The protocol biopsy strategy has been criticized because of risks and marginal utility. We tested the risk.
We performed 1171 protocol biopsies in 508 patients at 6, 12 and 26 weeks after renal transplantation, as well as 499 biopsies as indicated in 429 transplant patients. Biopsies were done as an outpatient procedure using an 18- or 16-gauge automated biopsy needle followed by 4 h bed rest.
Complications were: gross hematuria 3.5%, perirenal hematomas 2.5%, arterio-venous fistulas 7.3% and vasovagal reactions 0.5%. Major complications requiring invasive procedures such as blood transfusions or urinary catheter were seen in 1% of cases. The hospitalization rate for observation was 1.9%. According to the Banff criteria of specimen adequacy, biopsies with 18-gauge needles yielded >7 glomeruli and at least one artery in 53% of cases. Changing the needle size in October 2003, those biopsies done with 16-gauge needles yielded >7 glomeruli and at least one artery in 76% of cases, while the rate of major complications did not change.
In conclusion, transplant protocol biopsies with 16-gauge needles provide better utility and similar risk as biopsies with 18-gauge needles. A 4-h recovery after biopsy appears adequate for discharge.
Protocol renal biopsies may reveal hidden pathology and allow early treatment interventions; however, the strategy has been criticized because of potential risks and unproved utility. We established an outpatient protocol biopsy program with three biopsies per patient during the 6 months after renal transplantation (1). Since our program is homogeneous and involves a small number of transplant nephrologists from a single center, our experience permits assessing these issues. We compared our outpatient protocol biopsy program to the biopsies performed in our unit for specific indications.
Patients and Methods
Since November 2000, we began a protocol biopsy program with routine renal biopsies at 6, 12 and 26 weeks. Since then, all routine biopsies in our center were performed as outpatient procedures, including patients biopsied for specific indications, generally creatinine increase or proteinuria. Our Internal Review Board (Ethical Committee) reviewed the protocol, and written informed consent was obtained from all patients. The program has been well accepted, and 94% of patients have agreed to participate in the protocol biopsy program.
Patients attend the transplant outpatient clinic in the morning. Coagulation parameters and blood counts are monitored either during the week before biopsy or the same morning. Patients receiving warfarin or heparin, stop these medications before biopsy (warfarin is stopped at least 1 week before biopsy and is replaced by low-weight molecular heparin; the latter is stopped at least 24 h before biopsy; both are restarted 24 h after biopsy provided that no bleeding complication occurred). Patients with a known coagulation disorder or patients on warfarin who cannot stop the medication days before biopsy are excluded from the protocol biopsy program or, if indicated, the patient is admitted to hospital for biopsy. However, patients receiving low-dose aspirin (100 mg/day) for severe coronary heart disease or combined pancreas-kidney transplantation continue this medication.
The biopsy is done using an 18- or 16-gauge automated biopsy needle with ultrasound guidance (18-gauge needle from October 2000 until September 2003; 16-gauge needle since October 2003). Two cores are usually obtained in protocol biopsies and one core in biopsies on indication. One of the two cores of the protocol biopsies, the smaller one, is snap-frozen for scientific investigation without histologic evaluation. After the puncture, the specimen is checked for adequacy (presence of glomeruli) under the microscope. The patient is confined to bed rest for 4 h with blood pressure and urine monitoring. Ultrasound and Doppler sonography are done before and after biopsy and after the 4 h bed rest as well. During the bed rest period, the preliminary biopsy result is obtained. Gross hematuria in more than the first post-biopsy urine or hematoma in post-biopsy ultrasound greater than 3 cm diameter or any other sign of discomfort are regarded as indications for hospitalization at least overnight. Otherwise, the patients are sent home on their own recognizance.
We performed 1171 protocol biopsies in 508 patients between November 2000 and August 2004; 402 biopsies were done after 6 weeks, 410 biopsies after 12 weeks and 359 biopsies after 26 weeks. Additionally, 499 biopsies were performed in 429 outpatients as indicated during this time. There was no difference in the biopsy complication rate between protocol biopsies and biopsies on indication (Table 1).
Table 1. Biopsy complications in protocol biopsies and biopsies on indication; the differences were not statistically significant
Protocol biopsies n = 1171
Biopsies on indication n = 499
Thrombosis of the femoral vein
Not followed up
Gross hematuria was the most common complication (n = 58, 3.5%), followed by perirenal hematoma (n = 41, 2.5%) and vasovagal reaction (n = 9, 0.5%). Arterio-venous fistulae were detected in 122 cases (7.3%).
Of the 58 patients with gross hematuria, 21 were observed in the hospital overnight. Twelve required a urinary catheter. One patient received two units of packed erythrocytes. Out of the 41 patients with perirenal hematoma, eight were observed in hospital overnight. Four of these patients received transfusions. Three patients received two units each and one patient was given four units as well as two units of fresh frozen plasma. That patient had been admitted after biopsy for a non-biopsy-related reason. He was to receive a CT scan the subsequent day. During the night, he developed chest pain and was given a heparin infusion by the on-call physician for a suspected acute coronary syndrome. Subsequently, he developed a retroperitoneal hematoma that did not require operative drainage. The patient would probably have done better if he had been sent home after biopsy.
Patients on low-dose aspirin (234 of 1670, 14%) had slightly more often gross hematuria and perirenal hematoma than patients without (gross hematuria, 5.8% vs. 3%, p = 0.046; perirenal hematoma, 3.2% vs. 2.2%, ns).
Vasovagal reaction was observed in nine patients with the need for hospitalization in three cases. One patient had vomiting with slight hematemesis; Mallory-Weiss lesion was found at gastroscopy. Five patients had a transient bradycardia and hypotension after biopsy with syncope in one 21-year-old man and severe abdominal pain in another 68-year-old woman. Three patients had transient abdominal discomfort without any signs of bleeding.
One patient had an uncomplicated biopsy; however, 3 days later, his family doctor noticed thrombosis of the femoral vein on the same side as the transplanted kidney. Possibly, a pressure bandage that was applied for 4 h and a subsequent 3 h automobile trip during the same day predisposed the patient to this complication.
We detected 122 arterio-venous fistulae after transplant biopsy. They were usually very small (Figure 1). Forty-eight of 60 had resolved spontaneously when followed up by Doppler sonography several weeks later (77%). The other 62 fistulae were not followed up. Only three arterio-venous fistulae were sufficiently large to suggest possible hemodynamic consequences. However, the serum creatinine remained unchanged in these three patients.
Specimen adequacy of allograft biopsies is defined in the Banff Classification (2). At least seven glomeruli, one artery and two separate areas of cortex per specimen are needed for a biopsy to be considered as ‘adequate’ for histopathological evaluation. Specimen adequacy was a problem using the 18-gauge biopsy needle, since 44% of protocol biopsies and 45% of biopsies on indication contained <7 glomeruli, and 12% and 15% no artery. Altogether, 47% of protocol biopsies and 49% of biopsies on indication were ‘not adequate’ regarding the number of glomeruli and/or arteries according to the Banff Classification. We therefore changed to a 16-gauge needle in October 2003. Since then, the number of inadequate biopsy specimens in protocol biopsies has decreased to 23%, and in biopsies on indication to 27%. The biopsy complication rate has not changed significantly since that time (Table 2); just the number of hematoma increased without other consequences. Indication biopsies are repeated, if the biopsy yielded <3 glomeruli and/or no artery. The same is done in protocol biopsies as well, if the clinical course or the preceding protocol biopsy shows any abnormality.
Table 2. Comparison in biopsy complications between 18- and 16-gauge needle (*p = 0.04)
18-gauge needle n = 1222
16-gauge needle n = 448
Percutaneous renal biopsies using the aspiration technique were started in 1944 by Nils Alwall (3), who biopsied 13 patients but lost his enthusiasm and stopped doing so after one patient died of hemorrhage. Iverson and Brun (4) published their report on seven patients with renal biopsy in 1951, which encouraged Alwall to publish his biopsy series from 1944. After these reports, larger series of biopsies were done (5–10). Over the years, several severe complications were reported including 15 deaths (14 deaths in 20157 biopsies, 0.07%) (3,9–16).
The method became safer after the introduction of ultrasound in renal biopsy (17). Further progress was achieved by the use of an automated core biopsy system instead of the manual needle system (18). The essential advantage was that the depth of organ penetration and of obtaining tissue became standardized (19). During the following years, most nephrology centers changed their manual technique with the use of large needles (tru-cut needle, Travenol; modified Menghini needle; Franklin-Silverman or Vim-Silverman or Franklin-Vim-Silverman needle) for the automated gun biopsy needle with an 18-gauge needle.
Some authors, realizing the changing complication profile of biopsies, tried to compare the manual and automated biopsy technique in evaluating and comparing their complications. Some of the studies showed no difference in complication rate (19–21), while some showed more complications with the manual than with the automated method (22–25). Other centers reported that allograft biopsies had fewer biopsy complications than native kidney biopsies (15,23). In a principally retrospective multicenter protocol biopsy study of four transplant centers (including our center) analyzing symptomatic complications, the rate of serious complications was as low as <1% (26). In a recent study of orthotopic renal biopsies, gender, age and baseline partial thromboplastin time showed a positive predictive value for post-biopsy bleeding complications of native kidneys (27).
Renal biopsy as an outpatient procedure in transplant as well as in native kidneys has been reported in small series, mainly in pediatric patients (28–32).
However, no homogeneous group of ambulant allograft biopsies has been reported, done prospectively and sequentially three times per patient by the same team with the same biopsy technique during a relatively short period of time (4 years). The data were not only analyzed regarding symptomatic complications, but also regarding subclinical sonographic changes (number of hematomas, arterio-venous fistulae), and specimen adequacy in relation to needle size and complications.
The important findings in our study are that protocol transplant biopsies can be performed on an outpatient basis, that they engender no more complications than transplant biopsies done for specific indications, and that, taking one core, an 16-gauge biopsy needle is necessary to satisfy the Banff criteria for specimen adequacy in nearly 80% of cases.
The study as well as data from the literature cannot answer the question if taking two cores with an 18-gauge instead of one with an 16-gauge biopsy needle would result in the same specimen adequacy and, even more important, the same low biopsy complication rate. Patients on aspirin have slightly more hematomas, but no more serious complications. The conclusion is that aspirin should be stopped 1 week before biopsy with the exception of severe coronary heart disease and/or the fact of combined pancreas-kidney transplantation during the first 6 months post-transplantation.
The major complications can be defined as those requiring an invasive procedure above and beyond observation, such as blood transfusion, embolization or operation. After renal biopsy, all major complications consist of bleeding, either gross hematuria or hematoma. In our patients, the hospitalization rate was 1.9%. However, we saw only 1% major complications. The invasive procedures required were blood transfusions (0.3%) and urinary catheter placement (0.7%) to drain any clots that developed.
Comparing protocol biopsies with biopsies done on indication principally had the same complication rate in our study (Table 1); however, since in protocol biopsies two cores were taken (one snap-frozen for scientific purpose) and only one in biopsies done on indication, the complication rate in biopsies done on indication was relatively somewhat higher than in protocol biopsies. This is understandable, since biopsy of older, more fibrotic kidneys (biopsies on indication) may cause more bleeding than kidneys during the first 6 months after transplantation (protocol biopsies).
Arterio-venous fistulae were frequent, but generally harmless. Rates of arterio-venous fistulae formation after renal biopsy ranging 5–8% were observed by others (23,33). Astonishingly, the fact that arterio-venous fistulae may resolve spontaneously was observed as early as 1965, when Bennett and Wiener described nine patients with arterio-venous fistulae after native renal biopsy proven by angiography. Six of the patients were followed up several months later at which time four of the six fistulae had disappeared (34). Since protocol biopsies are done sequentially, we had the opportunity to follow most of the arterio-venous fistulae and noticed that 77% had disappeared.
Specimen adequacy is related to needle size and the number of specimen cores. Because of low specimen adequacy, we changed the 18-gauge for a 16-gauge needle in October 2003. Changing to a larger needle increased specimen adequacy significantly but biopsy complications only slightly. This state-of-affairs has also been observed by others (35,36). There was no difference in increasing specimen adequacy by changing needle size between protocol biopsies and biopsies on indication. Of course, specimen adequacy could have been increased as well by taking another core. Our specimen adequacy is difficult to compare to that cited in the literature, since most authors do not explain their definition of adequacy. Also, they do not use the definition of the Banff Classification. Specimen inadequacy mostly was due to the lack of glomeruli, not arteries.
Reviewing the complications in 1670 outpatient allograft biopsies, we conclude that ambulant renal transplant biopsies are feasible, since the number of major complications was low and those that occurred were of a benign nature. Furthermore, the complications we observed usually became apparent during the observation time of 4 h after biopsy. That means that protocol biopsies have a low risk and can be done as an outpatient procedure.