Eculizumab Dosing Intervals Longer than 17 Days May Be Associated with Greater Risk of Breakthrough Hemolysis in Patients with Paroxysmal Nocturnal Hemoglobinuria

Abstract

Eculizumab given bi-weekly is widely recommended for the treatment of paroxysmal nocturnal hemoglobinuria (PNH). We undertook a retrospective analysis on the medical records of 763 dosings of 14 PNH patients to investigate whether a threshold would exist in dosing intervals associated with breakthrough hemolysis. We identified 12 events of breakthrough hemolysis in 4 patients. Multivariate logistic regression and receiver operating characteristics (ROC) analysis revealed a significant association between increased risk of breakthrough hemolysis and prolonged dosing intervals of 17 days or more and concomitant inflammation: odds ratios (OR) and 95% confidence intervals (CIs) were 1.6 (1.3–2.0, p<0.01) and 5.5 (1.3–22.8, p=0.02), respectively. ROC analysis showed that the best cut-off dosing interval discriminating breakthrough hemolysis was 16.5 days. We consider that eculizumab dosing intervals longer than 17 days may be associated with an increased risk for developing breakthrough hemolysis in patients with PNH.

Lack of the complement inhibitor CD59 on the erythrocyte surface is largely responsible for clinical manifestation of paroxysmal nocturnal hemoglobinuria (PNH). Such patients are susceptible to chronic uncontrolled complement activation resulting in sustained intravascular hemolysis and morbidities including fatigue, dyspnea and abdominal pain.^1,2) Eculizumab is a humanized monoclonal antibody that specifically binds human C5 complement protein with high affinity, blocking generation of the potent pro-inflammatory C5a and the cell lytic terminal complement complex C5b-9,^1,3) thereby ameliorating most of the clinical symptoms of PNH. Its clinical efficacy and tolerability have been demonstrated in clinical trials.^4–10)

While eculizumab is effective in most patients, breakthrough hemolysis occurs in some patients during the maintenance phase. Risk factors for breakthrough hemolysis remain largely unknown. Biweekly dosing of eculizumab at 900 mg is recommended in the maintenance period, according to clinical trial data.^6,7) However, rigorous adherence to the recommended dosing interval may be difficult to implement in reality due to many practical reasons.^11,12) To the best of our knowledge, it remains unclear whether inevitable prolongation of the dosing interval may increase the risk of breakthrough hemolysis. We retrospectively reviewed the medical records of 14 patients with PNH who received long-term eculizumab therapy in a single medical center.

PATIENTS AND METHODS

Data Collection

The medical records of PNH patients aged 20 years or older who received intravenous infusions of eculizumab between February 1, 2010 and July 31, 2014 at NTT Medical Center Tokyo were reviewed. Dosing schedule of eculizumab was as follows: 600 mg by intravenous infusion over 35 min or longer weekly for the initial 4 weeks, followed by a 900 mg infusion after one week (initial period), and thereafter 900 mg biweekly (recommended interval 14±2 d) until the end of the study (maintenance period). Patients’ demographic data; eculizumab dosing interval; blood biochemistry including lactate dehydrogenase (LDH), total bilirubin, aspartate aminotransferase (AST) and alanine aminotransferase; complete blood cell counts and concomitantly administered drugs were collected. Eculizumab was administered after blood sampling for biochemistry and complete blood cell counts. All patients were vaccinated with a meningococcal vaccine at least 14 d before the commencement of eculizumab therapy. Clinical data obtained from patients while they underwent high-dose steroid therapy (i.e., equivalent to prednisolone 11 mg/d or higher) were excluded from analysis. For 11 patients who started the eculizumab therapy at NTT Medical Center Tokyo, the data obtained when the maintenance dose of eculizumab was commenced were considered as baseline data. For 3 of 14 patients who were referred to our hospital after they had begun eculizumab therapy at other hospitals, the data obtained immediately after their referral were considered as baseline data. The protocol of the present study was approved by Ethics Committee of NTT Medical Center Tokyo prior to the study.

Definition of Breakthrough Hemolysis

Since no consensus was reached regarding the definition of breakthrough hemolysis in patients with PNH, in the present study we adopted the criteria of breakthrough hemolysis taking into account the laboratory data reported previously⁵⁾: simultaneous elevations of serum LDH and AST levels to 1000 U/L or higher and 90 IU/L or higher, respectively. We reviewed the clinical courses of the patients and retrieved the dosing intervals preceding and those not preceding breakthrough hemolysis events. When breakthrough hemolysis developed, subsequent data of dosing intervals were excluded from analysis until LDH levels returned to baseline values or lower. In addition, patients were diagnosed as having systemic inflammation when they showed relevant clinical symptoms and C-reactive protein (CRP) levels 0.4 mg/dL or higher. We assumed that patients’ adherence to eculizumab therapy was acceptable when dosing intervals were between 12 and 16 d. We calculated the percentage of dosing intervals with acceptable adherence to all dosing intervals in each patient.

Statistical Analysis

We performed univariate analyses to screen covariates, including dosing interval, that were associated with the risk of developing breakthrough hemolysis, and variables with p values less than 0.10 were considered eligible for multivariate analysis. When a significant covariate was detected by the multivariate analysis, receiver operating characteristics (ROC) analysis was performed to calculate its sensitivity and specificity. Correlation between dosing interval and LDH level was examined by Pearson’s correlation analysis. Data are expressed as the mean±standard deviation (S.D.) or median and range as appropriate. A p value less than 0.05 was considered significant throughout the study. Analyses were performed using IBM SPSS Statistics 21.0 (IBM Japan Inc., Tokyo, Japan).

RESULTS

Table 1 shows the demographic and relevant clinical data of the 14 patients. At baseline, median (range) of serum LDH was 246 (154–388) U/L, which was significantly lower than those observed before initiation of eculizumab therapy (data not shown). During the maintenance period, serum LDH levels showed fluctuations between 76 and 2648 U/L in all subjects throughout the study period. Concomitant medications included warfarin in 7 and prednisolone in 7 patients. Dosing intervals of eculizumab during the maintenance period ranged from 4 to 24 d (Table 2). Among the data of 968 doses collected, 763 met the inclusion criteria and 12 events of breakthrough hemolysis were observed (1%). The distributions of all dosing intervals were as follows; 12 (2%) were 11 d or less, 88 (12%) were either 12 or 13 d, 450 (59%) were 14 d, 97 (13%) were either 15 or 16 d, 116 (15%) were 17 d or longer. The mean (±S.D.) dosing interval that preceded breakthrough hemolysis was 19±3 d and individually 2 were 14 d, 1 was 17 d, 1 was 18 d, 2 were 19 d, 4 were 21 d, 2 were 23 d. Adherence to the recommended dosing interval between 12 and 16 d was achieved in 83% of all dosing intervals. Although 10 of 14 patients maintained stable LDH levels (76–675 U/L), 4 male patients experienced 12 episodes of abrupt elevation of serum LDH level diagnosed as breakthrough hemolysis (Table 2). In addition, a weak, positive correlation (r=0.243, p<0.001) was observed between serum LDH and length of dosing interval for the 763 dosing intervals. One patient (Patient 7) showed very low adherence (22%) to the recommended dosing interval. The four patients who developed breakthrough hemolysis had dosing interval adherence rates of 55 to 96%.

Table 1. Baseline Demographic Clinical Characteristics of Patients with PNH Receiving Eculizumab

Patient	Age	Gender	Height (cm)	Weight (kg)	Disease duration (years)	LDH (U/L)	Cre (mg/dL)	AST (IU/L)	ALT (IU/L)	T-Bil (mg/dL)	RBC (×104/µL)	Hb (g/dL)
1	38	M	181	70	22	212	0.86	22	30	2.8	381	13.7
2	83	F	146	43	2.2	258	1.69	19	31	1.7	258	8.9
3	23	M	173	69	3.3	233	0.62	12	9	2.6	240	8.3
4	43	M	175	74	4.5	328	0.77	12	8	4.9	265	8.3
5	35	F	151	57	17	225	0.62	35	100	1.0	341	10.8
6	60	M	173	63	33	388	1.02	60	78	1.5	203	7.4
7	44	M	171	76	5.9	312	0.75	18	29	2.4	417	14.4
8	70	F	157	43	4.5	306	0.65	20	12	0.7	223	7.3
9	28	M	178	90	1.5	154	0.73	17	19	2.1	205	6.4
10	42	M	175	78	8.5	276	0.94	25	26	1.4	376	11.5
11	39	M	168	70	3.9	209	0.69	22	27	1.7	337	8.6
12	66	M	166	53	36	161	1.15	30	50	2.6	319	8.9
13	43	M	NA	103	10	258	1.08	52	85	1.5	382	13.0
14	42	M	169	62	14.2	227	0.91	24	32	2.9	331	12.9

Abbreviations: M, male; F female; Cre, serum creatinine concentration; LDH, lactate dehydrogenase; PNH, paroxysmal nocturnal hemoglobinuria; AST, aspartate aminotransferase; ALT, alanine aminotransferase; T-Bil, total bilirubin; RBC, red blood cell; Hb, hemoglobin; NA, not available.

Table 2. Clinical Characteristics and Distributions of Dosing Intervals during Eculizumab Maintenance Therapy

Patient	Duration of eculizumab therapy (years)	Total number of doses	Number of dosing intervals [number of breakthrough hemolysis events]							Dosing interval (d)	Number of inflammation episodes
			≤11 d	12–13 d	14 d	15–16 d	≥17 d	Adherence (%)	LDH (U/L)
			12 (2%)	88 (12%)	450 (59%)	97 (13%)	116 (15%)
1	6.5*	78	0	3	48 [1]a)	6	21[2]	74	202 (161–1415)	16±3	1
2	4.0	95	1	16	66	10	2	97	212 (76–456)	14±1	73
3	3.8	47	1	8	24	12	2	94	238 (208–585)	14±2	3
4	3.7	51	0	8	36	7	0	100	251 (200–364)	14±1	5
5	3.3	78	1	11	47	13	6	91	202 (164–345)	15±2	6
6	3.5	75	3	19	32 [1]b)	16	5[2]	89	420 (351–2529)	14±2	13
7	3.5	62	0	2	8	3	49	21	277 (235–472)	19±3	5
8	2.3	41	0	0	39	1	1	98	336 (240–675)	14±1	4
9	2.4	28	0	3	21	3	1[1]	96	180 (151–2648)	14±2	8
10	2.1	51	1	8	28	7	7	84	262 (233–393)	15±2	2
11	2.0	38	1	0	12	9	16[5]	55	256 (195–2196)	17±3	15
12	1.8	40	1	2	32	4	1	95	201 (174–232)	14±1	7
13	4.5*	74	3	6	55	5	5	89	209 (152–334)	14±1	6
14	3.8*	5	0	2	2	1	0	100	265 (222–316)	14±1	1

Abbreviations: LDH, lactate dehydrogenase. Data for LDH are median (range). a) Erythema annulare centrifugum and b) common cold were observed. * Patients who started eculizumab therapy before they were referred to NTT Medical Center.

Univariate analysis showed that age, body weight, dosing interval and inflammation may be associated with breakthrough hemolysis. However, multivariate analysis identified only the length of dosing interval and concomitant inflammation as independent variables significantly associated with breakthrough hemolysis: odds ratios (OR) and 95% confidence intervals (CIs) were 1.6 (1.3–2.0, p<0.01) and 5.5 (1.3–22.8, p=0.02), respectively (Table 3). The contribution of dosing intervals to the overall risk of breakthrough hemolysis was smaller than that of inflammation. Furthermore, ROC analysis showed that the best cut-off dosing interval discriminating the presence and absence of breakthrough hemolysis was 16.5 d, with sensitivity and specificity of 0.83 and 0.86, respectively.

Table 3. Multivariate Analysis for Risk of Breakthrough Hemolysis

Variable	Odds ratio	95% Confidence interval	p Value
Age	1.0	0.9–1.0	0.48
Body weight	1.0	0.9–1.1	0.88
Dosing interval	1.6	1.3–2.0	<0.01
Systemic inflammation	5.5	1.3–23.4	0.02

DISCUSSION

In the present study, we found that a delay of 3 d or longer from the recommended dosing interval (14 d) may be associated with an increased risk of breakthrough hemolysis. No significant associations were observed among for demographic characteristics and baseline biochemistry. Our data that dosing intervals longer than 17 d would pose a risk of breakthrough hemolysis appear compatible with the description of recommended dosing interval of 14±2 d in the prescribing information of eculizumab in U.S.A.¹³⁾ To our knowledge, the present study is the first retrospective cohort study to support the appropriateness of the eculizumab dosing interval, 14±2 d, based on previous clinical trials.^6–8) While Hillmen et al. did not clearly describe the threshold dosing interval beyond which the risk of breakthrough hemolysis increases, they reported that all patients developing breakthrough hemolysis were successfully treated by either shortening interval or increasing eculizumab doses.^2,9)

Our finding of a significant association between inflammation and development of breakthrough hemolysis is compatible with previous finding that continuous complement activation by infections may trigger hemolysis.^1,2) Nevertheless, we cannot categorically rule out the possibility that breakthrough hemolysis developed by chance in our 4 patients during the observation period, because we collected data from only 14 patients. We propose that the dosing interval of eculizumab during the maintenance phase should not be delayed to longer than 17 d, until further convincing data is available, particularly in patients with a history of breakthrough hemolysis.

Caution should be exercised for generalizing our findings. First, the subjects in the present study appeared to be good responders to eculizumab, because all showed reductions of serum LDH to 154–388 (U/L) after the commencement of eculizumab therapy (Table 1). Our findings may not be relevant for poor responders to eculizumab as reported by Nishimura et al.¹⁴⁾ In addition, while surgery, trauma and pregnancy were suggested to be risk factors of breakthrough hemolysis,^1,2) these covariates were excluded from our multivariate logistic analysis mainly because they were associated with only a small number of doses (8 and 16 dosing intervals for trauma and pregnancy, respectively).

We cannot draw a definitive conclusion about the relationship between plasma eculizumab and breakthrough hemolysis, because we did not monitor the drug concentrations in the present study. However, we consider it possible to explain that breakthrough hemolysis might have been caused by low plasma concentrations of eculizumab due to insufficient dosage or augmented plasma clearance or prolonged dosing intervals. While the reported mean plasma elimination half-life of eculizumab is 11.3±3.4 d,¹⁾ there is a large intra and interindividual variability in eculizumab plasma concentration–time courses in previous clinical trial and some patients exhibited plasma drug concentrations close or lower than the plasma drug levels required to suppress C5 activation.^4,5,15) Recently, Peffault de Latour et al. reported obesity (body mass index 29) might have been associated with low plasma eculizumab concentrations (<50 µg/mL) and frequent hemolytic attacks in a patient.¹⁵⁾ However, we found no significant association between body weight of patients and the development of breakthrough hemolysis. Although Hillmen et al. observed breakthrough hemolysis in 10% of 195 PNH patients receiving eculizumab,⁹⁾ they did not analyse the associated between the preceding dosing intervals of the drug and occurrence of the events. Therefore, further studies are required to individualize eculizumab dosage for a safety administration.

There are a number of limitations in the present study. We conducted statistical analysis assuming that there is little interindividual difference in the relationship between preceding dosing interval and risk of breakthrough hemolysis in all patients, although this assumption may not be totally true. Indeed, Patients 6 and 9 had breakthrough hemolysis even though their average dosing intervals were 14 d with high adherence rate of approximate 90% (Table 2). In addition, when Patients 1 and 6 unexpectedly developed breakthrough hemolysis at dosing interval of 14 d, they developed erythema annulare centrifugum and common cold, respectively. On the other hand, 8 patients did not develop breakthrough hemolysis even though their dosing intervals were prolonged longer than 17 d. We consider that we need more patients to draw definitive answer on this issue. However, it would be practically difficult to recruit a sufficient number of patients because PNH is a rare disease. In addition, plasma eculizumab concentrations, serum hemolytic activity and total amount of serum C5 levels were not measured. While anti-complement activity of eculizumab correlates directly with plasma drug concentration, and essentially complete blockade of hemolytic activity is achieved by plasma trough levels >35 µg/mL.⁴⁾ Ideally, the accurate relation between breakthrough hemolysis and prolongation of dosing interval should be determined using pharmacokinetics and pharmacodynamics analysis. However, the measurement of plasma eculizumab concentrations is routinely not possible under clinical conditions. In addition, the standard dosing intervals may not be applied to all PNH patients, mainly due to a large interindividual variability of the complement activation triggered by inflammations including infections and trauma. Recently, Peffault de Latour et al. reported the usefulness of a 50% hemolytic complement (CH50) assay for assessing the residual functional activity of C5. Because this test is available as one of the routine clinical laboratory tests, it would be useful for exploring reasons why some patients do not develop hemolysis with dosing intervals longer than 14 d.¹⁵⁾ Despite these limitations, our data support the hypothesis that an inevitable prolongation of eculizumab dosing interval is associated with breakthrough hemolysis in eculizumab responders.

In conclusion, the present study suggests that eculizumab dosing intervals longer than 17 d may have a greater risk for developing breakthrough hemolysis in patients with PNH. Further studies are warranted to validate our findings.

Conflict of Interest

KU received a research grant from Alexion Pharmaceuticals, Inc. The remaining authors declare no conflicts of interest.

REFERENCES

• 1) Rother RP, Rollins SA, Mojcik CF, Brodsky RA, Bell L. Discovery and development of the complement inhibitor eculizumab for the treatment of paroxysmal nocturnal hemoglobinuria. Nat. Biotechnol., 25, 1256–1264 (2007).
• 2) Röth A, Dührsen U. Treatment of paroxysmal nocturnal hemoglobinuria in the era of eculizumab. Eur. J. Haematol., 87, 473–479 (2011).
• 3) Thomas TC, Rollins SA, Rother RP, Giannoni MA, Hartman SL, Elliott EA, Nye SH, Matis LA, Squinto SP, Evans MJ. Inhibition of complement activity by humanized anti-C5 antibody and single-chain Fv. Mol. Immunol., 33, 1389–1401 (1996).
• 4) Hillmen P, Hall C, Marsh JC, Elebute M, Bombara MP, Petro BE, Cullen MJ, Richards SJ, Rollins SA, Mojcik CF, Rother RP. Effect of eculizumab on hemolysis and transfusion requirements in patients with paroxysmal nocturnal hemoglobinuria. N. Engl. J. Med., 350, 552–559 (2004).
• 5) Hill A, Hillmen P, Richards SJ, Elebute D, Marsh JC, Chan J, Mojcik CF, Rother RP. Sustained response and long-term safety of eculizumab in paroxysmal nocturnal hemoglobinuria. Blood, 106, 2559–2565 (2005).
• 6) Hillmen P, Young NS, Schubert J, Brodsky RA, Socié G, Muus P, Röth A, Szer J, Elebute MO, Nakamura R, Browne P, Risitano AM, Hill A, Schrezenmeier H, Fu C, Maciejewski J, Rollins SA, Mojcik CF, Rother RP, Luzzatto L. The complement inhibitor eculizumab in paroxysmal nocturnal hemoglobinuria. N. Engl. J. Med., 355, 1233–1243 (2006).
• 7) Brodsky RA, Young NS, Antonioli E, Risitano AM, Schrezenmeier H, Schubert J, Gaya A, Coyle L, de Castro C, Fu CL, Maciejewski JP, Bessler M, Kroon HA, Rother RP, Hillmen P. Multicenter phase 3 study of the complement inhibitor eculizumab for the treatment of patients with paroxysmal nocturnal hemoglobinuria. Blood, 111, 1840–1847 (2008).
• 8) Kelly RJ, Hill A, Arnold LM, Brooksbank GL, Richards SJ, Cullen M, Mitchell LD, Cohen DR, Gregory WM, Hillmen P. Long-term treatment with eculizumab in paroxysmal nocturnal hemoglobinuria: sustained efficacy and improved survival. Blood, 117, 6786–6792 (2011).
• 9) Hillmen P, Muus P, Röth A, Elebute MO, Risitano AM, Schrezenmeier H, Szer J, Browne P, Maciejewski JP, Schubert J, Urbano-Ispizua A, de Castro C, Socié G, Brodsky RA. Long-term safety and efficacy of sustained eculizumab treatment in patients with paroxysmal nocturnal haemoglobinuria. Br. J. Haematol., 162, 62–73 (2013).
• 10) Hillmen P, Elebute M, Kelly R, Urbano-Ispizua A, Hill A, Rother RP, Khursigara G, Fu C, Omine M, Browne P, Rosse W. Long-term effect of the complement inhibitor eculizumab on kidney function in patients with paroxysmal nocturnal hemoglobinuria. Am. J. Hematol., 85, 553–559 (2010).
• 11) Marin D, Bazeos A, Mahon F, Eliasson L, Milojkovic D, Bua M, Apperley JF, Szydlo R, Desai R, Kozlowski K, Paliompeis C, Latham V, Foroni L, Molimard M, Reid A, Rezvani K, de Lavallade H, Guallar C, Goldman J, Khorashad JS. Adherence is the critical factor for achieving molecular responses in patients with chronic myeloid leukemia who achieve complete cytogenetic responses on imatinib. J. Clin. Oncol., 28, 2381–2388 (2010).
• 12) Chan KE, Thadhani RI, Maddux FW. Adherence barriers to chronic dialysis in United States. J. Am. Soc. Nephrol., 25, 2642–2648 (2014).
• 13) “Prescribing information. Soliris^® (eculizumab) Concentrated solution for intravenous infusion, Alexion Pharmaceuticals, Inc., Cheshire, U.S.A.”: 〈http://soliris.net/sites/default/files/assets/soliris_pi.pdf〉, cited 31 October, 2015.
• 14) Nishimura J, Yamamoto M, Hayashi S, Ohyashiki K, Ando K, Brodsky AL, Noji H, Kitamura K, Eto T, Takahashi T, Masuko M, Matsumoto T, Wano Y, Shichishima T, Shibayama H, Hase M, Li L, Johnson K, Lazarowski A, Tamburini P, Inazawa J, Kinoshita T, Kanakura Y. Genetic variants in C5 and poor response to eculizumab. N. Engl. J. Med., 370, 632–639 (2014).
• 15) Peffault de Latour R, Fremeaux-Bacchi V, Porcher R, Xhaard A, Rosain J, Castaneda DC, Vieira-Martins P, Roncelin S, Rodriguez-Otero P, Plessier A, Sicre de Fontbrune F, Abbes S, Robin M, Socié G. Assessing complement blockade in patients with paroxysmal nocturnal hemoglobinuria receiving eculizumab. Blood, 125, 775–783 (2015).