A histopathology report with a final conclusion of ‘INADEQUATE FOR DIAGNOSIS / NOT DIAGNOSTIC / INAPPROPRIATE FOR DIAGNOSIS’ with a recommendation to re-biopsy is unpleasant for clinicians (pathology and non-pathology) and the patient. At least on one occasion, a patient has threatened to sue the reporting pathologist for deceit and it is not uncommon for the surgeon to leave a note of caution that the biopsy ‘process’ was not an easy one and that a repeat is undesirable.
Of course, as a pathologist, I am not amused if a patient threatens me for no wrong done or a surgeon suggests I make do with whatever tissue I have even if it is rotten or inadequate. Anytime I have been challenged in this manner I have asked myself questions; Must I make a diagnosis based on nothing? are histopathological diagnoses not based on certain quality and quantity criteria? or is it just an ‘affirmation’ of a non-pathology clinician’s diagnosis and thus arbitrary enough just to be made on any tissue presented irrespective of content or microscopic appearances? Without criteria, is there really a need to even present tissue for histopathological assessment? Should the non-pathology clinician not just go ahead and treat what is suspected and not bother the poor pathologist? in order to improve diagnostic outcome, can we (clinicians ) have a discussion before a biopsy is done for difficult to diagnose lesions?
To the histopathologist, the tissue is what the patient is to the physician or surgeon. It forms the basis of histopathological diagnosis. In this regard, tissue presented to the pathology laboratory must be adequate for diagnosis. In order to fulfill this adequacy criterion, tissue must be of good quality and of a certain minimum quantity. It is important to note that making a diagnosis of cancer requires typing, grading, staging, and treatment-related parameters and these are dependent on adequate tissue. Thus, the best tissue for histopathological analysis is one that is viable, without handling ‘obscuring’ artifacts (of good quality) and large enough to be representative of the viable part of the lesion being investigated. This means that sampling and fixation are key steps in obtaining the best tissue for histopathological diagnosis. Clinicians must always remember that tumors, though monoclonal are inherently heterogeneous in morphology both macroscopically and microscopically.
Sampling is the single determinant of what the pathologist reports other than the histopathological diagnosis. In other words, depending on the type and nature of sample submitted, one may end up without a diagnosis, with only a diagnosis (the type of lesion), or with a ‘useable’ diagnosis together with other prognostic indices (grade, stage, immunohistochemical markers, etc). It is important that a clinician understands sampling to optimize the diagnostic outcome for his/her patients. Pathology clinicians are aware that sometimes, for fear of the patient bleeding, clinicians may resort to taking very superficial biopsies or sampling necrotic areas just so the patient does not bleed to death in an attempt to obtain a diagnosis. Though this option may be ‘easier’ for the clinician, grossly necrotic tissue will certainly be necrotic microscopically. Again, necrosis does not allow for proper histopathological assessment. For this reason, necrotic tissue will certainly be signed out by the pathologist as necrotic and not diagnostic or inappropriate for diagnosis. It helps to remember that though the necrotic part looks most malignant, it yields the poorest result in histopathology. Even if a diagnosis of ‘suspicious’ is made, it requires another biopsy and only serves to reinforce an already existing suspicion. Then again if a daring pathologist decides to call an inadequate biopsy malignant then there will still be the need to obtain more representative tissue for typing or for additional studies such as immunohistochemistry.
To the pathologist, options in sampling include but are not limited to the following broad categories;
- PART OF THE LESION
- WHOLE OF THE LESION
The results obtained from these two types of biopsies are different. The whole biopsies yield a lot more relevant information including critical information on margins and stage. Again, the whole biopsies may be curative and may ultimately be of more benefit to the patient.
- PART OF A LESION: these include;
- Core biopsies in which a core/strand of the lesion is taken. These are usually done with a core biopsy needle.
- Punch biopsies are just bits of relevant tissue bitten off with an instrument that can grip and be pulled to tear tissue for analysis.
- Incision biopsies, involve cutting into and obtaining a part of the lesion of interest. When wedge-shaped, they are referred to as wedge biopsies.
- Curetting: these are obtained by scraping the lesion or cavity of interest. The term is usually used in relation to the endocervical or endometrial canal
- Endoscopic biopsies: these are named for the procedure during which the biopsies are obtained. They rely on an endoscope with essentially a grip and tear mechanism that can bite tissue during the endoscopic procedure. They may be gastric, colonic etc.
The common thing about all these types of biopsies is that only a part of the lesion is biopsied. Usually, this is only a small fraction with little attempt at reducing the size of the lesion under investigation. The main reason for this type of biopsy is to obtain a useable diagnosis (type/grade of the lesion) to guide management. On occasion when some other relevant histopathological information can be provided, it will be done but when not provided, this does not mean, that information is not present in the tissue left behind (in the patient). This type of biopsy however has some downsides. The most important of which is;
High sampling error;
Since the lesion may not be visualized during the biopsy, the lesion of interest may not be included in the biopsy. Instead, an irrelevant part of the lesion may rather be sampled and result in an unexpected diagnosis. This is a regular occurrence with endometrial biopsies, breast core biopsies for small lumps and bumps, digitally (finger) guided prostate biopsies, and many other difficult to reach lesions. Blind biopsies may miss lesions altogether resulting in a diagnosis of benign tissue when a clinician believes the clinical picture of the patient suggests malignancy. It is important to remember and I repeat that ‘though tumors are monoclonal, they are inherently heterogeneous’ and thus have areas that are necrotic associated with areas that are hemorrhagic, those that are in situ carcinoma, and those that are frankly malignant and of various grades. It is important thus to keep this in mind when sampling to ensure that the true nature of a tumor is represented in the submitted tissue.
Again, even if the lesion is biopsied, the limited nature of tissue submitted means that the material may only be adequate for diagnostic purposes. Again, prognostic indicators such as grade may be subjective because only a part of the lesion is biopsied. On occasion, molecular studies may not also be representative of the entire lesion. As an example, a core biopsy of the breast diagnosed as colloid carcinoma with IHC showing ER, PR positivity and Her-2 negativity may finally turn out to be an invasive carcinoma of No Special Type (NST) with focal colloid secretion (the part represented in the initial biopsy). A repeat Her-2 IHC may turn out to be strongly positive for Her-2. This only goes to show that part biopsies may not be representative of the entire lesion. Finally, there can be no pathological information on staging and margins for part biopsies.
How to decrease sampling error when sampling part of a lesion.
- Improve visualization of the lesion. This is done with the aid of imaging. Thus, to say the biopsy should be done under image guidance. All manner of imaging methods ranging from ultrasound to Computer Tomography scans may be employed. Interventional radiologists are skilled at these procedures and may be consulted in this regard.
- Increase the number of cores (passes) while sampling from different sites; The number of cores taken can be increased while the attending physician samples from different sites. This ensures that the inherent heterogeneity that exists in lesions is well represented in the biopsy. It is good to remember that in each organ involved by a disease process, there will be different stages of the same disease in the organ. In effect, a sextant is recommended for the prostate. Again, sampling different parts of the bowel in patients suspected of having inflammatory bowel disease is what clinches the diagnosis.
- Move away from ulcers, necrotic tissue, and blood (if it looks dead grossly, trust me it will be worse microscopically)
- Repeat if negative and clinical suspicion is high. This is what non-pathology clinicians fail to recognize. Considering all the problems with sampling a part of a lesion as stated above, if a ‘part biopsy’ is reported as negative and the clinical suspicion of malignancy for a given lesion is high, then it is important that the biopsy is repeated with all the procedures stated above employed to increase the chances of success.
The following figures show the difficulties that may be associated with obtaining the right material for diagnosis when only a part of the lesion is biopsied.
Fig A: A renal cell carcinoma with near-complete loss of tumor tissue, something that is unlikely to yield a diagnosis of malignancy. For such a case, the diagnosis can only be made when the entire lesion is submitted to the histopathology laboratory after an excision.
Fig B: another renal cell carcinoma showing a largely solid appearance with areas of necrosis, cystic degeneration, hemorrhage, normal tissue, and viable malignant tumor. The part sampled by the core biopsy is what will inform the histopathological diagnosis.
Fig C: shows a polypoid tumor arising from the fundus of the uterus and not reaching the lower uterine segment. Two previous biopsies which certainly failed to reach the lesion because of the narrowed, distorted endocervix were reported as negative for malignancy. After an excision was done, a diagnosis of carcinosarcoma was made.
- EXCISION (WHOLE OF LESION)
This is therapeutic and involves the removal of a tumor/mass from a site where it is ‘not wanted’. Excisions received at the histopathology laboratory include;
- Excision with safety margin (rim of normal tissue)
- Excision without safety margin (usually fragmented)
- Excision of free organ –ovary, appendix, uterus, thyroid etc.
Excisions can be done with or without regional lymph node dissection depending on the lesion being excised. The Ideal scenario for malignant lesions, is an excision with safety margins with added regional lymph node dissection. This ensures that the patient derives the best benefit from the procedure in relation to disease-free survival/cure. The others, though they may not offer the same benefits are chosen by surgeons based on histopathological diagnosis, stage, grade, complexity of surgical procedure, patient fitness, etc. It is however important that the pathologist reporting this type of specimen provides certain pertinent information based on the availability of tissue from which that information can be derived. As an example, an excision without lymph node dissection cannot be used as a basis for assessing lymph node involvement, and thus staging in this scenario will only be based on tumor size (T). Again, excised lesions presented in fragments with no safety margins are not appropriate if an assessment of resection margin is required. Furthermore, fragmented specimen present challenges with the measurement of tumor size, and thus the ‘T’ of tumor staging cannot be objectively assessed pathologically.
In conclusion, a histopathological diagnosis must be accurate, complete, timely, and useable. To obtain histopathological results with these attributes clinicians (pathology and non-pathology) must work together to identify the best site and sampling methods that will yield quality/adequate tissue that allows the pathologist to generate a report with these desired attributes. This collaboration is especially important in assessing lesions /organs with stricter diagnostic criteria.