The inside-out technique is safe and effective for thoracic central venous obstruction

Thoracic central venous obstruction (TCVO) presents a challenging scenario for patients requiring central venous access. The inside-out technique for crossing occluded veins has been described; however, to date, case series have reported on a limited number of patients. The purpose of the present study was to evaluate the indications for, efficacy of, and outcomes with the inside-out technique at a single tertiary academic center, with close attention to the severity of TCVO using the Society of Interventional Radiology (SIR) TCVO classification.

Patients who had undergone central venous access using the inside-out technique were identified from August 2007 to May 2021. The patient demographics, procedure indication, procedural details, SIR TCVO classification, outcomes, and procedure-related complications were recorded. Statistical analysis was performed using analysis of variance.

A total of 221 patients (109 men [49.3%] and 112 women [50.6%]) had undergone 338 inside-out procedures. Of the 221 patients, 49 had undergone the procedure multiple times (25 two times, 11 three times, 13 more than three times). The average patient age was 54.7 ± 14.8 years. The indications (n = 362) for the procedure included dialysis access (n = 230; 63.5%), infusion of parenteral nutrition, antibiotics, chemotherapy, or other medication (n = 81; 22.3%), cardiac access (n = 39; 10.8%), and other (n = 12; 3.3%); more than one indication for 20 procedures. Type 1 SIR TCVO was found during 147 procedures (43.5%), followed by type 4 for 142 (42.0%), type 2 for 36 (10.7%), type 3 for 6 (1.8%), and unable to determine for 7 (2.0%). The access site was the right femoral vein for 322 procedures (95.3%), left femoral vein for 14 (4.1%), and transhepatically for 2 (0.6%). The exit site location was the right supraclavicular region for 274 (81.3%), right subclavicular for 52 (15.4%), left supraclavicular for 3 (0.9%), left subclavicular for 6 (1.8%), and not defined for 2 (0.6%). Types 3 and 4 were associated with longer fluoroscopy times and more contrast compared with types 1 and 2. The median follow-up and device duration was 56.0 days and 76.5 days, respectively. No differences were found in device duration between the SIR TCVO types. Removal of a catheter was documented for 166 patients. The indications for removal included infectious causes (non-catheter-related bacteremia, catheter-related infection or bacteremia) for 70 (42.1%), catheter malfunction for 34 (20.5%), new hemodialysis access for 19 (11.5%), no longer needed for 19 (11.5%), patient removal of the catheter by 13 (7.8%), and replacement of a temporary catheter with a tunneled device for 11 (6.6%). No procedural complications were associated with the inside-out technique or catheter removal.

For a variety of indications, we have shown that the inside-out technique is safe and effective for establishing central venous access in patients with TCVO and can be performed repeatedly. More complex obstructive patterns were associated with longer fluoroscopy times and greater contrast administration. Durability was primarily limited by infectious complications.