To the best of our knowledge, this is the first case report suggesting that an abrupt increase in serum creatinine was associated with deferoxamine during the period of colistin therapy after CBT. It remains unclear whether deferoxamine is indicated for hyperferritinemia within the first few months after CBT. In addition, the frequency of deferoxamine-associated AKI is unclear. Deferoxamine infusion as chelation treatment is considered particularly useful for hyperferritinemia patients after HSCT due to significant reduction of total body iron burden [11]. In the present case, intravenous deferoxamine was administered to avoid continuous excess iron overload, with ferritin levels fluctuating between approximately 6000 and 13000 ng/mL during peri-CBT period.

A total of 217 records was identified through the database. Based upon inclusion criteria, 4 case report and 2 case series were eligible. As shown in Table 2, deferoxamine has been associated with AKI in 10 cases, including adult aged ≥ 18 years in 7 and less than 18 years in 3, in individual case reports and small case series studies [7, 12,13,14,15,16]. Eight of 10 cases developed AKI with increase in serum creatinine more than 2 folds compared with baseline value or reduction of urine volume less than 0.5mL/kg/hr. Development of AKI within 24 hours was reported in 3 identified cases [12, 13, 16]. In the 3 cases, sudden increases in serum creatinine were caused by accidental administration of high dose deferoxamine.

In this case, AKI developed during the period of treatment for GVHD. Indeed, GVHD is suggested as a common risk factor for AKI after HSCT [17, 18]. In this regard, treatment with steroid was continued, resulting in the amelioration of GVHD. Consequently, no symptoms of dehydration due to diarrhea caused by GVHD were observed during the period of deferoxamine administration.

In ADR assessment using Naranjo probability scale, the scores of deferoxamine and colistin indicated “possible” ADR, respectively. Therefore, no definitive conclusion has been reached. In this regard, since idiosyncratic ADRs would not have high scores indicating a definite or probable ADR using the Naranjo scale, it has not been validated for use in patients with critical illness or suffering specific organ toxicity [19]. In addition, the clinical manifestations of colistin-associated nephrotoxicity are the following; occurrence within the first 5 days of colistin treatment, recovery in serum creatinine within 1 to 2 weeks after the discontinuation in moderate AKI, and development of cylinduria [20,21,22,23]. However, this case was discordant with these properties. Furthermore, the following various risk factors of colistin-associated nephrotoxicity were suggested; dose ≥5 mg/kg/day, length of colistin treatment, receiving concomitant multiple nephrotoxic agents, age, gender, hypoalbuminemia less than 2.0 or 3.2 g/dL, hyperbilirubinemia more than 5 mg/dL in total-bilirubin, and severity of the illness. [6, 23, 24]. In this regard, while the length of colistin exposure and gender were relevant in this case, the duration of colistin therapy included an interruption of 10 days. Therefore, we consider the length of exposure to have had less influence. By contrast, in deferoxamine-associated AKI, the duration required for recovery to the baseline value in serum creatinine was 2 weeks to 1 month [16]. The time course in this case was concordant with the report. In addition, plausibility is an indispensable criterion in assessing the causal relationships [25, 26]. For the plausibility of development of AKI, we considered the rapid creatinine elevation to be a result of administration of deferoxamine rather than colistin.

While the approved infusion rate of deferoxamine was less than 15 mg/kg/hr, the standard intravenous administration is 40–50 mg/kg/day over 8–12 hours in adults [27]. Deferoxamine was administrated at a high infusion rate (approximately 12 mg/kg/hr) in this case. In 5 of 10 cases with AKI, deferoxamine with infusion rate more than 10 mg/kg/hr was administrated intravenously (Table 2). Koren et al demonstrated that in an animal model, AKI was caused by the administration of deferoxamine of 5 mg/kg/hr [16]. The rapid intravenous infusion of deferoxamine may be associated with AKI. In addition, there is a limitation in the procedure of the literature review, due to the search strategy focused upon the change of creatinine and/or urine volume. Clearly, further studies of the relation between AKI and infusion rate are required.

In our case, a sudden rise in serum creatinine was observed immediately following initiation of deferoxamine. Colistin is considered to cause acute tubular necrosis due primarily to tubular epithelial cell membrane permeability [28]. Meanwhile, deferoxamine reduces renal perfusion via inhibition of prostanoid synthesis [29]. This synergic nephrotoxicity causing tubular damage is a possible explanation for the AKI observed in our case.

Causality was not rigorously proven due to a number of limitations of the assessment for kidney function, using only serum creatinine as well as ADR scoring. Despite the limitations, in this post-CBT patient with hyperferritinemia and MDRP, deferoxamine use may have resulted in nephrotoxicity. Therefore, careful monitoring of kidney function is required in recipients of HSCT treated with deferoxamine.