Although generally considered safe, there is a morbidity and small, but measurable, mortality associated with the procedure, so it is imperative to subject only those patients in whom there will be a potential benefit to these risks. Indications for renal biopsy may vary from one center to another, but accepted indications are listed in Box 25.1 . The significant complications related to the procedure are hemorrhage, development of arteriovenous fistulas, and to a lesser extent sepsis. First, bleeding with macroscopic hematuria and the development of perinephric hematomas may be minor and self-resolving or major and require intervention in the form of blood transfusions, intravascular embolization, or rarely surgery (1.2%). Second, there is a risk of formation of arteriovenous fistulas (up to 14%), ^, which may be asymptomatic and spontaneously resolve or lead to a significant vascular steal syndrome, compromising the rest of the kidney through ischemia. Finally, there is the risk of sepsis following the procedure, through the introduction of a septic focus or its dissemination. Overall, the risks of complication vary from center to center and between practitioners but can be estimated between 3.5% and 13%, with the majority being minor complications (approximately 3%–9%), although this appears to have decreased in recent years. ^, Mortality from the procedure is generally a result of undiagnosed bleeding with significant hematoma formation and was reported in up to 0.2% of cases from some of the larger biopsy series, although other studies suggest that it represents an extremely rare adverse event. Some degree of bleeding is common, as approximately half of patients have a small drop in hemoglobin post biopsy, and one third will develop some hematoma, but only in a minority (≤7%) will bleeding be significant and require intervention. Complications appear to be more common in native than transplant kidneys and in patients with more advanced renal impairment, with prolonged bleeding times, or with lower hemoglobin. ^, In a recent retrospective study (1995–2015), complications after 1705 percutaneous native and transplant kidney biopsy were compared revealing a greater drop in hemoglobin (0.97 vs. 0.73 d/dL), a higher complication rate (6.5% vs. 3.9%), and higher transfusion rate (5.2% vs. 3.3%) after native kidney biopsies. One prospective study identified the only risk factors for bleeding complications as being female, younger (35 ± 14.5 years vs. 40.3 ± 15.4 years), and having a prolonged partial thromboplastin time. This finding was recently confirmed in a retrospective analysis of the Boston Kidney Biopsy Cohort, which revealed that the post-biopsy transfusion risk was associated with baseline hemoglobin level and female sex; estimated GFR and baseline hemoglobin were independently associated with a larger drop in hemoglobin. Other retrospective univariate analyses have reported that blood pressure of 160/100 mm Hg or higher or a serum creatinine of greater than 2 mg/dL more than doubled the risk of bleeding. Overall, however, no effective means has been established to identify those individuals at risk of developing “clinically” significant complications, while risk calculators (e.g.,

Bleeding after kidney biopsy

) have been suggested to assess the personalized risk, their use has not been evaluated in RCTs. In one small series, post-biopsy ultrasound within an hour had a 95% negative predictive value for predicting clinically significant hemorrhagic complications, meaning that the absence of a hematoma on the post-biopsy scan was suggestive of an uncomplicated clinical course.

Box 25.1

Indications for renal biopsy

• •

  Proteinuria (significant proteinuria (>1 g/day or protein-to-creatinine ratio >100 mg/mmol)

• •

  Acute nephritic syndrome (microscopic hematuria with any degree of proteinuria)

• •

  Unexplained reduction in kidney function (native or transplant kidney)

• •

  Renal manifestations of systemic disease

Debate continues regarding the routine use of DDAVP to counteract uremic bleeding tendencies. In part, this is because its use was previously reserved for only those patients with prolonged bleeding times, and numerous studies have since demonstrated that complication rates are no different if bleeding time estimation is omitted from the preoperative assessment, ^, as it does not predict clinical complications. However, more recent data from a randomized double-blinded trial suggested a significant benefit in preventing bleeding complications with few adverse events. A total of 162 low-risk adult patients undergoing biopsy were enrolled and randomized to subcutaneous DDAVP (0.3 μg/kg) or placebo. The patients were normotensive and had preserved renal function with serum creatinine of <1.5 mg/dL (estimated glomerular filtration rate >60 mL/min). Those who received DDAVP evidenced a significant reduction in incidence of post-biopsy bleeds (30.5% vs. 13.7%; relative risk 0.45), a significant reduction in hematoma size in those who did bleed, and a reduction in duration of hospital stay. However, hemoglobin drop after biopsy was minimal and there were no major complications, leading some to question the benefit of reduction in clinically unimportant hematomas, which can be frequently found following biopsy if looked for. No thrombotic, hyponatremic, or cardiovascular events were recorded. Whether these data, in patients with preserved renal function, could be translated to those higher-risk patients with greater renal impairment is unclear and is a question worthy of a further randomized trial.

Many centers stop antiplatelet therapy prebiopsy for elective procedures, but recent data suggest that bleeding rates may be no different in those taking aspirin or stopping a week beforehand, and continuing treatment may avoid the increased risk of cardiovascular events following aspirin withdrawal. In a center that does not routinely stop aspirin (but does stop clopidogrel), a retrospective analysis of 2563 biopsies revealed a major bleeding complication in only 2.2%, and in those in whom a complete drug record was available, no significant difference was noted on or off aspirin. Limited data are available on biopsies performed on patients taking clopidogrel.

Along with developing procedure-related complications, there is the chance that an inadequate core of tissue is obtained for diagnosis, containing too few glomeruli or insufficient cortical material, and this is reported in between 1% and 5% of cases. The size requirements for accurate diagnosis are discussed later.

Some absolute contraindications preclude percutaneous biopsy, whereas relative contraindications ( Table 25.1 ) may be tolerated depending on the importance of the biopsy, operator’s experience, and supportive facilities available. Ideally, all efforts should be made to address relative contraindications; however, in the context of acute kidney injury and rapidly progressive glomerulonephritis or suspected transplant rejection, this may not always be possible. With modern techniques, evidence is emerging that previously perceived high-risk factors such as obesity, plasma cell dyscrasias such as myeloma, or amyloidosis are not actually associated with higher rates of bleeding complications. The critical preoperative steps are to ensure that blood pressure is controlled, the patient does not have a bleeding diathesis or a urinary tract infection, and the kidneys are suitably imaged, with no evidence of obstruction, widespread cystic disease, or malignancy (although percutaneous biopsy is increasingly used to diagnose the nature of renal masses). As a result, preprocedural assessment should allow those patients unsuitable for percutaneous biopsy to be referred for an alternative approach ( Table 25.2 ). In these patients, there are other means of obtaining renal tissue, which include open biopsies, laparoscopic biopsies, or transjugular biopsies. Each is associated with specific complications and has particular merits depending on the clinical scenario (see Table 25.2 ). These alternative approaches are generally only required for a minority of patients requiring biopsies.

The safe duration for observation following renal biopsy has been investigated in numerous studies, which suggest that early discharge (after only 4-hour observation) will result in a number of missed complications, with many more occurring between 8 and 24 hours post procedure. Even after 8 hours, 23% to 33% of complications will be missed. However, an overnight stay will allow an extra 20% of complications to be identified before discharge with between 85% and 95% of complications being identified at 12 hours and 89% and 98% following 24-hour observation. ^, Some units practice a policy of day biopsies with a minimum 6-hour bed rest period, which is extended only if there is evidence of bleeding, and this appears to be associated with no increased complication rates. Vigilant observation of blood pressure, pulse rate, and evidence of hematuria is required in all cases.

Detailed descriptions of methods of handling biopsies can be found in many publications. A full assessment of the renal biopsy requires examination by light microscopy, immunohistochemistry, and electron microscopy (EM, also referred to as “triple diagnostics”), with the use of other tests in some circumstances. Light microscopy and immunohistochemistry are required for all cases, whereas EM is required for some diagnosis but could be omitted in other cases (see later). Therefore it is necessary to divide the biopsy tissue for each of these methods of examination. During this process, it is extremely important that the tissue is not damaged by handling, or by drying, and that the tissue is fixed in an appropriate fixative as quickly as possible, ideally within minutes. This is best achieved by dividing the biopsy at the bedside. Examination of the biopsy with a dissecting microscope allows the cortex, containing glomeruli, to be distinguished from medulla and nonrenal tissue (such as fat) ( Fig. 25.1 ). This facilitates assessment of the adequacy of the cores and division of the biopsy so that glomeruli are present in the samples for each modality of examination. If a dissecting microscope is not available, then a standard light microscope can be used with the biopsy placed in a drop of normal saline on a microscope slide. If it is not possible to examine the biopsy in this way, then a standard approach to obtain material for EM is to take small fragments (≈1 mm in length) from each end of each core. In that way, if there is cortex in the core, glomeruli should be sampled. The remainder of the cores can be either all fixed or divided for light microscopy and immunofluorescence (IF). The part of the biopsy for light microscopy is then placed in appropriate fixative, most often formalin, and the part for IF is either snap frozen or transported to the laboratory in a suitable transport medium such as that described by Michel and colleagues. Tissue placed in this medium can remain for several days at room temperature without loss of antigens. During division of the biopsy, it is important not to introduce artifacts due to crushing or stretching. Forceps should not be used to pick up the specimen; small polyester cleaning swabs are useful, but if not available, gentle use of a needle or a small wooden stick such as a toothpick is also possible. The biopsy should be cut using a fresh scalpel. If the biopsy must be taken to the histology laboratory for division, this should be done as quickly as possible with the biopsy wrapped in saline-moistened gauze or tissue culture medium. Artifacts may be produced if the biopsy is placed on dry gauze or gauze moistened with water or if it is placed in ice-cold saline. Some nephropathology laboratories use immunohistochemistry on formalin-fixed tissues, not requiring sample division, with a single biopsy core being sufficient for the diagnosis in most cases. It is also possible to take samples for EM from formalin-fixed biopsies such that the whole biopsy core is sent in one vial with formalin and all further workup is done by the nephropathology laboratory. This makes the process easier for the biopsy-performing centers, particularly those without in-house pathology.

If the amount of material obtained at biopsy is limited, then it may be necessary to adapt the way in which it is divided. The decision as to how this is done must depend on the clinical question. This procedure is preferentially performed by the nephropathology laboratory due to experience in tissue handling. In most cases, it is possible to omit frozen material for IF and instead perform immunohistochemistry on paraffin sections. However, if there is a suspicion of crescentic glomerulonephritis due to antiglomerular basement membrane (anti-GBM) disease, IF is more reliable in detecting the linear capillary wall staining. It may be possible to omit EM or perform EM on material reprocessed from the paraffin block. However, accurate measurements of glomerular capillary membrane thickness are not reliable after this procedure. Cost and method availability should also be considered in the biopsy workup (e.g., EM is not available in all pathology laboratories and is costly but not required for diagnosis in most transplant biopsies). Therefore the need for EM can be considered on a case-to-case basis.

Most renal pathologists employ several stains for light microscopy. Commonly used stains are periodic acid–Schiff (PAS) reaction, hematoxylin-eosin (H&E), silver methenamine, and trichrome. It is not necessary to perform all these stains. PAS and H&E are used by most laboratories. The PAS stain, also referred to as “H&E for the kidney,” is the most widely used and useful staining in nephropathology, permitting a good overview and visualization of details on kidney architecture.

Kidney biopsies must be examined systematically, first at low power to identify parts of the kidney (or other structures in some cases), whether there is cortex and/or medulla, the amount of chronic nephron damage with tubular atrophy and interstitial fibrosis, and the presence of interstitial inflammatory infiltrates. This will also permit assessment of interstitial expansion, most commonly due to either edema or fibrosis, but occasionally due to infiltration by, for example, amyloid. Although seldom seen in nephrology-indicated kidney biopsies, the presence of neoplastic nonmalignant or malignant lesions should also be assessed. Examination should then proceed by studying all other compartments in more detail (i.e., glomeruli, tubules, interstitium, and vessels including arteries, arterioles, and veins). Features that should be looked for in glomeruli and tubules are detailed in Boxes 25.2 and 25.3 . Arterioles should be examined for the presence of hyalinosis, thrombosis, and necrosis. Arteries should be assessed for intimal thickening and whether it is accompanied by reduplication of the internal elastic lamina, thrombosis, necrosis, inflammation, and cholesterol emboli.

Box 25.2

Features to be assessed by light microscopy in glomeruli

• •

  Globally sclerotic glomeruli

• •

  Size

• •

  Cellularity: increased cells (hypercellularity) in mesangium (mesangial hypercellularity; NOTE: normal mesangial areas contain two to three cells), in capillary lumens (endocapillary hypercellularity), or in Bowman space (extracapillary hypercellularity or crescents)

• •

  Capillary wall thickness and form (e.g., wrinkled, with double contours, spikes, spicules, craters, and holes; NOTE: best seen in periodic acid–Schiff or silver stain)

• •

  Mesangial expansion (mesangial sclerosis) with/without nodules

• •

  Deposition of abnormal material (e.g., amyloid)

• •

  Segmental sclerosis

• •

  Presence of thrombi, pseudothrombi, necrosis

• •

  The distribution of the previously mentioned lesions (i.e., focal vs. diffuse and segmental vs. global)

Box 25.3

Features to be assessed by light microscopy in tubules

• •

  Percentage atrophy

• •

  Signs of acute damage (e.g., dilatation, loss of brush border, epithelial flattening, mitoses, and necrosis)

• •

  Tubulitis

• •

  Casts: Granular casts suggest acute tubular injury; eosinophilic fractured casts suggest myeloma; neutrophil casts suggest acute pyelonephritis, pigmented cast (e.g., bile acids and myoglobin)

• •

  Crystals (e.g., oxalate and calcium phosphate)

• •

  Viral inclusions (e.g., BK virus)

• •

  Distribution of the abovementioned lesions (i.e., focal vs. diffuse)

The involvement of glomeruli by a pathologic process can be defined by the percentage of glomeruli involved by a lesion and by whether the lesion involves all or only part of any individual glomerulus. A lesion that involves all or nearly all glomeruli is described as “diffuse,” whereas one that involves some but not all glomeruli is described as “focal.” In the definitions given in the World Health Organization (WHO) atlas of glomerular diseases, it was suggested that the cutoff for focal versus diffuse should be 80% of glomerular involvement. However, in recent classifications of lupus glomerulonephritis and immunoglobulin A (IgA) nephropathy, the cutoff is defined as 50%. If a lesion involves only part of a glomerulus (i.e., with some capillary lumens remaining uninvolved), it is called “segmental,” whereas if it involves the whole glomerulus, it is called “global.” In the classifications of lupus glomerulonephritis and IgA nephropathy, the cutoff is set at 50% glomerular tuft involvement, except for segmental sclerosis in IgA nephropathy, in which any area of sclerosis that leaves some of the glomerulus unaffected is defined as segmental. Specific terminology is also used for tubulointerstitial and vascular lesions. A consensus on terminology of kidney lesions and their definitions has been recommended by the Renal Pathology Society.