Adding cyclosporin A to eltrombopag for aplastic anemia secondary to chemotherapy for solid cancers.

Abstract

To the Editor: Myelosuppression is one of the most common side effects of chemotherapy. Some patients would develop long-lasting aplastic anemia (AA) even after stopping chemotherapy for over three months, due to persistent destruction of the bone marrow. As a thrombopoietin-receptor agonist, eltrombopag (EPAG) has been widely used in patients with thrombocytopenia from various origins,[1] including those with AA secondary to chemotherapy. However, both EPAG and cyclosporin A (CsA) have been reported to possibly induce the expansion of abnormal clones or controversially increase the risk of neoplasms.[2] Thus, the challenge was to treat severe cytopenia while preventing cancer relapse. Our study reported the efficacy and safety of EPAG with or without CsA for patients with AA secondary to chemotherapy, in an attempt to find solutions to these difficulties. This retrospective study was approved by the Ethics Committee of Peking Union Medical College Hospital (PUMCH, No. S-K1976). Patients were informed of this study and gave their written informed consent to participate. We reviewed data from patients with solid tumors treated with chemotherapy at PUMCH, who developed AA from January 1, 2018, to November 20, 2021. Participants were deemed potentially eligible if they had clear diagnosis of solid cancers without previous hematologic malignancies and persistent cytopenia for at least six months after the last chemotherapy (radiation) treatment. The detailed inclusion criteria are shown in the [Supplementary materials, https://links.lww.com/CM9/B453]. Data were censored on May 21, 2022. EPAG was started at 25 mg/day and increased by 25 mg/day to the maximum dosage of 150 mg/day. CsA was started at 3 to 5 mg·kg−1·day−1 and adjusted monthly to achieve 100 to 200 ng/mL. The detailed methods of regimens, data collection, follow-ups, response criteria, statistical analysis and adverse events were presented in the [Supplementary materials, https://links.lww.com/CM9/B453]. Responses were assessed as complete response (CR), partial response (PR), and no response (NR).[3] Comparison between groups was conducted using Fisher's exact test or the Kruskal–Wallis test as appropriate. Sixty-two patients were reviewed. We excluded 17 patients (reasons are shown in the Supplementary materials, https://links.lww.com/CM9/B453), and 45 eligible patients were finally included. Group 1 included 32 patients who received EPAG and CsA simultaneously from the onset, namely EPAG + CsA group. Group 2 included 13 patients who received EPAG alone. The baseline characteristics of the two groups have been summarized in Supplementary Table 1 [https://links.lww.com/CM9/B453], including the baseline hemoglobin, neutrophils, and platelets. There were no statistically significant differences between patients in two groups at baseline. Group 1 patients were administered the combination regimen for 15 (3–37) months. The overall response rates (ORRs) at 1, 2, 3, and 6 months were 28% (9/32), 56% (18/32), 72% (23/32), and 100% (26/26), corresponding to 9% (3/32), 9% (3/32), 13% (4/32), and 42% (11/26) of CR, respectively [Table 1]. The median response time was 2 (1–5) months. Twenty-one patients who continued the combination regimen for over 12 months achieved at least PR (5 with PR and 16 with CR). A median follow-up of 18 (3–52) months revealed an ORR of 78% (25/32) before switching to other regimens, including 53% (17/32) of CR and 25% (8/32) of PR. Following at least three months of treatment, non-responders (patients did not respond at the time of evaluation, n = 9) either continued the treatment for a longer period (n = 4), discontinued the treatment and switched to other drugs (n = 2), received supportive care (n = 2), or were lost to follow-up (n = 1). During treatment with EPAG + CsA, three patients relapsed with AA, whose details have been reported in the [Supplementary materials, https://links.lww.com/CM9/B453]. Table 1 - Responses of two cohorts cancer patients with chemotherapy-induced AA. Characteristics EPAG + CsA (n = 32) EPAG (n = 13) Total (n = 45) OR (95% CI) P value Time to response (months) 2 (1–5) 2 (1–2) 2 (1–5) – 0.30 Response at one month 9 (28) 2 (15) 11 (24) 2.1 (0.3–23.4) 0.47∗ CR 3 1 4 0.5 (0.005–52.2) 1.00† PR 6 1 7 – – Non-responders 23 11 34 – – Average cumulative dose (mg) 1922 ± 790 1933 ± 808 1925 ± 786 – 0.97 Median cumulative dose (mg) 1500 (750–4500) 1500 (750–3000) 1500 (750–4500) – 0.88 Response at two months 18 (56) 4 (31) 22 (49) 2.8 (0.6–15.3) 0.19∗ CR 3 1 4 0.6 (0.03–41.9) 1.00† PR 15 3 18 – – Non-responders 14 9 23 – – Average cumulative dose (mg) 4078 ± 1602 4125 ± 1516 4092 ± 1560 – 0.93 Median cumulative dose (mg) 3938 (1500–9000) 3750 (1500–6000) 3750 (1500–9000) – 0.83 Response at three months 23 (72) 4 (31) 27 (60) 5.5 (1.2–31.2) 0.02∗ CR 4 1 5 0.6 (0.04–41.4) 1.00† PR 19 3 22 – – Non-responders 9 9 18 – – Average cumulative dose (mg) 6246 ± 2478 6433 ± 2229 6300 ± 2385 – 0.81 Median cumulative dose (mg) 6000 (2250–13500) 6750 (2250–9000) 6000 (2250–13500) – 0.62 Response at six months 26 (100)‡ 2 (33)‡ 28 (88)‡ Inf (4.4–Inf) <0.01∗ CR 11 1 12 0.7 (0.009–62.7) 1.00† PR 15 1 16 – – Non-responders 0 4 4 – – Average cumulative dose (mg) 11723 ± 5643 12500 ± 5254 11869 ± 5498 – 0.76 Median cumulative dose (mg) 12000 (4050–27000) 12750 (4500–18000) 12000 (4050–27000) – 0.59 Data are expressed as n, n (percentage), median (range) or mean ± standard deviation.∗For comparing the overall response rate.†For comparing the proportion of CR patients among responders between EPAG + CsA and EPAG.‡Six patients in the EPAG + CsA group and seven patients in the EPAG group did not use their regimens for ≥ 6 months. AA: Aplastic anemia; EPAG: Eltrombopag; CI: Confidence interval; CsA: Cyclosporin A; CR: Complete response; Inf: Infinite; OR: Odds ratio; PR: Partial response; –: Not available. The cancer-related treatment without recurrences during EPAG + CsA was as follows: Twenty patients did not need cancer-related treatment; two patients received endocrine therapy; three patients re-started surgery, chemotherapy, or radiotherapy after CR/PR; one patient switched to an investigational drug and underwent surgery followed by chemoradiation; one patient was scheduled with four additional chemotherapy cycles; one patients received supportive care, but cancer therapy was required. Cancer recurrences were confirmed in three patients (two with ovarian cancer and one with lung cancer) or suspected in one patient (breast cancer) during follow-up at 10, 14, 14, and 5 months after EPAG + CsA administration, respectively. They were all re-evaluated and actively treated [Supplementary materials, https://links.lww.com/CM9/B453]. Supplementary Figure 1, https://links.lww.com/CM9/B453 presented the detailed timeline of patients with cancer progression (metastasis and recurrence). Patients in Group 2 were exposed to EPAG for 5 (3–17) months. The ORRs at 1, 2, 3, and 6 months were, respectively, 2/13, 4/13, 4/13 and 2/6, with 1/13, 1/13, 1/13, and 1/6 of CR, respectively. The median time to response was 2 (1–2) months (P = 0.30 in comparison with Group 1) [Table 1]. Two patients who received EPAG monotherapy for more than 12 months achieved PR and CR, respectively. After a median of 16 (3–20) months of follow-up, the final ORR in Group 2 was 4/13, with 1/13 of CR. Non-responders at 3 months either continued EPAG (n = 6), underwent supportive therapy (n = 2), or were lost to follow-up (n = 1). One patient developed AA relapse [Supplementary materials, https://links.lww.com/CM9/B453]. The cancer-related therapies during the course of EPAG monotherapy were as follows: eight patients did not require any cancer-related treatment; one patient had supportive care though requiring treating cancer; and one patient was waiting on the results of gene tests. Cancer progression was observed in three patients [Supplementary Figure 1, https://links.lww.com/CM9/B453]: one with upper gastrointestinal cancer at 17 months, one with lung cancer at 14 months, and one with lower gastrointestinal cancer at 18 months; clonal evolution (JAK2 mutation) was detected in one patient at 10 months of EPAG. Careful re-assessment and active treatment were provided [Supplementary materials, https://links.lww.com/CM9/B453]. The ORRs of EPAG + CsA were statistically significantly higher than those of EPAG alone at 3 and 6 months (P = 0.02 and <0.01, respectively), but not statistically at 1 and 2 months, with similar cumulative exposure doses of EPAG at respective timepoints. The two groups had similar time to response and CR rates at 1, 2, 3, and 6 months [Table 1]. The overall survival (OS) and progression-free survival (PFS) from EPAG were not statistically significantly different between the two groups (P = 0.15 and 0.33, respectively). There were no statistically significant association between overall response at three months and all baseline characteristics, (all P ≥0.05). The OS and PFS from EPAG were not statistically significantly different between responders and non-responders (P = 0.62 and 0.78, respectively) but was significantly different between stages II and III (P = 0.04 and 0.02, respectively). Supplementary Table 2, https://links.lww.com/CM9/B453 provides a summary of adverse events in the two groups. Collectively, patients in Group 1 experienced more adverse events than Group 2, but the difference was not statistically significant (P = 0.10). Most events were mild and manageable. Herein, we first compared the efficacy of EPAG and EPAG + CsA in patients with chemotherapy-induced AA secondary to treatment for solid tumors. Previous reports showed 26% to 58% CR rate and 80% to 94% ORR in more severe types of AA.[4] We observed a similar CR rate (42%) but a slightly higher ORR (100%). The reason might be that patients were less severe refractory than those in previous studies. It is rare to find data on the first-line clinical practice of EPAG + CsA for non-severe AA, especially in real-life situations. A prospective clinical trial is currently underway to evaluate EPAG + CsA for moderate AA (NCT02773225),[1] and the results may shed light on the effectiveness of the combination therapy. The comparison for the first six months revealed statistically significant differences between the two groups at 3 and 6 months on the basis of similar EPAG doses. Such outcome suggested that CsA played a pivotal role in treating chemotherapy-induced AA. Our previous pilot study demonstrated that CsA monotherapy, after failure of EPAG, can achieve 36% and 64% ORR at 3 and 6 months, respectively, hence benefiting these AA patients. Different from transient myeloid suppression, chemotherapy-induced AA is likely due to a persistent immune response, and therefore stimulating hematopoiesis by itself is insufficient. Therefore, CsA is important not only for reawakening hematopoiesis but also for maintaining efficacy over time. Moreover, many responders of EPAG + CsA stopped EPAG but continued CsA and achieved a durable response in their later follow-up period. Hence, our results indicated that EPAG in combination with CsA was superior to EPAG alone. The primary concern regarding the use of CsA to patients with solid tumors is the risk of cancer recurrence or clonal evolution.[1] Neither OS nor PFS significantly differed between the two groups. The time from cancer diagnosis to relapse in seven patients was similar to the average time reported in the literature for cancer patients without AA.[5] In addition, the stage and type of cancer also contribute to the progression of the disease. Six of the seven recurrent cases were initially at stage III, which posed a high risk of progression. All seven cases of newly recurrent cancers were types with high recurrence risks. There were four ovarian cancers and one lower gastrointestinal cancer that recurred prior to the AA diagnosis. Their PFS was similar to that after receiving AA regimens. These two types of cancer were also involved in cancer progression after EPAG. As a result, we hypothesized that cancer relapse following an AA treatment may be correlated with tumor characteristics, such as type and stage. However, the follow-up time is still a limitation to drawing definite conclusions. The main limitation of this study was the imbalanced number of patients between the two groups. Second, the number of patients and follow-up time are limited for drawing a robust conclusion. Another limitation was the retrospective nature. Future studies could compare the two regimens in prospective studies. In conclusion, adding CsA to EPAG statistically significantly improved hematologic response in patients with AA following chemotherapy for solid cancers. Cancer progression was not statistically significantly different with or without CsA, at least in the short term. Funding This study was supported by the Chinese Acadmy of Medical Sciences Innovation Fund for Medical Sciences (CIFMS 2021-1-I2M-003) and PUMCH National High-Level Hospital Clinical Research Funding (Nos. 2022-PUMCH-C-026, 2022-PUMCH-D-002, 2022-PUMCH-B-046). Conflicts of interest None.