- Email: [email protected]
Sequence of Splenectomy and Rituximab for the Treatment of Steroid-Refractory Immune Thrombocytopenia: Does It Matter?
ORIGINAL ARTICLE
Sequence of Splenectomy and Rituximab for the Treatment of Steroid-Refractory Immune Thrombocytopenia: Does It Matter? William A. Hammond, MD, FACP; Prakash Vishnu, MD, FACP; Elisa M. Rodriguez; Zhuo Li, MS; Bhagirathbhai Dholaria, MBBS; Amanda J. Shreders, MD; and Candido E. Rivera, MD, FACP Abstract Objective: To evaluate the impact of the sequence of treatment with rituximab and/or splenectomy on time to relapse for patients with steroid-refractory immune thrombocytopenia (ITP). Patients and Methods: Patients 18 years or older with steroid-refractory immune thrombocytopenia who underwent treatment with splenectomy or rituximab from January 1, 2002, through December 31, 2015, at Mayo Clinic. Evaluation included freedom from relapse (FFR) and response rates after treatment with rituximab or splenectomy as single or sequential interventions. Results: A total of 218 eligible patients with ITP who were treated according to standard of care were included in this analysis. Patients failing steroids treated with splenectomy had a higher 5-year FFR than did those treated with rituximab (67.4% vs 19.2%; P<.001, propensity-score matched). Patients who failed splenectomy and were then treated with rituximab had a 2-year FFR similar to that of patients who failed rituximab and were then treated with splenectomy (73.4% vs 59.9%; P¼.52). Patients treated with rituximab after splenectomy had a longer 2-year FFR than did patients treated with rituximab as a second-line treatment (73.4% vs 29.0%; P<.001). Conclusion: For patients with ITP that relapse after treatment with steroids, splenectomy provides longer FFR than rituximab as a second-line therapy. Among patients who fail second-line treatment with splenectomy or rituximab, those who end up receiving sequential splenectomy-rituximab or rituximab-splenectomy therapy seem to derive similar benefit in the long term. Patients who received rituximab after splenectomy seem to derive superior benefit than do those who are treated with rituximab with an intact spleen. ª 2019 Mayo Foundation for Medical Education and Research
I
mmune thrombocytopenia (ITP) is an acquired disorder of isolated thrombocytopenia (platelet count, <100 109/L) characterized by immune-mediated platelet destruction and/or decreased platelet production. There are multiple mechanisms leading to the development of ITP, including production of antiplatelet autoantibodies leading to opsonization and increased clearance by the splenic reticuloendothelial system; however, more recent discoveries indicate that immune dysregulation can also lead to platelet underproduction.1 Adults with ITP are initially treated with glucocorticoids with or without intravenous
n
Mayo Clin Proc. 2019;94(11):2199-2208
immunoglobulin or anti-D immunoglobulin, but as many as 80% of these patients will eventually have a relapse and become steroid-refractory.2 There is no universal agreement on second-line therapy for steroid-refractory immune thrombocytopenia (sr-ITP). Consensus guidelines and recent appraisal of the evidence for the role of splenectomy in patients with ITP recommend either rituximab or splenectomy as acceptable and potentially curative options and that splenectomy should continue to be offered to patients failing initial treatments in the absence of increased surgeryrelated risks.3,4 Other treatment options
Mayo Clin Proc. n November 2019;94(11):2199-2208 n https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org n ª 2019 Mayo Foundation for Medical Education and Research
For editorial comment, see page 2161 From the Division of Hematology and Medical Oncology (W.A.H., P.V., B.D., A.J.S., C.E.R.) and Division of Biomedical Statistics and Informatics (Z.L.), Mayo Clinic, Jacksonville, FL; University of Puerto Rico at Mayagüez (E.M.R.), Mayagüez; Division of Affiliations continued at the end of this article.
2199
MAYO CLINIC PROCEEDINGS
such as thrombopoietin receptor agonists exist, but require continuous therapy to maintain ongoing platelet response.5 It has long been known that removal of the spleen can produce lasting responses in sr-ITP, but the reasons for successful treatment and relapse are not fully known.6 Long-term responses to splenectomy in patients with sr-ITP have been reported around 60% to 70%, indicating that thrombocytopenia is induced by processes other than increased clearance by splenic macrophages in as many as 40% of patients.7 In addition, risk of infection and bleeding complications after splenectomy, particularly in older patients, make this treatment less desirable for some patients with ITP.8-10 Rituximab, an anti-CD20 monoclonal antibody, has been found to be a safe and effective treatment of sr-ITP, although its efficacy as monotherapy has been brought into question recently.11,12 A pilot study of 60 patients with newly diagnosed or relapsed ITP with an intact spleen comparing adjuvant rituximab with placebo (saline) did not report a significant difference in the rate of treatment failure between the 2 cohorts.13 This trial used clinical end points of the frequency of any platelet count below 50 109/L after treatment, significant bleeding, or need for rescue treatment, and saw no difference between rituximab and saline infusion. In patients with previously untreated immune thrombocytopenia, 2 studies comparing the efficacy of rituximab plus dexamethasone with that of dexamethasone alone reported a significantly higher rate of sustained response at 6 months with the combination therapy.14,15 Another small study that compared low-dose rituximab with dexamethasone with dexamethasone alone, mostly in previously untreated ITP, also reported a higher response rate with the combination therapy.16 The long-term responses to rituximab in adult patients with sr-ITP have been reported to be a modest 21% at 5 years.17 It is also unknown whether the sequence of treatments after the first relapse affects the response to subsequent therapies. Previous studies have suggested that rituximab is equally effective in terms of patient response, 2200
Mayo Clin Proc.
n
regardless of whether they have had splenectomy; however, to our knowledge, there are no studies specifically addressing this question.11,18 In this retrospective analysis, we evaluated the impact of sequence of therapy with splenectomy and rituximab in patients treated in a single health system according to what was considered standard of care at the time to determine whether patients experience longer freedom from relapse (FFR) after splenectomy followed by rituximab or after rituximab followed by splenectomy. PATIENTS AND METHODS Patient Identification and Data Abstraction We identified all adult patients at Mayo Clinic (Scottsdale, Arizona, Jacksonville, Florida, and Rochester, Minnesota) who had a diagnosis of ITP and were treated with rituximab or splenectomy by using diagnostic, billing, and procedure codes. We searched billing records for patients with any diagnostic code involving thrombocytopenia (International Statistical Classification of Diseases and Related Health Problems, Ninth revision codes 287.3, 287.30, 287.39, 287.4, and 287.5) associated with a Healthcare Common Procedure Coding System code for treatment with rituximab (Healthcare Common Procedure Coding System code J9310) or a Current Procedural Terminology code for splenectomy (Current Procedural Terminology codes 38100, 38101, and 38102) from January 1, 2002, through December 31, 2015. This search strategy identified patients who received treatments during this time range; however, some patients were diagnosed before this search date and some were 18 years or younger at the time of diagnosis, but all were older than 18 years at the time of treatment with splenectomy or rituximab (ie, patients may have been observed or treated with steroids before turning 18 and were still included in the analysis if they were older than 18 years when they received a second-line therapy). Patients had to have a clinical diagnosis of ITP requiring therapy for that diagnosis. Study data were collected and managed using REDCap (Research Electronic Data
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org
SPLENECTOMY VS RITUXIMAB IN THE TREATMENT OF ITP
Capture) electronic data capture tools hosted at Mayo Clinic.19 This retrospective review of patient records was approved by the Mayo Clinic Institutional Review Board. Inclusion and Exclusion Criteria Patients were included if they met the criteria for diagnosis of primary ITP by international criteria.2 Those with Evans syndrome, including patients with active ITP and a positive Coombs test result and/or evidence of hemolysis, indicating immunemediated hemolysis (whether compensated or not), were included only if the treatment was primarily for thrombocytopenia and not for symptomatic hemolytic anemia. No patient with suspected Evans syndrome was treated with combination chemoimmunotherapy, and only those patients treated with splenectomy or rituximab monotherapy for thrombocytopenia were included. Patients were excluded if they did not meet criteria of primary ITP, including having active autoimmune conditions (such as systemic lupus erythematosus), suspected drug-induced thrombocytopenia, confirmed heparin-induced thrombocytopenia, underlying hematologic malignancy, or uncontrolled infections, such as viral hepatitis, human immunodeficiency virus infection, Helicobacter pylori infection. These were checked at the discretion of the treating hematologist, so not done in all cases. Patients were included only if they met inclusion criteria and had sufficient data to determine time frames for the end points of initiation of therapy and relapse or ongoing response. Given that patients were not treated according to a prespecified protocol, the last time point with documented platelet response was used as the time for ongoing response. This would mean that patients may have relapsed after, and this would not have been included in the analysis. Where specific data were not available, those data points were omitted from the collection and counted as missing. Given that no specific steroid regimen has been found to be superior to any other, all previous treatments with steroids were considered sufficient for first-line therapy. Mayo Clin Proc. n November 2019;94(11):2199-2208 www.mayoclinicproceedings.org
n
No specific definition for steroid-refractory was used, as treatment was at the discretion of the treating clinician at the time and there is no universally agreed upon definition. Patients were included regardless of the time from diagnosis to treatment with secondline therapy with splenectomy or rituximab, so not all patients met criteria for chronic ITP. Rituximab was administered as 375 mg/m2 per week for 4 weeks. However, patients were included with intention to treat regardless of whether they completed 4 weekly treatments. Definition of Terms Primary ITP and secondary ITP were defined according to an international working group.20 Presence of detectable antinuclear antibodies or antiphospholipid antibodies on their own without a concomitant clinical syndrome of the disease, such as lupus erythematosus, did not qualify as secondary ITP as indicated by guidelines.2 Bleeding severity scale is not uniformly defined in ITP, so we used the term major bleeding to include patients with fatal hemorrhage, intracranial bleeding of any type, or bleeding that led to hospitalization or required transfusion of red blood cells.21 Minor bleeding was considered any other skin or mucosal bleeding that was self-limited, not requiring hospitalization or transfusion. Response was defined by a platelet count of 30 109/L or higher in the absence of other treatment of ITP (including corticosteroids), and complete response was a platelet count of 100 109/L or higher.20 Response in patients with Evans syndrome was defined by the same criteria, regardless of the response of hemolysis. No response was defined as the platelet count remaining below 30 109/L, and relapse was defined as a platelet count below 30 109/L after having achieved a response. Outcomes The primary aim of our study was to measure the duration of FFR after treatment with rituximab or splenectomy as single or sequential interventions in patients with ITP failing steroids. A secondary aim was
https://doi.org/10.1016/j.mayocp.2019.05.024
2201
MAYO CLINIC PROCEEDINGS
412 Included for review
8 Refused inclusion in research efforts
65 Excluded because of: 26 Younger than 18 y 2 With insufficient data 13 With platelet count not <30x109/L 5 With hepatitis C 1 With drug-induced thrombocytopenia 18 With splenomegaly
194 Excluded after chart review
121 Excluded because of: 107 With active malignancy 32 With lymphoma 36 With CLL 12 With MDS 7 With myelofibrosis 7 With acute leukemia 4 With CMML 1 With plasma cell dyscrasia 1 With PNH 1 With Thymoma 14 With autoimmune condition 1 With aplastic anemia 1 With Sjogren syndrome 10 With lupus 1 With CVID 1 With TTP
218 Included
FIGURE 1. CONSORT flow diagram of the selection process of patients included in the study. CLL ¼ chronic lymphocytic leukemia; CMML ¼ chronic myelomonocytic leukemia; CVID ¼ common variable immunodeficiency; MDS ¼ myelodysplastic syndrome; PNH ¼ paroxysmal nocturnal hemoglobinuria; TTP ¼ thrombotic thrombocytopenic purpura.
to assess initial response rates after splenectomy or rituximab alone and in sequence. Statistical Analyses Descriptive statistics for categorical variables are reported as frequency (percentage), while continuous variables are reported as mean SD and median (range). A propensity-score matched sensitivity analysis was performed using a logistic regression analysis to compensate for potential confounding baseline variables. Patients with sr-ITP treated with rituximab were matched to patients treated with splenectomy on the propensity score for having rituximab treatment. The caliper used was 0.24 (0.25*std of logit propensity scores) for the probability of rituximab treatment as a second-line therapy. For patients treated with rituximab, a patient receiving splenectomy was randomly selected from the potential pool of controls defined by the calipers. The propensity for rituximab treatment was estimated using a logistic 2202
Mayo Clin Proc.
n
regression model, with the response variable being rituximab treatment at second-line treatment and the independent variables being age at second-line treatment, sex, body mass index (calculated as the weight in kilograms divided by the height in meters squared), platelet count at diagnosis, bleeding symptom at diagnosis, regardless of whether a bone marrow biopsy was done, time to first-line treatment, as well as time to second-line treatment. This sensitivity analysis was feasible for the second-line treatment cohort. Similar sensitivity analysis for the third-line treatment cohort was deemed unreliable as the number of patients in this cohort was too small, and they were not matched. Categorical variables were compared between treatments using the chi-square test or Fisher exact test and continuous variables were compared using the 2-sample t test or Wilcoxon rank-sum test, where appropriate. Given that response rates were defined as best response at a single point in time and
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org
SPLENECTOMY VS RITUXIMAB IN THE TREATMENT OF ITP
TABLE 1. Patient Characteristicsa,b Characteristic Age at diagnosis (s) Mean SD Median (range)
Rituximab (n¼103)
Splenectomy (n¼115)
Total (N¼218)
56.3 18.4 59.0 (15.0-93.0)
47.7 19.0 48.0 (11.0-82.0)
52.8 19.2 55.5 (11.0-93.0)
58.0 18.7 61.0 (19.0-96.0)
51.1 18.3 52.0 (18.0-82.0)
54.4 18.8 57.5 (18.0-96.0)
55 (53.4%) 48 (46.6%)
60 (52.2%) 55 (47.8%)
115 (52.8%) 103 (47.2%)
29.1 6.7 29.0 (17-47.1)
30.1 7.0 29.0 (17-60.9)
29.6 6.8 29 (17-60.9)
P value .002
Age at second-line treatment (y) Mean SD Median (range)
.01
Sex
.86
Male Female BMI
.41
Mean SD Median Platelet count at diagnosis (109/L) Mean SD Median (range)
.44 22.2 25.0 12.0 (1.0-119.0)
19.6 23.0 12.0 (0.0-100.0)
49 (51.0%) 42 (43.8%) 5 (5.2%) 7
52 (47.3%) 49 (44.5%) 9 (8.2%) 5
20.8 24.0 12.0 (0.0-119.0)
Bleeding symptoms/signs at diagnosis None Minor Major Data not available
.66 101 (49.0%) 91 (44.2%) 14 (6.8%) 12
Concomitant hemolytic anemia No Yes Data not available
.31 80 (85.1%) 14 (14.9%) 9
89 (89.9%) 10 (10.1%) 16
169 (87.6%) 24 (12.4%) 25
BMI ¼ body mass index. Data are presented as No. (percentage) unless otherwise indicated.
a
b
not time dependent (ie, not duration of response); these were analyzed as categorical variables (chi-square test or Fisher exact test) and not Kaplan-Meier variables. The Kaplan-Meier method was used to estimate FFR. A log-rank test was performed to compare cumulative rates between the groups. All statistical tests were 2-sided, with an a level set at .05 for statistical significance. For patients who had a relapse, comparisons of long-term responses between splenectomy and rituximab in the third-line setting were made in a similar manner.
Analyses were performed with SAS version 9.2 (SAS Institute Inc). RESULTS Patient Characteristics After appropriate exclusions, we identified 218 patients who met the criteria sufficient for a diagnosis of sr-ITP. The CONSORT flow diagram of the selection process of patients included in the study is outlined in Figure 1. Patient characteristics are summarized in Table 1. Treatment characteristics
TABLE 2. Kaplan-Meier Method Estimates of 1- and 5-y FFR After Second-Line Treatment of SteroidRefractory Immune Thrombocytopenia (Propensity-Score Matched) Second-line treatment
Total, n
Relapse, n
1-y FFR (95% CI) (%)
5-y FFR (95% CI) (%)
P value <.001
Splenectomy
67
16
71.8 (60.3-85.5)
67.4 (54.3-83.6)
Rituximab
67
49
33.8 (23.7-48.1)
19.2 (10.7-34.6)
FFR ¼ freedom from relapse.
Mayo Clin Proc. n November 2019;94(11):2199-2208 www.mayoclinicproceedings.org
n
https://doi.org/10.1016/j.mayocp.2019.05.024
2203
MAYO CLINIC PROCEEDINGS
Freedom from relapse (%)
were similar in all groups, except patients treated initially by splenectomy were younger at the time of second-line treatment (median age, 52 years (range, 18 to 82 years) vs 61 years (range, 19 to 96 years); P¼.01). Platelet count at diagnosis and frequency of mucosal bleeding did not differ between the groups. The median follow-up time after second-line treatment was 14.5 months (range, 0-293 months) in the splenectomy
100 90 80 70 60 50 40 30 20 10 0
Log-rank, P<.001
0 Rituximab 103 Splenectomy 115
3 2 Follow-up time (y) 16 12 26 19
1 31 48
4
5
9 13
7 11
Type of second-line treatment Rituximab Splenectomy
Freedom from relapse after third treatment (%)
A
B
100 90 80 70 60 50 40 30 20 10 0
Log-rank, P=.52
0.0
0.5
R to S 72 S to R 24
35 18
1.0 Follow-up time (y) 25 16 Treatment order R to S
1.5
2.0
17 15
15 13
S to R
FIGURE 2. Freedom from relapse after (A) second-line treatment and (B) third-line treatment of steroid-refractory immune thrombocytopenia. R to S ¼ rituximab to splenectomy; S to R ¼ splenectomy to rituximab.
2204
Mayo Clin Proc.
n
group and 18.9 months (range, 0.9-133.9 months) in the rituximab group. Outcomes We analyzed the entire cohort for the primary and secondary aims of the study as defined above. Patients treated with splenectomy as a second-line treatment had a significantly higher FFR at 5 years than did those treated with rituximab (67.4% [95% CI, 54.3%-83.6%] vs 19.2% [95% CI, 10.7%34.6%]; P<.001) (Table 2 and Figure 2A). Patients who were treated with postsplenectomy rituximab exhibited no difference in 2-year FFR when compared with those who underwent post-rituximab splenectomy (73.4% [95% CI, 57.2%-94.2%] vs 59.9% [95% CI, 47.5%-75.6%]; P¼.52) (Table 3 and Figure 2B). Five-year FFR analysis was not possible for patients who received sequential therapy, as the number of patients with available follow-up data at that time point was too small. At 5 years after third-line therapy, there were only 13 patients with available data, including 9 who were treated initially with rituximab followed by splenectomy and 4 treated with splenectomy followed by rituximab, with a median follow-up time of 14 months (range, 0-168 months). This is the reason for the shorter 2-year follow-up analysis. Comparing the long-term duration of FFR in patients treated with rituximab before or after splenectomy, those patients who had previous splenectomy achieved a significantly longer FFR than did those treated initially with rituximab with an intact spleen (Figure 3). The 2-year FFR for rituximab as a second-line treatment vs post-splenectomy rituximab as a third-line treatment was 29.0% (95% CI, 20.7%40.6%) vs 73.4% (95% CI, 57.2%-94.2%) (P<.001). For the overall cohort, patients treated with splenectomy as a second-line therapy achieved a significantly higher initial complete response rate than did those treated with rituximab (90.4% [95% CI, 83.4%-95.1%] vs 49.5% [95% CI, 39.4%-59.6%]; P<.001) (Table 4). Almost all patients achieved a high initial complete response after third-line
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org
SPLENECTOMY VS RITUXIMAB IN THE TREATMENT OF ITP
TABLE 3. Kaplan-Meier Method Estimates of 1- and 2-Year FFR After Third-Line Treatment of SteroidRefractory Immune Thrombocytopenia Third-line treatment
Total, n
Relapse, n
1-y FFR (95% CI) (%)
2-y FFR (95% CI) (%)
P value .52
Splenectomy to rituximab
24
10
73.37 (57.16-94.17)
73.37 (57.16-94.17)
Rituximab to splenectomy
72
24
63.65 (52.09-77.76)
59.90 (47.45-75.61)
FFR ¼ freedom from relapse.
DISCUSSION This study investigates the long-term outcomes of patients with sr-ITP on the basis of the sequence of using potentially curative second- and third-line treatments with splenectomy and/or rituximab. To our knowledge, this is the first analysis addressing whether sequence of therapy affects duration of response. Splenectomy induced a significantly higher FFR than did rituximab in patients with sr-ITP. In patients who needed subsequent therapy (splenectomy followed by rituximab vs rituximab followed by splenectomy), there was no difference in FFR in either sequence of treatment, suggesting that splenectomy is still effective after rituximab. When looking at responses to rituximab with or without an intact spleen, patients who received post-splenectomy rituximab had a higher response rate and FFR than did those who received rituximab with an intact spleen, suggesting that removal of the spleen may enhance response to rituximab. There have been no randomized studies that define the optimal order of treatment of sr-ITP, and controversy remains over whether rituximab is an appropriate splenectomy-sparing treatment.12,22 Moulis et al9 retrospectively compared rituximab and splenectomy in the treatment of srITP with the primary intent of assessing safety and survival and did not find any overall difference in safety when adjusted for a propensity score. Mayo Clin Proc. n November 2019;94(11):2199-2208 www.mayoclinicproceedings.org
n
The finding that rituximab appears to have a longer-lasting response after splenectomy than when given with an intact spleen raises important questions about the mechanism of action of rituximab in the treatment of sr-ITP. Presumably, those receiving rituximab after splenectomy are more “treatment resistant” than those receiving this as a second-line therapy, so intuitively one might expect a lesser response given the paradigm that earlier lines of therapy typically provide more robust responses. In addition, we found that patients who had a relapse after second-line treatment with rituximab or splenectomy still have substantial benefit from further therapy with subsequent splenectomy or rituximab, respectively.
100 Log-rank, P<.001
90 Freedom from relapse (%)
treatment with post-splenectomy rituximab or post-rituximab splenectomy (87.5% [95% CI, 67.6%-97.3%] vs 84.7% [95% CI, 74.3%92.1%]; P¼.16).
80 70 60 50 40 30 20 10 0 0.0
0.5
R in 2nd line 103 R in 3rd line 24
45 18
1.0 Follow-up time (y) 31 16
1.5
2.0
23 15
16 13
Line R in 2nd line
R in 3rd line
FIGURE 3. Freedom from relapse after second-line vs third-line treatment with rituximab of steroid-refractory immune thrombocytopenia. R in 2nd line ¼ rituximab in the second-line treatment; R in 3rd line ¼ rituximab in the third-line treatment.
https://doi.org/10.1016/j.mayocp.2019.05.024
2205
MAYO CLINIC PROCEEDINGS
TABLE 4. Comparison of Response Rates With Splenectomy vs Rituximab in the Treatment of Steroid-Refractory Immune Thrombocytopeniaa,b Response to the second-line treatment Degree of response
Splenectomy (n¼115)c
Response to the third-line treatment P value
Rituximab (n¼103)d
Splenectomy (n¼72)
Rituximab (n¼24)
CR
103 (90.4%; 95% CI, 83.4%- 50 (49.5%; 95% CI, 39.4%95.1%) 59.6%)
PR
6 (5.3%; 95% CI, 2%-11.1%) 32 (31.7%; 95% CI, 22.8%- <.001 11 (15.3%; 95% CI, 7.9%41.7%) 92.1%)
NR
5 (4.4%; 95% CI, 1.4%-9.9%) 19 (18.8%; 95% CI, 11.7%27.8%)
ORR (%)
94.8
P value
61 (84.7%; 95% CI, 74.3%- 21 (87.5%; 95% CI, 67.6%92.1%) 97.3%) 2 (8.3%; 95% CI, 1%-27%)
0 (0%; 95% CI, 0%-5%)
1 (4.2%; 95% CI, 0.1%21.1%)
100
95.8
79.6
.16
CR ¼ complete response; NR ¼ no response; ORR ¼ overall response rate; PR ¼ partial response. Data are presented as No. (percentage) unless otherwise indicated. c One patient with these data unavailable. d Two patients with these data unavailable. a
b
There have been attempts at explaining the differences in response to splenectomy vs rituximab in patients with ITP. Historically, it has been assumed that platelets are cleared primarily by splenic reticuloendothelial cells. However, a recent report has suggested that platelets may be cleared from circulation in locations other than the spleen, such as Kupffer cells in the liver, and the pathogenesis of ITP may be more complex than just antibody-mediated platelet destruction.23 Although rituximab has an effect on other secondary lymphoid organs by altering the type of B-cell populations present, it is unclear whether it has an impact at these other reticuloendothelial sites of platelet clearance.24,25 T cells also influence the pathogenesis of ITP. Splenic T-cell populations are skewed in patients with ITP, with increased T helper type 1 cells and decreased suppressive regulatory T cells in the blood and spleen.26 This skewed population was found to correlate with increased splenic macrophage phagocytosis, and antiplatelet antibody production can be controlled by T cells and their cytokines.27,28 It is not known whether rituximab would have a lasting impact on these pathogenic processes of ITP. Lastly, the spleen contains both IgMþ memory B cells and CD20 plasma cells, 2 potential key effectors of ongoing autoimmune 2206
Mayo Clin Proc.
n
propagation.29-31 Rituximab will not have any effect on either CD20 plasma cells or T cells, so removing the spleen may provide seemingly necessary depletion if treatment with rituximab alone cannot. Notably, this study is limited by its retrospective nature. Despite broadly inclusive search criteria, it is possible that some patients meeting inclusion criteria were not captured. In addition, many patients treated with splenectomy at our institution had short follow-up before returning to their referring provider, explaining the shorter median follow-up for patients treated with splenectomy than for those treated with rituximab, which is the opposite of what has typically been reported in previous series.12,17 Moreover, as this study was retrospective and treatment decisions were made as per the treating clinician’s judgment, some patients treated with rituximab may not have been allowed sufficient time to respond and may have proceeded to surgery prematurely.3 This may account, in part, for the higher percentage of “no response” in the rituximab group and likely reflects what commonly occurs in the treatment of patients with sr-ITP. An additional limitation is the small sample sizes for the third-line treatments. This limited the ability to perform sensitivity analysis and propensity scoring for the
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org
SPLENECTOMY VS RITUXIMAB IN THE TREATMENT OF ITP
comparisons of the 2 treatment groups and must be noted when considering the significance of the results for the third-line treatment analyses. Lastly, complications and adverse outcomes from the treatments are not reported, as these data have been well established in the literature with regard to both the relative safety of splenectomy8-10 and a comparison directly between rituximab and splenectomy.9 Regardless of these limitations, this analysis provides important “realworld” clinical outcome information about responses of patients to what would be considered standard potentially curative treatments where randomized data are lacking. CONCLUSION Our analysis of a large longitudinal cohort of patients with sr-ITP treated within a single health system reveals that splenectomy is a more effective therapy in terms of both achieving a higher response rate and resulting in a longer duration of response than rituximab as the second-line treatment. Also, for those patients requiring third-line treatment, FFR appeared equivalent in those receiving sequential splenectomy-rituximab and those receiving sequential rituximabsplenectomy therapy. Interestingly, patients who had a relapse of sr-ITP after splenectomy and received rituximab as a subsequent line of treatment had a longer 2-year FFR than did those who received rituximab with an intact spleen, suggesting that previous splenectomy may induce a better response to rituximab. In patients with sr-ITP whose most important goal of treatment is having the most effective therapy in terms of achieving and maintaining a long-term platelet response without continuous treatment, these data support a recommendation for splenectomy before rituximab. ACKNOWLEDGMENT We thank Ronald Go, MD (Mayo Clinic, Rochester, MN), for his expertise and assistance with the review of the data and manuscript. Mayo Clin Proc. n November 2019;94(11):2199-2208 www.mayoclinicproceedings.org
n
Abbreviations and Acronyms: FFR = freedom from relapse; ITP = immune thrombocytopenia; sr-ITP = steroid-refractory immune thrombocytopenia Affiliations (Continued from the first page of this article.): Hematology and Medical Oncology, Baptist MD Anderson Cancer Center, Jacksonville, FL (W.A.H., A.J.S.); and Department of Hematology-Oncology, Vanderbilt University Medical Center, Nashville, TN (B.D.).
Potential Competing Interests: The authors report no competing interests. Data Previously Presented: These data were presented at the 58th American Society of Hematology Annual Meeting in San Diego, CA, on December 5, 2016. Correspondence: Address to Candido E. Rivera, MD, FACP, Division of Hematology and Medical Oncology, Mayo Clinic, 4500 San Pablo Rd, Jacksonville, FL 32224 ([email protected]).
REFERENCES 1. Audia S, Mahévas M, Samson M, Godeau B, Bonnotte B. Pathogenesis of immune thrombocytopenia. Autoimmun Rev. 2017; 16(6):620-632. 2. Provan D, Stasi R, Newland AC, et al. International consensus report on the investigation and management of primary immune thrombocytopenia. Blood. 2010;115(2):168-186. 3. Neunert C, Lim W, Crowther M, Cohen A, Solberg L Jr, Crowther MA; American Society of Hematology. The American Society of Hematology 2011 evidence-based practice guideline for immune thrombocytopenia. Blood. 2011; 117(16):4190-4207. 4. Rodeghiero F. A critical appraisal of the evidence for the role of splenectomy in adults and children with ITP. Br J Haematol. 2018;181(2):183-195. 5. Zeng Y, Duan X, Xu J, Ni X. TPO receptor agonist for chronic idiopathic thrombocytopenic purpura. Cochrane Database Syst Rev. 2011;7:CD008235. 6. Dameshek W. The “humors” and idiopathic thrombocytopenic purpura. Blood. 1951;6(10):954-955. 7. Kojouri K, Vesely SK, Terrell DR, George JN. Splenectomy for adult patients with idiopathic thrombocytopenic purpura: a systematic review to assess long-term platelet count responses, prediction of response, and surgical complications. Blood. 2004;104(9):2623-2634. 8. Vianelli N, Galli M, de Vivo A, et al. Efficacy and safety of splenectomy in immune thrombocytopenic purpura: long-term results of 402 cases. Haematologica. 2005;90(1):72-77. 9. Moulis G, Sailler L, Sommet A, Lapeyre-Mestre M, Derumeaux H, Adoue D. Rituximab versus splenectomy in persistent or chronic adult primary immune thrombocytopenia: an adjusted comparison of mortality and morbidity. Am J Hematol. 2014;89(1):41-46. 10. Bagrodia N, Button AM, Spanheimer PM, Belding-Schmitt ME, Rosenstein LJ, Mezhir JJ. Morbidity and mortality following elective splenectomy for benign and malignant hematologic conditions: analysis of the American College of Surgeons National Surgical Quality Improvement Program data. JAMA Surg. 2014; 149(10):1022-1029. 11. Cooper N, Stasi R, Cunningham-Rundles S, et al. The efficacy and safety of B-cell depletion with anti-CD20 monoclonal antibody in adults with chronic immune thrombocytopenic purpura. Br J Haematol. 2004;125(2):232-239.
https://doi.org/10.1016/j.mayocp.2019.05.024
2207
MAYO CLINIC PROCEEDINGS
12. Ghanima W, Khelif A, Waage A, et al; RITP Study Group. Rituximab as second-line treatment for adult immune thrombocytopenia (the RITP trial): a multicentre, randomised, double-blind, placebo-controlled trial. Lancet. 2015;385(9978): 1653-1661. 13. Arnold DM, Heddle NM, Carruthers J, et al. A pilot randomized trial of adjuvant rituximab or placebo for nonsplenectomized patients with immune thrombocytopenia. Blood. 2012;119(6): 1356-1362. 14. Zaja F, Baccarani M, Mazza P, et al. Dexamethasone plus rituximab yields higher sustained response rates than dexamethasone monotherapy in adults with primary immune thrombocytopenia. Blood. 2010;115(14):2755-2762. 15. Gudbrandsdottir S, Birgens HS, Frederiksen H, et al. Rituximab and dexamethasone vs dexamethasone monotherapy in newly diagnosed patients with primary immune thrombocytopenia. Blood. 2013;121(11):1976-1981. 16. Li Z, Mou W, Lu G, et al. Low-dose rituximab combined with short-term glucocorticoids up-regulates Treg cell levels in patients with immune thrombocytopenia. Int J Hematol. 2011; 93(1):91-98. 17. Patel VL, Mahévas M, Lee SY, et al. Outcomes 5 years after response to rituximab therapy in children and adults with immune thrombocytopenia. Blood. 2012;119(25): 5989-5995. 18. Cooper N, Evangelista ML, Amadori S, Stasi R. Should rituximab be used before or after splenectomy in patients with immune thrombocytopenic purpura? Curr Opin Hematol. 2007; 14(6):642-646. 19. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)da metadatadriven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009; 42(2):377-381. 20. Rodeghiero F, Stasi R, Gernsheimer T, et al. Standardization of terminology, definitions and outcome criteria in immune thrombocytopenic purpura of adults and children: report from an international working group. Blood. 2009;113(11):2386-2393.
2208
Mayo Clin Proc.
n
21. Neunert C, Arnold DM. Severe bleeding events in adults and children with primary immune thrombocytopenia: a systematic review: reply. J Thromb Haemost. 2015;13(8):1522-1523. 22. George JN. Sequence of treatments for adults with primary immune thrombocytopenia. Am J Hematol. 2012;87(suppl 1): S12-S15. 23. Kuwana M, Okazaki Y, Ikeda Y. Splenic macrophages maintain the anti-platelet autoimmune response via uptake of opsonized platelets in patients with immune thrombocytopenic purpura. J Thromb Haemost. 2009;7(2):322-329. 24. Kamburova EG, Koenen HJ, Borgman KJ, ten Berge IJ, Joosten I, Hilbrands LB. A single dose of rituximab does not deplete B cells in secondary lymphoid organs but alters phenotype and function. Am J Transplant. 2013;13(6):1503-1511. 25. Wiseman GA, White CA, Stabin M, et al. Phase I/II 90Y-Zevalin (yttrium-90 ibritumomab tiuxetan, IDEC-Y2B8) radioimmunotherapy dosimetry results in relapsed or refractory non-Hodgkin’s lymphoma. Eur J Nucl Med. 2000;27(7):766-777. 26. Chen X, Chen S, Li C, Zhu Y, Peng B. Skewed T-cell subsets and enhanced macrophages phagocytosis in the spleen of patients with immune thrombocytopenia failing glucocorticoids. Int J Hematol. 2011;94(3):248-254. 27. Kuwana M, Ikeda Y. The role of autoreactive T-cells in the pathogenesis of idiopathic thrombocytopenic purpura. Int J Hematol. 2005; 81(2):106-112. 28. Kuwana M, Okazaki Y, Kaburaki J, Kawakami Y, Ikeda Y. Spleen is a primary site for activation of platelet-reactive T and B cells in patients with immune thrombocytopenic purpura. J Immunol. 2002;168(7): 3675-3682. 29. Cameron PU, Jones P, Gorniak M, et al. Splenectomy associated changes in IgM memory B cells in an adult spleen registry cohort. PLoS One. 2011;6(8):e23164. 30. Weller S, Braun MC, Tan BK, et al. Human blood IgM “memory” B cells are circulating splenic marginal zone B cells harboring a prediversified immunoglobulin repertoire. Blood. 2004;104(12):3647-3654. 31. Mahévas M, Patin P, Huetz F, et al. B cell depletion in immune thrombocytopenia reveals splenic long-lived plasma cells. J Clin Invest. 2013;123(1):432-442.
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org
Sequence of Splenectomy and Rituximab for the Treatment of Steroid-Refractory Immune Thrombocytopenia: Does It Matter? William A. Hammond, MD, FACP; Prakash Vishnu, MD, FACP; Elisa M. Rodriguez; Zhuo Li, MS; Bhagirathbhai Dholaria, MBBS; Amanda J. Shreders, MD; and Candido E. Rivera, MD, FACP Abstract Objective: To evaluate the impact of the sequence of treatment with rituximab and/or splenectomy on time to relapse for patients with steroid-refractory immune thrombocytopenia (ITP). Patients and Methods: Patients 18 years or older with steroid-refractory immune thrombocytopenia who underwent treatment with splenectomy or rituximab from January 1, 2002, through December 31, 2015, at Mayo Clinic. Evaluation included freedom from relapse (FFR) and response rates after treatment with rituximab or splenectomy as single or sequential interventions. Results: A total of 218 eligible patients with ITP who were treated according to standard of care were included in this analysis. Patients failing steroids treated with splenectomy had a higher 5-year FFR than did those treated with rituximab (67.4% vs 19.2%; P<.001, propensity-score matched). Patients who failed splenectomy and were then treated with rituximab had a 2-year FFR similar to that of patients who failed rituximab and were then treated with splenectomy (73.4% vs 59.9%; P¼.52). Patients treated with rituximab after splenectomy had a longer 2-year FFR than did patients treated with rituximab as a second-line treatment (73.4% vs 29.0%; P<.001). Conclusion: For patients with ITP that relapse after treatment with steroids, splenectomy provides longer FFR than rituximab as a second-line therapy. Among patients who fail second-line treatment with splenectomy or rituximab, those who end up receiving sequential splenectomy-rituximab or rituximab-splenectomy therapy seem to derive similar benefit in the long term. Patients who received rituximab after splenectomy seem to derive superior benefit than do those who are treated with rituximab with an intact spleen. ª 2019 Mayo Foundation for Medical Education and Research
I
mmune thrombocytopenia (ITP) is an acquired disorder of isolated thrombocytopenia (platelet count, <100 109/L) characterized by immune-mediated platelet destruction and/or decreased platelet production. There are multiple mechanisms leading to the development of ITP, including production of antiplatelet autoantibodies leading to opsonization and increased clearance by the splenic reticuloendothelial system; however, more recent discoveries indicate that immune dysregulation can also lead to platelet underproduction.1 Adults with ITP are initially treated with glucocorticoids with or without intravenous
n
Mayo Clin Proc. 2019;94(11):2199-2208
immunoglobulin or anti-D immunoglobulin, but as many as 80% of these patients will eventually have a relapse and become steroid-refractory.2 There is no universal agreement on second-line therapy for steroid-refractory immune thrombocytopenia (sr-ITP). Consensus guidelines and recent appraisal of the evidence for the role of splenectomy in patients with ITP recommend either rituximab or splenectomy as acceptable and potentially curative options and that splenectomy should continue to be offered to patients failing initial treatments in the absence of increased surgeryrelated risks.3,4 Other treatment options
Mayo Clin Proc. n November 2019;94(11):2199-2208 n https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org n ª 2019 Mayo Foundation for Medical Education and Research
For editorial comment, see page 2161 From the Division of Hematology and Medical Oncology (W.A.H., P.V., B.D., A.J.S., C.E.R.) and Division of Biomedical Statistics and Informatics (Z.L.), Mayo Clinic, Jacksonville, FL; University of Puerto Rico at Mayagüez (E.M.R.), Mayagüez; Division of Affiliations continued at the end of this article.
2199
MAYO CLINIC PROCEEDINGS
such as thrombopoietin receptor agonists exist, but require continuous therapy to maintain ongoing platelet response.5 It has long been known that removal of the spleen can produce lasting responses in sr-ITP, but the reasons for successful treatment and relapse are not fully known.6 Long-term responses to splenectomy in patients with sr-ITP have been reported around 60% to 70%, indicating that thrombocytopenia is induced by processes other than increased clearance by splenic macrophages in as many as 40% of patients.7 In addition, risk of infection and bleeding complications after splenectomy, particularly in older patients, make this treatment less desirable for some patients with ITP.8-10 Rituximab, an anti-CD20 monoclonal antibody, has been found to be a safe and effective treatment of sr-ITP, although its efficacy as monotherapy has been brought into question recently.11,12 A pilot study of 60 patients with newly diagnosed or relapsed ITP with an intact spleen comparing adjuvant rituximab with placebo (saline) did not report a significant difference in the rate of treatment failure between the 2 cohorts.13 This trial used clinical end points of the frequency of any platelet count below 50 109/L after treatment, significant bleeding, or need for rescue treatment, and saw no difference between rituximab and saline infusion. In patients with previously untreated immune thrombocytopenia, 2 studies comparing the efficacy of rituximab plus dexamethasone with that of dexamethasone alone reported a significantly higher rate of sustained response at 6 months with the combination therapy.14,15 Another small study that compared low-dose rituximab with dexamethasone with dexamethasone alone, mostly in previously untreated ITP, also reported a higher response rate with the combination therapy.16 The long-term responses to rituximab in adult patients with sr-ITP have been reported to be a modest 21% at 5 years.17 It is also unknown whether the sequence of treatments after the first relapse affects the response to subsequent therapies. Previous studies have suggested that rituximab is equally effective in terms of patient response, 2200
Mayo Clin Proc.
n
regardless of whether they have had splenectomy; however, to our knowledge, there are no studies specifically addressing this question.11,18 In this retrospective analysis, we evaluated the impact of sequence of therapy with splenectomy and rituximab in patients treated in a single health system according to what was considered standard of care at the time to determine whether patients experience longer freedom from relapse (FFR) after splenectomy followed by rituximab or after rituximab followed by splenectomy. PATIENTS AND METHODS Patient Identification and Data Abstraction We identified all adult patients at Mayo Clinic (Scottsdale, Arizona, Jacksonville, Florida, and Rochester, Minnesota) who had a diagnosis of ITP and were treated with rituximab or splenectomy by using diagnostic, billing, and procedure codes. We searched billing records for patients with any diagnostic code involving thrombocytopenia (International Statistical Classification of Diseases and Related Health Problems, Ninth revision codes 287.3, 287.30, 287.39, 287.4, and 287.5) associated with a Healthcare Common Procedure Coding System code for treatment with rituximab (Healthcare Common Procedure Coding System code J9310) or a Current Procedural Terminology code for splenectomy (Current Procedural Terminology codes 38100, 38101, and 38102) from January 1, 2002, through December 31, 2015. This search strategy identified patients who received treatments during this time range; however, some patients were diagnosed before this search date and some were 18 years or younger at the time of diagnosis, but all were older than 18 years at the time of treatment with splenectomy or rituximab (ie, patients may have been observed or treated with steroids before turning 18 and were still included in the analysis if they were older than 18 years when they received a second-line therapy). Patients had to have a clinical diagnosis of ITP requiring therapy for that diagnosis. Study data were collected and managed using REDCap (Research Electronic Data
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org
SPLENECTOMY VS RITUXIMAB IN THE TREATMENT OF ITP
Capture) electronic data capture tools hosted at Mayo Clinic.19 This retrospective review of patient records was approved by the Mayo Clinic Institutional Review Board. Inclusion and Exclusion Criteria Patients were included if they met the criteria for diagnosis of primary ITP by international criteria.2 Those with Evans syndrome, including patients with active ITP and a positive Coombs test result and/or evidence of hemolysis, indicating immunemediated hemolysis (whether compensated or not), were included only if the treatment was primarily for thrombocytopenia and not for symptomatic hemolytic anemia. No patient with suspected Evans syndrome was treated with combination chemoimmunotherapy, and only those patients treated with splenectomy or rituximab monotherapy for thrombocytopenia were included. Patients were excluded if they did not meet criteria of primary ITP, including having active autoimmune conditions (such as systemic lupus erythematosus), suspected drug-induced thrombocytopenia, confirmed heparin-induced thrombocytopenia, underlying hematologic malignancy, or uncontrolled infections, such as viral hepatitis, human immunodeficiency virus infection, Helicobacter pylori infection. These were checked at the discretion of the treating hematologist, so not done in all cases. Patients were included only if they met inclusion criteria and had sufficient data to determine time frames for the end points of initiation of therapy and relapse or ongoing response. Given that patients were not treated according to a prespecified protocol, the last time point with documented platelet response was used as the time for ongoing response. This would mean that patients may have relapsed after, and this would not have been included in the analysis. Where specific data were not available, those data points were omitted from the collection and counted as missing. Given that no specific steroid regimen has been found to be superior to any other, all previous treatments with steroids were considered sufficient for first-line therapy. Mayo Clin Proc. n November 2019;94(11):2199-2208 www.mayoclinicproceedings.org
n
No specific definition for steroid-refractory was used, as treatment was at the discretion of the treating clinician at the time and there is no universally agreed upon definition. Patients were included regardless of the time from diagnosis to treatment with secondline therapy with splenectomy or rituximab, so not all patients met criteria for chronic ITP. Rituximab was administered as 375 mg/m2 per week for 4 weeks. However, patients were included with intention to treat regardless of whether they completed 4 weekly treatments. Definition of Terms Primary ITP and secondary ITP were defined according to an international working group.20 Presence of detectable antinuclear antibodies or antiphospholipid antibodies on their own without a concomitant clinical syndrome of the disease, such as lupus erythematosus, did not qualify as secondary ITP as indicated by guidelines.2 Bleeding severity scale is not uniformly defined in ITP, so we used the term major bleeding to include patients with fatal hemorrhage, intracranial bleeding of any type, or bleeding that led to hospitalization or required transfusion of red blood cells.21 Minor bleeding was considered any other skin or mucosal bleeding that was self-limited, not requiring hospitalization or transfusion. Response was defined by a platelet count of 30 109/L or higher in the absence of other treatment of ITP (including corticosteroids), and complete response was a platelet count of 100 109/L or higher.20 Response in patients with Evans syndrome was defined by the same criteria, regardless of the response of hemolysis. No response was defined as the platelet count remaining below 30 109/L, and relapse was defined as a platelet count below 30 109/L after having achieved a response. Outcomes The primary aim of our study was to measure the duration of FFR after treatment with rituximab or splenectomy as single or sequential interventions in patients with ITP failing steroids. A secondary aim was
https://doi.org/10.1016/j.mayocp.2019.05.024
2201
MAYO CLINIC PROCEEDINGS
412 Included for review
8 Refused inclusion in research efforts
65 Excluded because of: 26 Younger than 18 y 2 With insufficient data 13 With platelet count not <30x109/L 5 With hepatitis C 1 With drug-induced thrombocytopenia 18 With splenomegaly
194 Excluded after chart review
121 Excluded because of: 107 With active malignancy 32 With lymphoma 36 With CLL 12 With MDS 7 With myelofibrosis 7 With acute leukemia 4 With CMML 1 With plasma cell dyscrasia 1 With PNH 1 With Thymoma 14 With autoimmune condition 1 With aplastic anemia 1 With Sjogren syndrome 10 With lupus 1 With CVID 1 With TTP
218 Included
FIGURE 1. CONSORT flow diagram of the selection process of patients included in the study. CLL ¼ chronic lymphocytic leukemia; CMML ¼ chronic myelomonocytic leukemia; CVID ¼ common variable immunodeficiency; MDS ¼ myelodysplastic syndrome; PNH ¼ paroxysmal nocturnal hemoglobinuria; TTP ¼ thrombotic thrombocytopenic purpura.
to assess initial response rates after splenectomy or rituximab alone and in sequence. Statistical Analyses Descriptive statistics for categorical variables are reported as frequency (percentage), while continuous variables are reported as mean SD and median (range). A propensity-score matched sensitivity analysis was performed using a logistic regression analysis to compensate for potential confounding baseline variables. Patients with sr-ITP treated with rituximab were matched to patients treated with splenectomy on the propensity score for having rituximab treatment. The caliper used was 0.24 (0.25*std of logit propensity scores) for the probability of rituximab treatment as a second-line therapy. For patients treated with rituximab, a patient receiving splenectomy was randomly selected from the potential pool of controls defined by the calipers. The propensity for rituximab treatment was estimated using a logistic 2202
Mayo Clin Proc.
n
regression model, with the response variable being rituximab treatment at second-line treatment and the independent variables being age at second-line treatment, sex, body mass index (calculated as the weight in kilograms divided by the height in meters squared), platelet count at diagnosis, bleeding symptom at diagnosis, regardless of whether a bone marrow biopsy was done, time to first-line treatment, as well as time to second-line treatment. This sensitivity analysis was feasible for the second-line treatment cohort. Similar sensitivity analysis for the third-line treatment cohort was deemed unreliable as the number of patients in this cohort was too small, and they were not matched. Categorical variables were compared between treatments using the chi-square test or Fisher exact test and continuous variables were compared using the 2-sample t test or Wilcoxon rank-sum test, where appropriate. Given that response rates were defined as best response at a single point in time and
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org
SPLENECTOMY VS RITUXIMAB IN THE TREATMENT OF ITP
TABLE 1. Patient Characteristicsa,b Characteristic Age at diagnosis (s) Mean SD Median (range)
Rituximab (n¼103)
Splenectomy (n¼115)
Total (N¼218)
56.3 18.4 59.0 (15.0-93.0)
47.7 19.0 48.0 (11.0-82.0)
52.8 19.2 55.5 (11.0-93.0)
58.0 18.7 61.0 (19.0-96.0)
51.1 18.3 52.0 (18.0-82.0)
54.4 18.8 57.5 (18.0-96.0)
55 (53.4%) 48 (46.6%)
60 (52.2%) 55 (47.8%)
115 (52.8%) 103 (47.2%)
29.1 6.7 29.0 (17-47.1)
30.1 7.0 29.0 (17-60.9)
29.6 6.8 29 (17-60.9)
P value .002
Age at second-line treatment (y) Mean SD Median (range)
.01
Sex
.86
Male Female BMI
.41
Mean SD Median Platelet count at diagnosis (109/L) Mean SD Median (range)
.44 22.2 25.0 12.0 (1.0-119.0)
19.6 23.0 12.0 (0.0-100.0)
49 (51.0%) 42 (43.8%) 5 (5.2%) 7
52 (47.3%) 49 (44.5%) 9 (8.2%) 5
20.8 24.0 12.0 (0.0-119.0)
Bleeding symptoms/signs at diagnosis None Minor Major Data not available
.66 101 (49.0%) 91 (44.2%) 14 (6.8%) 12
Concomitant hemolytic anemia No Yes Data not available
.31 80 (85.1%) 14 (14.9%) 9
89 (89.9%) 10 (10.1%) 16
169 (87.6%) 24 (12.4%) 25
BMI ¼ body mass index. Data are presented as No. (percentage) unless otherwise indicated.
a
b
not time dependent (ie, not duration of response); these were analyzed as categorical variables (chi-square test or Fisher exact test) and not Kaplan-Meier variables. The Kaplan-Meier method was used to estimate FFR. A log-rank test was performed to compare cumulative rates between the groups. All statistical tests were 2-sided, with an a level set at .05 for statistical significance. For patients who had a relapse, comparisons of long-term responses between splenectomy and rituximab in the third-line setting were made in a similar manner.
Analyses were performed with SAS version 9.2 (SAS Institute Inc). RESULTS Patient Characteristics After appropriate exclusions, we identified 218 patients who met the criteria sufficient for a diagnosis of sr-ITP. The CONSORT flow diagram of the selection process of patients included in the study is outlined in Figure 1. Patient characteristics are summarized in Table 1. Treatment characteristics
TABLE 2. Kaplan-Meier Method Estimates of 1- and 5-y FFR After Second-Line Treatment of SteroidRefractory Immune Thrombocytopenia (Propensity-Score Matched) Second-line treatment
Total, n
Relapse, n
1-y FFR (95% CI) (%)
5-y FFR (95% CI) (%)
P value <.001
Splenectomy
67
16
71.8 (60.3-85.5)
67.4 (54.3-83.6)
Rituximab
67
49
33.8 (23.7-48.1)
19.2 (10.7-34.6)
FFR ¼ freedom from relapse.
Mayo Clin Proc. n November 2019;94(11):2199-2208 www.mayoclinicproceedings.org
n
https://doi.org/10.1016/j.mayocp.2019.05.024
2203
MAYO CLINIC PROCEEDINGS
Freedom from relapse (%)
were similar in all groups, except patients treated initially by splenectomy were younger at the time of second-line treatment (median age, 52 years (range, 18 to 82 years) vs 61 years (range, 19 to 96 years); P¼.01). Platelet count at diagnosis and frequency of mucosal bleeding did not differ between the groups. The median follow-up time after second-line treatment was 14.5 months (range, 0-293 months) in the splenectomy
100 90 80 70 60 50 40 30 20 10 0
Log-rank, P<.001
0 Rituximab 103 Splenectomy 115
3 2 Follow-up time (y) 16 12 26 19
1 31 48
4
5
9 13
7 11
Type of second-line treatment Rituximab Splenectomy
Freedom from relapse after third treatment (%)
A
B
100 90 80 70 60 50 40 30 20 10 0
Log-rank, P=.52
0.0
0.5
R to S 72 S to R 24
35 18
1.0 Follow-up time (y) 25 16 Treatment order R to S
1.5
2.0
17 15
15 13
S to R
FIGURE 2. Freedom from relapse after (A) second-line treatment and (B) third-line treatment of steroid-refractory immune thrombocytopenia. R to S ¼ rituximab to splenectomy; S to R ¼ splenectomy to rituximab.
2204
Mayo Clin Proc.
n
group and 18.9 months (range, 0.9-133.9 months) in the rituximab group. Outcomes We analyzed the entire cohort for the primary and secondary aims of the study as defined above. Patients treated with splenectomy as a second-line treatment had a significantly higher FFR at 5 years than did those treated with rituximab (67.4% [95% CI, 54.3%-83.6%] vs 19.2% [95% CI, 10.7%34.6%]; P<.001) (Table 2 and Figure 2A). Patients who were treated with postsplenectomy rituximab exhibited no difference in 2-year FFR when compared with those who underwent post-rituximab splenectomy (73.4% [95% CI, 57.2%-94.2%] vs 59.9% [95% CI, 47.5%-75.6%]; P¼.52) (Table 3 and Figure 2B). Five-year FFR analysis was not possible for patients who received sequential therapy, as the number of patients with available follow-up data at that time point was too small. At 5 years after third-line therapy, there were only 13 patients with available data, including 9 who were treated initially with rituximab followed by splenectomy and 4 treated with splenectomy followed by rituximab, with a median follow-up time of 14 months (range, 0-168 months). This is the reason for the shorter 2-year follow-up analysis. Comparing the long-term duration of FFR in patients treated with rituximab before or after splenectomy, those patients who had previous splenectomy achieved a significantly longer FFR than did those treated initially with rituximab with an intact spleen (Figure 3). The 2-year FFR for rituximab as a second-line treatment vs post-splenectomy rituximab as a third-line treatment was 29.0% (95% CI, 20.7%40.6%) vs 73.4% (95% CI, 57.2%-94.2%) (P<.001). For the overall cohort, patients treated with splenectomy as a second-line therapy achieved a significantly higher initial complete response rate than did those treated with rituximab (90.4% [95% CI, 83.4%-95.1%] vs 49.5% [95% CI, 39.4%-59.6%]; P<.001) (Table 4). Almost all patients achieved a high initial complete response after third-line
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org
SPLENECTOMY VS RITUXIMAB IN THE TREATMENT OF ITP
TABLE 3. Kaplan-Meier Method Estimates of 1- and 2-Year FFR After Third-Line Treatment of SteroidRefractory Immune Thrombocytopenia Third-line treatment
Total, n
Relapse, n
1-y FFR (95% CI) (%)
2-y FFR (95% CI) (%)
P value .52
Splenectomy to rituximab
24
10
73.37 (57.16-94.17)
73.37 (57.16-94.17)
Rituximab to splenectomy
72
24
63.65 (52.09-77.76)
59.90 (47.45-75.61)
FFR ¼ freedom from relapse.
DISCUSSION This study investigates the long-term outcomes of patients with sr-ITP on the basis of the sequence of using potentially curative second- and third-line treatments with splenectomy and/or rituximab. To our knowledge, this is the first analysis addressing whether sequence of therapy affects duration of response. Splenectomy induced a significantly higher FFR than did rituximab in patients with sr-ITP. In patients who needed subsequent therapy (splenectomy followed by rituximab vs rituximab followed by splenectomy), there was no difference in FFR in either sequence of treatment, suggesting that splenectomy is still effective after rituximab. When looking at responses to rituximab with or without an intact spleen, patients who received post-splenectomy rituximab had a higher response rate and FFR than did those who received rituximab with an intact spleen, suggesting that removal of the spleen may enhance response to rituximab. There have been no randomized studies that define the optimal order of treatment of sr-ITP, and controversy remains over whether rituximab is an appropriate splenectomy-sparing treatment.12,22 Moulis et al9 retrospectively compared rituximab and splenectomy in the treatment of srITP with the primary intent of assessing safety and survival and did not find any overall difference in safety when adjusted for a propensity score. Mayo Clin Proc. n November 2019;94(11):2199-2208 www.mayoclinicproceedings.org
n
The finding that rituximab appears to have a longer-lasting response after splenectomy than when given with an intact spleen raises important questions about the mechanism of action of rituximab in the treatment of sr-ITP. Presumably, those receiving rituximab after splenectomy are more “treatment resistant” than those receiving this as a second-line therapy, so intuitively one might expect a lesser response given the paradigm that earlier lines of therapy typically provide more robust responses. In addition, we found that patients who had a relapse after second-line treatment with rituximab or splenectomy still have substantial benefit from further therapy with subsequent splenectomy or rituximab, respectively.
100 Log-rank, P<.001
90 Freedom from relapse (%)
treatment with post-splenectomy rituximab or post-rituximab splenectomy (87.5% [95% CI, 67.6%-97.3%] vs 84.7% [95% CI, 74.3%92.1%]; P¼.16).
80 70 60 50 40 30 20 10 0 0.0
0.5
R in 2nd line 103 R in 3rd line 24
45 18
1.0 Follow-up time (y) 31 16
1.5
2.0
23 15
16 13
Line R in 2nd line
R in 3rd line
FIGURE 3. Freedom from relapse after second-line vs third-line treatment with rituximab of steroid-refractory immune thrombocytopenia. R in 2nd line ¼ rituximab in the second-line treatment; R in 3rd line ¼ rituximab in the third-line treatment.
https://doi.org/10.1016/j.mayocp.2019.05.024
2205
MAYO CLINIC PROCEEDINGS
TABLE 4. Comparison of Response Rates With Splenectomy vs Rituximab in the Treatment of Steroid-Refractory Immune Thrombocytopeniaa,b Response to the second-line treatment Degree of response
Splenectomy (n¼115)c
Response to the third-line treatment P value
Rituximab (n¼103)d
Splenectomy (n¼72)
Rituximab (n¼24)
CR
103 (90.4%; 95% CI, 83.4%- 50 (49.5%; 95% CI, 39.4%95.1%) 59.6%)
PR
6 (5.3%; 95% CI, 2%-11.1%) 32 (31.7%; 95% CI, 22.8%- <.001 11 (15.3%; 95% CI, 7.9%41.7%) 92.1%)
NR
5 (4.4%; 95% CI, 1.4%-9.9%) 19 (18.8%; 95% CI, 11.7%27.8%)
ORR (%)
94.8
P value
61 (84.7%; 95% CI, 74.3%- 21 (87.5%; 95% CI, 67.6%92.1%) 97.3%) 2 (8.3%; 95% CI, 1%-27%)
0 (0%; 95% CI, 0%-5%)
1 (4.2%; 95% CI, 0.1%21.1%)
100
95.8
79.6
.16
CR ¼ complete response; NR ¼ no response; ORR ¼ overall response rate; PR ¼ partial response. Data are presented as No. (percentage) unless otherwise indicated. c One patient with these data unavailable. d Two patients with these data unavailable. a
b
There have been attempts at explaining the differences in response to splenectomy vs rituximab in patients with ITP. Historically, it has been assumed that platelets are cleared primarily by splenic reticuloendothelial cells. However, a recent report has suggested that platelets may be cleared from circulation in locations other than the spleen, such as Kupffer cells in the liver, and the pathogenesis of ITP may be more complex than just antibody-mediated platelet destruction.23 Although rituximab has an effect on other secondary lymphoid organs by altering the type of B-cell populations present, it is unclear whether it has an impact at these other reticuloendothelial sites of platelet clearance.24,25 T cells also influence the pathogenesis of ITP. Splenic T-cell populations are skewed in patients with ITP, with increased T helper type 1 cells and decreased suppressive regulatory T cells in the blood and spleen.26 This skewed population was found to correlate with increased splenic macrophage phagocytosis, and antiplatelet antibody production can be controlled by T cells and their cytokines.27,28 It is not known whether rituximab would have a lasting impact on these pathogenic processes of ITP. Lastly, the spleen contains both IgMþ memory B cells and CD20 plasma cells, 2 potential key effectors of ongoing autoimmune 2206
Mayo Clin Proc.
n
propagation.29-31 Rituximab will not have any effect on either CD20 plasma cells or T cells, so removing the spleen may provide seemingly necessary depletion if treatment with rituximab alone cannot. Notably, this study is limited by its retrospective nature. Despite broadly inclusive search criteria, it is possible that some patients meeting inclusion criteria were not captured. In addition, many patients treated with splenectomy at our institution had short follow-up before returning to their referring provider, explaining the shorter median follow-up for patients treated with splenectomy than for those treated with rituximab, which is the opposite of what has typically been reported in previous series.12,17 Moreover, as this study was retrospective and treatment decisions were made as per the treating clinician’s judgment, some patients treated with rituximab may not have been allowed sufficient time to respond and may have proceeded to surgery prematurely.3 This may account, in part, for the higher percentage of “no response” in the rituximab group and likely reflects what commonly occurs in the treatment of patients with sr-ITP. An additional limitation is the small sample sizes for the third-line treatments. This limited the ability to perform sensitivity analysis and propensity scoring for the
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org
SPLENECTOMY VS RITUXIMAB IN THE TREATMENT OF ITP
comparisons of the 2 treatment groups and must be noted when considering the significance of the results for the third-line treatment analyses. Lastly, complications and adverse outcomes from the treatments are not reported, as these data have been well established in the literature with regard to both the relative safety of splenectomy8-10 and a comparison directly between rituximab and splenectomy.9 Regardless of these limitations, this analysis provides important “realworld” clinical outcome information about responses of patients to what would be considered standard potentially curative treatments where randomized data are lacking. CONCLUSION Our analysis of a large longitudinal cohort of patients with sr-ITP treated within a single health system reveals that splenectomy is a more effective therapy in terms of both achieving a higher response rate and resulting in a longer duration of response than rituximab as the second-line treatment. Also, for those patients requiring third-line treatment, FFR appeared equivalent in those receiving sequential splenectomy-rituximab and those receiving sequential rituximabsplenectomy therapy. Interestingly, patients who had a relapse of sr-ITP after splenectomy and received rituximab as a subsequent line of treatment had a longer 2-year FFR than did those who received rituximab with an intact spleen, suggesting that previous splenectomy may induce a better response to rituximab. In patients with sr-ITP whose most important goal of treatment is having the most effective therapy in terms of achieving and maintaining a long-term platelet response without continuous treatment, these data support a recommendation for splenectomy before rituximab. ACKNOWLEDGMENT We thank Ronald Go, MD (Mayo Clinic, Rochester, MN), for his expertise and assistance with the review of the data and manuscript. Mayo Clin Proc. n November 2019;94(11):2199-2208 www.mayoclinicproceedings.org
n
Abbreviations and Acronyms: FFR = freedom from relapse; ITP = immune thrombocytopenia; sr-ITP = steroid-refractory immune thrombocytopenia Affiliations (Continued from the first page of this article.): Hematology and Medical Oncology, Baptist MD Anderson Cancer Center, Jacksonville, FL (W.A.H., A.J.S.); and Department of Hematology-Oncology, Vanderbilt University Medical Center, Nashville, TN (B.D.).
Potential Competing Interests: The authors report no competing interests. Data Previously Presented: These data were presented at the 58th American Society of Hematology Annual Meeting in San Diego, CA, on December 5, 2016. Correspondence: Address to Candido E. Rivera, MD, FACP, Division of Hematology and Medical Oncology, Mayo Clinic, 4500 San Pablo Rd, Jacksonville, FL 32224 ([email protected]).
REFERENCES 1. Audia S, Mahévas M, Samson M, Godeau B, Bonnotte B. Pathogenesis of immune thrombocytopenia. Autoimmun Rev. 2017; 16(6):620-632. 2. Provan D, Stasi R, Newland AC, et al. International consensus report on the investigation and management of primary immune thrombocytopenia. Blood. 2010;115(2):168-186. 3. Neunert C, Lim W, Crowther M, Cohen A, Solberg L Jr, Crowther MA; American Society of Hematology. The American Society of Hematology 2011 evidence-based practice guideline for immune thrombocytopenia. Blood. 2011; 117(16):4190-4207. 4. Rodeghiero F. A critical appraisal of the evidence for the role of splenectomy in adults and children with ITP. Br J Haematol. 2018;181(2):183-195. 5. Zeng Y, Duan X, Xu J, Ni X. TPO receptor agonist for chronic idiopathic thrombocytopenic purpura. Cochrane Database Syst Rev. 2011;7:CD008235. 6. Dameshek W. The “humors” and idiopathic thrombocytopenic purpura. Blood. 1951;6(10):954-955. 7. Kojouri K, Vesely SK, Terrell DR, George JN. Splenectomy for adult patients with idiopathic thrombocytopenic purpura: a systematic review to assess long-term platelet count responses, prediction of response, and surgical complications. Blood. 2004;104(9):2623-2634. 8. Vianelli N, Galli M, de Vivo A, et al. Efficacy and safety of splenectomy in immune thrombocytopenic purpura: long-term results of 402 cases. Haematologica. 2005;90(1):72-77. 9. Moulis G, Sailler L, Sommet A, Lapeyre-Mestre M, Derumeaux H, Adoue D. Rituximab versus splenectomy in persistent or chronic adult primary immune thrombocytopenia: an adjusted comparison of mortality and morbidity. Am J Hematol. 2014;89(1):41-46. 10. Bagrodia N, Button AM, Spanheimer PM, Belding-Schmitt ME, Rosenstein LJ, Mezhir JJ. Morbidity and mortality following elective splenectomy for benign and malignant hematologic conditions: analysis of the American College of Surgeons National Surgical Quality Improvement Program data. JAMA Surg. 2014; 149(10):1022-1029. 11. Cooper N, Stasi R, Cunningham-Rundles S, et al. The efficacy and safety of B-cell depletion with anti-CD20 monoclonal antibody in adults with chronic immune thrombocytopenic purpura. Br J Haematol. 2004;125(2):232-239.
https://doi.org/10.1016/j.mayocp.2019.05.024
2207
MAYO CLINIC PROCEEDINGS
12. Ghanima W, Khelif A, Waage A, et al; RITP Study Group. Rituximab as second-line treatment for adult immune thrombocytopenia (the RITP trial): a multicentre, randomised, double-blind, placebo-controlled trial. Lancet. 2015;385(9978): 1653-1661. 13. Arnold DM, Heddle NM, Carruthers J, et al. A pilot randomized trial of adjuvant rituximab or placebo for nonsplenectomized patients with immune thrombocytopenia. Blood. 2012;119(6): 1356-1362. 14. Zaja F, Baccarani M, Mazza P, et al. Dexamethasone plus rituximab yields higher sustained response rates than dexamethasone monotherapy in adults with primary immune thrombocytopenia. Blood. 2010;115(14):2755-2762. 15. Gudbrandsdottir S, Birgens HS, Frederiksen H, et al. Rituximab and dexamethasone vs dexamethasone monotherapy in newly diagnosed patients with primary immune thrombocytopenia. Blood. 2013;121(11):1976-1981. 16. Li Z, Mou W, Lu G, et al. Low-dose rituximab combined with short-term glucocorticoids up-regulates Treg cell levels in patients with immune thrombocytopenia. Int J Hematol. 2011; 93(1):91-98. 17. Patel VL, Mahévas M, Lee SY, et al. Outcomes 5 years after response to rituximab therapy in children and adults with immune thrombocytopenia. Blood. 2012;119(25): 5989-5995. 18. Cooper N, Evangelista ML, Amadori S, Stasi R. Should rituximab be used before or after splenectomy in patients with immune thrombocytopenic purpura? Curr Opin Hematol. 2007; 14(6):642-646. 19. Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)da metadatadriven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009; 42(2):377-381. 20. Rodeghiero F, Stasi R, Gernsheimer T, et al. Standardization of terminology, definitions and outcome criteria in immune thrombocytopenic purpura of adults and children: report from an international working group. Blood. 2009;113(11):2386-2393.
2208
Mayo Clin Proc.
n
21. Neunert C, Arnold DM. Severe bleeding events in adults and children with primary immune thrombocytopenia: a systematic review: reply. J Thromb Haemost. 2015;13(8):1522-1523. 22. George JN. Sequence of treatments for adults with primary immune thrombocytopenia. Am J Hematol. 2012;87(suppl 1): S12-S15. 23. Kuwana M, Okazaki Y, Ikeda Y. Splenic macrophages maintain the anti-platelet autoimmune response via uptake of opsonized platelets in patients with immune thrombocytopenic purpura. J Thromb Haemost. 2009;7(2):322-329. 24. Kamburova EG, Koenen HJ, Borgman KJ, ten Berge IJ, Joosten I, Hilbrands LB. A single dose of rituximab does not deplete B cells in secondary lymphoid organs but alters phenotype and function. Am J Transplant. 2013;13(6):1503-1511. 25. Wiseman GA, White CA, Stabin M, et al. Phase I/II 90Y-Zevalin (yttrium-90 ibritumomab tiuxetan, IDEC-Y2B8) radioimmunotherapy dosimetry results in relapsed or refractory non-Hodgkin’s lymphoma. Eur J Nucl Med. 2000;27(7):766-777. 26. Chen X, Chen S, Li C, Zhu Y, Peng B. Skewed T-cell subsets and enhanced macrophages phagocytosis in the spleen of patients with immune thrombocytopenia failing glucocorticoids. Int J Hematol. 2011;94(3):248-254. 27. Kuwana M, Ikeda Y. The role of autoreactive T-cells in the pathogenesis of idiopathic thrombocytopenic purpura. Int J Hematol. 2005; 81(2):106-112. 28. Kuwana M, Okazaki Y, Kaburaki J, Kawakami Y, Ikeda Y. Spleen is a primary site for activation of platelet-reactive T and B cells in patients with immune thrombocytopenic purpura. J Immunol. 2002;168(7): 3675-3682. 29. Cameron PU, Jones P, Gorniak M, et al. Splenectomy associated changes in IgM memory B cells in an adult spleen registry cohort. PLoS One. 2011;6(8):e23164. 30. Weller S, Braun MC, Tan BK, et al. Human blood IgM “memory” B cells are circulating splenic marginal zone B cells harboring a prediversified immunoglobulin repertoire. Blood. 2004;104(12):3647-3654. 31. Mahévas M, Patin P, Huetz F, et al. B cell depletion in immune thrombocytopenia reveals splenic long-lived plasma cells. J Clin Invest. 2013;123(1):432-442.
November 2019;94(11):2199-2208
n
https://doi.org/10.1016/j.mayocp.2019.05.024 www.mayoclinicproceedings.org