Thermoregulatory catheter–associated inferior vena cava thrombus

Abstract

The use of thermoregulatory catheters (TRCs) in critically ill patients has become increasingly popular. TRCs have been shown to be effective in regulating patient body temperature with improved outcomes. Critically ill patients, especially multitrauma patients and those with femoral catheters, are at high risk for deep vein thrombosis (DVT). Among patients for whom chemical DVT prophylaxis is not an option, inferior vena cava (IVC) filters are often placed prophylactically. The development of intravascular ultrasound (IVUS) has allowed placement of IVC filters at the bedside for patients who are too ill for transport to the operating room or cardiac catheterization lab. After encountering several patients with occult DVT of the IVC during bedside IVC filter placement, we performed a retrospective review to determine the incidence of DVT or pulmonary embolus (PE) in patients who had been treated with a TRC at Baylor University Medical Center at Dallas. Since 2008, IVC filters have been deployed at the bedside with the use of IVUS at Baylor University Medical Center. During that same time period, 83 patients had a TRC placed for either intravascular warming or cooling during their resuscitation. Forty-seven out of 83 patients who had a TRC placed survived their injuries. Ten of 47 patients (21%) were diagnosed with DVT or PE, and 6 of these 10 (60%) were found to have caval thrombus. We present this case series as evidence that undiagnosed IVC thrombus associated with TRCs may be higher than previously suspected, given that 5 out of 10 patients who had IVUS of their IVC for prophylactic IVC filter placement, as well as one patient diagnosed with PE, were found to have caval thrombus.

Within the last 15 years, the use of thermoregulatory catheters (TRCs) has gained popularity. They have been used to induce hypothermia to improve outcomes in cases of cardiopulmonary arrest and to reverse the harmful effects of hypothermia in the trauma patient by providing a means for rapid rewarming (1–3). Over the last 3 years at our institution, the trauma service has been utilizing the Alsius catheter and Coolguard Icy thermoregulatory system to aid in resuscitation of hypothermic patients as well as in the cooling of patients with fever and traumatic brain injury. During that same period of time, the vascular surgery service has been placing bedside inferior vena cava (IVC) filters in critically ill patients using intravascular ultrasound (IVUS) (4, 5). Some of the patients who had TRCs used during their hospital course also had IVC filters placed either for prophylaxis or after a diagnosis of deep venous thrombosis (DVT) or pulmonary embolus (PE). Surprisingly, caval thrombus was found in several of these patients undergoing placement of an IVC filter with IVUS. Currently, only one series has examined the risk of iliofemoral DVT (6), and only one case has been reported of vena cava thrombus associated with the use of a TRC (7). We performed a retrospective review to examine whether trauma patients who have been exposed to a TRC are at additional risk for iliocaval DVT in addition to the risk of DVT associated with femoral vein catheterization.

MATERIALS AND METHODS

Institutional review board approval was obtained for our study. Patient records were obtained from the trauma registry at Baylor University Medical Center at Dallas to identify patients who had TRCs placed beginning in 2008. Catheterization lab records were reviewed to identify patients receiving IVC filters during the same time period. The Student's t test was used for statistical comparison of age and injury severity score (ISS).

RESULTS

Since 2008, 83 trauma patients have had a TRC placed as part of their postinjury care: 47 of those patients, with an average age of 41 years and an ISS of 20.9, survived their initial injuries, and 32 patients had no diagnosis of DVT/PE, nor were they selected for prophylactic IVC filter placement. Fifteen of the 47 were referred for IVC filter placement. Five of these 15 patients were diagnosed with either DVT or PE prior to vascular referral. Four patients received an IVC filter; of them, three underwent filter placement under fluoroscopy, and the other patient who received a filter secondary to PE had it placed with IVUS guidance. This demonstrated a caval thrombus. The fifth patient was found to have a femoral DVT and was treated with anticoagulation.

Ten patients underwent prophylactic placement or attempted placement of an IVC filter at bedside with IVUS, four of whom were discovered to have IVC thrombus at the time of filter placement. A fifth patient was diagnosed with caval thrombus by IVUS, but had too extensive a thrombus to allow for IVC filter placement. He was treated with anticoagulation. The remaining five patients did not have caval thrombus detected by IVUS (Figure 1).

Figure 1.

Diagnosis of DVT in patients receiving a thermoregulatory catheter. DVT indicates deep vein thrombosis; IVC, inferior vena cava; IVUS, intravascular ultrasound; PE, pulmonary embolus.

The average age for patients with DVT, PE, or vena cava thrombus was 28 years. This was less than the average age of the overall group (41) but was not statistically different (P > 0.05). The ISS of the patients in the DVT, PE, or vena cava thrombus subgroup was 33. This was much higher than in the overall group (ISS = 21) treated with TRCs who survived their initial injuries, and this difference reached statistical significance (P = 0.039).

DISCUSSION

TRCs have been shown to be an effective clinical tool, whether to improve outcomes when used for corporal cooling after cardiac arrest or to efficiently reverse hypothermia (1–3). However, there is little reported evidence addressing potential complications. Specifically, the potential to form venous thromboembolism (VTE) has only been reported in one study (6) and in a single case report (7).

The patient population examined in this study was at high risk for DVT/VTE. Because of their associated injuries, most of these patients could not receive chemical DVT prophylaxis. Critical illness with contraindications to chemoprophylaxis carries a 7% DVT risk (8). DVT/VTE associated with femoral vein catheters ranges from 10% to 25% (9, 10). In looking at our institutional experience with TRCs by the trauma service, we found that the overall rate of DVT/VTE formation was 21% (10/47). This is at the high end of the spectrum, but not out of line with what has been previously published. A previous study of DVT formation in patients with TRCs showed a 50% rate of DVT formation (6).

We found occult caval thrombus in 60% of patients ultimately diagnosed with DVT. Given the high incidence of occult caval DVT discovered at the time of prophylactic filter placement, we feel that the 21.3% rate of DVT (specifically caval DVT) in this study may be an underestimation. Only 10 of 47 patients underwent cavography or IVUS. The status of the IVC was not evaluated in 32 of the 47 patients in this series, and a significant number of caval DVT cases could have gone undetected.

We did find a higher ISS in patients with DVT/PE (33 vs. 21, P = 0.039) compared to the overall group. This could indicate that injury severity contributes to DVT formation. However, we feel that this possibly represents a selection bias, as the more critically ill patients are typically selected to receive prophylactic IVC filters. Only by studying every patient receiving a TRC can the true rate of DVT and caval DVT be determined and the effect of age and ISS be honestly assessed.

There are two points of concern regarding the DVT/VTE associated with these catheters. The first is the fact that half of the caval DVTs were found in asymptomatic patients. Therefore, the actual rate of DVT formation could be as high as 50% in our studied patient population (similar to the experience of Simosa et al). Since 60% of patients having prophylactic IVC filters placed were found to have IVC DVT, the use of TRCs may increase the incidence of DVT significantly higher than the baseline rates for critically ill patients who typically develop DVT in the lower extremity veins or in the femoral vein if a catheter has been placed there. The second point of concern is that these proximal DVTs are much more dangerous, as they are associated with a higher mortality rate when compared with more distal DVTs (11). They are also less likely to be detected by surveillance, symptoms, or noninvasive imaging. The increased incidence of caval DVT with TRCs identified by our study is therefore quite worrisome.

Although there has not been a comparison of IVUS and venography for thrombotic occlusion, IVUS has been shown to be more sensitive than venography in detecting nonthrombotic lesions (12, 13). This increased sensitivity may translate to the diagnosis of thrombotic lesions as well. Caval DVT in patients from TRCs can be nonocclusive and attached to the wall of the IVC and may not be obvious on venography. However, IVUS readily identifies these clots (Figure 2). Notably, the three patients in our series who had fluoroscopy only when having their IVC filters placed could have had missed caval thrombus. We believe that the use of IVUS during the placement of IVC filters increased the sensitivity in identifying occult caval DVTs in asymptomatic patients. Therefore, we recommend the use of IVUS to detect occult DVT in patients with a history of femoral TRCs.

Figure 2.

Intravascular ultrasound image depicting inferior vena cava thrombus.