Effectiveness of lymphoplasmapheresis compared with therapeutic plasma exchange for thrombotic thrombocytopenic purpura: a retrospective evaluation

ABSTRACT Objectives Thrombotic thrombocytopenic purpura (TTP) is an acute life-threatening disease usually treated with therapeutic plasma exchange (TPE), but some patients are refractory to TPE. The study aimed to compare lymphoplasmapheresis (LPE), an innovative treatment for TTP based on plasma exchange, with TPE in TTP treatment. Methods This retrospective study included patients with TTP treated at Xiang-Ya Hospital in China during 2009-2018. All patients with microangiopathic hemolysis and thrombocytopenia who received either LPE or TPE were included. The treatment outcomes were the number of sessions, volume of plasma, time in hospital, hospital costs, and rates of remission and relapse. All patients attended the hospital for follow-up. Results Forty-five patients were included in the study; 18 received TPE and 27 LPE. There were no significant differences in sex, etiology of TTP, initial platelet count, schistocyte, LDH, and bilirubin between the two groups. At the time of discharge, patients treated with TPE required more treatment sessions (4.5 vs. 2, P=0.04) and higher plasma volume (7300 vs. 3100 ml, P=0.01) than patients treated with LPE. The proportions of remission (P=0.197) and relapse (P=0.257) were not significantly different between the two groups. The time to remission from admission (P=0.75) and the time to remission from first therapy (P=0.53) were also not significantly different between the two groups. Conclusion Compared with TPE, LPE reduced the number of treatment sessions and plasma volume needed to treat TTP. Therefore, we propose that LPE might be a suitable treatment for TTP.


Introduction
Thrombotic thrombocytopenic purpura (TTP) is an acute, life-threatening disease. It has an estimated annual incidence ranging from 3.7-11.5 cases/million [1,2]. The main cause of this disease is the deficiency of the von-Willebrand factor (vWF) cleavage metalloprotease, called ADAMTS13. In the absence of ADAMTS13, ultra-large multimers (ULMM) attract platelets in circulation to form platelet thrombi. Consumption of platelets in platelet-rich thrombi results in thrombocytopenia, while thrombi appear to be responsible for renal and cerebral lesions, and this often results in damage to other organ systems [3][4][5][6].
If untreated, mortality from TTP may exceed 90% [1,2]. Once recognized, the accepted standard of care for TTP is daily therapeutic plasma exchange (TPE). Since the introduction of TPE in the 1970s, the mortality rate associated with TTP has gradually decreased to 10%-20% [7]. Despite improved overall survival, some patients with acquired idiopathic autoimmune disease are more likely to take longer to achieve a normal platelet count and require more plasma therapy. Plasma exchanges are not always effective, some patients are refractory to this therapy, and the relapse rate remains high at 30%-50% [6,8]. The reported incidence of patients who do not respond to TPE and corticosteroids and require additional therapy varies between 10% and 42% [9][10][11]. In recent guidelines on TTP treatment, the American Association of Blood Banks (AABB) recommends daily plasma exchange until a platelet count above 150 × 10 9 /L is attained for 2-3 consecutive days. The American Society for Apheresis (ASFA) recommends daily TPE until the LDH is normal or near-normal, and the platelet count has risen above 100×10 9 /L and continues to rise following cessation of TPE [7]. As a result, TPE-treated patients need large amounts of plasma, which requires long-term hospitalization to achieve a normal platelet count and complete recovery.
Lymphoplasmapheresis (LPE) is an innovative treatment based on plasma exchange. This process uses a blood cell separator to remove lymphocytes and mononuclear cells. Furthermore, hemolyzed red blood cells (RBCs), vWF, and vWF-multimers, close to the lymphocytes and mononuclear cells, can also be removed. In a previous study, our group demonstrated that LPE was an appropriate method for treating patients with immune diseases, such as treating Guillain-Barre syndrome (GBS) and myasthenia gravis (MG) patients by removing immunoglobulin, complement, monocytes, and fibrinogen as well as regulating lymphocyte subsets [12]. Still, no report has been published about TTP patients treated with LPE until now. There is also a lack of accurate information about the efficacy and prognosis of patients with TTP treated with LPE.
In our center, both TPE and LPE have been available for TTP patients, where the choice between these two options was made before the first treatment session at the physician's discretion. There was no specific orientation to guide this choice, performed individually by the physician responsible for each patient's care. Therefore, the primary aim of this study was to compare the effectiveness of TPE with LPE and determine whether using LPE was a suitable treatment for TTP. It is hoped that this study will provide direction for clinical decisions on the effective treatment of TTP.

Study design and patients
This observational retrospective study was performed in compliance with the Helsinki Declaration II and the Chinese Standards for Good Clinical Practice. The study protocol was approved by the ethics committee of the XXX South University, Changsha, China. The need for individual consent was waived by the committee.
This retrospective study included consecutive newly diagnosed TTP patients between February 2009 and December 2018 in the Departments of Hematology and Rheumatology at Xiang-Ya Hospital.
The inclusion criteria were (1) thrombocytopenia, (2) schistocytes were detected, (3) TTP were diagnosed for the first time, and (4) the diagnosis of TTP was based on clinical presentation and laboratory findings, mainly consistent with consumptive thrombocytopenia, hemolytic anemia, and/or the presence of schistocytes in the peripheral blood smear [13]. The patients with other thrombotic microangiopathies, including disseminated intravascular coagulation (DIC), cancer, preeclampsia, autoimmune hemolysis/ Evans syndrome, hemolytic uremic syndrome (diarrhea positive/negative), catastrophic antiphospholipid syndrome, malignancy, or vasculitis were excluded. Patients who switched between TPE and LPE for any reason were also excluded.
The patients were grouped according to the treatment they received (TPE or LPE groups). Their treatment was determined prior to their first treatment session according to the preference of the attending clinician after discussion with the patient. The patient provided consent to treatment.

LPE and TPE procedures
An apheresis instrument (COBE Spectra, Terumo BCT, Lakewood, CO, USA) was for LPE. LPE was performed with a white cell channel installed. The lymphoid cells were collected using density gradient centrifugation and photoelectricity technology in a manual program. Then, the feedback channel of patients' autologous plasma was clipped to fill the collection bag. Along with peripheral blood mononuclear cells (PBMCs), the lymphocytes were removed, while fresh frozen plasma was injected from the clamp at the top left of the heparin cap. It was reinfused to the patient's body after the confluence of the patient's red blood cells. Approximately 70%-80% of total blood volume was exchanged during each procedure. The volume was replaced with frozen plasma or plasma without cryoprecipitate. LPE was stopped when platelets were >100×10 9 /L for a minimum of 2 days.
TPE was performed using standard procedures until complete remission, defined as a platelet count above 100×10 9 /L for a minimum of 2 days.
Patients' data such as hemoglobin (Hb) and platelet levels, bilirubin concentrations, LDH levels, and clinical symptoms were collected for 24-48 h after the last TPE or LPE treatment. If symptoms recurred, TPE or LPE treatment was given again until the patients' condition was stable.

Follow-up and data collection
Patients were followed up until discharged from the hospital. During in-hospital follow-up, the curative effect of the two therapeutic procedures was evaluated.
The rates of remission, relapses, urticaria, and hypocalcemia, the number of treatment sessions, the total volume of plasma received, days of hospitalization, cost of hospitalization, time to remission from admission, and time to remission from first therapy were calculated. The plasma volumes and numbers of sessions were calculated from the start of treatments to stable remission. Clinical remission was defined as an increase in platelets accompanied by improvement of clinical symptoms such as improvement of hemolysis, fever, kidney damage, and mental acuity. Recurrence was defined as an increase then fall in the number of platelets accompanied or not by the appearance of clinical symptoms.
Data of patient characteristics, clinical and laboratory test results, number of LPE or TPE treatments, and total volume of plasma received until remission were recorded from the patients' medical records and blood transfusion electronic registries. Data on days of hospitalization, hospital costs, and relapses were obtained from the medical records. Routine laboratory tests such as peripheral blood cell counts, reticulocyte count, coagulation profile, serum lactate dehydrogenase (LDH), bilirubin, serum creatinine, cardiac enzymes, and urinalysis were performed. Only some patients underwent ADAMTS13 enzyme testing, as our hospital did not carry out this test before 2015. Laboratory tests were performed with standard laboratory procedures. Peripheral blood smears were routinely prepared from all TTP patients.

Statistical analysis
All statistical analyses were performed using SPSS 22.0 for Windows (IBM Corp., Armonk, NY, USA). Analysis showed that all data except initial LDH and schistocyte were distributed normally. Baseline demographic characteristics were calculated in frequencies and percentages. All continuous variables were expressed as means, standard deviations, and ranges. Differences between groups were investigated according to the data distribution, using the two-sample t-test and Wilcoxon rank test. A Mann-Whitney test was used to assess the effectiveness of treatment based on the number of sessions, total volume of plasma, days of hospitalization, and hospital costs. Yates' chi-squared analysis compared the effectiveness of treatment based on the remission and relapse rates.

Results
Forty-five patients were identified with newly diagnosed TTP and received TPE or LPE treatment and were included in this analysis. The baseline characteristics of the 45 patients are summarized in Table 1.
The population age ranged from 13 to 71 years and included 32 females and 13 males. Twenty-eight (62.2%) patients were diagnosed with primary TTP and 17 (37.8%) with secondary TTP. Thirty-seven (82.2%) patients presented with neurologic features (headache, stroke, coma, and seizures), 25 (55.5%) patients had fever (temperature >37.5°C), 20 (44.4%) patients had renal impairment, and 39 (86.7%) patients manifested hemolytic anemia. The characteristics were similar in the two groups at baseline except for the hemoglobin levels, which were higher in the LPE group than in the TPE group (63.0 (42-104) vs. 62.5 (44-78) g/L, P=0.003). There were only nine patients with ADAMTS13 activity test data, all of which were <5% (<2.5%, 3.8%, 0%, <5%, <5%, 0%, <5%, <5%, <5%). Table 2 shows that treatment was effective for both groups because compared with the pre-treatment results, Hb and platelet count were significantly higher while total and direct bilirubin concentrations and LDH levels were significantly reduced after TPE or LPE treatment (P<0.001). However, there were no significant differences in the changes in these parameters before and after treatment between the two groups. Table 3 shows the outcomes from the in-hospital follow-up. In terms of treatment effectiveness, patients treated with TPE required more treatment sessions and larger plasma volumes than patients treated with LPE (all P<0.05). The rates of remission and relapse were similar between the two groups. The number of treatment sessions, volume of plasma, days spent in hospital, and cost of hospitalization are compared in Supplementary Figure 1. In the TPE group, the median number of TPE treatments until remission was 4.5 (interquartile range (IQR), 1.75-6.75). In the LPE group, the median number of LPE treatments until remission was 2 (IQR, 1-3), significantly less than in the TPE group (P=0.04). TPE-treated patients were also exposed to a larger volume of plasma (TPE: 7300 (IQR, 2175-15,073) vs. LPE: 3100 (IQR, 1760-6439) ml, P=0.01). Compared with the LPE group, TPE-treated   patients also spent more money in the hospital and more days in the hospital, but these differences did not reach statistical significance (P=0.08 and P=0.30). The time to remission from admission (P=0.75) and time to remission from first therapy (P=0.53) were also not significantly different between the two groups. As Supplementary Figure 1 presents a patient who appears to have a requirement for extensive therapy, the analysis was performed after excluding the extreme values. After excluding the two extreme values, the difference was still statistically significant (Supplementary Figure 2).
Mild allergic reactions were the most common treatment-related adverse events in both groups, representing >70% of all adverse events. Several patients had urticaria on the face or breast, accompanied by skin itch. These symptoms disappeared after receiving corticosteroids or immunosuppressive agents. No other serious adverse reactions such as hemolysis, anaphylactic shock, circulatory overload, hypothermia, and hemorrhage were observed during and after LPE treatment. As shown in Table 3, the numbers of urticaria and hypocalcemia events were not significantly different between the two groups (P>0.05).

Discussion
This study aimed to compare the effectiveness of an innovative method (LPE) with the standard method (TPE) during hospitalization. The results should provide evidence for whether LPE can be used as an effective treatment for TTP. Of the 45 patients enrolled in the study, 18 received TPE and 27 LPE. Patients treated with TPE required more treatment sessions and higher plasma volume than patients treated with LPE, but the numbers of remissions and relapses  were not significantly different between the two groups. Therefore, we conclude that LPE was effective in treating TTP and might be a suitable treatment method.
Our study showed that TPE and LPE were both effective for TTP patients, and both could bring about temporary clinical improvements. LPE reduced the number of treatment sessions and volume of plasma required for effective treatment. Therefore, this suggests that LPE might become a preferred method for TTP treatment. Previous studies have shown that the relapse rate after treatment for TTP is between 30%-50% [6,8]. This study showed that the relapse rates were quite low; 18.5% of patients treated with LPE relapsed while 33.3% of patients treated with TPE relapsed. Still, the differences between the two groups were not significant, and larger studies might be needed to evaluate the benefits of using LPE fully.
The mechanisms for both TPE and LPE might be similar. These include removing abnormal vWF multimers released by endothelial cells, scavenging ADAMTSl3 autoantibodies, replenishing deficient ADAMTSl3, and restoring the normal degradation of circulating vWF. In addition, these processes might supplement prostacyclin I in plasma, scavenge some of the pathogenic antibodies in plasma, remove various endothelial-stimulating cytokines that damage endothelial cells and activate platelets, and scavenge circulating inflammatory factors, including tumor necrosis factor-α, interleukin (IL)-6, and IL-8 to prevent multiple organ dysfunction syndromes (MODS) [14].
However, apheresis does not appear to change B or T cell counts in normal donors [15] since this type of blood purification technique involves extracting the plasma [16]. Therefore, it cannot remove activated inflammatory cells, such as sensitized T and B lymphocytes and other immunocompetent cells that produce pathological substances. The decline in antibody titers in the blood circulation after TPE results in the removal of the feedback inhibition of lymphocytes and soon causes a rebound production of pathological materials. On the other hand, LPE not only removes the pathological materials in the blood but also eliminates activated T and B cells by removing PBMCs, consequently preventing the production of antibodies and inflammatory cytokines. It might be that lymphocyte removal during the most severe period of rejection provides enhanced immunosuppression. After removing T and B cell populations, feedback inhibition on clones of antibody producing cells might decrease, providing a novel concept for the treatment of patients with TTP. Severe deficiency of ADAMTS13 is present in most cases of acquired TTP primarily because of the presence of IgG antibodies to ADAMTS13. Therefore, adding lymphocyte removal will reduce the level of anti-ADAMTS-13 antibodies. We hypothesize that the long-term effects of LPE therapy in patients with TTP included slow regeneration of memory B cell subsets and persistently reduced BAFF-R expression across all B cell subpopulations. It might delay both the selection and differentiation of autoreactive anti-ADAMTS-13 B cells, resulting in a relatively long time to relapse after LPE therapy [17].
The underlying mechanisms for LPE's therapeutic effects might also be similar to lymphokinetics [18]. The circulating lymphocyte pool includes both short-lived and long-lived populations, some living as long as 20 years. T cells comprise roughly 90% of the lymphocytes in the thoracic duct, 65% in the blood. Most circulating lymphocytes enter the blood via the thoracic duct and spend most of their lives percolating through tissues and recirculating. Thus, it is reasonable to assume that lymphopheresis would have effects similar to thoracic duct drainage. LPE can cause circulating lymphopenia by removal of both T and B cell populations. The addition of lymphocyte removal might provide augmented suppression of cell-mediated immune function [19].
The present study has several limitations. First, the patients in the two groups were not randomly selected to receive either LPE or TPE, so there might be some bias in the grouping. Although the two groups were similar in most baseline characteristics, there were significant differences in hemoglobin levels. Second, the levels of ADAMTS13 and anti-ADAMTS-13 antibodies could not be analyzed, as our hospital did not carry out this test before 2015. Third, the retrospective design of the current study is associated with inherent limitations. Finally, there was no follow-up data after discharge. Larger randomized prospective studies are needed to evaluate the efficacy of LPE for TTP treatment fully.
In summary, we retrospectively analyzed 45 patients with TTP and found that most of them were successfully treated by TPE or LPE. We concluded that TPE and LPE were both effective for TTP patients. Compared with TPE, LPE can reduce the number of treatment sessions and volumes of plasma needed for effective treatment. Therefore, we propose that LPE treatment for TTP should be a suitable procedure for further investigation.

Funding
This work was supported by the Youth Science Foundation of Xiangya Hospital 2019.

Data availability statement
Data sharing does not apply to this article as no new data were created or analyzed in this study.