In renal transplantation the method most often used to obtain cytological specimens from the transplant is a modification of Franzen’s FNA.⁷

FNAs are taken using a 20–22 gauge spinal needle connected to a 20 mL syringe which contains 5 mL of RPMI-1640 tissue culture medium supplemented with 5% human serum albumin, 50 IU/mL heparin and 1% Hepes buffer. An FNA pistol may be used, but is not necessary for the procedure. The transplant is usually located easily by palpation. When necessary, ultrasound guidance can be used.

The FNA is performed percutaneously without local anaesthesia in aseptic conditions. The needle is inserted into the renal cortex, full suction is applied and the needle is moved back and forth three or four times through a distance of 1–2 cm. In this way, the needle traverses the entire cortex and reaches several periglomerular and perivascular areas. Sampling is complete when the colour of cellular fluid can be seen in the needle hub. The needle is then rapidly withdrawn after releasing suction and the sample of 10–50 μL inside the needle is flushed immediately with tissue culture medium to get the whole sample into the syringe. The syringe with sample inside is then sent to the laboratory for immediate processing. If necessary, samples can be kept overnight in syringes containing culture medium in a refrigerator.

To compare the relative leucocyte distribution between peripheral blood and graft, simultaneous samples of blood, usually 2–3 drops, are taken from the fingertip into another syringe containing 5 mL of the same RPMI medium.

The FNA and blood samples are processed in parallel. The cells are centrifuged, resuspended, counted and 100–200 μL aliquots are spun onto microscope slides using a cytocentrifuge. The smears are air dried for routine diagnostic use and stained with May–Grünwald Giemsa (MGG). Parallel preparations can be used for other cytological stains or immunocytochemical staining techniques.

MGG-stained cytospin smears of the graft and blood are examined microscopically and the findings reported on a standard report form (Table 15.1), derived from the First International Workshop on Transplant Aspiration Cytology in Munich in 1982. The three most important features to be evaluated are specimen adequacy, the leucocyte differential and the morphological features of the parenchymal cells.

Table 15.1 FNA report evaluating specimen adequacy, leucocyte differential and morphological features of parenchymal cells

To be evaluated, the aspirate must be representative. The inevitable but variable contamination of renal and liver aspirates by blood makes assessment of specimen adequacy of utmost importance in interpretation of the findings.

Since the cellular infiltration of acute rejection is always most pronounced in the renal cortex, representative specimens must include sufficient cortical material. Cytological samples taken from the medulla are usually not diagnostic,⁸ as is also the case in histology. The presence of glomeruli cannot be used as a criterion for adequacy since glomeruli do not always appear in aspirates and many of them are lost in processing. Instead, adequacy of representation is assessed by calculating the ratio of tubular cells to inflammatory leucocytes.

Cytological criteria for representative samples and reproducibility of transplant aspirate specimens have been established by analysing duplicate aspirates,⁹ where a correlation coefficient of 0.95 was obtained if both specimens contained at least seven tubular cells per 100 inflammatory leucocytes. When the ratio of tubular cells to leucocytes in either sample fell below this, the correlation coefficient fell accordingly. Other groups have reported similar results for double aspirate biopsy analysis.¹⁰

Most of the inflammatory leucocytes in cytological preparations can be identified readily according to standard haematological criteria^4,7 in routine MGG-stained smears (Fig. 15.1). These include small lymphocytes, activated lymphocytes with increased cytoplasmic basophilia, and large granular lymphocytes, which are the morphological form of natural killer cells.¹¹ The appearance of blast cells is diagnostic of acute rejection.

Fig. 15.1 Inflammatory cells in kidney and liver transplants during rejection. Giemsa-stained cytocentrifuge preparations. (A) A lymphoid blast cell with immature nucleus and cytoplasmic basophilia; (B) a blast cell with plasmacytoid morphology and a large granular lymphocyte with cytoplasmic granulation; (C) an activated lymphocyte; (D) a monocyte with irregular nuclear contours; (E) macrophages with cytoplasmic vacuolation and elongated nuclei; (F) a thrombocyte aggregate with a monocyte.

Lymphoid blast cells are characterised by their large size (15–25 μm in diameter), their large immature nuclei and intense cytoplasmic basophilia. Approximately 50% of the lymphoid blast cells are B blasts containing intracytoplasmic immunoglobulins, as identified by immunofluorescence staining. The other half of the blast cells are T blasts, characterised by T-cell surface antigens, and identified with monoclonal antibodies using immunoperoxidase staining.

Different forms of monocyte–macrophage cells can also be seen, including small monocytes, large monocytes with irregular, multilobed nuclei and macrophages in different stages of maturation. Macrophages are large cells, up to 60 μm in diameter, usually showing vacuolated cytoplasm and pyknotic elongated nuclei lying at the periphery of the cell. They are usually seen in abundance in the later phases of severe and irreversible rejection and in acute vascular rejection in the early phase.

Eosinophils are usually more frequent at the beginning of immunological reactivation, and are seen both in aspirates and in peripheral blood, indicating generalised immune response to the graft.¹² Platelets are also seen during rejection in excess in the graft.¹³ Small loose platelet aggregates disappear during successful rejection treatment, but large aggregates on endothelial cells seem to indicate a worse prognosis.¹³ Neutrophils are usually seen in excess in the graft only when there is irreversible rejection with necrotic changes. In cases with bacterial infection of the graft, neutrophil aggregates with intracellular bacteria can be seen.

Aspiration biopsy specimens consist mainly of single parenchymal cells, although clumps of renal tubular cells or even parts of tubules and whole glomeruli may often be encountered in the specimens. The parenchymal cells most commonly seen are tubular cells from different parts of the nephron, and endothelial cells from the renal vascular endothelium, usually from capillary blood vessels (Fig. 15.2).

Fig. 15.2 Renal parenchymal cells in Giemsa-stained cytopreparations. (A) A group of normal tubular cells; (B) swelling, degeneration and vacuolation in tubular cells in ATN; (C) pronounced isometric vacuolation and swelling in tubular cells during CyA toxicity; (D) necrotic tubular cells in graft necrosis.

For more detailed characterisation of the parenchymal cells, immunocytochemistry can be used. With monoclonal antibodies different parenchymal cell types can be identified precisely. Cytokeratin antibodies are used for tubular cell characterisation, since these antibodies do not stain endothelial cells or leucocytes. Endothelial cells can be stained with antibodies to Factor VIII-related antigens or with vimentin antibodies, which do not stain tubular cells.¹⁴

Changes in the parenchymal cells are scored from 1–4, 1 being normal and 4 representing necrosis, and this score is recorded in the report (Table 15.1). Although typical findings in parenchymal cells can be seen in several different graft complications, the findings are basically non-specific. Degenerative changes can be seen in tubular cells in acute tubular necrosis (Fig. 15.2), in advancing rejection, in CyA nephrotoxicity and also in urological complications. The information derived from graft parenchymal cell morphology is useful in the differential diagnosis of graft complications, but interpretation of the changes requires concomitant evaluation of the inflammation and also knowledge of the clinical data.

In acute tubular necrosis (ATN) the tubular cells are swollen, with cytoplasmic degeneration and irregular vacuolation. In very severe cases necrotic tubular cells may also be seen, but usually only a few. These changes are due to prolonged cold ischaemia and return to normal with improving graft function in 1–2 weeks. In pure ATN there are no signs of immunoactivation.

Similar, but more pronounced changes are seen in the tubular cells in acute CyA toxicity. The cells are swollen, with increased cytoplasmic basophilia and prominent isometric vacuolation.^15,16 Toxic isometric vacuolation is quite typical of acute CyA toxicity, although it is a non-specific phenomenon. It has also been reported in experimental kidney transplantation models.¹⁷ The deposits of CyA and its metabolites can be demonstrated by specific monoclonal antibodies and immunofluorescence techniques.¹⁸ Isometric vacuolation is not seen in chronic CyA toxicity, where tubular atrophy and degeneration with interstitial fibrosis are the dominating features. Today, with triple drug immunosuppressive treatment and rather low CyA doses, acute CyA toxicity is not a very common complication.

At the beginning of acute rejection, tubular and endothelial cells usually have normal morphology and also retain normal morphology in short, easily treated rejections. In severe and prolonged rejection, on the other hand, there are progressive degenerative changes in tubular cells and even necrosis in irreversible rejection.

Severe necrotic changes in tubular cells are also seen in graft infarction,¹⁸ which is usually due to renal vein thrombosis. This clinical complication is often fulminant with rapidly progressing necrosis, but fortunately it is not a frequent event.