Evidence-based management of ITP (Proceedings)
Dogs with immune-mediated thrombocytopenia (ITP) usually present with platelet counts low enough to be considered life-threatening, although overt signs of bleeding are uncommon. Owners usually only note anorexia or lethargy, although in some cases epistaxis, cutaneous bruising/petecchiae/echymoses, or oral bleeding may be noted. Very rarely patients will present with complications that require immediate intervention such as pleural effusion. As a specialist, the most common reason for referral to me of ITP patients is inadequate or no response to appropriate immunosuppression. As with IMHA, this often is due to unfamiliarity with some of the treatment options beyond glucocorticoids, or a difficult-to-diagnose underlying disease. This presentation will briefly review the more common causes of secondary ITP and the diagnostic tests which I routinely consider prior to instituting therapy, and then discuss in-depth treatment options for dogs with this disease.
ITP: DIAGNOSTIC WORK-UP
Primary versus Secondary causes of ITP
Development of anti-platelet antibodies may be idiopathic (primary ITP), or may occur secondary to a number of infectious or neoplastic diseases, or certain drugs. The diseases listed below can cause thrombocytopenia through other mechanisms, without the presence of anti-platelet antibodies. Differentiation can be difficult, as there are no widely available tests which allow detection of anti-platelet antibodies. An anti-megakaryocyte antibody test is available through some specialty laboratories, but this must be performed on bone marrow aspirate samples. As a very general rule primary ITP results in much lower peripheral platelet counts than ITP secondary to concurrent diseases. Primary ITP dogs usually present with platelet counts less than 20,000; dogs with secondary ITP often have platelet counts higher than this. However clinicians should remember that these are general rules only—I have seen many cases of secondary ITP (particularly with RMSF or neoplastic diseases) where the peripheral platelet count is very low, in the 'primary ITP' range. Intense history-taking and physical examination are required in every case, as well as screening tests to ensure that there are no concurrent diseases. Diseases which are definitely associated with thrombocytopenia include:
• Acute severe hemorrhage (e.g. gastrointestinal bleeding, rodenticide toxicity)
• Vasculitis (e.g. sepsis/endotoxemia; tick-borne diseases)
• Portal hypertension
• Splenomegaly (barbiturate or acepromazine overdose; hypersplenism, splenic infarction)
• Splenic torsion
• Myelophthisis (bone marrow infiltration by tumors)
• Estrogen excess (Sertoli cell tumor; iatrogenic)
• Infections: Ehrlichia sp., Leishmania infantum; Histoplasma capsulatum; FeLV
Increased destruction (Immune-mediated thrombocytopenia)
• Drugs: sulfonamides, other antibiotics, anti-thyroidal drugs
• Recent vaccination?
• Neoplasia: lymphoma, malignant histiocytosis
• Infectious diseases: Ehrlichia sp., Babesia sp.; RMSF; Bartonella sp.?,
• Other autoimmune diseases: SLE
Most patients with ITP that present to first-line practitioners are relatively stable. Although immunosuppression in most cases is appropriate and often rapidly leads to normalization of platelet count, diagnostic testing to determine if there are concurrent diseases causing secondary ITP or thrombocytopenia by non-immune-mediated mechanisms is always recommended first. If the history or physical examination reveal any clinical signs or findings that do not fit with a 'classic' case of ITP then these should be pursued as appropriate. Unlike dogs with hemolytic anemia, I am less willing to institute immunosuppressive therapy without thorough diagnostic testing, because it is much more common for me to find secondary causes of thrombocytopenia than anemia. Additionally, primary ITP is often much more difficult to treat than IMHA, and thus finding any cause of secondary disease is very much to the patient's advantage.
I always perform a full minimum database—complete blood count, full serum chemistry panel, and urinalysis. I only recommend a urine culture if the urinalysis suggests an infection may be present. Thoracic and abdominal radiographs and abdominal ultrasound are ideal. I consider a fundic examination to be part of a normal physical examination; evidence of vasculitis (suggesting rickettsial disease or hematologic malignancies), granulomas (fungal lesions), or neoplastic cells (particularly lymphoma) may be detected. Careful palpation of lymph nodes is done multiple times, and if there is any hint of lymphadenopathy, I aspirate and at a minimum look at the slide myself. If in the appropriate geographic region, I submit serum for tick-borne disease titers, and start doxycycline therapy while waiting for results.
The final diagnostic test I perform in every case of ITP is a bone marrow aspirate. Severe thrombocytopenia is not considered a contraindication for bone marrow aspiration, as bleeding is limited by the bone itself, and severe bleeding from this site does not endanger the patient's life. As with IMHA, lymphoma is a common cause of secondary ITP, and this disease must be diagnosed prior to instituting immunosuppression. All the drugs used to treat ITP, particularly glucocorticoids, result in lysis of lymphocytes; therefore treatment may put these dogs into remission. Although this may be an apparent advantage, prednisone alone is not the optimal therapy for lymphoma—dogs with lymphoma treated with prednisone alone have a median survival of only 3 months, as opposed to longer survival rates with combination chemotherapy. Additionally administration of prednisone prior to the diagnosis of lymphoma worsens long-term prognosis, as glucocorticoids induce multi-drug resistance against many chemotherapeutics. That being said, the bone marrow aspiration does not provide information which influences diagnosis or case outcome in the vast majority of dogs with thrombocytopenia.
Glucocorticoids are only clearly indicated in primary ITP (in other words, if concurrent diseases are not identified). Because secondary causes of ITP usually result in platelet counts that are higher than the range which typically puts patients at risk for spontaneous bleeding, I do not routinely start immunosuppression in these cases. Additionally, treatment of the true cause of the secondary ITP should theoretically result in rapid normalization of the platelet count in most cases. However, there are always exceptions; I have seen several cases of ITP likely secondary to Ehrlichia sp. or RMSF in which I needed to treat with anti-inflammatory to immunosuppressive doses of prednisone in order to normalize platelet count, even after appropriate courses of doxycycline.
Once I have decided to immunosuppress my patient with glucocorticoids, I begin prednisone at 1 mg/kg PO q12h (in other words, immunosuppressive doses—2 mg/kg/day). Although other clinicians have advocated higher doses of prednisone during the 'induction' period, or feel that higher doses offer greater immunosuppression, I have not appreciated this; in my experience increases above this dose only worsen signs of hyperadrenocorticism and lead to more adverse side-effects. There are only rare instances where I choose to use alternative glucocorticoids. In animals that are vomiting I opt for intravenous dexamethasone. If this is necessary, recall that dose adjustments are needed to account for the difference in potency: dexamethasone has a potency of approximately 5 to 7 times that of prednisone, so instead of 1.0 mg/kg in one dose, I give 1.0 mg/kg÷ 5 = 0.2 mg/kg IV q12h. Some clinicians do feel that those dogs that do not respond to one glucocorticoid may respond to an alternate one; these clinicians will start oral dexamethasone or methylprednisolone after 3-4 weeks of unsuccessful immunosuppressive prednisone. Again, I am not an advocate of this, although I agree in principal that this may be true. Instead I administer vincristine (see below), repeat basic screening tests to look for secondary diseases that may have been missed, and then add a non-glucocorticoid immunosuppressive (azathioprine). I usually hospitalize dogs with orders for strict cage rest until platelet count is clearly rising, and has rapidly climbed above approximately 50,000.
Azathioprine is a purine analog; its biologic structure is similar to the basic building blocks of DNA, but when incorporated into new DNA during cell replication it results in slowed cell growth and death through mutations or apoptosis. Activated lymphocytes are the primary target of this drug. Although few controlled studies exist on the use of azathioprine in veterinary patients, it is the non-GC immunosuppressant with which we as a profession have the most anecdotal experience. The primary benefit of azathioprine is that because it has a different mechanism of action than GCs, simultaneous use may allow lower doses of GCs or more rapid tapering.
Use of azathioprine in IMHA is well accepted, and studies have supported a short- and long-term benefit in patients who are administered this drug as an adjunct to prednisone. Unfortunately, the same studies have not been performed in dogs with ITP. Unlike with IMHA I do not routinely start azathioprine administration at the time of diagnosis. Instead I begin this drug once patients have failed the first attempt to taper prednisone, or if I am unable to remove them from glucocorticoid therapy completely, or if they are particularly sensitive to the side-effects of prednisone.
Prednisone alone with or without azathioprine is generally considered to be the first-line therapy for ITP. Unlike with IMHA, cyclosporine has not been reported in the literature as an alternative or adjunctive therapy for ITP. I have never used this drug for treatment of ITP, but think it may be reasonable to consider if dogs do not respond to glucocorticoids within 2-3 weeks, and vincristine has been attempted.
As discussed in these proceedings under 'Evidence-based management of IMHA,' cyclophosphamide has fallen out of favor for IMHA. Because this drug affects all dividing cells, bone marrow stem cells are also suppressed, and thus erythrocyte regeneration is slowed in dogs with IMHA; prognosis is actually worse. For this reason I do not recommend attempting use of cyclophosphamide in ITP, under the assumption that platelet production would be blunted as well.
Whole blood transfusion (and platelet-rich plasma)
Most dogs with primary ITP usually present with peripheral platelet counts less than 20,000, and often less than 10,000. Although these dogs are theoretically at risk for spontaneous bleeding, in truth this is rarely noted. As mentioned above petechiae or ecchymoses or epistaxis are not uncommon, but bleeding into the lungs, pleural space, or CNS is very rare without concurrent trauma. Routine transfusion of ITP dogs with ultralow platelet counts is not recommended for several reasons. First, the amount of blood which would be required to increase a patient's peripheral platelet count is huge, and unrealistic. Second, transfused platelets are likely destroyed more rapidly than the patient's own cells because in addition to being targeted by the anti-platelet antibodies causing the ITP, they are also inherently more antigenic (we do not cross-match for platelet compatibility). Finally, platelets have a relatively short half-life, and repeated transfusions would be necessary to maintain the peripheral platelet count above 20,000. Despite these limitations, transfusions are definitely indicated in those rare patients with ITP that present with or develop life-threatening hemorrhage. Whole blood or platelet-rich plasma is required in these cases.
Vincristine's antineoplastic effects occur secondary to inhibition of the mitotic spindle assembly that must occur for cell division to proceed. By perhaps similar mechanisms vincristine also induces premature fragmentation of megakaryocytes and increases peripheral platelet count. It is unclear however if these immature platelets have normal function, or if the beneficial effect of vincristine is partially due to macrophages engulfing vincristine-loaded platelets which then inhibit further phagocytosis.
The use of vincristine has been evaluated in a single prospective case series. Vincristine administration was associated with more rapid normalization of platelet count and shorter duration of hospitalization. No adverse effects were seen, and there was no increased incidence of spontaneous bleeding that would imply a platelet function defect. Interestingly the authors of this study did not report long-term survival of the two different treatment groups, and it was the individual clinician's preference as to whether a given patient would receive vincristine or not. Regardless of these study limitations, I generally give vincristine to those dogs with primary ITP and platelet counts less than 20,000 who have not responded to standard immunosuppression within 3 to 5 days.
Human-derived concentrated immunoglobulin (IVIG) is an expensive, sporadically available treatment option for dogs with IMHA. The mechanism of action of IVIG is unclear, but may be by blockade of macrophage receptors which would normally bind to erythrocyte-bound antibody. Prospective controlled studies have not been performed as of yet, but a case series examining the use of IVIG for treatment of dogs with presumptive primary ITP reported more rapid normalization of platelet count and shorter duration of hospitalization.
Long term monitoring and therapy
Those animals with ITP that are successfully in remission should be tapered slowly; in general platelet count should be within the normal range for a minimum of 1 month, and preferably longer, prior to the first decrease in prednisone dose. It is my impression that animals that are tapered too soon may become resistant to the immunosuppressive effects of prednisone, and thus can become more difficult to treat. Additionally, unlike IMHA where clinical signs are often sensed by the owner when mild to moderate anemia recurs, patients with ITP have rapid declines in their platelet count, and re-present with ultralow counts. Beginning dose reduction may be less risky after shorter periods of remission in those cases where autoimmune disease has developed secondary to a clearly identified concurrent disease that has been adequately resolved. An example is ITP secondary to rickettsial diseases; in these cases I will attempt to decrease the dose after only two weeks of remission while receiving immunosuppressive doses of prednisone.
Dose decreases are usually done by either decreasing the mg/kg dose in half or by extending the interval between doses without changing the amount given. The most common tapering technique is to decrease the dose in half from immunosuppressive to anti-inflammatory doses in one or two steps, and then switch to alternate day dosing of the anti-inflammatory dose. Each step should be at least 2, and preferably 4 weeks apart. ITP dogs should have normal platelet counts prior to each taper step—a slightly decreased platelet count should be considered lack of remission. Clinicians should not rely on estimations of platelet count if the platelets are obviously clumped; a repeat platelet count should be performed prior to dose reduction. Dogs usually show minimal side-effects of chronic glucocorticoid administration once they are receiving alternate day dosing, so I tend to keep my patients on this dose for several months before withdrawing glucocorticoids completely.