Thrombocytopenia due to low-dose colchicine therapy: A possible drug interaction with nivolumab and implications for supportive care.

Hyperuricemia and gout often accompany hematologic and oncologic disorders. Colchicine, long used to manage gout, is rarely associated with hematologic toxicity. Infrequently, pancytopenia accompanied by gastrointestinal abnormalities may follow high dose or intravenous colchicine administration in subjects with underlying renal or hepatic dysfunction [1 – 3]. A case of thrombocytopenia during low-dose colchicine administration for gout prophylaxis is now described and a drug interaction with the investigational PD-1 pathway inhibitor, nivolumab is suggested [4].


To the Editor,
Hyperuricemia and gout often accompany hematologic and oncologic disorders. Colchicine, long used to manage gout, is rarely associated with hematologic toxicity. Infrequently, pancytopenia accompanied by gastrointestinal abnormalities may follow high dose or intravenous colchicine administration in subjects with underlying renal or hepatic dysfunction [1 -3]. A case of thrombocytopenia during low-dose colchicine administration for gout prophylaxis is now described and a drug interaction with the investigational PD-1 pathway inhibitor, nivolumab is suggested [4].

Case report
A 70-year-old male was diagnosed with malignant melanoma metastatic to brain, lung, rib, liver, mesentery and small intestine in February 2013. He received gamma knife treatment of brain metastases on two occasions; a fi ve-day course of IL-2 during April 2013 (without a tumor response); and a single intravenous dose of ipilumab (3 mg/kg) combined with nivolumab (1 mg/kg) on 7 June 2013. He then developed diarrhea; C. difi cile colitis; small bowel obstruction with gastrointestinal bleeding (requiring ileal resection on 31 July 2013); and asymptomatic portal and left common iliac vein thromboses on 26 August 2013 (treated with 1.5 mg/kg/day of enoxaparin subcutaneously).
Intravenous nivolumab (3 mg/kg) every two weeks was begun on 29 August 2013. The fi rst dose was followed by seven days of fever (temperatures between 37.9 o C and 39.2 o C). Eleven days after beginning nivolumab, podagra developed and oral colchicine therapy was begun at a dose of 1.8 mg per day with rapid improvement. The serum uric acid level was 7.6 mg/dl two days after the onset of symptoms (normal range ϭ 3.5 -7.0 mg/dl) but had been 5.9 mg/dl seven weeks previously. Colchicine was discontinued after three days but podagra recurred 13 days later.
Colchicine was reinstituted for one month. However, acute bilateral knee pain and swelling developed six days after colchicine was discontinued. The serum uric acid level had increased to 8.6 mg/dl and remained Ͼ 7.5 mg/dl thereafter. Joint symptoms again rapidly resolved with the reinstitution of colchicine, which was then continued for 12 weeks. The platelet count was 160 000/mm 3 (normal range ϭ 150 000 -350 000/mm 3 ) when the third course of colchicine was instituted but fell to 125 000/mm 3 eight weeks later and reached a nadir of 115 000 mm 3 after 12 weeks ( Figure 1a). The absolute neutrophil count (ANC) also decreased during this time to a nadir of 2200/mm 3 ( Figure 1b) but hemoglobin values ranged from 13.9 g/dl to 14.8 g/dl. Serum BUN, creatinine, alanine aminotransferase, aspartate aminotransferase, bilirubin, alkaline phosphate and gamma glutamyl transferase values were all normal. By six weeks after colchicine was discontinued, the platelet count returned to normal and the ANC increased to 3200/mm 3 . Platelet counts remained between 132 000/mm 3 and 151 000/mm 3 during an additional nine months of follow-up on continued nivolumb therapy with all but three of 19 values Ն 145 000/mm 3 while the ANC remained Ն 3200/ mm 3 . Biweekly nivolumab was continued throughout the course and daily enoxaparin was continued until six weeks after colchicine was discontinued. The patient never experienced nausea, vomiting or diarrhea. Platelet counts on four occasions during the year prior to the development of metastatic disease were 139 000/ mm 3 to 175 000/mm 3 (mean ϭ 159 000/ mm 3 ).

Discussion
In this subject, treatment with colchicine was instituted for recurrent episodes of gout since both nonsteroidal anti-infl ammatory agents and steroids were contraindicated by ongoing treatment with enoxaparin and nivolumab, respectively, and since concerns about drug interactions precluded the use of allopurinol [5].
Thrombocytopenia developed eight weeks after beginning the third course of colchicine ( Figure 1a) and resolved 4 -6 weeks after colchicine was discontinued. More profound thrombocytopenia may have occurred had not blood counts been obtained every two weeks on the investigational protocol. The ANC also decreased during colchicine therapy but did not reach neutropenic levels (Figure 1b). This time course suggests that thrombocytopenia was due to non-immune marrow damage, a fi nding consistent with previous reports [1,2]. Neither enoxaparin nor nivolumab alone are likely responsible for the fall in the platelet count or ANC since recovery occurred despite continued use of both these agents.
Although hematologic toxicity due to colchicine has been noted with high doses given intravenously to subjects with renal or hepatic dysfunction [1,2], no cases of hematotoxicity occurred in two studies of the long-term ( Ͼ 2 year) use of low-dose colchicine (1 -1.8 mg/day) in a total of 293 patients for the prophylaxis of either gout or familial Mediterranean fever [1,2,5,6]. It is of note then, that colchicine metabolism involves both CYP3A4 and P-glycoprotein and that interactions between colchicine and other drugs metabolized by these pathways have been described [3,7]. Although metabolism of nivolumab is undefi ned, other monoclonal antibodies (i.e. rituximab and cetuximab) block P-glycoprotein activity [8,9]. Moreover, nivolumab and other PD-1 antagonists may impair CYP-mediated metabolism [10]. Thus, a drug interaction between colchicine and nivolumab can explain the thrombocytopenia observed in this subject. Since enoxaparin does not interact with either P-glycoprotein or the CYP450 system, it is unlikely that thrombocytopenia resulted from an interaction of colchicine and enoxaparin [11].
Many antineoplastic drugs and symptom-directed medications also interact with CYP3A4 and P-glycoprotein (see Supplementary Table I, to be found online at http://informahealthcare.com/doi/ abs/ 10.3109/0284186X.2014.1002572) [12 -18] and the meta bolism of other agents in both these categories remain undefi ned. Thus, similar interactions may not be infrequent in oncologic settings. However, while the possibility of drug-drug interactions involving supportive care medications in oncology patients has been suggested [18], there are no previous reports of hematologic toxicity resulting from the concurrent use of anti-neoplastic and symptom-directed medications. Thus, these interactions may go unrecognized with cytopenias attributed solely to the chemotherapeutic agents, leading to erroneous dose reduction or discontinuation of anti-neoplastic treatments. This possibility is of particular importance with the earlier implementation of palliative care approaches in the setting of advanced malignancies [19].

Declaration of interest:
The author has no competing interests to disclose and there were no sources of funding for this study. The author is solely responsible for all aspects of this study and for manuscript preparation. Figure 1. Platelet and neutrophil counts related to oral colchicine therapy. Dashed line indicates lower normal platelet count level of 150 000/mm 3 . Numbered boxes on x-axis indicate courses of oral colchicine therapy as follows: 1.8 mg/day for 3 days, 1.8 mg/ day for 3 days; then 1.2 mg/day for 11 days; then 0.6 mg/day for 14 days, 1.8 mg/day for 1 day; then 1.2 mg/day for 7 days; then 0.6 mg/day for 65 days, ANC, absolute neutrophil count.

To the Editor,
Technological advances in radiotherapy have led to increased interest in the use of high-dose treatments given in only a few fractions to try and exploit favorable anti-tumor effects of hypo-fractionation. The class of treatments combining extra-cranial location, high dose, more extreme hypo-fractionation, organ at risk avoidance/sparing and a high level of accuracy is often referred to as stereotactic body radiotherapy (SBRT) [1]. It requires appropriate equipment, the relevant personnel with adequate training and expertise, and careful patient selection. Despite several previous surveys [2 -4], a substantial number of unknowns remain concerning the contemporary practice of SBRT, especially within Europe. There is, for example, limited information about the numbers of patients that have been treated with SBRT, approaches to implementation, attitudes towards the evidence base, resource utilization and reimbursement. We therefore performed a survey of SBRT practice in six selected European countries.