Periprocedural safety and outcome after pump implantation for intravenous treprostinil administration in patients with pulmonary arterial hypertension

A standardized protocol for PAH patients undergoing LenusPro® pump implantation has been established at our center in order to reduce perioperative risk as much as possible. All patients underwent extensive preoperative risk assessment including transthoracic echocardiography, 6MWD, WHO-FC, capillary blood gas analysis and measurement of NT-proBNP in addition to common anesthesia-related preoperative assessments including duplex sonography of the upper body vein status.

Preoperative anesthesiological evaluation of PAH patients was performed according to current recommendations and consensus statements [17, 18] and a comprehensive cardiopulmonary assessment (i.e. history and examination, medication, functional status, 6MWD, pulmonary function, blood gas analysis, laboratory tests (incl. creatinine, GFR, NT-proBNP), electrocardiogram, chest X-ray, and echocardiography [11].

All patients were admitted to our hospital for intravenous up-titration of treprostinil via a peripherally inserted central venous catheter prior to pump implantation. Treprostinil dose was increased every 8–12 h depending on patient’s tolerability, and pump implantation surgery was performed once a sufficient individual dose was reached.

Whenever possible, surgery was conducted during concious sedation, i.e., in local anesthesia combined with low-dose analgosedation as needed. Standard monitoring includes electrocardiogram, non-invasive blood pressure measurement, pulse oximetry, respiratory rate, and approximation of end-tidal carbon dioxide (etCO₂). During the procedure, oxygen (4–8 l/min) was applied via face mask allowing sampling of exhaled carbon dioxide during administration of supplementary oxygen (Flexicare Dual Mask™, Flexicare Medisize Germany, Siegburg, Germany). Oxygen was delivered to maintain peripheral oxygen saturation (SpO₂) at 92–94 %. All patients received 1–2 peripheral intravenous lines (> G18). Analgosedation was achieved by continuous infusion of remifentanil (concentration: 0.5–1 mg/50ml; infusion rate 0.02–0.15 µg/kg/min) with or without co-administration of midazolam boli (0.025–0.05 mg/kg) until the end of surgery. Depth of sedation aimed for a Richmond Agitation and Sedation Score (RASS) of 0 to – 2 (“conscious sedation“). Convective temperature management system was used to achieve normothermia (TWINWARM MoeckWarmingSystem™, Moeck&Moeck GmbH, Hamburg, Germany).

Furthermore, close attention was paid to avoid circulatory dysregulation and hypothermia as well as hypoxemia, hypercapnia or pain spikes/stress as all of these can affect pulmonary hemodynamics [19, 20]. PAH therapy was continued in all patients during the perioperative period.

Operations were carried out by an experienced cardiothoracic surgeon and performed in supine position in compliance with usual hygiene regulations. After extensive disinfection and sterile covering of the operating area, the target areas were subcutaneously infiltrated by local anesthetics (prilocaine 1 %, 30 ml), and intravenous antibiotics (intra-venous single-shot cefuroxime 1.5 g) were administered. After adequate exposure time, skin incision was made caudally to the left or right clavicle (Fig. 1d) with subsequent preparation of the subcutaneous adipose tissue towards the subclavian vein (left sided implantation was preferred whenever possible in order to avoid kinking of the pump line). The subclavian vein was punctured and a probe was inserted via Seldinger’s technique and moved forward fluoroscopy-guided to the superior vena cava / right atrium junction. Following this, a second skin incision was made under the left costal arch (Fig. 1a) and a subcutaneous slot was created (Fig. 1b). From there, a tunnelizer was brought through the subcutaneous tissue to the subclavian area and the pump line was pulled through the tunnel and connected to the probe using an appropriate connector (Fig. 1c). Subsequently, LenusPro® pump was moistened by vancomycin and gentamycin solution, introduced to the slot, and fixed by sutures to avoid pump rotations in the further course. Choice of pump size depended on patient’s constitution, pump pocket location, and the surgeon’s appraisal. Smaller pumps (20 ml reservoir) were used for lean patients or atypical pump positions while larger pumps (40 ml reservoir with normal or high septum) were used for all other patients. Every choice was made individually and no BMI cut-off existed. Finally, adequate position of all components was controlled by fluoroscopy (Fig. 2) followed by wound closure.

In patients undergoing lung transplantation, right lung is always transplanted before left lung. After right-sided transplantation is completed, the surgeon makes a left-sided thoracical incision and prepares towards the left internal thoracic artery. During this procedure, the pump line becomes visible and the surgeon attaches a clamp to interrupt treprostinil administration. Pump and pump line are removed at the end of the procedure.

The first filling procedure was performed intraoperatively to avoid punction of the freshly operated site. In-hospital pump refills were performed by an experienced physician. Outpatients received refills at home by a trained nurse.

Under aseptic conditions, localization of the septum was usually done manually with support by a special localization device or ultrasound, if necessary. Correct position was verified by reflux of fluid from the pump and by checking reflux in the syringe during injection. After application of treprostinil, NaCl was injected to avoid local reactions resulting from residual amounts of treprostinil in/on the needle.

Given that our data was observational, retrospective and post-hoc, our results were depicted in a descriptive manner without formal statistical analyses.