A pathological diagnosis is central to the treatment of patients whenever one is required. If cancer is the issue then tissue is the answer. This all-important answer informs the treating doctor about the type of cancer, its likely behavior, to what extent it has affected the patient, whether all of it has been removed and what medications have or will work best for a patient with that cancer. Many people wonder how their histopathological diagnosis was arrived at. For many the results should be ready in an hour or two and most are surprised when asked to come back in 3 days. For most people including doctors, histopathological diagnosis is confusing and their relationship with the pathologist (histopathologist) is a rather complicated one.
The main subject of medical science (patient) has autonomy and that sets medical science apart from all other sciences. Unlike other sciences that rely on non-human material, laboratory medicine relies on human material (tissue, cells, etc) to study diseases and guide treatment of patients with these diseases. All the disciplines of medicine have laid down procedures (science and art) that are followed in order to arrive at an accurate, complete and useable diagnosis in a timely fashion. The fundamentals of these procedures are taught during undergraduate medical education and reinforced during residency training (professional post graduate medical training). Pathologists are expected to know these procedures before they start reporting histopathology slides (slides) on their own. Short of that, any attempt to report slides is dangerous and may result in needless treatment followed by death. An entire clinical team can be misled by the wrong histopathology report with resulting fatality in the end. It is thus crucial that the process of making a histopathological diagnosis is understood by all who request and use histopathology reports.
Disease that manifest in the flesh (organs) are characterized by changes that start at the molecular level, and finally manifest at the cellular, then tissue, then at the organ system level. This means that whenever a patient complains of an ailment, the signs and symptoms are supported at the cellular level by changes in cellular architecture or function. Histopathology relies on cellular / tissue architecture and increasingly on molecular alterations to diagnose diseases. This means that each disease has a set of features that define it. For most of these diseases, these features are clear at the microscopic level if the right tissue is submitted for microscopic examination. Cancer diagnosis has become the most important area that relies on histopathological examination and diagnosis. Cancers have morphological features that define them and determine their behavior. These features are well documented though evolving and pathologists rely on these to arrive at a diagnosis. To be able to recognize these features, light microscopic slides are prepared from tissue and stained (colored) with general, routine stains (Hematoxylin and Eosin), special stains (specific for certain cellular contents) and immunohistochemical stains (depend on the presence of antibodies of interest). These stained slides form the bases of pathology diagnosis.
To start with, quality and quantity criteria must be met. This means that the quality of tissue submitted for examination must be optimal. It must allow for assessment and interpretation of histopathological morphology. Then again, the quantity of diagnostic material must be adequate for making not just a diagnosis but a useable diagnosis. Once these basic criteria are met, pathologists have traditionally arrived at pathological diagnosis using a few time-honored approaches:
- Pattern recognition
- Deduction (heuristic)
- A combination of pattern recognition and deduction.
Cells, tissue and organs have normal histological architecture. When affected by disease, the normal histological or cellular architecture of cells, tissues and organs is altered. Histopathologists rely on this altered architecture to make histopathological diagnosis. The architectural changes may be easily identified and form the bases of what pathology clinicians refer to as a ‘spot diagnosis’. In a significant proportion of cases however, the change in architecture does not result in a recognizable pattern that is easily identified as belonging to a particular disease process. Essentially, in the pattern recognition approach, a pathologist relies on a mental atlas (learnt over time) of familiar patterns which represent specific lesions (disease entities) to make a diagnosis. It explains names such as tubular (forming tubules) adenoma and alveolar (possessing an alveolar pattern) rhabdomyosarcoma.
With this approach, patterns unique to some lesions can be spotted at a glance. Many early carrier pathologists find this approach appealing and may make use of this approach to render a diagnosis at the start of their carrier. The downside for a beginner however, is that because you usually have not seen a large number of cases as an early carrier pathologist, you usually don’t have an adequate number of patterns in your ‘mental atlas’ (you are still building this atlas through work experiences). One may thus make the wrong diagnosis because one is unaware a particular pattern is not uniquely diagnostic of a single entity or that patterns may not be diagnostic in themselves. As an example, an ‘alveolar’ pattern defines alveolar soft part sarcoma and alveolar rhabdomyosarcoma, two tumors that require different management approaches. Again, there are ‘true’ and ‘false’ patterns that are so identical that one cannot separate the true pattern from its ‘false’ look alike. Examples, of these patterns are epitomized in terms such as rosettes vs pseudo-rosettes, squamoid vs. squamous, sarcoma vs. sarcomatoid and epithelioid vs. epithelial. These may be confusing for any pathologist who relies on pattern alone to render histopathological diagnosis. It is important that clinicians realize that the microscopic descriptions in their reports are mainly of these patterns and eventually culminate in the conclusions that are drawn by the pathologist. For many pathologists however, these patterns only become a lead but not sole criteria for making a conclusive diagnosis of any particular lesion. They learn to rely on additional criteria to arrive at the right diagnosis in an objective way.
A histopathological diagnosis that is based on a ‘pathognomonic’ pattern may ignore clinical history and examination findings. Such scenarios may be responsible for the belief that the pathologist sees it all ‘at the altar of the microscope’ and does not need clinical history to bias him. This is a rather unfortunate perception. Bias is a necessary part of medical practice, without which we go on a wild goose chase that may be rather expensive for the patient or whoever pays the hospital bills. Bias (clinical history, examination findings, results of investigations etc) allows for a quick assessment of what is likely and unlikely; setting the stage for a useful focused pathological assessment in a timely fashion.
Heuristic approach (deductive approach)
This relies on deduction and combines disease epidemiology, clinical history, examination findings and pathological findings. It is an approach that is usually less used on its own considering the evidence on the slide is what ultimately determines what diagnosis can be made. As an example, a clinical history and investigation suggesting prostate cancer but without malignant cells on the slides will not result in a diagnosis of prostate cancer but rather a diagnosis of benign tissue if indeed that is what is present on the slide or of inadequate tissue if that is what the slide shows. No amount of deduction will alter the diagnosis in such a scenario. This approach is however a more medically reasonable approach to diagnosis since it depends on clinical findings. It allows the recognition of entities such as ‘sarcomatoid carcinoma’, ‘epithelioid sarcoma’ and the like where a carcinoma masquerades as a sarcoma or a sarcoma masquerades as a carcinoma. Deduction is however only of value if combined with the histopathological features seen on the slide. In my opinion, this approach is what prevents pathologists from jumping to conclusions based on a perceived pattern being present. A deductive approach to slides, makes it less likely that a pathologist will be misled by a pattern. Using this approach, a pathologist is unlikely to diagnose a jaw lesion in a child as an osteosarcoma. Such a pathologist is likely to explore other differentials that are more likely at that age. It forms the bedrock of the histopathology practice of more experienced pathologists.
Combination of pattern-based approach and deduction-based approaches.
This is the integrated approach used by many pathologists and relies on a firm understanding of clinical presentation of diseases, their epidemiology and histopathological morphology. It ensures that a diagnosis is not made out of nothing but rather follows logical medical reasoning combined with ‘near exact’ science. Using this approach, few lesions escape correct diagnosis with most of those that can’t be diagnosed requiring molecular studies. In my practice I have realized that this approach relies on adequate clinical history and examination findings and requires the pathologist to realize and play his clinical role actively without assuming he is a ‘pure scientist’. In my opinion it is the method of choice for the practice of histopathology and clinicians should be aware of this so they provide adequate clinical history, examination findings and results of other ancillary investigations such as radiology. In that regard, the pathologist should be in constant communication with the clinician regarding a case rather than trying to rely on patterns to make a hasty diagnosis. A hasty diagnosis based on questionable patterns, scanty clinical history or hearsay usually turns out wrong when the clinician is contacted and the right history, examination findings, findings of ancillary investigations and adequate and representative diagnostic histopathological material are obtained.
These pathological maneuvers described early on are engaged in at varying levels of expertise in a diagnostic pathology department; residents, specialists, consultants and subspecialists. There is an inbuilt quality control and quality assurance process that ensures that there is consultation between pathologists of different grades/ experience and expertise in the same department or in different departments. Using this approach, many lesions (disease conditions) are diagnosed and classified (typed) leaving a minority of lesions that remain untyped. For even these difficult and unresolved cases, the issues are usually beyond the conclusion of benign or malignant but rather are to do with the ‘histogenetic’ type of a lesion in question. For such diagnostic problems the level of certainty or the objectivity of reviews between pathologists may become questionable if not unreliable (realm of seniority/ eminence). It is recommended in such cases that objective review (relevant immunohistochemistry or other molecular studies) is employed in addition to second opinion in order to arrive at a conclusion.
Finally, the processes described earlier are carried out on routine Hematoxylin & Eosin stained slides as a first step. This is followed by supportive stains such as special stains and immunohistochemistry. By their nature, special stains and immunohistochemical stains are somewhat more specific to cells and thus are used mostly in the setting of typing. Other molecular techniques such as genetics are now available for typing of cancers when the stains fail to conclusively type a tumor. In all this, it is good to remember that the material on the microscope slides (what you took from the patient) is the only acceptable scientific medical evidence of histopathological diagnosis and is the only reason the pathologist made that expected or unexpected diagnosis.