ARTICLE |
a Pediatric Hematology/Oncology
b Neonatology, Children's Hospitals and Clinics of Minnesota, St Paul, Minnesota
c Division of Pediatric Nephrology, Department of Pediatrics, University of Minnesota, Minneapolis, Minnesota
| ABSTRACT |
|---|
|
|
|---|
PATIENTS, INTERVENTIONS, AND OUTCOME MEASURES. Twenty-eight newborns with renal venous thrombosis were treated at Children's Hospitals and Clinics of Minnesota and Fairview University Medical Center from 1991 to 2001.
RESULTS. Unilateral involvement was noted in 25 neonates (89%) and bilateral involvement in 3 (11%). Unilateral renal venous thrombosis affected mostly term infants, whereas 2 of 3 infants with bilateral renal venous thrombosis were <32 weeks' gestational age (birth weight: 745–1505 g). One mother had antiphospholipid syndrome. Of 11 neonates evaluated for congenital thrombophilia, 1 had the factor V Leiden mutation. Ten neonates received either unfractionated or low molecular weight heparin for 3 days to 7 months. Three infants with unilateral renal venous thrombosis treated with heparin alone did not seem to benefit from such therapy. Seven neonates were treated with fibrinolytics and unfractionated heparin (4 neonates with unilateral renal venous thrombosis and 3 with bilateral renal venous thrombosis). Treatment with fibrinolytics did not result in restoration of renal function in the 4 neonates with unilateral renal venous thrombosis but was associated with return of almost normal function in the 2 neonates with bilateral renal venous thrombosis who received fibrinolytics immediately after renal venous thrombosis diagnosis. Two neonates treated with fibrinolytics suffered bleeding complications at the area of adrenal hemorrhage. Two neonates (not treated for renal venous thrombosis) died as a result of underlying disease.
CONCLUSIONS. Most neonatal renal venous thrombosis is unilateral and does not respond to fibrinolytic therapy and heparin. The use of fibrinolytics may prevent chronic renal failure in neonates with bilateral renal venous thrombosis if begun within 24 hours of diagnosis. Fibrinolytic therapy, however, is associated with a risk of bleeding, specifically if there is an associated adrenal hemorrhage.
Key Words: renal venous thrombosis neonate fibrinolytics heparin
Abbreviations: RVT—renal venous thrombosis rtPA—recombinant tissue plasminogen activator
Renal venous thrombosis (RVT) in neonates has been associated clinically with hyperosmolar state after angiocardiography, coagulopathy, sepsis, shock, acute blood loss, diarrhea, dehydration, maternal diabetes, sickle cell disease, birth asphyxia after maternal ingestion of thiazides, and neonatal infection.1 Although the factor V 1691G
A (factor V Leiden) mutation has been associated with RVT2,3 and with the development of RVT in utero,3 and maternal lupus anticoagulant has been associated with RVT of the fetus,4 the underlying mechanism for the development of RVT is unknown.1 A major issue in neonatal RVT is preservation of renal function. Both heparin and fibrinolytics have been used with varying success. Heparin was found to be effective in preserving renal function in a large retrospective study,3 but this was not confirmed in a small prospective study.5 Fibrinolytics such as urokinase and recombinant tissue plasminogen activator (rtPA) have been effective in several case reports of RVT,6,7 but results from a French series failed to show an improvement in outcome.8
With the availability of fibrinolytic agents, there has been a greater interest in treating RVT with these drugs. The recent experience with RVT in a cohort of patients reported here suggest that fibrinolytic therapy failed to rescue the affected kidney in patients with unilateral RVT, perhaps because of a delay in diagnosis and treatment. In contrast, patients with bilateral RVT are generally diagnosed early and, thus, have the potential to respond to fibrinolytics with recovery of renal function.
| PATIENTS AND METHODS |
|---|
|
|
|---|
Patients were either diagnosed with RVT in the NICUs of the above-named institutions or were transferred for additional management after diagnosis at an outside institution. In all patients, the diagnosis of RVT and associated clots was confirmed with ultrasound/Doppler using previously described criteria.9 Patients were divided into 2 groups: unilateral RVT and bilateral RVT. Because criteria for response to heparin or fibrinolytic therapy are poorly established, we evaluated the effect of therapy on renal outcome at the last documented follow-up with renal ultrasound and, for some patients, renal scintigraphy. The use of central venous and arterial catheters before the diagnosis of RVT was recorded. Anticoagulant activity levels of antithrombin III, protein S, and protein C were performed using established methods at the Special Hematology Laboratory of the University of Minnesota. Factor V 1691G
A (factor V Leiden) and Prothrombin G-A 20210 mutation were evaluated using the polymerase chain reaction as previously reported.10
The decision to use fibrinolytics and/or heparin was made by the attending physician for each child, in consultation with the hematologist. Only 10 patients received therapy. In the absence of data, fibrinolytic therapy was initiated in an attempt to salvage the involved kidney. Heparin was used to prevent propagation of clot and/or in an attempt to improve renal outcome. Similarly, the treating team determined the length of anticoagulation therapy. Specific criteria used to determine the length of therapy were not available.
| RESULTS |
|---|
|
|
|---|
|
Unilateral RVT affected the left kidney 1.5 times more frequently than the right kidney. Concomitant adrenal hemorrhage occurred in 9 (32%) of 28 patients (Table 1). It was associated more commonly with unilateral left RVT (5 of 9), although 2 patients had bilateral adrenal hemorrhage (both associated with unilateral RVT). An inferior vena cava clot was detected in 11 (44%) of the patients with unilateral RVT; inferior vena cava clots occurred in 2 patients with bilateral RVT; and a clot in the aorta was noted in 1 patient with bilateral RVT (Table 1).
Predisposing Conditions: Maternal
Maternal antiphospholipid antibody was associated with 1 case of unilateral RVT. Maternal diabetes was present in 5 mothers (18%). Delivery for 16 patients was by cesarean section because of fetal distress (7 patients), placenta abruptus (2 patients), and other unspecified reasons (7 patients). Delivery was by vacuum extraction for 2 patients (1 unilateral, 1 bilateral RVT).
Thrombophilia
Only 1 of 8 patients evaluated for the presence of factor V 1691G
A (factor V Leiden) was heterozygote. This patient was diagnosed with unilateral RVT in utero; 2 other patients diagnosed in utero were not evaluated for thrombophilia. Prothrombin G-A 20210 mutation was not found in any of the 5 patients evaluated. Protein C and S deficiencies were evaluated in 11 patients: 1 infant had low factor C (chromogenic = 13%; [reference: 35% ± 9%]; antigen = 34%), but no follow-up levels were obtained. None of the 10 patients evaluated were found to have antithrombin III deficiency.
Comorbid Conditions
Unilateral RVT diagnosed early in life (in utero or within 3 days of birth) was associated in 5 patients with severe comorbid conditions: primary pulmonary hypertension of the newborn, congestive heart failure, severe respiratory distress syndrome, and hydrops fetalis. Of the 5 infants who developed unilateral RVT >5 days after birth, 3 had severe comorbid conditions including primary pulmonary hypertension of the newborn, congestive heart failure, severe respiratory distress syndrome, and multiorgan failure leading to death in 2. These were the only deaths in this series. Bilateral RVT occurred in 1 patient with severe respiratory distress syndrome, chronic lung disease on diuretics, and dehydration.
Before the diagnosis of RVT, central vascular lines were in place in 7 (28%) of the patients with unilateral RVT and in 2 of the patients who developed bilateral RVT. These included umbilical arterial catheters in 4 patients, umbilical venous and arterial catheters in 2 patients, femoral lines into the inferior vena cava in 2 patients, and multiple central lines in 1 patient on extracorporeal membrane oxygenation.
Treatment
Fibrinolytic Therapy
Fibrinolytics were used in 7 patients, as shown in Table 2. Of the 4 patients with unilateral RVT, 3 received urokinase and 1 received rtPA. In these patients, therapy was initiated within 24 hours of identifying the RVT. One patient (U-10) was diagnosed with RVT on day 12; at that time the inferior vena cava clot was already calcified and there was no response to rtPA, indicating an old thrombosis. Fibrinolytic therapy was associated with restoration of renal blood flow in 2 patients and improvement in blood flow in 1 patient, as measured by ultrasound/Doppler. However, renograms on all 4 patients revealed abnormal kidneys 2 to 6.5 years after treatment (Table 2).
|
Bleeding complications were seen in 2 of the 7 treated patients. The first had unilateral RVT treated with urokinase and suffered a localized bleed to the ipsilateral adrenal gland and perirenal fascia and a small peritoneal bleed. The second, with bilateral RVT treated with rtPA, suffered a retroperitoneal bleed in the area of the right adrenal gland. The bleeding in both patients resolved spontaneously on discontinuation of fibrinolytics, requiring no surgical intervention, fresh-frozen plasma, or packed red blood cell transfusion.
Heparin Therapy
Heparin was used in 10 patients (36%; Table 2). The 7 patients treated with fibrinolytics also received low doses (15–25 U/kg per hour) of unfractionated heparin. Three patients were treated with heparin but received no fibrinolytics; therapy was initiated within 24 hours in 2 patients and only after 10 days in 1 patient (Table 2). Unfractionated heparin dose was monitored with a target thrombin time of 4:1 normal or activated prothrombin time of 60 to 85 seconds. When the anti-Xa assay became available, unfractionated heparin was given with a target level of 0.35 to 0.7 U/mL. Low molecular weight heparin was monitored using the anti-Xa assay; when administered as 1.5 mg/kg per 12 hours, the treatment target was 0.5 to 1, and lower doses targeted a prophylactic level of 0.2 to 0.4. Low molecular weight heparin was used either alone or after unfractionated heparin for up to 7 months. As noted in Fig 1, heparin therapy failed to improve renal outcome in the patients with unilateral RVT (Fisher's exact test, 2-sided: P = .52). Of the 19 patients with unilateral RVT not treated with anticoagulation, only the patient with maternal antiphospholipid antibody experienced a neonatal stroke; no other embolic phenomena occurred. Although 2 of these 19 unilateral RVT patients ultimately died as a result of primary pulmonary hypertension, none suffered from extension of the clot in the cava or other vascular obstruction outside the kidney.
|
|
Overall survival was 93%. Two deaths occurred in patients diagnosed with unilateral RVT at 5 and 28 days of life. These deaths were ascribed to the underlying comorbid conditions rather than to RVT.
| DISCUSSION |
|---|
|
|
|---|
RVT affects twice as many males as females1,11–14; most patients have unilateral involvement.3,13–15 The left kidney seems to be most commonly affected,3,14,15 although there is disparity in the literature.1,13 We and others found that adrenal hemorrhage is predominantly associated with RVT affecting the left kidney. This predilection is most likely because the left adrenal vein drains into the left main renal vein, whereas the right adrenal vein drains directly into the inferior vena cava. Almost two-thirds of the neonates with unilateral RVT were delivered by cesarean section, suggesting an etiologic relation to stress immediately preceding or associated with the procedure. The contribution of central vascular catheters to the development of RVT is not clear. Twenty-eight percent of our patients had central catheters, mostly umbilical arterial catheters. The Canadian registry noted central venous lines in 17% of 72 infants with RVT.14
All neonates with RVT should be screened for prothrombotic risks, including maternal lupus inhibitor. In the largest series of patients with RVT evaluated for thrombophilia, 40 (67%) of the 56 patients had at least 1 prothrombotic risk factor, compared with 11.9% in the control group (odds ratio: 15.6; 95% confidence interval: 7.2–34.2).9 Factor V 1691G
A was found in 37.3% of the patients with RVT, compared with 5.9% in the control group, and multivariate analysis retained an association only with factor V heterozygous mutation and Lipoprotein (a).9 A second study found a high correlation between factor V 1691G
A and RVT in 31 newborns (odds ratio: 10.9; 95% confidence interval: 3.85–31.1; P < .0001).16 Activated protein C resistance was found in 8 (38%) of 21 newborns. Fifty percent of the patients screened for prothrombotic risks were positive.15 In our study, 1 patient (12.5%) of 8 was heterozygous for factor V 1691G
A mutation, similar to the reported mutation rate of 3% to 12% in the white population.17
The use of fibrinolytics in neonatal RVT is controversial. Beaufils et al8 reported that fibrinolytics used in 18 neonates for 24 to 96 hours resulted in a "cure" in 3 (37.5%) of 8 with unilateral RVT and 2 (20%) of 10 with bilateral RVT. In contrast, a second study reported no significant effect on outcome with intravenous therapy in 11 neonates.9
In the present study, renal blood flow improved with the use of fibrinolytics in 3 of our unilateral patients with RVT, and renal function was restored in 2 of our bilateral patients with RVT (Table 2), suggesting that the doses of fibrinolytics used and length of therapy were adequate. In all patients treated with fibrinolytics, therapy was initiated within 1 day of the diagnosis of RVT. However, in 1 case (U-10), RVT was identified at 12 days of life at which point the IVC thrombus was calcified, suggesting that the RVT was identified too late for effective therapy. Similarly, patient B-3 was born with renal failure and the RVT was identified on day 2; it is likely that therapy was initiated too late to be effective. Renal outcome was poor in both of these patients.
Our data suggest that fibrinolytic therapy has no effect on the long-term outcome of the affected kidney in patients with unilateral RVT, because 3 of our patients had an atrophic kidney and 1 had a small kidney on follow-up. We speculate that most cases of unilateral RVT occur either in utero or at birth and are not recognized early enough to allow prompt initiation of fibrinolytic therapy in an attempt to prevent permanent renal damage. It is also possible that the location of the thrombus or specific physiologic factors prevent fibrinolytic therapy from altering the outcome. In contrast, the 2 patients with bilateral RVT regained renal function after fibrinolytic therapy, although treatment was associated with residual loss of parenchyma. The 7th American College of Chest Physicians' Conference on Antithrombotic and Thrombolytic Therapy in Children has recommended therapy with unfractionated heparin and thrombolytics only for patients with bilateral RVT and various degrees of renal failure.18 (recommendation 1.3.3)
Fibrinolytic therapy was associated with minor bleeding complications in 2 of our patients, requiring no intervention and resolving spontaneously after discontinuation of fibrinolytics. Interestingly, in both cases the bleeding was related to adrenal hemorrhage. Zenz et al19 report that fibrinolytic therapy was associated with bleeding complications in 13% of preterm infants, and in up to 25% of neonates treated during the first week of life. Intraparenchymal and intraventricular hemorrhage was reported in a preterm infant with bilateral RVT treated unsuccessfully with fibrinolytic and anticoagulation therapy.20
Heparin therapy has been recommended for neonates with unilateral RVT and clot formation in the inferior vena cava to prevent pulmonary embolism and extension of the thrombosis.18,21 In the present study, 3 patients with unilateral RVT received no anticoagulation but recovered spontaneously, whereas all unilateral RVT patients treated with heparin with or without fibrinolytics had a poor renal outcome (Table 3, Fig 1). Although a protective effect for heparin has been reported,3 heparin therapy in this study and others5,9,15,22 did not preserve renal function. Moreover, heparin therapy may increase bleeding complications.5
| CONCLUSIONS |
|---|
|
|
|---|
| ACKNOWLEDGMENTS |
|---|
We thank Gretchen Williams for her assistance in the preparation of this report.
| FOOTNOTES |
|---|
Address correspondence to Yoav Messinger, MD, 345 N Smith Ave, St Paul, MN 55102. E-mail: yoav.messinger{at}childrensmn.org
The authors have indicated they have no financial relationships relevant to this article to disclose.
| REFERENCES |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||