite, the baby started losing weight. When she arrived at the pain clinic, her weight was 6700 g, significantly less than the mean normal weight at her age. The parents described the pain as excruciating and no one could touch her skin, especially around the right knee up to the gluteus area and down to the big toe.

During the first week they consulted the pediatrician who initially prescribed paracetamol 10 mg/kg oral syrup/6 hourly PRN. He then added ibuprofen syrup, 6 mg/kg every 8 h, and referred her to a pediatric orthopedic specialist. The specialist examined her with marked difficulty as severe allodynia and partial contraction of the right knee were present. He could not extend the patient’s leg. Hence, he ordered an X-ray image of the right lower limb and knee to exclude any bone lesions. All of the X-ray images were normal. She was suspected of having CRPS and was referred to the pediatric neurology department. Pediatric neurologists agreed on the suspect of CRPS, but as it is an exclusion syndrome, they did not have the possibility of an EMG that is usually performed in such a condition, to exclude nerve injuries. The exam would have been extremely difficult for the young patient without general anesthesia, which could have affected the results. Neurologists prescribed oral gabapentin 25 mg/day that was gradually escalated to 100 mg on three divided doses/day. They also added amitriptyline 5 mg that was increased to 10 mg at bedtime. The drugs' oral regimen represented the maximum possible dose for that age. The patient showed some improvement in sleep, and the pain appeared reduced and more localized around the knee. Still, she could not extend her knee, which was a little swollen and in a persistent condition of contraction at 45 degrees. The patient was referred for rehabilitation that could not be performed because of marked allodynia. The neurologist referred the patient to the pain clinic for a potential interventional treatment.

Therefore, the patient came to our attention after a journey of 2 months in the described conditions. She appeared emaciated and underweight (her body weight was 6700 g). She was also apprehensive, and showed an exaggerated fear of strangers, especially those in white coats. The pain score at rest was 8 out of 10 utilizing the FLACC scale [10]. Her right knee was semi-flexed at 45 degrees. She had marked allodynia in the right lower limb but mostly around the knee. There were not many dysautonomic signs (color and temperature changes) apart from mild edema around the knee. The parents could recall color changes in the form of redness, hotness, and edema, which accompanied the breakthrough bouts of pain. The right knee was permanently in a state of contracture at 45 degrees, without any possibility of passive extension. She could contract it more to some degree but could not extend it at all. She lost the ability to walk, as she could not perform any weight-bearing on her partially flexed right leg. There was no other abnormality. After discussing the risks and potential benefits of the planned intervention and the limitations in the literature regarding similar extremely young cases, the parents approved the trial procedure to save the limb of their baby girl and restore its function. Written informed parental consent was obtained for the intervention and the use of the patient’s anonymized data for the publication of this case report.

The patient was introduced to the preparation room by her mother. An intravenous line was established after local anesthesia using EMLA cream. Sedation with intravenous midazolam (0.05 mg/kg) was initiated as per our hospital protocol. Standard monitors were used, including pulse oximetry, noninvasive blood pressure measurements, and ECG. The patient was transferred to the intervention room accompanied by her mother. Premedication with i.v. 0.02 mg/kg atropine was administered followed by intravenous ketamine at a dose of 2 mg/kg to initiate general anesthesia. A face mask was used with sevoflurane 1–2%, followed by laryngeal airway mask (LAM) application. At this point, the mother was requested to leave the room. Prone position with a soft roll under the pelvis was achieved and a skin temperature probe was applied to the right calf muscle area. Spontaneous breathing of 50% oxygen mixed with air and sevoflurane 1% was aided by manually assisted ventilation, if needed.

Under strict aseptic conditions and fluoroscopy guidance, L3 vertebrae was visualized (Fig. 1). The C-arm was placed in the oblique plane until the end of the transverse process was aligned with the anterior aspect of the vertebral body (Fig. 2). For the intervention, we used a 4-cm-long trocar of a 22G cannula, which was shorter than any available spinal needle. A needle was introduced and tunneled vision was obtained (Fig. 3). Once the trajectory was confirmed, a lateral view was obtained to determine the depth of the needle (Fig. 4). Needle advancement was checked every 1–2 mm depth, and we aimed to keep contact with the vertebral body as much as possible and slipped along the vertebral body to ensure correct placement. When the required depth was achieved, an A-P view was obtained to confirm the correct position of the needle tip no deeper than the mid-pedicular position. After suction, 1 ml of iohexol (180 mgI/ml) over 20 s was injected to ensure proper needle placement (Figs. 5, 6). For the anesthetic block, we used a mixture of bupivacaine 2 mg/kg with dexamethasone 0.2 mg/kg, confirmed by adequate spread of the dye by fluoroscopy (Fig. 7).

A-P view of lumbar vertebrae

Oblique view with the end of the transverse process aligned with the anterior aspect of the vertebral body

Needle introduced in tunneled vision

Lateral view obtained to determine the depth of the needle

Proper needle position, with dye (iohexol) injection to confirm accurate needle position in the antero-posterior (A-P) view

Proper needle position, with dye (iohexol) injection to confirm accurate needle position in the lateral view

Adequate dye spread after injecting the local anesthetic and steroid

There are no established guidelines recommending the volume of injectate for this technique in pediatrics. Therefore, we used the guidelines for epidural dosing in pediatrics as a proportional reference. If we know that the lumbar epidural volume for pediatrics ranges from 0.25 to 0.0.5 ml/kg [11], we agreed to have a reduced volume for our case because the point of injection is not in a closed space like the epidural space. We injected a total volume of 2 ml, confirmed by the dye spread after injection, which was satisfactory with no evidence of epidural, intravascular, or lateral spread.

The radiation dose from recorded images and number of images recorded during the procedure were reduced to attain optimum clarity and quality of the images without ignoring the safety of the patient or the procedure.

Accumulated values measured by the machine were as follows: radiation time: 00.00.38 (38 s); dose area product (DAP): 12.67 µGy∙m2 (reference air kerma (mGy) 0.30.

At this point, the patient was turned in a supine position. The general anesthesia was terminated, and the LMA was removed after securing adequate reflexes. The lower limb temperature was checked upon arrival in the recovery room. An elevation of 2.9 °C was achieved, confirming an adequate sympathetic block. The hemodynamic values showed no significant differences. The mother was invited to join the daughter in the recovery room. The first general impression was that there was no allodynia, and the child was trying to extend her knees. After complete recovery, she was discharged from the recovery room, and after another 3 h, she was discharged, and referred to the pediatric rehabilitation department.

After 3 and 8 weeks, in a follow-up visit, the parents reported overall improvement. Pain and allodynia were reported to be improved by 80%, and 70% respectively, with noticeable improvement in sleep, mood, and activity. On examination, there was almost no allodynia. However, the knee was still contracted. The patient had gained 900 g of body weight at the 3-week follow-up visit, and the parents reported a better appetite. During a follow-up phone call 4 months later, the patient’s mother expressed her satisfaction for the significant improvement of pain and sleep, indicating these were no longer issues. However, she did note a delay in rehabilitation progress. On a positive note, the patient can now step over her feet, has started walking, and is able to bear weight on her affected knee.