Real-world Experience of Rituximab in Immune Thrombocytopenia

Abstract

Immune thrombocytopenia (ITP) is a relapsing–remitting disease often requiring more than one line of therapy. Rituximab is a recommended second-line therapy, but the real-world data on its efficacy and safety from resource constraint settings is limited. We aimed to analyze the safety and efficacy of rituximab in ITP. This is a single-center, retrospective study. This study was conducted at a tertiary care hospital in Northern India from 2005 to 2019. On audit of medical records, all patients of ITP (n-513) who had received rituximab (n-81) were screened for inclusion. Patients whose response assessment was not possible were excluded. Finally, 66 patients were analyzed using statistical packages of Python v3.7. The cumulative incidence of overall response on day 20 was 30.61%, and day 30 was 51.72%. The median time to response was 28 day (range 21–51 day). Cumulative incidence of complete response was 16.67%, and partial response 37.88%. After a median follow-up of 789 day (range 181–5260 day), the cumulative incidence of relapse was 30.32%, 36.12%, and 56.57% at 1, 2, and 5 years respectively. There was no effect of age, sex, duration of disease, lines of therapy received, and platelet count on either cumulative incidence of overall response or relapse. ANA positivity was significantly related to the better cumulative incidence of overall response (p = 0.012), but not with relapse. Infusion-related reactions were the commonest adverse event noted (n-4, grade ≥ 3 CTCAEv4). Rituximab and its generic version are safe and effective second line agent in ITP with a good overall response and sustained response.

Introduction

Immune thrombocytopenia (ITP) is characterized by increased destruction of platelets with or without low production. Clinically, it has varied presentation ranging from mild mucocutaneous bleeding (petechiae/ecchymoses) to life-threatening bleeding (e.g., intracranial bleeding) [1]. The bleeding risk in such patients correlates with the platelet count and function [2, 3]. The American Society of Hematology has recommended treatment of all ITP patients with platelet count < 30,000/μL, irrespective of presence or absence of symptoms [4]. The corticosteroids are the first-line therapy, but relapse following a brief initial response is common; about half of these patients require second-line therapy. Splenectomy, rituximab, and thrombopoietin receptor agonists (TPO-RAs) are the three most widely used therapeutic approaches in the setting of ITP relapse [1, 4–7].

Rituximab is a chimeric monoclonal antibody against CD20 antigen present on B-lymphocytes. It reversibly depletes the B-lymphocytes with long term remission in ITP. A fixed duration of therapy for 4 to 6 weeks, along with an overall response rate (ORR) of 50–60%, are the advantages like using rituximab [8, 9]. Though the widespread use of innovator rituximab is limited by cost constraints, it has been partially overcomed by the availability of “generic” rituximab in the resource-constraint settings [10–12]. The literature on safety and efficacy of rituximab in ITP from resource-constraint settings is limited. We aim to study the safety profile, response rate, and long-term sustained response in patients with ITP following rituximab therapy.

Methods

This is a retrospective, single-center study conducted at a tertiary care hospital in northern India. The study was approved by the Institutional Review Boards (AHRR IEC44/2020). Written consents were obtained from the patients. The patients’ medical treatment details were accessed from the outpatient clinics, and inpatient hospital admission/discharge notes. After careful screening of the documents, data were captured and entered in the password protected excel sheet. The missing data were minimized by making phone calls to the patient or the family. ITP patients who were diagnosed as per the International Working Group Criteria (IWG standardized criteria), and treated with generic rituximab between 2005 and 2019 were included in this study [1]. All these patients had received at least one prior line of therapy before rituximab. All patients received intravenous rituximab (dose per infusion: 375 mg/m²). Those who had received additional drug with rituximab, and missing follow-up information, were excluded from the study (Fig. 1).

Fig. 1.

Flow diagram of the study design

The short-term (up to 1 year) and long-term adverse events were noted. Adverse events were graded as per the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE)v4.0 [13].

Definitions for the study [14, 15].

Partial response (PR): A platelet count of ≥ 30,000/μL or double from the baseline, with no active bleeding manifestations.

Complete response (CR): A platelet count of ≥ 100,000/μL with no active bleeding manifestations.

Overall response rate (ORR): CR and PR combined.

No response: Platelet count < 30,000/μL or < double from the baseline, or having active bleeding manifestations requiring another line of therapy.

Relapse: Any new-onset bleeding manifestation necessitating therapy or drop in platelet counts below 30,000/μL or patients who were started on another line of therapy or underwent splenectomy for low platelet counts.

Response duration: Time duration from the day of receiving rituximab to the day of documented relapse.

Statistical methods: The statistical analysis was done with statistical packages (scipy [16], pandas [17], and statsmodels [18]) in Python v3.7 (www.python.org). Continuous variables were summarised using the mean/median and standard deviation/interquartile range. Categorical variables were summarised using proportions. Time to overall and complete response (OR/CR) and relapse post rituximab were analyzed non-parametrically using Kaplan–Meier analysis. Log-rank analysis was used to compare time to event data among various variables. The p value < 0.05 was considered statistically significant, and 95% confidence intervals were used wherever statistical inference was drawn.

Result

Demographics: A total of 81 patients received rituximab from 2005 to 2019. Fifteen patients were excluded due to lack of treatment details or lost to follow-up. A total of 66 patients (27 males and 39 females) who received rituximab were included in this study. There were newly diagnosed (n = 9), persistent (n = 26), and chronic (n = 31) ITP in our study cohort. Generic rituximab was given to 92% (61/66) patients. The baseline demographic characteristics and pre-rituximab parameters are summarised in Table 1. The median age of the patient cohort was 24.5 years, the youngest was of 2.5 years, and the oldest was of 64 years. There were 20 children (age ≤ 12 years) included in this study. More than one-third of the patients had received at least two prior lines of therapy (24/66, 36.36%), followed by those who had received single or three prior lines of therapies. Four, five, and six lines of therapy were received by five, two, and one patient, respectively. The mean duration from ITP diagnosis to rituximab therapy was 792 days (range 21–6922 days). Eight patients (8/66, 12.12%) had undergone splenectomy prior to receiving rituximab. The mean platelet count at the time of receiving rituximab was 11,069/µL (range 1000–30,000/µL). Thirteen out of 54 patients(24.07%) were tested positive for antinuclear antibody (ANA) at the time of ITP diagnosis. All the patients were tested negative for Australian antigen (HBsAg) for hepatitis B and had a normal chest X-ray prior to the rituximab therapy. For pediatric patients (age ≤ 12 years) immunoglobuloin profile was also done before giving rituximab. All patients underwent, bone marrow evaluation before second line agent. Whenever suspected, flow cytometric analysis for presence of alpha beta double-negative T (DNT) cells was done. All patients were given rituximab via intravenous route at a weekly dose of 375 mg/m². All patients continued for four or more cycles of rituximab except for 10 patients. Four of these 10 discontinued due to adverse events, while remaining six could not continue due to withdrawal of consent to continue rituximab therapy.

Table 1.

Demographic profile and baseline clinical characteristics of the ITP patients who received generic rituximab

Variable	Median (range) or n (%)
Evaluable pateints	66
Females	39 (59.1%)
Age (years)	24.5 (2.5–64)
Interval from ITP diagnosis to rituximab (days)	792 (21–6922)
Number of prior therapies	3.5 (1–6)
One line	18 (27.27%)
Two lines	24 (36.36%)
Three lines	16 (24.24%)
Four lines	5 (7.57%)
Five lines	2 (3.03%)
Six lines	1 (1.51%)
Splenectomy
Yes	8 (12.12%)
Mean Platelet count (/μL)	11,069 (1000–30,000)
ANAa
Positive	13 (24.07%)
Negative	41 (75.93%)
Dose of Rituximab (weekly)	375 mg/m2
No. of rituximab doses
One	1
Two	3
Three	6
Four	54
Five	0
Six	3

^aANA was available for 54 pateints before they received rituximab

Response rate: The OR on day 20 was 30.61% (n = 20), and day 30 was 57.72% (n = 32). Six patients responded after day 30 of rituximab therapy. The median time to response was 28 days (range 21–51 days). An OR, PR, and CR was seen in 36 (54.54%), 25 (37.88%), and 11 (16.67%) patients respectively (Fig. 2). The response assessment is summarised in Table 2.

Fig. 2.

Kaplan–Meier curve showing the cumulative incidence of complete response and overall response

Table 2.

Response to generic rituximab in ITP patients

Variable	n (%) or median (range)	95% CIs
Time to response (median)	28 days (21–51 days)
Cumulative incidence of overall response
At day 20 of rituximab	20 (30.61%)	[19.42–41.82%]
At day 30 of rituximab	32 (51.72%)	[38.96–64.47%]
PR	25 (37.88%)
CR	11 (16.67%)
Time to relapse (days)	1139 days
	(384-not reached)
Cummulative incidence of relapse
At 1 year	10 (30.32%)	[14.73–45.92]
At 2 years	11 (36.12%)	[19.52–52.71]
At 5 years	17 (56.57%)	[38.4–74.74]

Factors affecting the response to rituximab: We studied parameters like demographics, disease duration, number of lines of therapy, splenectomy status, baseline platelet count, and ANA status for their impact on response to rituximab. Patients who were ANA positive had a better response than the ANA negative patients (p = 0.012) (Table 3). There was no clinically significant effect of parameters like sex, duration of disease, prior lines of therapy received, splenectomy, and platelet count on the OR to rituximab therapy in our study population.

Table 3.

Univariate analysis for response assessment at day + 20 and day + 30 after generic rituximab

	n	Overall response
		n	Cumulative incidence of overall response [95% CIs]	p
Total patients	66	36
On day + 20		18	30.61% [19.42–41.82%]
On day + 30		30	51.72% [38.96–64.47%]
Male	27	16		0.623
On day + 20			33.33% [15.55–51.11]
On day + 30			68.89%
Female	39	20
On day + 20			28.73% [14.36–43.11]
On day + 30			48.97% [31.94–66]
Age ≤ 18 years	24	10		0.218
On day + 20			25.23% [7.75–42.71%]
On day + 30			49.52% [21.36–77.14%]
Age > 18 years	42	26
On day + 20			33.64% [19.26–48.02%]
On day + 30			58.42% [42.7–74.14%]
Duration of ITP (Log day) ≤ 5.56	34	16		0.264
On day + 20			30.12% [14.45–45.78%]
On day + 30			46.35% [27.53–65.18%]
Duration of ITP (Log day) > 5.56	32	20
On day + 20			31.25% [15.19–47.3%]
On day + 30			59.78% [41.98–77.57%]
Paltelets ≤ 15,000/µL	47	9		0.673
On day + 20			28.13% [15.14–41.13]
On day + 30			–
Paltelets > 15,000/µL	16	25
On day + 20			48.86% [33.56–64.17]	0.373
On day + 30			31.25% [8.54–53.96]
≤ 3 prior lines of therapy	42	20
On day + 20			29.03% [15.16–42.9%]
On day + 30			47.21% [30.58–63.85%]
>3 prior lines of therapy	24	16
On day + 20			33.33% [14.47–52.19%]
On day + 30			64.45% [43.21–85.69%]
Splenectomy; Yes	8	3		0.297
On day + 20			23.44% [0–52.56%]
On day + 30			37.98% [1.76%–74.19%]
Splenectomy; No	58	33
On day + 20			33.17% [20.95–45.39%]
On day + 30			56.03% [42.08–69.97%]
ANA Positive	13	11
On day + 20			46.15% [19.05–73.25%]
On day + 30			79.12% [54.26–100%]
ANA Negative	41	17
On day + 20			24.85% [11.46–38.23%]	0.012
On day + 30			48.3%

Response evaluation at follow up: After a median follow-up of 789 days (range 181–5260 days), 17 (25.76%) patients continued to maintain the response (PR/CR). By the end of 1, 2, and 5 years, 10 (30.32%), 11 (36.12%) and 17 (56.57%) patients lost their response. Two patients lost their response after 5 years of rituximab therapy (Fig. 3). There was no effect of age, sex, duration of disease, lines of therapy received, ANA status, and platelet count on the loss of response (Table 4 and Fig. 3).

Fig. 3.

Kaplan-Meier curve showing the cumulative incidence of relapse following rituximab

Table 4.

Univariate analysis for relapse assessment

	n	1y Cumulative incidence of relapse [95% CI]	2y Cumulative incidence of relapse [95% CI]	5y Cumulative incidence of relapse [95% CI]	p
Total patients	19
Male	9	31.69% [8.33–55%]	39.06% [14.46–63.66%]
Female	10	33.05% [11.05–55.05%]	36.47% [13.59–59.35%]	50.64% [25.54–75.75%]	0.729
Age ≤ 18 years	4	28.74% [0.19–57.27%]	33.26% [3.02–63.5%]	39.16% [7.15–71.917]
Age > 18 years	15	33.22% [12.2–52.24%]	37.99% [18.01–57.98%]	69.37% [47.25–91.49%]	0.168
Plat ≤ 15,500/	15	36.05% [16.68–55.41%]	42.37% [22.12–62.61%]	64.13% [43.69–84.58%]	0.483
Plat > 15,500	4	25.33% [0–55.49%]	38.67% [2.12–75.22%]	58.39%
ITP Duration
≤ 5.94 Log day	7	43.89% [14.09–73.69%]	56.09% [23.72–88.48%]		0.179
> 5.94 Log day	12	25.42% [7.26–43.58%]	31.83% [12.19–51.47%]	49.93% [28.08–71.78%]
No of lines of therapy
< 3 line	10	38.46% [15.63–61.29%]	45.62% [21.25–45.62%]	50.62% [24.44–76.79%]	0.48
≥ 3 lines	09	22.98% [22.28–43.69%]	31.46% [8.27–54.64]
ANA positive	6	38.69% [9.63–67.74%]	43.52% [13.16–73.88%]		0.8222
ANA negative	10	36% [11.35–60.65%]	48.27% [22.24–74.3%]
Response: PR	14	33.2% [14.46–51.93%]	36.8% [17.59–56.03%]	52.12%	0.743
Response: CR	5	29.09% [0–61.48%]	48.95% [11.05–86.84%]	[31.81–72.42%]

Side effects to rituximab therapy: Infusion-related reactions or IRRs (chills, fever, hypotension, and breathlessness) were the most common adverse event noted. Four patients developed grade ≥ 3 IRRs (2 developed hypotension, and 2 developed respiratory distress with hypoxemia), necessitating hospitalization and permanent discontinuation of the drug. IRRs were noted during the first, second and third infusion in 16, 5 and 1 patients, respectively. All grade ≥ 3 were observed during the first infusion of rituximab. All patients had received premedication with chlorpheniramine, paracetamol, and ranitidine injections before receiving rituximab infusion as per the institutional protocol. During the extended follow-up, none of the patients developed progressive multifocal leukoencephalopathy (PMLE) or reactivation of hepatitis B. The adverse events are summarised in Table 5.

Table 5.

Short and long-term adverse events noted with generic rituximab therapy

	Total events	Grade ≥ 3 AEs
Short-term adverse events
1. IRRb (Chills, fever, hypotension, breathlessness) 2. Cytopenias	22 3	4
3. Infections Tuberculosis Others	4 1 3	Nil Nil
Long-term Adverse events (Hepatitis B reactivation or PMLEb)	Nil	Nil

^aIRR infusion-related reaction

^bPMLE progressive multifocal leucoencephalopathy

Discussion

The current consensus recommends using rituximab as a second-line therapy in patients with ITP [14, 15, 19, 20]. In developing countries, financial restrictions poses a challenge for the easy accessibility of rituximab to a large group of needy patients [12]. The clinicians working in resource-constraint settings have explored the utility of low-dose rituximab, albeit with a slightly inferior response [10]. Moreover, with the scarcity of health care facilities for populations living in far-flung areas, timely and regular follow-up of patients is another limitation for the treating clinicians and hematologists. The fixed duration therapy with rituximab cuts down the follow up visits, and the cost-effectiveness of its generic variants make it an economically viable option for the treatment of relapsed and refractory ITP in resource-limited settings.

We observed an ORR of 30.61% on day 20 of rituximab therapy and 51.72% on day 30 of rituximab therapy. Our results are comparable to the previous studies reporting an ORR of rituximab in ITP ranging between 40 and 73% [9, 21–26]. The median time to response was 28 days (95% CI 21–51 days). We noted a CR rate of 16.67% (n = 11). The CR rate reported by earlier studies for rituximab in ITP patients ranges from 20 to 54% (Table 6) [9].

Table 6.

Comparison of results from index study with the published studies on utility of rituximab as second-line therapy in ITP

Study	Patel et al. 2012 [19]	Khellaf et al. 2014 [28]	Marangon et al. 2016 [22]	Hindilerden et al. 2016 [18]	Present study
Study design	Retrospective	Retrospective	Retrospective	Retrospective	Retrospective
No of patients (n)	72	248	103	15	66
Mean age, years (range)	39 (18–78)	51 (21–71)	46 (18–72)	29.6 ± 15.8	24.4 (2.5–64)
Male: female	25:47	89:159	42:61	2:15	27:39
Median interval from ITP diagnosis to rituximab (months/days)	3y(1mo-39 years)	≥ 1 year, 146 (59%)	20(1–403)	204 ± 106.2	792 days (21–6922)
Median number of prior therapies	Information not available	Information not available	≥3 17(16%)	3.6 ± 1.04	3.5 (1–6)
Splenectomy	23(32%)	25(10%)	11(11%)	15%)(100%)	8 (12.12%)
Rituximab dose	@375 mg/m2	@375 mg/m2	@375 mg/m2	@375 mg/m2	@375 mg/m2
Number of rituximab infusions	4	4	4	4	4
Overall response rate	41 (57%)	152 (61%)	57(55%)	7(46.7%)	36(54.54%)
Median follow-up duration	5 years	24 (12–30)	59 months	89.7 ± 19.4	789 days (181–5260)
Sustained response
At 1 year	38%		36% (48 months)	26.7% (> 6 months)	39%
At 2 years	31%		31% (72 months		38%
At 5 years	21%	96 (39%)			29%
Short-term AEs ≥grade 3	Nil	IRR: 3 (1.%)	Nil	–	IRR: 4 (6%)
Long-term ADRs	Nil	Nil	–	–	Nil

The ORR in our study was lesser than the observation of the largest published RCT of rituximab in ITP, the RITP trial. The RITP trial had an overall response rate of 81% in the rituximab arm [27]. Not many studies till date have been able to reproduce similar success with rituximab. In another study, Zaja and colleagues reported a superior ORR of 73%. This ORR was better than their previous published study with ORR of 65%, and they hypothesized that early use of rituximab in ITP confers a higher ORR [28, 29].

On univariate analysis of the subgroups, it was observed that age and sex of the patients, duration of ITP, number and types of prior lines of therapy, splenectomy, and pre-treatment platelet count did not affect response rate in patients treated with rituximab. This finding is consistent with the previous studies [21, 30–32]. However, contrary to results from the study by Marangon and colleagues, female sex and younger age did not confer any advantage with regard to achievement of ORR with rituximab in our study [25, 33].

ANA was tested positive in 24.07% (13/54) of patients in our study group. The finding was comparable to the reported incidence of 30% patients of primary ITP being ANA positive by Wang et al. [34]. The overall response rate in patients with ANA positive ITP patients was significantly better in comparison to ANA negative patients. In a recently published article, Wang and colleagues reported that ANA positive ITP cases have a statistically significant better ORR than ANA negative ITP patients [34]. However, the long-term response rate was inferior in ANA positive patients. Though there is no significance of ANA in the absence of other disease defining events of other connective tissue diseases, the better ORR not translating into a better-sustained response hints towards a possible secondary cause and needs further research [34–36]. The long term follow-up is also necessary for this group as the studies have also shown a predilection for chronicity in ITP patients who are ANA positive [35, 36].

The inferior response to the rituximab in patients who had ITP for a long duration and who had already received multiple lines of therapy has led to the recommendation of early use of rituximab in ITP [25, 37]. However, the duration of disease with ITP before the rituximab therapy and the baseline platelet counts before the rituximab therapy had no statistically significant effect on the response rate in our study. The response rate in patients with functional spleen or having splenectomy was also not statistically significant. Previous reports also found a similar observation [32, 37].

A total of 19 out of 36 patients, who responded earlier relapsed during the median follow up of 789 days (181–5260 days). The relapse rate was 30.32% [10], 36.12% [11], and 56.57% [17] at 1 year, 3 years, and 5 years of follow-up, respectively. Two patients relapsed after 5 years of receiving rituximab.

Overall, 25.76% (17/66) patients from the study cohort continue to be in remission at the time of completion of this study. The patients who responded (PR/CR) to rituximab had 47.22% (17/36) drug-free remission rate. Patel and colleagues observed a 21% sustained response at 5 years of observation, while Marangon and colleagues observed a 30% response rate at 59 months post rituximab [19, 22]. In the largest randomised control trial of rituximab in ITP, Ghanima and colleagues observed a sustained remission of 42% after 72 weeks of rituximab therapy [27]. Rituximab for ITP in our study had a comparable sustained response rate with published research (Table 6) [27].

In contrast to earlier studies, univariate analysis of the subgroups failed to demonstrate a better-sustained response rate with female sex, younger age, lower duration of ITP, and CR rate [9, 25, 38]. However, our findings were comparable to one of the largest and latest article published by Deshayes and colleagues [8]. After univariate and multivariate analysis, age, sex, and disease duration did not predict long-term sustained response.

Approximately, 1/3rd patients (22/66, 33.33%) developed IRR following rituximab despite administering the pre-infusion prophylaxis with ranitidine, paracetamol, and chlorpheniramine. After discontinuing the infusion due to IRR, the therapy could be re-started and completed in all except four patients (4/66, 6%) who had grade ≥ 3 IRR. There was no mortality, and all of them recovered without any residual disability. During the first follow-up year, the infection was the second commonest adverse event noted, and tuberculosis was seen in one of the patients after 6 months of therapy. All the infectious complications were managed successfully with anti-microbial and anti-tubercular drugs. There were other contributing factors like co-morbid conditions and steroid use in these patients, putting up a question on attributability to the rituximab use.

The reports of adverse events associated with rituximab in ITP have been diverse. The two challenging clinical scenarios while managing ITP are (A) managing infections in patients with ITP, and (B) managing thrombocytopenia in patients with ITP and active infection [39, 40]. Studies have reported infection as a severe concern in patients receiving rituximab leading to reduced tolerance [41]. However, Hindilerden and colleagues recorded no serious adverse events with rituximab in ITP [18]. No increased risk of infections and excellent safety profile has also been reported by other studies, including one meta-analysis [42].

The increased IRR related with rituximab in our study may be attributed to widespread use of generic molecules. Generic rituximab may cause more IRR in comparison to the original molecules as was earlier reported for lymphoma [12]. However, our study had a lower rate of IRR in comparison to the published study. Increasing age is one of the risk factors associated with poor tolerability to rituximab; therefore, we hypothesise that a younger age group in our study might have contributed to the safer outcome.

There was no long-term adverse events like reactivation of hepatitis B or PMLE. The safety profile is comparable to the earlier published data of rituximab [12, 21, 22].

Limitations of the study: Our study has limitations related to its retrospective nature, including a lack of some of the data in the documents. The minimization of the same was attempted by individually calling the patients and filling the missing data. However, the long-term follow up is a strong aspect of our study.

Conclusion

Rituximab as a second-line therapy in ITP is a safe and effective alternative with an excellent overall response rate as well as a long-term sustained response in up to one-fourth of the patients. The safety profile is comparable to the published research on rituximab. As a short duration, cost-effective therapy with long-lasting remission or cure potential, generic rituximab can be an excellent alternative to the other available second-line therapy. This study also calls for a prospective study, possibly RCT, with other second-line therapies, which can further substantiate our findings.

Funding

This research received no specific grant from any funding agency, commercial or not-for-profit sectors.

Compliance with Ethical Standards

Conflict of Interest

There is no conflict of interest between the authors.

Informed Consent

Informed signed written consent was taken from the patient involved.

Ethical Standards

The study has been sanctioned by the Institutional Ethical committee (AHRR IEC44/2020).

Ethical Approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.

Human and Animals Rights

No animals were involved in the study.