When trying to convince an owner to opt for surgical biopsy, the mostpowerful argument is that biopsy often offers the greatest chance of makinga diagnosis relevant to therapy and/or prognosis. It probably rates asone of life's more embarrassing moments, therefore, when the report comesback with very inconclusive findings. I thought it might be useful to reviewsome of the conditions for which biopsy interpretation has greater-than-usuallikelihood of being inconclusive. As you will see, the reasons vary a greatdeal, but nonetheless most pathologists would agree with most of my selectionsfor "the 10 toughest diagnoses in surgical pathology"!

1. Atopy. There is no diagnostic histologic lesion for canineallergic dermatitis, be it inhalant or food allergy. The histologic lesionis a very mild superficial dermal edema with a few extra mononuclear leucocytesaround dilated dermal blood vessels, but this picture is also present ina wide variety of other inflammatory skin diseases. There is even somesuspicion that these lesions of atopy are not actually the result of theallergic condition, but nonspecific cutaneous responses to pruritus. Wewill sometimes back in to a diagnosis of "compatible with atopy"if there are no other markers pointing towards other diseases, such asthe presence of pustules (pyoderma) or clusters of eosinophils (flea allergy).A particularly frequent request is for me to determine whether a dog hasallergic disease underlying a current problem with pyoderma. That is, quitefrankly, an impossible task because the robust lesions of pyoderma easilymask the very subtle changes that might be attributable to atopy. In theend, the utility of skin biopsy in cases of suspected allergic diseaserests more with my ability to exclude other possibilities, hopefully leavingyou with a much shorter list of alternative diagnoses than you had priorto biopsy.
2. Sarcoptic mange. Incredible though it may seem, the histologiclesions associated with canine sarcoptic mange are very nonspecific. Ideally,there is dramatic epidermal hyperplasia and a subtle, diffuse, and uniformeosinophilic perivascular dermatitis. Somewhat sheepishly, I must admitthat these time-honored criteria have recently come under assault becauseI have had cases of confirmed Sarcoptes (with mites visible even in histologicsection) that do not have an eosinophilic infiltrate, and other cases whichseemed absolutely classical which failed to respond to appropriate medicaltherapy. I think the bitter truth is that any hyperplastic pruritic dermatitiscould be sarcoptic mange, and biopsy is probably of very limited help inmaking that diagnosis. If you cannot find mites on scraping, the chancesof seeing any in biopsy are extremely remote.
3. Splenic hemangiosarcoma. While it may be frustratingthat biopsy is of limited utility in diagnosing the above allergic skindiseases, it is much more painful when I am unable to confirm a diagnosisof suspected splenic hemangiosarcoma. It is, in the minds of many pathologists,the most difficult and frustrating diagnosis in all of surgical pathologybecause of the "needle in a haystack" phenomenon. The problemis that the neoplastic endothelial cells frequently represent only a smallpercentage of the large mass, the majority of which is hematoma resultingfrom the extreme fragility of the neoplastic vascular channels. In somespecimens, good fortune results in an easy diagnosis in the very firstsection, while in other cases multiple sections reveal only hematoma andnecrosis. If there are omental or mesenteric implantations in the formof tiny red nodules, submitting these as well as the spleen will increasethe probability of diagnosis because these small nodules are metastasesof "pure" tumor. In years gone by, I used to make quite a fewdiagnoses of splenic hematoma. I now make very few such diagnoses becausetoo many of my hematomas were proven, later, to have metastasized!
4. Sudden death. Spontaneous or anaesthesia-associated suddendeath is particularly emotional and painful for all parties concerned,and it would be comforting to know exactly what happened. Unfortunately,no more than one case in ten has microscopic lesions to explain the death.There may be several reasons for this, but the most logical candidate isthat truly sudden death simply offers no time for the tissue reactionsto occur that are recognizable to pathologists as tissue necrosis. Experimentalevidence with ischemic heart disease, for example, clearly demonstratesthat the myocardium remains histologically normal for at least six hoursafter lethal myocardial ischemic injury. In an animal that truly "dropsdead", there will be no histologic lesions to see in the vast majorityof cases. There may, however, be macroscopic changes detectable at necropsy,such as dramatic laryngeal edema, excessive pulmonary edema, myocardialpallor, and so on. The most frustrating of all "sudden deaths"is that associated with anaesthesia. Except for those few cases that havesome undetected pre-existent disease, it is extremely rare to find a lesionto explain the intra-operative or post-operative death with apparentlysuitable anaesthetic protocols.
5. Missing or disfigured nails. This enigmatic but surprisinglyfrequent complaint seems to defy diagnosis by any means other than histologicexamination of amputated toes. Even then, the specimen requires carefulprocessing in order to obtain well-oriented sections through the germinalepithelium of the digit. I have, over the years, tried to make the diagnosison all kinds of less dramatic and disfiguring biopsies, but have neverbeen successful. Examining the shed toe nail itself is, in my experience,a waste of time. The elusive microscopic lesion is an interface dermatitis(single cell necrosis, hydropic change, pigmentary dispersal) within thegerminal layer of the follicular epithelium, very similar to what is seenwith discoid lupus, with dermatomyositis of shelties, or with ischemicskin damage. Whether it represents some "autoimmune" phenomenonor ischemic damage has yet to be determined. There may be some way to obtaina deep biopsy that will capture this epithelium, but I have not yet receivedsuch a sample!
6. Atrophic myositis. For reasons that I do not understandat all, it has proven extremely frustrating to confirm a diagnosis of immune-mediatedor eosinophilic myositis in samples from obviously atrophic temporal orother facial muscles. I admit to feeling like an idiot when, faced witha convincing history of dramatic muscle atrophy, I send back a histologicdiagnosis of "apparently normal muscle". I even have difficultyconfirming the atrophy, suggesting to me that at least part of the macroscopicchange is related to drop out of entire fibers, rather than a reductionin the size of those fibers remaining. The inflammatory infiltrate tendsto be very widely scattered, and you will have better luck confirming thediagnosis with multiple random small biopsies instead of one or two largewedges. I have even suggested that multiple 3 mm skin punch biopsies, inwhich you "lean on" the punch to get as deep a sample as possible,offers a reasonable method of obtaining such samples. Another good idea,with all muscle biopsies, is to not be too hasty in dropping them intothe fixative. Such fixation creates tremendous contraction artifact. Lettingthem sit for a few minutes to desensitize, or letting them stick to a pieceof cardboard for a few minutes before putting them into formalin, willgreatly reduce the artifact and will result in superior sections.
7. Bone core biopsies. I find it particularly distressingnot to be able to offer a positive diagnosis, one way or the other, oncore biopsies taken from suspected bony malignancies. In many tumors, theneoplastic cells are found only within and adjacent to the lytic core ofthe tumor, and the proliferative compartment that surrounds the centerof the tumor is made up of reactive new bone and fibrous tissue. Becausethese tumors tend to bleed terribly during the biopsy attempt, it is commonthat the core instrument, or even the wedge biopsy, not be extended deeplyenough into the tumor. I am left with a view only of periosteal new boneand periosteal fibroplasia, and you are left with blood all over the room,no diagnosis, and a dissatisfied owner. Even worse, you must try to convincethe owner to go for yet another surgery! The key to success is courage!You must go right to the radiographically-lytic center of the tumor, andthe ideal biopsy is a core that contains portions of everything from theperiosteal reaction to the lytic center. Somewhere along that core I willfind the tumor, if indeed there is one to be found. Incidentally, exactlythe same "faint of heart" problem applies to endoscopic nasalbiopsies, which also tend to bleed a great deal in the process of gettingsamples of adequate size and depth.
8. Inflammatory bowel disease (lymphocytic-plasmacytic, or eosinophilic,enteritis). One of the great myths of our time is that endoscopicbiopsy can diagnose inflammatory bowel disease. Perhaps an even greatermyth is that inflammatory bowel disease is a specific syndrome that ischaracterized by too many leucocytes within the intestinal lamina propria.Neither assumption has been validated in any scientifically- acceptableway. The whole basis for the diagnosis is the observation that some dogsor cats with chronic diarrhea have a lot of leucocytes within the intestinallamina propria. The prevalent assumption that these cells are somehow causingthe diarrhea or vomiting suffers from several serious problems. Firstly,the range of normal for the number of various leucocytes within the laminapropria of any part of the bowel has never been defined. Presumably, thereis wide individual variation that may depend on diet, parasite burden,time of the year, and who knows what else! Secondly, there is no inherentreason to believe that an increase in lymphocytes and plasma cells withinthe gut - the body's largest lymphoid organ - necessarily is the causeof disease. To cite a parallel example, when increased lymphoid activityin a lymph node is seen, we merely assume that the lymph node is respondingto an antigen, and is therefore doing its normal, physiologic, job. Whyis it, then, that we assume that the intestinal lymphoid organ is diseasedwhen it does exactly the same thing? I continue to examine a large numberof endoscopic biopsies, and continue to offer extremely subjective opinionsabout whether the cellularity within the lamina propria is "normal"or not. At the same time, I continue to sample dogs or cats coming to postmortem without any complaint of intestinal disease, and am amazed to seeanimals with leucocyte populations well in excess of that "normal"range, with no apparent functional effects whatsoever. The jury is stillout on this one!
9. Any chronic inflammatory skin disease. The scenariois a familiar one: a dog or cat is rushed to you for a second opinion,with a complaint of three years of inflammatory skin disease that has beentreated with everything from chicken soup to garlic powder! You calmlylook the owner in the eye, and sagely advise skin biopsy! This, to me,is a somewhat sly way of passing the buck on to me, so that I have to admitignorance rather than you. This is truly a catch-22 situation, becausethe more chronic the skin disease, the less specific are the histologicmarkers. On one hand, you do not wish to rush every skin case to biopsybefore you have had a chance to try standard therapy for your leading clinicaldiagnosis; on the other hand, you do not want to delay so long that yourun the risk of the dreaded diagnosis of "end stage skin". Ifyou feel you have no choice but to biopsy such chronic skin cases, at leastprotect yourself from abuse by making only modest promises of the utilityof such biopsy, and selectively biopsy the most active-looking lesionsthat you can find. As an added reminder, if you find it necessary to biopsya dog that is still under the influence of corticosteroids (oral steroidswithin the last ten days, long acting injectables within the last six weeks),remember to tell me about the steroid therapy so I can attempt to factorthat into my interpretation.
10. Chronic mucoid or suppurative rhinitis in dogs or cats.While not a fatal disease, this is certainly an annoying and perplexingproblem for both owners and veterinarians. Biopsy is often the diagnosticprocedure of last resort when therapy has failed to resolve the problem,but again your expectations may be too lofty. The histologic change isvery mild and nonspecific, with a mixture of lymphocytes, plasma cells,and neutrophils within an edematous lamina propria below a normal respiratoryepithelium. The lesions carry no etiologic weight whatsoever, and nobodyreally believes that the presence of neutrophils predicts a bacterial pathogenesis.The biopsy may be useful in ruling out alternative diagnoses such as mycoticrhinitis, nasal carcinoma, or eosinophilic (allergic?) rhinitis, but itwill not offer a specific etiologic or even therapeutic insight.