A 74-year-old woman with severe back pain and fever was diagnosed with infectious spondylodiscitis at L2–3 and an early right psoas abscess. Contrast-enhanced magnetic resonance imaging (MRI) showed enhancement of the L2–3 disc space and adjacent endplates with a small abscess pocket in the right psoas muscle. She underwent transforaminal percutaneous endoscopic debridement and drainage (PEDD) under local anesthesia. Infected disc material was removed, and a drain was inserted under fluoroscopic guidance (Figure 1). Streptococcus constellatus was cultured, and both the patient’s symptoms and inflammatory markers improved after antibiotics treatment with ceftriaxone.

A 72-year-old man with a history of treated pulmonary tuberculosis presented with severe localized back pain and fever. Contrast-enhanced MRI revealed infectious spondylodiscitis at L3–4 with an epidural abscess and bilateral psoas abscesses with right-sided dominance (Figure 2). Combined transforaminal and psoas muscle PEDD was performed under local anesthesia, achieving complete debridement and drainage. The disc and psoas muscle granulation tissue both appeared grayish, and the psoas granulation demonstrated cheesy-like areas suggestive of caseous necrosis (Figure 3). Histology confirmed tuberculous spondylodiscitis, and the patient showed gradual improvement in symptoms and inflammatory markers with antituberculosis therapy.

In this video, we present 2 illustrative cases of PEDD for infectious spondylodiscitis performed under local anesthesia using a full-endoscopic spine surgery: one through a transforaminal approach, and another combining transforaminal PEDD with psoas muscle PEDD.

Infectious spondylodiscitis is becoming more common, particularly in elderly patients and those with multiple comorbidities [1]. Conventional diagnostic methods such as blood cultures and bone needle biopsy often fail to identify the causative organism, with reported yields of only 30%–50% [1,2]. Consequently, many patients remain culture-negative, which leads to prolonged use of empirical antibiotics, development of resistance, and, even treatment failure [1,3].

PEDD has emerged as a minimally invasive alternative [3-6]. This procedure can be safely performed under local anesthesia [2,8]. Clinical studies have demonstrated infection control rates exceeding 80%, and culture yields ranging from 67% to 90%, with complication rates below 10% [2-7]. In addition, PEDD enables direct drainage of associated paraspinal or psoas muscles, thereby expanding its therapeutic potential [5].

A 74-year-old female presented with severe lower back pain accompanied by fever. She had a history of L3 vertebroplasty for osteoporotic compression fracture and multiple acupuncture procedures. Laboratory investigations revealed marked leukocytosis and elevated C-reactive protein (CRP). On neurological examination, there was no motor weakness or radiating leg pain, but she complained of intractable localized back pain.

Her lumbar spine x-ray revealed old compression fractures at L2 and L3 with wedge-shaped deformity, resulting in segmental kyphotic change. The L3 vertebral body was in a post-vertebroplasty state. The L2–3 disc space appeared narrowed with irregular endplate changes, and overall lumbar alignment showed levoscoliosis with diffuse severe spondylosis. Her imaging also demonstrated S1 lumbarization. On nonenhanced MRI scan, the L2 and L3 nerve roots appeared intact without compression. The images also showed more prominent destructive changes of the L2–3 disc space and adjacent endplates on the right side compared with the left. Contrast-enhanced MRI scan showed infectious spondylodiscitis at L2–3 with contrast enhancement of the disc space and adjacent endplates, accompanied by an early-stage right psoas muscle abscess pocket. Based on these findings, the right-sided portion of the L2–3 disc space was selected as the surgical target, and a right L2–3 PEDD was planned. Computed tomography scan confirmed the planned entry trajectory. To access the selected target, the needle insertion site was determined approximately 7 cm lateral to the midline. Because the posterior margin of the vertebral body at this level was relatively narrow, drilling at the disc space entry point was considered.

The patient was positioned prone on a radiolucent operating table. The hips and knees were slightly flexed to reduce lumbar lordosis, and all the bony prominences were adequately padded. This positioning increased the size of the intervertebral foramen and reduced traction on the exiting nerve roots. The abdomen was also left free to minimize venous congestion and reduce epidural bleeding.

An 18-gauge spinal needle was inserted approximately 7 cm lateral to the midline under fluoroscopic guidance. After docking the needle at the lateral margin of the superior articular process (SAP) base, it was slightly advanced ventrally to approach through Kambin triangle. The correct needle position was confirmed under C-arm fluoroscopy. A selective root block for the right L3 exiting nerve root was then performed using a mixture of local anesthetic and radio-opaque dye. Following this, the needle was advanced slightly further into the disc space, and discography was carried out using indigo carmine mixed with contrast, which stained the infected disc material and delineated the annular defect. Subsequently, a guide wire was introduced, followed by sequential dilation, and finally, the working cannula and endoscope were inserted.

We use an endoscope with an outer diameter of 6.3 mm, a working channel of 3.7 mm, and a 30° viewing angle. Once the endoscope is inserted, it is essential to identify the anatomical landmarks, since necrotic tissue, infected granulation, and irregular bone formation often obscure the normal anatomy in spondylodiscitis. The curved tip of the radiofrequency probe can be used as a smooth dissector to gently separate tissue planes and assist in the exposure of landmarks such as SAP, pedicle. It is important to avoid injury to exiting nerve root when identifying reliable anatomical landmarks, including the right L3 SAP and the right L3 pedicle. At the planned disc entry point, the disc stained with indigo carmine is identified. Granulation tissue and the stained disc material are removed, while samples are simultaneously collected for biopsy and culture. If vertebroplasty cement or bony lesions prevent entry into the disc space, a high-speed drill is employed to secure the entry point. Once access has been established, discectomy is performed using disc forceps, flexible curved grasping forceps, and endoscopic punches. C-arm fluoroscopy is utilized throughout the procedure to confirm the working trajectory and depth, while taking care to avoid injury to the anterior longitudinal ligament and adjacent vascular structures. When the target area is reached, a guide wire is inserted under endoscopic visualization. After the endoscope is withdrawn, the distal part of a hemovac drain, which had been pretrimmed according to the length of the guide wire, is inserted over the guide wire into the target location. Final placement of the drain is verified with C-arm fluoroscopy.

Streptococcus constellatus was identified, and the patient was treated with ceftriaxone. The patient’s back pain improved, allowing ambulation with a thoracolumbosacral orthosis. She remained afebrile, and inflammatory markers gradually normalized without evidence of recurrence.

A 72-year-old male presented with severe back pain accompanied by fever. He had a history of pulmonary tuberculosis, which was cured 8 months earlier. Laboratory investigations revealed a white blood cell count of 9.4 ×10³/µL (neutrophils 62.4%, monocytes 9.4%) and CRP of 4.36 mg/dL. On neurological examination, there was no motor weakness or radiating pain, but he complained of severe localized back pain.

Contrast-enhanced MRI revealed infectious spondylodiscitis at L3–4 with an accompanying epidural abscess, as well as bilateral psoas muscle abscesses with right-sided dominance.

The largest psoas muscle abscess pocket was located between the L3 and L4 transverse processes, and based on MRI findings, the needle insertion site was determined approximately 7 cm lateral to the midline.

After needle insertion 7 cm lateral to the midline, the endoscope was introduced toward the right L3 transverse process. Using the curved tip of the radiofrequency probe, the caudal margin of the L3 transverse process was palpated, and the endoscope was advanced ventrally between the right L3 and L4 transverse processes. Whitish-to-grayish infected granulation tissue was encountered and progressively removed, which facilitated further advancement of the endoscope. A psoas muscle abscess pocket was subsequently visualized, and its ventral margin was confirmed with C-arm fluoroscopy. Thorough debridement of the granulation tissue was performed under direct endoscopic visualization. A guide wire was then placed through the working channel, and a pretrimmed hemovac drain was inserted over the wire into the abscess cavity. The final position of the drain was verified fluoroscopically, and the procedure was completed without complications.

Postoperative x-ray confirmed appropriate placement of the drainage catheters within the intradiscal space and the psoas muscle abscess pocket. The disc and psoas muscle granulation tissue both appeared grayish, and the psoas granulation demonstrated cheesy-like areas suggestive of caseous necrosis. Psoas muscle granulation tissue showed necrosis with chronic granulomatous inflammation, suggesting tuberculous infection. Polymerase chain reaction testing confirmed Mycobacterium tuberculosis positivity, establishing the diagnosis of tuberculous spondylodiscitis. The patient was started on standard 4-drug antituberculosis therapy, consisting of isoniazid, rifampin, ethambutol, and pyrazinamide. Following initiation of treatment, the previously elevated CRP levels gradually decreased, and the patient’s back pain gradually improved.

PEDD represents a safe and minimally invasive option for the management of infectious spondylodiscitis [2,8]. It provides a higher culture yield (67%–90%) than conventional diagnostic methods, while achieving infection control rates exceeding 80% with a low complication profile [1,2-7]. Furthermore, PEDD allows simultaneous drainage of paraspinal or psoas abscesses, thereby broadening its therapeutic applicability [5].