Tumor lysis syndrome in the era of novel and targeted agents in patients with hematologic malignancies: a systematic review

Abstract Effective new treatments are now available for pa- tients with hematologic malignancies. However, their propen- sity to cause tumor lysis syndrome (TLS) has not been sys- tematically examined. A literature search identified published Phase I–III clinical trials of monoclonal antibodies (otlertuzumab, brentuximab, obinutuzumab, ibritumomab, ofatumumab); tyrosine kinase inhibitors (alvocidib [flavopiridol], dinaciclib, ibrutinib, nilotinib, dasatinib, idelalisib, venetoclax [ABT-199]); proteasome inhibitors (oprozomib, carfilzomib); chimeric antigen receptor (CAR) T cells; and the proapoptotic agent lenalidomide. Abstracts from major congresses were also reviewed. Idelalisib and ofatumumab had no reported TLS. TLS incidence was ≤5 % with brentuximab vedotin (for anaplastic large-cell lympho- ma), carfilzomib and lenalidomide (for multiple myeloma), dasatinib (for acute lymphoblastic leukemia), and oprozomib (for various hematologic malignancies). TLS incidences were 8.3 and 8.9 % in two trials of venetoclax (for chronic Electronic supplementary material The online version of this article (doi:10.1007/s00277-015-2585-7) contains supplementary material, which is available to authorized users.
lymphocytic leukemia [CLL]) and 10 % in trials of CAR T cells (for B-cell malignancies) and obinutuzumab (for non- Hodgkin lymphoma). TLS rates of 15 % with dinaciclib and 42 and 53 % with alvocidib (with sequential cytarabine and mitoxantrone) were seen in trials of acute leukemias. TLS mitigation was employed routinely in clinical trials of alvocidib and lenalidomide. However, TLS mitigation strate- gies were not mentioned or stated only in general terms for many studies of other agents. The risk of TLS persists in the current era of novel and targeted therapy for hematologic ma- lignancies and was seen to some extent with most agents. Our findings underscore the importance of continued awareness, risk assessment, and prevention to reduce this serious poten- tial complication of effective anticancer therapy.

Tumor lysis syndrome (TLS) results from the rapid release of potassium, phosphorus, and nucleic acids following the lysis of malignant cells. This shift of intracellular constituents into the extracellular space may overcome normal homeostatic mechanisms, leading to TLS and life-threatening acute kidney injury, arrhythmias, and neurologic complications [1, 2]. TLS can occur spontaneously; however, it usually results from ini- tial anticancer treatment in association with rapid destruction of tumor cells.Prevention of clinical TLS depends on heightened aware- ness, risk stratification, and risk-based management. Although TLS sometimes occurs with solid tumors, hematologic malig- nancies pose the highest risk, especially bulky, chemosensitive disease [3]. Identification of at-risk patientsand stratification by degree of risk (low, intermediate, high) is critical [4, 5]. Vigorous hydration, close monitoring of elec- trolytes, uric acid and renal function, and control of uric acid, phosphorus, and potassium levels with allopurinol or rasburicase, phosphate binders, and potassium-lowering agents are the cornerstones of management [5].Treatments for hematologic malignancies have rapidly evolved in the past decade, but TLS incidence with contem- porary treatment regimens is not well understood.

Their im- proved antitumor activity may increase TLS risk; indeed, TLS emerged as an important, and sometimes fatal, toxicity during the early clinical development of several new agents [6–9]. We conducted a systematic literature review to examine the incidence of TLS with both recently approved and emerging novel and targeted agents.We evaluated selected recently approved agents and late-stage compounds for the treatment of hematologic malignancies: alvocidib (flavopiridol), brentuximab vedotin, chimeric anti- gen receptor (CAR) T cells, carfilzomib, dasatinib, dinaciclib, ibritumomab, ibrutinib, idelalisib, lenalidomide, nilotinib, obinutuzumab, ofatumumab, oprozomib, otlertuzumab, and venetoclax (ABT-199). Published clinical trials were identi- fied in the National Library of Medicine’s PubMed database from January 1, 2010 to October 31, 2014, using searcheswith Bagent (title) AND phase (any field)^. Identified abstracts were reviewed for inclusion/exclusion; articles that appeared to report Phase I to III clinical trials to treat hematologic ma-lignancy were selected. Articles were retrieved and reviewed for mention of TLS. Congress abstracts from the American Society of Clinical Oncology (ASCO) and American Society of Hematology (ASH) annual meetings were searched with Embase, using the same time period and search strategy. Congress abstracts were reviewed for reporting of TLS, in- cluding mention of a mitigation strategy, either preventative or therapeutic. The search strategy and cumulative number of abstracts and articles published each year about new antican- cer agents are shown in Figs. 1 and 2.The ASH/ASCO meeting abstracts were reviewed for the agents with fewer than five published articles. Congress ab- stracts and published articles were cross-checked to avoid du- plication of data stemming from the same study; if the same study had been reported as an abstract and in an article, only the article was used. If the same study was reported in >1 abstract, the most recent abstract was used. All articles and abstracts were reviewed by ≥2 authors to ensure accurate col- lection of TLS-related information. There were no discordant reviews, which would have been managed by additional re- viewers and discussion until consensus was achieved.

Our findings show an increasing number of new and emerging agents for hematologic malignancies (Fig. 2). There was sub- stantial heterogeneity regarding the number of assessable ar- ticles and/or abstracts. Most of the articles and abstracts de- scribed nonrandomized phase I or II studies (Table S1). Notably, only 23 % of articles and 11 % of abstracts discussed TLS (Table 1).Tumor lysis syndrome was not mentioned for the radioimmunotherapeutic monoclonal antibody ibritumomab, the anti-CD37 monoclonal antibody otlertuzumab, or the breakpoint cluster region-Abelson (BCR-ABL) inhibitor nilotinib (Table 1). Therefore, these agents are not discussed further. TLS was observed in patients with acute leukemia and high-grade non-Hodgkin lymphoma (NHL), mantle cell lym- phoma (MCL), indolent B-cell NHL, chronic lymphocytic leukemia (CLL), and multiple myeloma (MM; Tables 2, 3, 4, and 5).Reports of TLS for the broad cyclin-dependent kinase (CDK) inhibitor alvocidib were assessed in six published Phase I (n = 5) or Phase II (n = 1) clinical trials conducted in various hematologic malignancies (Table 2) [6, 10–14]. Among Phase I trials, three reported no TLS [12–14], one reported an inci- dence of 4.2 % (a single case in a patient with acute myeloid leukemia; AML) [11], and one trial of alvocidib combined with fludarabine and rituximab in patients with NHL, MCL, or CLL reported an incidence of 13 % (5/38) [10]. A Phase II trial in poor-risk AML, which assessed alvocidib in a sequen- tial regimen with cytarabine plus mitoxantrone, found a TLS incidence of 42 % [6].

The TLS prevention strategy in the Phase II trial was primarily oral allopurinol with the phosphate binder sevelamer [6, 15]. The Phase I trials described varying prevention strategies (Table 2). One trial, which reported no TLS among high-risk CLL, provided the most com- prehensive strategy: allopurinol for the entire trial dura- tion in conjunction with hydration, urine alkalinization, phosphate binder, and rasburicase with the first two doses [12]. Another Phase I trial with no incidence of TLS (alvocidib plus vorinostat in relapsed/refractory acute leukemias, mainly AML) attributed the lack of TLS to aggressive prophylaxis and monitoring as re- quired per protocol [14]. Only one study, a Phase I MCL/NHL/CLL study of alvocidib in combination with fludarabine and rituximab, used a prevention strategy involving modification of the alvocidib dosing regimen: alvocidib was initiated during cycle 2 (per a protocol amendment) to reduce TLS risk [10].
Among nine published articles, one mentioned TLS in asso- ciation with brentuximab vedotin, an antibody drug conjugate containing an anti-CD30 antibody plus the antimicrotubular agent monomethyl auristatin E [16]. This Phase II trial of single-agent brentuximab for anaplastic large-cell lymphoma reported TLS in 1/58 patients (1.7 %) [16]. TLS developed after the first dose of brentuximab vedotin and was managed by unspecified supportive care measures [16].One of seven published clinical trials using CD19-targeted CAR T-cell therapy contained information on TLS incidence [17]. In a Phase II trial of single-agent CAR T-cell therapy for persistent B-cell malignancy post allogeneic hematopoietic stem cell transplantation, one of ten patients developed TLS 8 days after the infusion. TLS emerged despite prophylactic administration of allopurinol initiated 1 day before infusion. Rasburicase was used to manage serum uric acid levels [17].

Four of 12 published articles [18–21] provided TLS informa- tion for the irreversible proteasome inhibitor carfilzomib (Table 3). The articles reporting results of single-agent carfilzomib in relapsed/refractory MM found a TLS incidence of 2.7 % in a Phase I/II trial [21], 4.3 and 0.4 % in two Phase II trials [18, 19], and 1 % in an integrated safety analysis that included four Phase II clinical trials [20]. In the Phase I/II trial and one of the Phase II trials, each of which reported a single case of TLS, the prevention strategy was allopurinol for at-risk and high-risk patients [19, 21]. The other studies did not report TLS prophylaxis or management strategies. Although no articles for the BCR-ABL and Src family kinase inhibitor dasatinib reported on TLS, two of three abstracts did provide information [22, 23]. One analyzed the relative effi- cacy and safety of dasatinib versus imatinib according to base- line medication use in patients with chronic-phase chronic myeloid leukemia. Approximately half of the patients re- ceived prophylactic allopurinol to prevent TLS, but its inci- dence was not reported [22]. The second abstract reported dasatinib in combination with low-intensity chemotherapy in elderly patients with Philadelphia chromosome–positive acute lymphoblastic leukemia (ALL), with dasatinib given during induction (with vincristine and dexamethasone), consolida- tion (sequentially with methotrexate and L-asparaginase and then cytarabine), and maintenance (alternating with 6-mercap- topurine/methotrexate and dexamethasone/vincristine). During induction, 3/71 patients (4.2 %) experienced renal fail- ure attributed to TLS [23].

One publication [24] and one abstract [25] provided TLS infor- mation for dinaciclib, a CDK1, 2, 5, and 9 inhibitor. The article reported results of a Phase II trial that evaluated single-agent dinaciclib in advanced leukemias (ALL or AML) [24]. Among 20 evaluable patients, 3 had grade ≥3 clinical TLS, for an inci- dence of 15 %. One patient with AML required hemodialysis but ultimately succumbed to complications of acute renal failure. This patient had laboratory evidence of TLS before study treat- ment and had received prophylaxis with rasburicase, calcium acetate, and sevelamer and aggressive management with insulin, calcium gluconate, and sodium polystyrene sulfonate. The ab- stract, describing a Phase I trial in relapsed/refractory CLL, re- ported a TLS incidence of 15 % (5/33 patients, 2 of whom required dialysis). These findings prompted plans for stepped- up dosing for prevention of TLS in future dinaciclib studies [25].None of 4 articles, but 2 of 21 abstracts, reported TLS informa- tion for the Bruton tyrosine kinase (BTK) inhibitor ibrutinib [26, 27]. A Phase II trial of single-agent ibrutinib in elderly patients with previously untreated or treated CLL/small lymphocytic lymphoma reported no TLS in the first 26 patients completing >2 cycles [26]. The second abstract, describing interim results of a Phase Ib/II study in patients with relapsed/refractory CLL re- ceiving ibrutinib, bendamustine, and rituximab, found a 6.7 % TLS incidence (2/30 patients) [27]. No mention was made of strategies for TLS prevention or management.

Three of five articles [28–30] provided TLS information re- garding idelalisib, a small molecule inhibitor of phos- phatidylinositol 3-kinase. The three studies, in relapsed/ refractory MCL, relapsed/refractory CLL, and previously treated indolent NHL, specifically noted that no patients de- veloped TLS. No information was provided for TLS preven- tion or management [28–30].Among 27 articles for lenalidomide, an immunomodulatory analog of thalidomide, 8 mentioned TLS [31–38] (Table 4). Six Phase II trials and one Phase I trial for CLL or various lymphomas were single-agent studies; one was a Phase II trial of lenalidomide in combination with rituximab [35, 36]. TLS was noted with regard to prophylaxis in only one study [31]; three trials reported no cases of TLS [32, 34, 37]; three trials reported one case each for incidences of 1.7, 1.9, and 4.0 % Four of 9 articles and 2 of 11 abstracts provided TLS informa- tion for the anti-CD20 (type II) monoclonal antibody obinutuzumab [39, 40]. Two Phase I studies and one Phase I/ II study reported single cases of TLS among patients with NHL or CLL, for incidences of 3.0, 4.5, and 4.8 %, respectively [41–43]. A Phase II trial in relapsed/refractory diffuse large B-cell lymphoma found an incidence of 5 % (2/40 patients) [44].

Only the Phase I/II study reported a prevention strategy: high-risk TLS patients were medicated with acetaminophen/ paracetamol and an antihistamine preinfusion [43]. One of 12 articles for the anti-CD20 monoclonal antibody ofatumumab reported TLS. TLS did not emerge among 81 patients in a Phase II trial of single-agent therapy for relapsed/ progressive diffuse large B-cell lymphoma [45]. Only 1 of 37 congress abstracts mentioned TLS; a Phase II trial of ofatumumab plus lenalidomide found no cases of grade 3/4 TLS among 34 patients with relapsed CLL who had prior treatment with a purine analog [46]. The abstract noted that patients received oral allopurinol 300 mg/day during the first 2 weeks of the first treatment cycle for prevention of TLS.Among three abstracts for oprozomib, a structural analog of carfilzomib (no relevant articles were available), one reported TLS incidence [47]. A Phase Ib/II single-agent study in pa- tients with hematologic malignancies who had failed ≥1 pre- vious line of therapy had 1 case of TLS (characterized as a dose-limiting toxicity) among 42 patients, an incidence of 2.4 %. No details were provided for TLS prevention or man- agement [47].Six abstracts were identified for the second-generation small- molecule B-cell lymphoma/leukemia 2 inhibitor venetoclax, but there were no published trials [7, 9, 48–51] (Table 5). All abstracts reported Phase I clinical trial results, and five of six reported ≥1 case of TLS, with incidences of 2.7, 3.2, 4.5, 8.3, and 8.9 %, respectively [7, 9, 48–51]. Three of these studies, in patients with relapsed/refractory CLL, reported one fatality each [7, 9, 51]. Limited information regarding TLS prevention was provided other than protocol-specified stepwise increases in the dosing to improve initial tolerance [9, 48–51].

In this systematic review, we characterize TLS associated with novel and targeted agents for hematologic malignan- cies, a setting considered at highest risk of this toxicity in the context of traditional cytotoxic chemotherapy. TLS has been largely underreported in the current era, as reflected in how few of the articles and abstracts mentioned TLS incidence (even if it was 0 %), or strategies for its pre- vention and management. The available evidence suggests that incidence varies by the treatment regimen (new agent or new combinations of agents), hematologic malignancy subtype, number of prior relapses (a surrogate for chemosensitivity), and supportive care employed.
ASCO American Society of Clinical Oncology, ASH American Society of Hematology, CLL chronic lymphocytic leukemia, DLBCL diffuse large B-cell lymphoma, MCL mantle cell lymphoma, NHL non-Hodgkin lymphoma, SLL small lymphocytic leukemia, TLS tumor lysis syndromeTumor lysis syndrome in association with innovative new treatments came to light during the early clinical development of some of the agents covered here. For venetoclax (ABT-199), TLS-related fatalities put a temporary halt to development despite strong signals of efficacy [52]. Stepped dosing of venetoclax was adopted for TLS prevention. Similarly, during early development of lenalidomide for CLL, enrollment into a clinical trial was suspended following two TLS-associated deaths among seven cases. The trial was ultimately resumed but with a modified dos- ing regimen [53]. One lenalidomide trial in relapsed/refractory CLL was initially a Phase II/III trial of 2 different doses, but it was later amended to a Phase I dose-escalation trial because of 4 cases of serious TLS among the first 18 patients. Subsequent trials adopted a prespecified TLS prevention strategy, with rou- tine use of prophylactic allopurinol. Overall, these strategies ap- pear to have been effective in controlling TLS incidence, al- though some breakthrough cases have occurred.

Considering the various hematologic malignancies for which data are available, venetoclax and lenalidomide exhibit the highest propensity
for inducing TLS in CLL (Tables 4 and 5). Alvocidib has notably high rates of TLS in patients with poor-risk AML, approaching or exceeding 50 % [6, 15]. A lower, yet still substantial, TLS incidence of 15 % has been seen with dinaciclib in advanced acute leukemias (AML and ALL) [24] and in CLL [25]. In MM, carfilzomib trials have reported one to two cases per study (Table 3). There were no cases of TLS in studies with idelalisib [28–30] or ofatumumab [45, 46]. Evidence for TLS was limited to single trials for brentuximab vedotin, CAR T cells, and oprozomib; the pau- city of available data and small numbers of patients who had received these agents at the time of this analysis preclude firm conclusions about their association with TLS, though caution is warranted based on the single trials and anecdotal reports.
In this era of innovative agents that act by novel (and not necessarily fully elucidated) mechanisms, we encourage reporting of the incidence of TLS and strategies for TLS pre- vention and management. When the per-protocol TLS preven- tion strategy was specified for the trials analyzed here, allopu- rinol was a typical component. Rasburicase is another established option, although our analysis found that it was rarely used prophylactically. The National Comprehensive Cancer Network (NCCN) recognizes both allopurinol and rasburicase as options, specifying the latter for patients with certain risk factors in NHL (presence of any high-risk feature, urgency to start therapy for high-bulk disease, inadequate hy- dration, acute renal failure) [54].

It should be noted that half of adverse events are dis- covered after a product comes to market. Postmarketing pharmacovigilance via the US Food and Drug Administration’s MedWatch program and other indepen- dent pharmacovigilance groups, as well as mandatory follow-up studies from drug manufacturers, are critical for ensuring patient safety. Furthermore, it is essential that clinicians stay current with TLS risks of not only newly approved agents but also traditional cytotoxic agents, for which new information may become available. For exam- ple, Stevens–Johnson syndrome and toxic epidermal necrolysis have emerged as serious and potentially fatal complications of concurrent use of bendamustine (which carries a known risk of TLS) and allopurinol, as noted in the bendamustine product labeling [55, 56]. Future re- search efforts should seek to determine if there are any nuances specific to TLS prophylaxis with newer noncytotoxic agents. Additional research is also needed to elucidate the underlying mechanism of TLS occurrence with biologic agents that pose relatively high risks. For example, a population pharmacokinetic/pharmacodynamic modeling study found that the only established alvocidib- associated TLS biomarker was the glucuronide metabolite, which may also apply to dinaciclib and any other agents for which metabolism and excretion involve glucuronidation [57]. Finally, clinicians must be aware that new agents are often studied in multiply relapsed patients who are less likely to respond briskly. TLS risk may be heightened when these agents are brought into frontline use or used in combination with other treatments.

The risk of TLS persists in the current era of novel and targeted therapy. Although TLS is primarily reported in patients with acute leukemia and high-grade NHL, it has also been observed in patients with other hematologic cancers, including MCL, indolent B-cell NHL, CLL, and MM. The rapid development of highly effective agents and new combination treatment reg- imens suggest that increasing numbers of patients will be at high risk for TLS. Our findings underscore the importance of heightened awareness of TLS, upfront risk assessment, and preventative intervention as we move to incorporate innova- tive, highly effective treatments Oprozomib for hematologic malignancies.