Routine Chest X-ray is not Mandatory after Fluoroscopy-Guided Totally Implantable Venous Access Device Insertion Theodoros Thomopoulos,1 Jeremy Meyer,1 Wojciech Staszewicz,1 Ilias Bagetakos,2 Max Scheffler,2 Antoine Lomessy,1 Christian Toso,1 Christoph D. Becker,2 and Philippe Morel,1 Geneva, Switzerland
Background: The aim of this study is to determine whether systematic postoperative chest X-ray is required after totally implantable venous access port device (TIVAD) placement under fluoroscopic control. Methods: A retrospective chart review of all consecutive patients with fluoroscopy-guided TIVAD insertion from July 10, 2009 to April 16, 2012 was conducted at the Geneva University Hospitals (n ¼ 927). Patients with an available postoperative chest X-ray were included, regardless of approach (open or percutaneous) and venous access site (subclavian, cephalic, jugular, etc.). Exclusion criteria were incomplete data and preexisting pneumothorax or hemothorax. Results: Eight hundred ninety-one patients were included. First-intention venous cutdown was performed in 878 patients (98.5%), with success rates of 79.4% and 88.2% when targeting the left and right cephalic veins, respectively. Percutaneous access was the chosen first-intention procedure for 12 patients (1.3%). Eight-hundred thirty-six (93.8%) insertions were performed only by the open approach and 53 (5.9%) implantations required at least one venous puncture. Two implantations were performed using previous central venous accesses. Immediate complications associated with TIVAD placement and detected on the postoperative chest X-ray consisted of 1 asymptomatic pneumothorax, 1 symptomatic hemothorax, and 2 malpositions of the catheter. One additional pneumothorax was discovered during the first night after TIVAD insertion in a patient who became symptomatic. Conclusions: The very low incidence of immediate complications detected by postprocedural chest X-ray suggests that such a control is not mandatory as a routine method after fluoroscopyguided TIVAD insertion mainly performed by venous cutdown. X-ray should be performed only in cases of clinical suspicion.
INTRODUCTION
The two first authors contributed equally to this study. 1 Division of Visceral and Transplantation Surgery, University Hospitals of Geneva, Switzerland. 2
Division of Radiology, University Hospitals of Geneva, Switzerland.
Correspondence to: Theodoros Thomopoulos, MD, Services de chirurgie viscerale et transplantation, Departement de chirurgie, H^opitaux Universitaires de Geneve, Rue Gabrielle-Perret-Gentil 4, 1211 Geneve 14, Suisse; E-mail: [email protected] Ann Vasc Surg 2014; 28: 345–350 http://dx.doi.org/10.1016/j.avsg.2013.08.003 Ó 2014 Elsevier Inc. All rights reserved. Manuscript received: April 1, 2013; manuscript accepted: August 22, 2013; published online: December 18, 2013.
First introduced in the 1980s, totally implantable venous access devices (TIVADs) have become essential in the management of a number of disorders, including chronic infection and cancer. They provide a reliable venous access for the administration of long-term antibiotics, chemotherapeutic agents, blood products, and total parenteral nutrition. The choice of the site of implantation is based on local recommendations and the physician’s skills. The most popular sites are the internal and external jugular and subclavian veins. Alternative locations include the cephalic vein in the deltopectoral groove, with the axillary vein and the lower extremity 345
346 Thomopoulos et al.
veins as last options. Catheter insertion can be performed using either an open or percutaneous approach, with or without imaging guidance.1 Short- and long-term complications associated with this proceduredoften seen as ‘‘minor’’dare a matter of concern and have a reported prevalence of 6e18%.2e6 They are often classified chronologically, with immediate complications occurring between the time of placement and the first use, and late complications occurring in the subsequent period. The most frequent immediate complications include catheter malposition, local hematoma, cardiac arrhythmia, arterial puncture, and pneumothorax, which have each been observed in up to 6% of TIVAD insertions.3e6 During the early eras, TIVADs were placed without intraoperative image guidance, using only anatomic landmarks. Later, postoperative chest X-ray was recommended to confirm the exact location of the catheter and to detect immediate complications, such as pneumothorax and hemothorax. More recently, ultrasound has been used for a more precise venous puncture, and fluoroscopy for guidance of the catheter’s tip placement. Despite these technical refinements, postoperative chest X-ray is still performed at many institutions, remaining part of the standard practice. However, use of systematic postoperative chest X-ray presents additional irradiation, is relatively expensive, may delay more important procedures, and can lead to a false feeling of safety because not all complications can be detected early after the procedure7 and not all visible complications are correctly diagnosed by nonspecialists.8 The aim of this study was to determine whether systematic postoperative chest X-ray is required after TIVAD placement under fluoroscopic control.
METHODS A retrospective chart review of all consecutive patients with fluoroscopy-guided TIVAD insertion from July 10, 2009 to April 16, 2012 was conducted at the Geneva University Hospitals. Patients with an available postoperative chest X-ray were included, regardless of approach (open or percutaneous) and venous access site (subclavian, cephalic, jugular, etc.). Exclusion criteria were incomplete data and preexisting pneumothorax or hemothorax. The study protocol was reviewed and accepted by the ethics committee at the Geneva University Hospitals. A single type of TIVAD (CelsiteÒ; B. Braun Medical) was used for all patients. This device consists of a single lumen composite access port
Annals of Vascular Surgery
with a titanium chamber, silicone membrane, and catheter. The procedure was performed under local anesthesia by general surgeons, always in the presence of an anesthesiologist for optional administration of intravenous sedation. Cefuroxime axetil was used as prophylaxis. Cephalic venous cutdown was the first choice, and subsequent implantation sites were left to the preference of the surgeon, often selecting the external jugular venous cutdown or the subclavian vein puncture. Second- and thirdchoice accesses were performed on the same side for practical reasons. The percutaneous approach was performed under ultrasonography. The catheter was systematically introduced under fluoroscopic guidance with assessment and perioperative correction of the position of its tip. The final fluoroscopic image was printed and attached to the patient chart. A cavity was then made on the pre-pectoral fascia to introduce and fix the chamber. A Gripper needle was left in the device at the end of the procedure. A postoperative chest X-ray was obtained within 30 minutes of the procedure. Interpretation was made by the surgeon. The patient was then discharged. For this study, charts were reviewed with special attention given to operative and postoperative data, looking for the nature of the access site, perioperative fluoroscopy results, and the occurrence of the three immediate complications detectable on chest X-ray: pneumothorax; hemothorax; and catheter malposition. All X-rays were retrospectively blindly reviewed by board-certified radiologists. In cases of a detected complication, the anesthesic records and the consecutive clinical notes were reviewed for a more detailed analysis. The catheters of the TIVAD were considered misplaced if they were not located within 5 cm at the junction of the superior vena cava and the right atrium. Pneumothorax and hemothorax were always considered as a complication, regardless of their dimension and or the presence of symptoms.
RESULTS Nine hundred twenty-seven TIVAD placements were performed during the study period, and 36 were excluded for the reasons listed in Table I. The remaining 891 procedures followed the implantation methods described in Figure 1. First-intention venous cutdown was performed in 878 patients (98.5%), mainly targeting the right cephalic vein, with success rates ranging from 79.4% to 92.3%. Percutaneous access was the chosen first-intention procedure for 12 patients (1.3%). Eight hundred
Vol. 28, No. 2, February 2014
Post-TIVAD chest X-ray 347
Table I. Main reasons for exclusion (36 patients were excluded) Reasons for exclusion
Number of patients
No postoperative radiography Incomplete file Preexisting pneumothorax
20 14 2
thirty-six (93.8%) insertions were performed only by the open approach and 53 (5.9%) implantations required at least one venous puncture. Two implantations were performed using the access of a previously inserted central venous catheter. Overall, immediate complications associated with TIVAD placement and detected on the postoperative chest X-ray consisted of 1 asymptomatic pneumothorax, 1 symptomatic hemothorax, and 2 malpositions of the catheter. One additional pneumothorax was discovered during the first night after TIVAD insertion in a patient who became symptomatic. It had not been detected on the immediate postoperative chest X-ray. Cases presenting with immediate complications after TIVAD insertion are described in what follows. Case 1 A 78-year-old woman, with squamous cell cancer of the right lung and a lobular carcinoma of the right breast, underwent a TIVAD placement by left cephalic vein cutdown. The insertion was performed without technical problems and fluoroscopic imaging did not reveal any complication. However, the immediate chest X-ray showed a left minimal apical pneumothorax of 7 mm. The patient remained asymptomatic. She was followed up in the recovery ward. Because of her desire to return home and while remaining nonsymptomatic, she was discharged the same day with instructions to return in case of respiratory symptoms or chest pain. Case 2 A 69-year-old man, with metastatic adenocarcinoma of the sigmoid to the lungs and the liver, was scheduled for a TIVAD insertion for palliative chemotherapy. The catheter was inserted by left cephalic vein cutdown under fluoroscopic guidance, with a good reflux at the end of the procedure. However, postoperative chest X-ray showed that the tip of the catheter was not located in the superior vena cava (SVC), but in one of its branches, probably the azygos vein. The patient was reoperated the same day for catheter reposition, with good results.
Case 3 A 58-year-old man, with lymph node recurrence of a right testicular seminoma after orchidectomy, was scheduled for a TIVAD placement for chemotherapy by right cephalic vein cutdown as first intention. At the end, the device was implanted on the third attempt by puncture of the right subclavian vein due to failure to locate the right cephalic and right external jugular veins via an open approach. The operative and radiologic records described no complication. Twelve hours after the intervention, the patient developed dyspnea and right chest pain. Chest X-ray showed a 7-cm pneumothorax. A chest tube was inserted, and the patient could thereafter be discharged after an uneventful 8-day course. Case 4 A 75-year-old man, with recurrence of a folliculartype B-cell lymphoma, underwent a TIVAD insertion by left cephalic venous cutdown on first intention. Failure to find the vein led to puncture of the left subclavian vein. The tip of the catheter was advanced to the junction of the superior vena cava (SVC) and the right atrium, but no reflux occurred after multiple attempts. The procedure was then interrupted because the patient complained of right shoulder pain. Subsequent chest X-ray demonstrated a large left pleural effusion. Computed tomography confirmed the presence of a 900-mL hemothorax without active bleeding. Due to the patient’s desire not to be drained, conservative treatment was undertaken. Stable hemoglobin values allowed discharge at day 7. Case 5 An 86-year-old woman, with ductal invasive cancer of the right breast, was scheduled for a TIVAD placement for chemotherapy. The catheter was inserted by right cephalic vein cutdown under fluoroscopic guidance, with a good reflux at the end of the procedure. Bad reflux was then reported during chemotherapy, thus prompting angiographic control. We observed malposition of the catheter, with tip placement in the left branchiocephalic vein. Due to the absence of a good reflux after a lysis protocol, the patient was reoperated for catheter repositioning. The TIVAD was replaced by left cephalic vein cutdown with a good result. A retrospective reading of the first postoperative X-ray indicated a short catheter placed very high in the SVC. This was not corrected during the first intervention, probably due to a misreading by the surgeon.
348 Thomopoulos et al.
Annals of Vascular Surgery
Fig. 1. Totally implantable venous access port device implantations performed from July 10, 2009 to April 16, 2012 in the 891 patients studied. Patients are distributed according to the chosen vascular access procedures. Primary subsets represent first-intention vascular access; secondary subsets represent second-intention vascular access; and tertiary subsets represent third-intention vascular access. Numbers in parentheses represent the proportion of patients who underwent the related procedure in relation to the preceding subset. SR indicates success rate for each implantation. *Left common femoral venous cutdown (n ¼ 1), right saphenous
venous cutdown (n ¼ 1), right internal jugular venous puncture (n ¼ 1), transcatheterization of a previous TIVAD in the right cephalic vein (n ¼ 1), left internal jugular venous puncture (n ¼ 2), and left subclavian venous puncture (n ¼ 2). **Left internal jugular venous cutdown (n ¼ 1), left external jugular venous puncture (n ¼ 1), and left internal jugular venous puncture (n ¼ 1). ***Right internal jugular venous cutdown (n ¼ 1), right internal jugular venous puncture (n ¼ 1), right external jugular venous puncture (n ¼ 1), and transcatheterization of a previous central venous access in the right internal jugular vein (n ¼ 1).
DISCUSSION
under fluoroscopy or ultrasound guidance.9e11 A similar observation was made in the pediatric population, with an incidence of pneumothorax of 1.6% in 824 patients undergoing 1039 central venous catheter implantations.12 Regarding TIVAD placement, similar low rates of immediate complications have been reported, suggesting that routine control X-ray is not mandatory. Pneumothorax has been seen in 0e3.4% of TIVAD insertions, most often performed with venous puncture.4,5,13,14 Catheter malposition has been observed in
Background: The aim of this study is to determine whether systematic postoperative chest X-ray is required after totally implantable venous access port device (TIVAD) placement under fluoroscopic control. Methods: A retrospective chart review of all consecutive patients with fluoroscopy-guided TIVAD insertion from July 10, 2009 to April 16, 2012 was conducted at the Geneva University Hospitals (n ¼ 927). Patients with an available postoperative chest X-ray were included, regardless of approach (open or percutaneous) and venous access site (subclavian, cephalic, jugular, etc.). Exclusion criteria were incomplete data and preexisting pneumothorax or hemothorax. Results: Eight hundred ninety-one patients were included. First-intention venous cutdown was performed in 878 patients (98.5%), with success rates of 79.4% and 88.2% when targeting the left and right cephalic veins, respectively. Percutaneous access was the chosen first-intention procedure for 12 patients (1.3%). Eight-hundred thirty-six (93.8%) insertions were performed only by the open approach and 53 (5.9%) implantations required at least one venous puncture. Two implantations were performed using previous central venous accesses. Immediate complications associated with TIVAD placement and detected on the postoperative chest X-ray consisted of 1 asymptomatic pneumothorax, 1 symptomatic hemothorax, and 2 malpositions of the catheter. One additional pneumothorax was discovered during the first night after TIVAD insertion in a patient who became symptomatic. Conclusions: The very low incidence of immediate complications detected by postprocedural chest X-ray suggests that such a control is not mandatory as a routine method after fluoroscopyguided TIVAD insertion mainly performed by venous cutdown. X-ray should be performed only in cases of clinical suspicion.
INTRODUCTION
The two first authors contributed equally to this study. 1 Division of Visceral and Transplantation Surgery, University Hospitals of Geneva, Switzerland. 2
Division of Radiology, University Hospitals of Geneva, Switzerland.
Correspondence to: Theodoros Thomopoulos, MD, Services de chirurgie viscerale et transplantation, Departement de chirurgie, H^opitaux Universitaires de Geneve, Rue Gabrielle-Perret-Gentil 4, 1211 Geneve 14, Suisse; E-mail: [email protected] Ann Vasc Surg 2014; 28: 345–350 http://dx.doi.org/10.1016/j.avsg.2013.08.003 Ó 2014 Elsevier Inc. All rights reserved. Manuscript received: April 1, 2013; manuscript accepted: August 22, 2013; published online: December 18, 2013.
First introduced in the 1980s, totally implantable venous access devices (TIVADs) have become essential in the management of a number of disorders, including chronic infection and cancer. They provide a reliable venous access for the administration of long-term antibiotics, chemotherapeutic agents, blood products, and total parenteral nutrition. The choice of the site of implantation is based on local recommendations and the physician’s skills. The most popular sites are the internal and external jugular and subclavian veins. Alternative locations include the cephalic vein in the deltopectoral groove, with the axillary vein and the lower extremity 345
346 Thomopoulos et al.
veins as last options. Catheter insertion can be performed using either an open or percutaneous approach, with or without imaging guidance.1 Short- and long-term complications associated with this proceduredoften seen as ‘‘minor’’dare a matter of concern and have a reported prevalence of 6e18%.2e6 They are often classified chronologically, with immediate complications occurring between the time of placement and the first use, and late complications occurring in the subsequent period. The most frequent immediate complications include catheter malposition, local hematoma, cardiac arrhythmia, arterial puncture, and pneumothorax, which have each been observed in up to 6% of TIVAD insertions.3e6 During the early eras, TIVADs were placed without intraoperative image guidance, using only anatomic landmarks. Later, postoperative chest X-ray was recommended to confirm the exact location of the catheter and to detect immediate complications, such as pneumothorax and hemothorax. More recently, ultrasound has been used for a more precise venous puncture, and fluoroscopy for guidance of the catheter’s tip placement. Despite these technical refinements, postoperative chest X-ray is still performed at many institutions, remaining part of the standard practice. However, use of systematic postoperative chest X-ray presents additional irradiation, is relatively expensive, may delay more important procedures, and can lead to a false feeling of safety because not all complications can be detected early after the procedure7 and not all visible complications are correctly diagnosed by nonspecialists.8 The aim of this study was to determine whether systematic postoperative chest X-ray is required after TIVAD placement under fluoroscopic control.
METHODS A retrospective chart review of all consecutive patients with fluoroscopy-guided TIVAD insertion from July 10, 2009 to April 16, 2012 was conducted at the Geneva University Hospitals. Patients with an available postoperative chest X-ray were included, regardless of approach (open or percutaneous) and venous access site (subclavian, cephalic, jugular, etc.). Exclusion criteria were incomplete data and preexisting pneumothorax or hemothorax. The study protocol was reviewed and accepted by the ethics committee at the Geneva University Hospitals. A single type of TIVAD (CelsiteÒ; B. Braun Medical) was used for all patients. This device consists of a single lumen composite access port
Annals of Vascular Surgery
with a titanium chamber, silicone membrane, and catheter. The procedure was performed under local anesthesia by general surgeons, always in the presence of an anesthesiologist for optional administration of intravenous sedation. Cefuroxime axetil was used as prophylaxis. Cephalic venous cutdown was the first choice, and subsequent implantation sites were left to the preference of the surgeon, often selecting the external jugular venous cutdown or the subclavian vein puncture. Second- and thirdchoice accesses were performed on the same side for practical reasons. The percutaneous approach was performed under ultrasonography. The catheter was systematically introduced under fluoroscopic guidance with assessment and perioperative correction of the position of its tip. The final fluoroscopic image was printed and attached to the patient chart. A cavity was then made on the pre-pectoral fascia to introduce and fix the chamber. A Gripper needle was left in the device at the end of the procedure. A postoperative chest X-ray was obtained within 30 minutes of the procedure. Interpretation was made by the surgeon. The patient was then discharged. For this study, charts were reviewed with special attention given to operative and postoperative data, looking for the nature of the access site, perioperative fluoroscopy results, and the occurrence of the three immediate complications detectable on chest X-ray: pneumothorax; hemothorax; and catheter malposition. All X-rays were retrospectively blindly reviewed by board-certified radiologists. In cases of a detected complication, the anesthesic records and the consecutive clinical notes were reviewed for a more detailed analysis. The catheters of the TIVAD were considered misplaced if they were not located within 5 cm at the junction of the superior vena cava and the right atrium. Pneumothorax and hemothorax were always considered as a complication, regardless of their dimension and or the presence of symptoms.
RESULTS Nine hundred twenty-seven TIVAD placements were performed during the study period, and 36 were excluded for the reasons listed in Table I. The remaining 891 procedures followed the implantation methods described in Figure 1. First-intention venous cutdown was performed in 878 patients (98.5%), mainly targeting the right cephalic vein, with success rates ranging from 79.4% to 92.3%. Percutaneous access was the chosen first-intention procedure for 12 patients (1.3%). Eight hundred
Vol. 28, No. 2, February 2014
Post-TIVAD chest X-ray 347
Table I. Main reasons for exclusion (36 patients were excluded) Reasons for exclusion
Number of patients
No postoperative radiography Incomplete file Preexisting pneumothorax
20 14 2
thirty-six (93.8%) insertions were performed only by the open approach and 53 (5.9%) implantations required at least one venous puncture. Two implantations were performed using the access of a previously inserted central venous catheter. Overall, immediate complications associated with TIVAD placement and detected on the postoperative chest X-ray consisted of 1 asymptomatic pneumothorax, 1 symptomatic hemothorax, and 2 malpositions of the catheter. One additional pneumothorax was discovered during the first night after TIVAD insertion in a patient who became symptomatic. It had not been detected on the immediate postoperative chest X-ray. Cases presenting with immediate complications after TIVAD insertion are described in what follows. Case 1 A 78-year-old woman, with squamous cell cancer of the right lung and a lobular carcinoma of the right breast, underwent a TIVAD placement by left cephalic vein cutdown. The insertion was performed without technical problems and fluoroscopic imaging did not reveal any complication. However, the immediate chest X-ray showed a left minimal apical pneumothorax of 7 mm. The patient remained asymptomatic. She was followed up in the recovery ward. Because of her desire to return home and while remaining nonsymptomatic, she was discharged the same day with instructions to return in case of respiratory symptoms or chest pain. Case 2 A 69-year-old man, with metastatic adenocarcinoma of the sigmoid to the lungs and the liver, was scheduled for a TIVAD insertion for palliative chemotherapy. The catheter was inserted by left cephalic vein cutdown under fluoroscopic guidance, with a good reflux at the end of the procedure. However, postoperative chest X-ray showed that the tip of the catheter was not located in the superior vena cava (SVC), but in one of its branches, probably the azygos vein. The patient was reoperated the same day for catheter reposition, with good results.
Case 3 A 58-year-old man, with lymph node recurrence of a right testicular seminoma after orchidectomy, was scheduled for a TIVAD placement for chemotherapy by right cephalic vein cutdown as first intention. At the end, the device was implanted on the third attempt by puncture of the right subclavian vein due to failure to locate the right cephalic and right external jugular veins via an open approach. The operative and radiologic records described no complication. Twelve hours after the intervention, the patient developed dyspnea and right chest pain. Chest X-ray showed a 7-cm pneumothorax. A chest tube was inserted, and the patient could thereafter be discharged after an uneventful 8-day course. Case 4 A 75-year-old man, with recurrence of a folliculartype B-cell lymphoma, underwent a TIVAD insertion by left cephalic venous cutdown on first intention. Failure to find the vein led to puncture of the left subclavian vein. The tip of the catheter was advanced to the junction of the superior vena cava (SVC) and the right atrium, but no reflux occurred after multiple attempts. The procedure was then interrupted because the patient complained of right shoulder pain. Subsequent chest X-ray demonstrated a large left pleural effusion. Computed tomography confirmed the presence of a 900-mL hemothorax without active bleeding. Due to the patient’s desire not to be drained, conservative treatment was undertaken. Stable hemoglobin values allowed discharge at day 7. Case 5 An 86-year-old woman, with ductal invasive cancer of the right breast, was scheduled for a TIVAD placement for chemotherapy. The catheter was inserted by right cephalic vein cutdown under fluoroscopic guidance, with a good reflux at the end of the procedure. Bad reflux was then reported during chemotherapy, thus prompting angiographic control. We observed malposition of the catheter, with tip placement in the left branchiocephalic vein. Due to the absence of a good reflux after a lysis protocol, the patient was reoperated for catheter repositioning. The TIVAD was replaced by left cephalic vein cutdown with a good result. A retrospective reading of the first postoperative X-ray indicated a short catheter placed very high in the SVC. This was not corrected during the first intervention, probably due to a misreading by the surgeon.
348 Thomopoulos et al.
Annals of Vascular Surgery
Fig. 1. Totally implantable venous access port device implantations performed from July 10, 2009 to April 16, 2012 in the 891 patients studied. Patients are distributed according to the chosen vascular access procedures. Primary subsets represent first-intention vascular access; secondary subsets represent second-intention vascular access; and tertiary subsets represent third-intention vascular access. Numbers in parentheses represent the proportion of patients who underwent the related procedure in relation to the preceding subset. SR indicates success rate for each implantation. *Left common femoral venous cutdown (n ¼ 1), right saphenous
venous cutdown (n ¼ 1), right internal jugular venous puncture (n ¼ 1), transcatheterization of a previous TIVAD in the right cephalic vein (n ¼ 1), left internal jugular venous puncture (n ¼ 2), and left subclavian venous puncture (n ¼ 2). **Left internal jugular venous cutdown (n ¼ 1), left external jugular venous puncture (n ¼ 1), and left internal jugular venous puncture (n ¼ 1). ***Right internal jugular venous cutdown (n ¼ 1), right internal jugular venous puncture (n ¼ 1), right external jugular venous puncture (n ¼ 1), and transcatheterization of a previous central venous access in the right internal jugular vein (n ¼ 1).
DISCUSSION
under fluoroscopy or ultrasound guidance.9e11 A similar observation was made in the pediatric population, with an incidence of pneumothorax of 1.6% in 824 patients undergoing 1039 central venous catheter implantations.12 Regarding TIVAD placement, similar low rates of immediate complications have been reported, suggesting that routine control X-ray is not mandatory. Pneumothorax has been seen in 0e3.4% of TIVAD insertions, most often performed with venous puncture.4,5,13,14 Catheter malposition has been observed in
