1
Before the Analysis: Rules for Interpretation of Hepatic Cytology

After several years of daily diagnostic practice, I have come to believe that, on the one hand, there is cytology based on classic criteria, which are listed in dedicated chapters of books on the subject, as well as being updated in specialized press articles and discussed at professional meetings. On the other hand, there is liver cytology, which is based on criteria that can frequently differ from standardized, shared, and consolidated concepts applied to several systems. The criteria I am referring to are represented by variations from classic, widely acknowledged issues and interpretations that cannot ignore the clinical context within which the liver disease is being investigated. I believe that the cytological diagnostic approach must be accompanied by in-depth knowledge of microanatomy, as well as cytological and histopathological aspects of the organs being investigated. This statement is particularly important when dealing with an organ such as the liver, since it justifies the differences from classic cytology. Aside from these mandatory bases, I firmly believe there are initial conditions, preexisting situations to the analysis and diagnostic obstacles that must be known in order to modulate one's diagnostic skills. Based on thousands of cases analyzed and diagnosed - frequently supported by histological assessment and further corroborated by clinical course, compatibility with data provided by laboratory and imaging diagnostics or response to therapy - I have been able to draw a few conclusions about the primary skills pathologist must develop to interpret this very complex organ. I have summarized them in a short collection of rules, which I believe should be memorized and taken into due consideration before producing any cytological diagnostic conclusion.

1.1 The Rules for Cytological Diagnosis of Hepatic Diseases

1.1.1 Rule 1

The diagnostic value of cytopathology in evaluation of hepatic diseases ranges from 30.3% to 82.1% agreement with histopathological diagnosis [1, 2]. This discrepancy is mostly due to the fact that the samples used for cytological investigation represent a very small percentage of a potentially pathological liver, and thus may not be indicative of some lesional processes (low diagnostic sensitivity). Moreover, according to Wang, the diagnostic agreement between cytology and histology should be high in cases of so-called "vacuolar hepatopathy," although this is, in my view, not a specific diagnosis and just morphological evidence of hepatocellular damage. Even a large number of cytological samples from a pathological liver may not detect any alteration, as in some cases of vascular disturbances; in others instances, the tip of the sampling needle collects cells which may not be affected by the primary pathological process occurring or may be affected by aspecific changes. Especially in the course of widespread pathological processes, it is always preferable to sample from many different parts of the liver, as this will increase the chance of collecting samples and therefore data that are morphologically useful for diagnostic purposes. Similarly, when evaluating nodular lesions, comparison between cells from the lesion site and those from the surrounding nonnodular parenchyma may give good diagnostic results (widely described in relevant chapters).

1.1.2 Rule 2

There are some pathological processes, such as amyloidosis or extrahepatocytic cholestasis, whose cytological identification is always extremely useful in terms of diagnosis (very high specificity), even if not corroborated by other tests. For example, amyloidosis in the cat may not be associated with a significant increase in serum amyloid [3]; similarly, cholestasis, in rare instances, may not necessarily be associated with an increase in the concentration of total bilirubin [4]. Indeed, some unmistakable morphological signs may be the result of focal phenomena in a progressive phase and therefore precede certain alterations of other diagnostic parameters.

1.1.3 Rule 3

Many pathological processes can only be successfully interpreted if a histopathological architectural context is available. Furthermore, cytology provides only nonspecific aspects of these processes; for example, fibrosis or inflammation does not provide any information about the causes, extent or distribution but they may highlight an important morphological aspect, which is a valid and sufficient reason to carry out an in-depth histological analysis.

1.1.4 Rule 4

Liver diseases are often not evaluable by cytology. They are often better assessed by histopathology and, in that case, recognition of a specific disease is the result of a morphological diagnosis based on a biopsy of tissue fragments, carried out according to established, accepted, and shared criteria [5]. Cytology is a diagnostic aid that may render histological examination unnecessary (for example, when recognizing amyloidosis or several neoplastic conditions, such as hepatic large cell lymphoma), but in most cases, the information it provides is nonspecific, as in many cases of mixed inflammation or aspecific reversible change. The role of cytology is often limited to excluding other potential suspected pathologies or to reducing or contextualizing the possible differential diagnoses, which must undergo histopathological evaluation.

1.1.5 Rule 5

Sometimes, it is hard not to feel defeated but I will always be determined to persuade clinical colleagues that a morphological diagnosis of cellular or tissue characteristics is, in many cases, impossible without a comparison with all data resulting from clinical and anamnestic investigations, collateral tests, laboratory and imaging diagnostics. The readers of this book will understand that a specific morphological characteristic may correspond to several different clinical conditions (each with its own therapy and prognosis); furthermore, if they have had firsthand experience in making a diagnosis through cytological morphology, they are also likely to understand the importance of being sufficiently informed about data relating to the lesion being analyzed. Given the above, I call on anyone reading this book to join me in this battle: to accept that collaboration between clinicians and pathologists is essential if we are to succeed in improving the management of a disease.

1.1.6 Rule 6

There is an urgent need for cytologists to translate every morphological characteristic of the sample into a diagnosis that is clinically useful in order to come to terms with the relatively scanty information that a liver sample can provide. Cytological samples must be approached with humility, refraining from drawing any diagnostic conclusion when the signs are insufficient or the correlation with clinical indications is incomplete or missing.

1.1.7 Rule 7

Romanowsky-type stains are represented by a group of different stains, such as Hemacolor®, Diff-Quik®, May Grünwald Giemsa and others [6]. These are normally used as routine stains in veterinary cytology, but chromatic results can differ from one stain to the other; consequently, what can stain deeply basophilic or red with one stain can appear as black or brown with another; for example, bile can appear variously as deeply basophilic, greenish or black on the base of the selected stain. Sometimes a color may appear darker or lighter due to the time the slide is exposed to the stains, mostly in cases when a stain procedure is not standardized; always remember that the colors frequently are subjective and must be interpreted carefully.

I believe these rules provide a solid base on which to start discovering the diagnostic secrets hidden in liver cytology. I have attempted to explore them in this book.

1.1.8 Rule 8

Always consider that, after the therapy is initiated according to cytological features, the pathological process can change and histopathologic examination, when done weeks or months after the cytological examination, can result in different findings. For example, an inflammatory process could initially be evident on cytologic preparations but disappear or appear attenuated on histopathologic examination if this latter is done after adequate therapy has started.

1.2 Diagnostic Approach to Liver Disease

In my view, the diagnosis of liver disease is the result of an algorithm that provides an evaluation of the patient based on:

• historical, clinical, and anamnestic signs
• hematochemical investigation
• ultrasonographic investigation
• cytological investigation
• histopathological investigation.

The above is summarized in the diagnostic pyramid shown in Figure 1.1, which clearly shows that only by going from the bottom up is it possible to refine a diagnostic investigation aimed at identifying the causes. Each step contains the indications necessary to proceed to the next diagnostic phase and, eventually, to reach the diagnostic perception of a specific liver disease. With the exception of the first step, which generally provides nonspecific clinical signs, each step can potentially contribute to acquiring information concerning the liver disease in question. In the final...