Therapeutic plasma exchange for the treatment of thrombotic thrombocytopenic purpura: A retrospective multicenter study

Transfusion and Apheresis Science 36 (2007) 57–67

intl.elsevierhealth.com/journals/tras

Therapeutic plasma exchange for the treatment of thromboticthrombocytopenic purpura: A retrospective multicenter study

Fevzi Altuntas a,¤, Ismet Aydogdu b, Sibel Kabukcu c, Ismail Kocyigit a,Kerim CÂkÂm b, Ismail SarÂ

a, M. Ali Erkut b, Bulent Eser a, Ahmet Ozturk d,Emin Kaya b, Mustafa Cetin a, Ali Keskin c, Ali Unal a

a Erciyes University, Faculty of Medicine, Department of Hematology and Hemapheresis Unit, 38039 Kayseri, Turkeyb

Ãnonu University, Faculty of Medicine, Department of Hematology and Hemapheresis Unit, Malatya, Turkeyc Pamukkale University, Faculty of Medicine, Department of Hematology and Hemapheresis Unit, Denizli, Turkey

d Erciyes University, Faculty of Medicine, Department of Statistics, Kayseri, Turkey

Received 6 April 2006; accepted 30 May 2006

Abstract

Background: Thrombotic thrombocytopenic purpura (TTP) is a rare disease that is fatal if it is not treated. Therapeuticplasma exchange (TPE) has resulted in excellent remission and survival rates in TTP patients.

Material and methods: We describe our experience with 52 TTP patients treated with TPE during the past eight years(65% of the patients were females; patient median ageD 34 years, range: 17–73). TPE was carried out 1–1.5 times plasmavolume. Fresh frozen plasma (FFP) or cryosupernatant plasma (CSP) was used as the replacement Xuid. TPE was per-formed daily until normalization of serum LDH and recovery of the platelet count to >150£ 109/dL; TPE was then slowlytapered. Clinical, laboratory data, the number of TPE, other given therapy modalities, treatment outcomes and survivalrate were evaluated retrospectively.

Results: Overall response (OR) and complete response (CR) rates were 77% and 60%, respectively. Response was excel-lent in 82.8% of the patients with primary TTP among whom 74.2% were CR. Additionally, there were statistical diVerencesin terms of CR rate between patients with primary TTP and secondary TTP (74.2% vs. 29.4%; pD 0.005). OR and CR rateswere 79% and 57.9% in patients on TPE alone and 75.8% and 60.6% in patients on TPE + prednisolone, respectively (pD 1and pD 0.8). Additionally, there were no statistical diVerences in terms of OR and CR rates between patients on TPE withFFP and CSP (pD 0.25 and pD 0.16, respectively). The presence of fever and the number of TPE were statistically impor-tant factors inXuencing the probability of response in multivariate logistic regression analysis (p < 0.01 and p < 0.01, respec-tively). Additionally, in multivariate Cox’s regression analysis, the probability of survival was higher in patients who wereresponsive to treatment compared to patients who were unresponsive (p < 0.001).

Conclusion: TPE is an eVective treatment for primary TTP; however, it may be used as adjunctive therapy for secondaryTTP until it is under control. The addition of steroids to TPE had no advantage compared to TPE alone. CSP as replace-ment Xuid is not superior compared to FFP. Fever appears to be a bad prognostic indicator. Therefore, prolonged treat-ment with TPE may be needed in patients with fever.© 2006 Elsevier Ltd. All rights reserved.

1473-0502/$ - see front matter © 2006 Elsevier Ltd. All rights reserved.doi:10.1016/j.transci.2006.05.014

* Corresponding author. Tel.: +90 352 437 49 37; fax: +90 352 437 93 48.E-mail address: [email protected] (F. Altuntas).

58 F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67

Keywords: Therapeutic plasma exchange; Thrombotic thrombocytopenic purpura; Treatment; Outcome

1. Introduction treatment, i.e., therapeutic plasma exchange, plasma

Thrombotic thrombocytopenic purpura (TTP) isa rare, sometimes fatal disease typically character-ized by a clinical pentad of thrombocytopenia,microangiopathic hemolytic anemia (MAHA), neu-rological involvement, renal dysfunction and fever[1–3]. The triad of thrombocytopenia, schistocytosis,and elevated LDH levels is often suYcient to suggestthe disorder [2].

Systemic microvascular aggregation of plateletshas been implicated in the pathogenesis of the dis-ease [4,5]. Recent studies have showed that vWFcleaving protease (vWF-CPase) activity is deWcientin patients with familial TTP and auto antibodiesinhibiting vWF-CPase are present in patients withnon-familial TTP [5–8]. Therefore, they act as a trig-ger for increased platelet aggregation [4,9]. TTP maybe idiopathic, but it may also occur secondary toautoimmune diseases, pregnancy, bone marrowtransplantation (BMT), malignancy, infection dis-eases and drugs [10–17].

Therapeutic plasma exchange (TPE) is theaccepted method of treatment in patients with TTP[1,3,18–20]. Although TPE has increased the survivalrate from 10% to greater than 90% [1,19–22], asubset of patients with resistant TTP still fails torespond to TPE or remains dependent on this proce-dure. Therefore, other treatment modalitiesincluding steroids, vincristine, cyclophosphamide,intravenous immunoglobulin (IVIG), rituximab,aspirin, dipyridamole and splenectomy have all beenused to control the disease [3,23–30]. However,because of the rarity of randomized clinical trials,there is scant data regarding the eYcacy and safetyof these agents.

This retrospective study was performed as ananalysis of our eight-year experience with TPE,other treatment modalities and treatment outcomesin adult patients with TTP.

2. Patients and methods

This is a retrospective analysis of TTP requiringTPE in three University Hospitals between 1997 and2005. The list of patients was obtained from the TPElog-book of the apheresis facility of our apheresisunits. Laboratory and clinical data, such as initialpresentation, concomitant diseases, drug intake,

infusion, adjuvant and/or alternative agents, andoutcome were also collected.

The diagnosis of TTP was made when the patientpresented with thrombocytopenia and MAHA withor without signs of renal dysfunction, neurologicalmanifestation or fever. Thrombocytopenia was deW-ned as a platelet count below 150£ 109/dL. MAHAwas deWned by anemia, presence of red cell fragmen-tation on peripheral blood smear, reticulocytosis,high indirect bilirubin and lactate dehydrogenase(LDH) serum levels. Renal failure was deWned asserum creatinine (Cr) level >1.2 mg/dL, blood ureanitrogen (BUN) level >25 mg/dL, proteinuria and/orhematuria as deWned by the normal ranges of ourlaboratory. Fever was deWned as a body temperatureabove 38 °C.

All patients underwent the following laboratoryinvestigations prior to treatment; complete bloodcount, reticulocyte count, haptoglobin level, periph-eral blood smear, direct antiglobulin test, biochemi-cal tests such as serum alanine aminotransferase,aspartate aminotransferase, LDH, direct and indi-rect bilirubin, BUN, Cr, serological tests for hepati-tis A, B and C virus, CMV, EBV, rubella,toxoplasma, brucella, tests for antinuclear antibod-ies, Anti-ds-DNA, RF and ANCA antibodies, coag-ulation studies for prothrombin time, activatedpartial thromboplastin time (APTT), Wbrinogen andd-dimer level, urine test, and blood and urine cul-tures and, if indicated, stool cultures. Laboratorytests were performed using standard laboratory pro-cedures. The VWF-CPase activity was not carriedout in our patients.

TPE was immediately performed when clinicaldiagnosis of TTP was conWrmed by a hematologist.The TPE was carried out 1–1.5 times at the pre-dicted plasma volume using Fresenius-AS 204 (Fres-enius AG, Germany) and Fenwal CS 3000+ (Baxter,Healthcare, DeerWeld, IL, USA) apheresis devices.Plasma volumes were calculated for each patientusing the patient’s body surface area (calculatedusing height and weight), sex, and hematocrit. Freshfrozen plasma (FFP) or cryosupernatant plasma(CSP) was used as the replacement product. Duringthe treatment period, the patient’s progress wasassessed based on clinical and laboratory parame-ters (hemoglobin (Hb), hematocrit, platelet count,LDH, peripheral blood smear for schistocytes, and

F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67 59

serum creatinine). TPE was performed daily untilthe disappearance of clinical symptoms, normaliza-tion of serum LDH and recovery of platelet countsto >150£ 109/dL; treatment was then slowlytapered. Infusions of FFP were carried out for main-tenance therapy until a response was maintained forat least 3 days. Platelet transfusions were not given ifthere was no severe bleeding, e.g. intracranial hem-orrhage. Red cell transfusions were given for severesymptomatic anemia (Hb6 8 g/dL).

Complete response (CR) to treatment was deW-ned by a platelet count >150£109/dL and normal-ization of LDH and the disappearance all clinicalsymptoms. Partial response (PR) was deWned as aplatelet count >50£ 109/dL or doubling of the ini-tial platelet count and no new clinical events.Patients who maintained a CR for at least 2 weekswithout treatment were considered to be in completeremission. Overall (OR) response was deWned as thetotal number of CR and PR. No response to treat-ment was considered in patients with no plateletresponse (platelet count <20£ 109/dL) or the plate-let increment was <100% or deterioration of thepatient’s clinical status. Exacerbation is deWned as adrop in platelet count or an increase in LDH afterinitial improvement during the treatment period orwithin 2 weeks from the last TPE. Relapse is deWnedas recurrence of any of the following after a com-plete remission: initial symptoms, MAHA, thromb-ocytopenia, deterioration of clinical status andlaboratory values.

2.1. Statistics

Data were expressed as the median, mean§SDor SE and the range. The Mann–Whitney U-testwas used to compare quantitative variables such asage, number of TPE sessions, serum BUN, Cr, Hband LDH level, duration of remission, follow-uptime and time to CR. Kaplan–Meier survival curveswere used to estimate cumulative frequency of anendpoint, and the log rank test was used to comparesurvival curves between groups. Pearson’s chi-squared test and Fisher’s exact tests were used todetermine the prognostic value of diVerent clinicaland laboratory parameters on treatment outcomes.A univariate analysis with Cox’s regression wasused to determine variables that had an eVect onsurvival. A multivariate analysis (Cox’s regression)was used to determine which variables had an inde-pendent eVect on survival. A univariate analysiswith logistic regression was used to determine vari-

ables that had no eVect on response (CR + PR). Amultivariate analysis with logistic regression wasused to determine which variables had an indepen-dent eVect on response. Statistical signiWcance wasset at P < 0.05.

3. Results

3.1. Clinical features

Of the 52 TTP patients, 18 (35%) were men and34 (65%) women. The median age was 34 (mean:38.4§16.5, range: 17–73) years. All of the patientshad anemia and thrombocytopenia upon presenta-tion. Neurological manifestations were the chief pre-senting symptom in 34 (65.4%) of the patients, mostfrequently as headache, confusion, unusual behav-iour, paresis, paresthesias, vertigo, ataxia, blurredvision, seizures and coma. Renal dysfunction wasdetected in 27 patients (52%). The most commonWnding of renal dysfunction was increased BUNlevel, serum Cr, microscopic hematuria and/or pro-teinuria. Three of 17 patients also required acutedialysis support. Fever was present in 16 of the 52patients (30.8%) and the highest temperature at pre-sentation was 38.9 °C (auxiliary measurement). Theclassical pentad picture was present only in 12 cases(23%). Twenty-nine (55.7%) of the patients had hem-orrhagic manifestations including petechiae, pur-pura and/or echymoses, microscopic hematuria,epistaxis, vaginal bleeding, haemoptysis, gastrointes-tinal hemorrhage and intracranial bleeding. Theclinical features and laboratory data at presentationare shown in Tables 1 and 2.

3.2. Patient characteristics

In the study group, 35 patients (67.3%) presentedwith idiopathic TTP. Secondary reason and/or

Table 1General patient characteristics and initial laboratory data

Normal laboratory value for creatinine (0.8–1.2 mg/dL), BUN(2–25 mg/dL), LDH (<450 U/L).

CharacteristicsMedian age (range) 34 (17–73)Female/male 34/18Mean hemoglobin level § SD (g/dL) 8.2§ 2.0 (4.6–12.8)Mean platelet count § SD (£109/L) 44.2 § 35.9 (4–137)Mean BUN level § SD (mg/dL) 41.9 § 33.6 (7–188)Mean creatinine level § SD (mg/dL) 1.6 § 1.3 (0.5–6.2)Mean LDH level § SD (U/L) 2727§ 1930 (834–11192)

60 F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67

co-morbidities were identiWed in 17 (32.7%) of thepatients: seven patients (13.5 %) were pregnant, fourpatients (7.7%) had malignancy two of whom werebreast cancer, one was pancreatic carcinoma andone was non-Hodgkin lymphoma, three patients(5.7%) had systemic lupus erythematosus (SLE), two(3.8%) had undergone BMT for which they receivedconditioning regimens with busulfan and Xudara-bine and were on cyclosporine-A (CsA) for prophy-laxis of graft versus host disease (GVHD) and onepatient (2%) had brucella infection.

3.3. Therapy characteristics

All patients received TPE as initial therapy. Pred-nisolone together with TPE was administered in 33patients (63.5%) at a standard dose of 1 mg/kg/day.FFP was used in 40 (76.9%) patients; the remaining12 (23.1%) patients received CSP as a replacementproduct for TPE. In addition to TPE, in 15 patientsa total of 18 other treatment modalities were usedincluding the following: high dose methylpredniso-lone (1 g/day for 3 days, IV, and then 1 mg/kg/day);in three (16.6%) cases, vincristine (2 mg/weekly for 2weeks, IV); in four (22.2%) patients, cyclophospha-mide (single 1-g, IV bolus); in seven (38.9%) cases,intravenous immunoglobulin (IVIG) (1 g/kg/day for3 days); in one (5.6%) patient; and, rituximab(375 mg/m2/week, 2 times) in one (5.5%) patient. Inaddition to these adjunctive therapies, one patient(5.6%) with brucellosis received TPE together withrifampicin (600 mg/day) and doxycycline (200 mg/day) for 3 weeks. Splenectomy was also performedin one (5.6%) refractory patient to TPE and othertreatment modalities.

3.4. Response to treatment

The patients underwent a median of Wve(rangeD1–45) sessions of TPE. In completeresponders, the median number of exchanges wasseven (rangeD 2–45).

Table 2Presenting symptoms of patients

Presenting symptoms Number of cases (n D 52)

Anemia 52 (100%)Thrombocytopenia 52 (100%)Neurological involvement 34 (65.4%)Renal dysfunction 27 (52%)Fever 16 (30.8%)Classical pentad 12 (23%)

Overall response to initial treatment was seen in40/52 patients (OR rateD77%): CRD31 (60%), andPRD9 (17%). No response to initial therapy wasseen in 12 patients (23%). Median time to achieve aresponse was 10 days (rangeD4–49 days); time toachieve CR was 10 days (rangeD5–49 days).

Overall responses to initial treatment were 75.8%(25/33) and 79% (15/19) in patients treated withTPE + prednisolone and TPE alone, respectively(pD1). Although TPE + prednisolone resulted in ahigher CR rate [60.6% (20/33) vs. 57.9% (11/19)] com-pared with TPE alone, the diVerence was not statisti-cally signiWcant (pD0.8). Median time to a CR waslonger in patients on TPE + prednisolone comparedto TPE alone, 11 days (rangeD5–49 days) vs. 9 days(rangeD7–12 days), respectively (pD0.3).

Overall responses to initial treatment were 91.6%(11/12) and 72.5% (29/40) in patients on TPE withFFP and CSP, respectively (pD0.25). Although TPEwith CSP resulted in a higher CR rate [83.3% (10/12)vs. 52.5% (21/40)] compared to TPE with FFP, thediVerence was not statistically signiWcant (pD0.16).Median time to a CR was 11 days (rangeD 5–48)and 10 days (rangeD5–49) in patients on TPE withCSP and FFP, respectively (pD0.8) (Table 3).

Although no statistical diVerences in terms of ORrate to initial treatment were noted between patientswith primary and secondary TTP [82.8% (29/35) vs.64.7% (11/17); pD0.17], there were statistical diVer-ences in terms of CR rate to initial treatmentbetween groups [74.2% (26/35) vs. 29.4% (5/17);pD0.005].

Overall response and CR were achieved in 10(55.5%) and six (33.3%) out of 18 other treatmentmodalities used with a median time to response of11 days (rangeD 9–47) in 15 patients. Response rateto other treatment modalities are shown in Table 4.

Six out of nine patients with primary TTP with-out a CR received other treatment modalitiesincluding high-dose methylprednisolone, vincristine,cyclophosphamide and rituximab; the remainingthree patients passed away. Three (50%) out of sixpatients achieved a CR; two were on vincristine andone who also did not achieve CR with vincristinewas on rituximab (nD 1) (Table 4). However, theremaining three patients remained unresponsive toother treatment modalities as well. These three unre-sponsive patients died.

Nine out of 12 secondary patients received othertreatment modalities including cyclophosphamide,IVIG, high-dose methylprednisolone, antibiotic andsplenectomy; remaining two out of three patients

F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67 61

passed away (one with BMT was on steroids). Three(33.3%) out of nine patients achieved CR; two wereon cyclophosphamide for SLE and one on antibioticfor brucellosis (Table 4). Of the other six patients,three remained PR and three were unresponsive totreatments. Five out of these seven patients who didnot achieved a CR died. Characteristics of thepatients with primary and secondary TTP areshown in Table 5.

Five of seven pregnant women had CR with TPEalone. Two other pregnant patients died in the early

stages of TPE therapy, due to intracranial bleeding.None of six patients who presented with the TTPsecondary to BMT and/or CsA and malignancyachieved CR with TPE + prednisolone or TPE aloneas initial therapy. Also, these patients did not show asigniWcant response to other treatment modalitiesincluding cyclophosphamide, high dose methylpred-nisolone, IVIG and splenectomy. Only one of thesepatients is still alive with PR. Three patients withSLE achieved a PR with TPE + prednisolone.Although two patients had CR after cyclophospha-

Table 3Initial treatment regimens and replacement Xuids: comparison of treatment outcomes in TTP patients

CR, complete response; OR, overall response; d, days; m, months; TPE, therapeutic plasma exchange; FFP, fresh frozen plasma; and CSP,cryosupernatant plasma.

a Pearson’s chi-squared test was used.b Mann–Whitney U-test was used.

TPE + steroid(n D 33)

TPE alone(n D 19)

P TPE with FFP(n D 40)

TPE with CSP(n D 12)

p

Median time to CR daysa (range) 11 (5–49) 9 (7–12) 0.3 10 (5–49) 11 (5–48) 0.8OR, number of patientsa (%) 25 (75.8) 15 (79) 1 29 (72.5) 11 (91.6) 0.25CR, number of patientsa (%) 11 (60.6) 20 (57.9) 0.8 21 (52.5) 10 (83.3) 0.16Mortality ratea (%) 10 (33.6) 5 (26.3) 0.1 14 (35) 1 (8.3) 0.14Median no. TPE sessionsb (range) 7 (1–45) 5 (1–11) 0.6 5 (1–45) 8 (1–45) 0.16Median follow-up timeb months (range) 9 (3d–72m) 42 (3d–73m) 0.02 16 (3d–73m) 18 (3d–42m) 0.5

Table 4Other treatment regimens received by the patients

SLE, systemic lupus erythematosus; IVIG, intravenous immunoglobulin; BMT, bone marrow transplantation; OR, overall response; andCR, complete response.

Diagnosis OR CR

Vincristine (nD 4) Primary (n D 4) 4 (100%) 2 (50%)Cyclophosphamide (n D 7) SLE (n D 3) 3 (42.8%) 2 (28.5%)

Malignancy (n D 3) 0 (0%) 0 (0%)Primary (n D 1) 0 (0%) 0 (0%)

IVIG (n D 1) BMT 0 (0%) 0 (0%)High-dose methylprednisolone (n D 3) Malignancy (n D 1) 0 (0%) 0 (0%)

Primary (n D 1) 0 (0%) 0 (0%)SLE (n D 1) 1 (33.3%) 0 (0%)

Rituximab (nD 1) Primary (n D 1) 1 (100%) 1 (100%)Splenectomy (n D 1) Malignancy (n D 1) 0 (0%) 0 (0%)Antibiotic (n D 1) Brucella (n D 1) 1 (100%) 1 (100%)

Table 5Comparison of initial treatment outcomes in primary and secondary TTP patients

OR, overall response; CR, complete response; d, days; and m, months.a Pearson’s chi-squared test was used.b Mann–Whitney U-test was used.

Primary TTP (n D 35) Secondary TTP (nD 17) p

Median time to CR, daysa (range) 11 (5–49) 9 (7–17) 0.45OR, number of patientsa (%) 29 (82.8%) 11 (64.7%) 0.17CR, number of patientsa (%) 26 (74.2%) 5 (29.4%) 0.005Mortality ratea (%) 8 (22.9%) 7 (41.2%) 0.3Median no. TPE sessionsb (range) 7 (1–45) 5 (1–13) 0.52Median follow-up timeb, months (range) 20 (3d–73m) 10 (3d–49m) 0.36

62 F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67

mide, the third patient died in PR status due to sep-sis after high dose methylprednisolone andcyclophosphamide. One patient with brucellosis hadCR with TPE and antibiotic treatment.

3.5. Relapse and exacerbation

In one of the primary TTP patient who had CRafter daily TPE therapy for 12 days, exacerbation ofthe disease symptoms, increased LDH level anddecreased platelet level occurred during FFP main-tenance therapy, TPE was reintroduced twice a dayand a CR was achieved following eight session ofTPE. Three (8.1%) of 37 complete responders expe-rienced a relapse. All three patients relapsed wereprimary TTP. One presented with a combination ofthrombocytopenia, MAHA and renal failure within39 months after discharge and achieved a CR afterseven sessions of TPE + prednisolone. Other twopatients presented with thrombocytopenia andMAHA within 12 and 34 months after discharge:One achieved a CR after 12 sessions ofTPE + prednisolone; another achieved a CR afterFFP infusion + prednisolone at another center.

3.6. Follow up characteristics

The median follow-up period was 40 months(range: 3 days–73 months). Currently, 37/52 (71.1%)patients are still alive; 15/52 (28.8%) patients died.The causes of death were related to multiple factors:sepsis§disseminated intravascular coagulation(DIC)§ multiple organ failure§GVHD§primarydiseaseD 5, DICD 1, hemorrhageD 2, cardiacreasonD 3, sepsisD2 and strokeD 1. There were nostatistical diVerences in terms of mortality ratebetween patients with secondary and primary TTP[41.2% (7/17) vs. 22.9% (8/35); pD 0.3]. 3/7 (46.6%)with primary TTP and 4/7 with secondary TTP (twowere pregnant and two had malignancy) passedaway during the Wrst acute episode before complet-ing the course of TPE. Four out of eight patientswith primary TTP and 4/8 with secondary TTP (tworeceived BMT, CsA and steroid, one had malig-nancy, one had SLE) passed away at 3 days–72months after initial treatment.

No statistical diVerences in terms of mortalityrate were noted between patients on TPE alone andTPE + prednisolone [26.3% (5/19) vs. 30.3% (10/33),pD0.1] and patients on TPE using replacementproducts of FFP and CSP [35% (14/40) vs. 8.3% (1/12);pD0.14]. Although a 30-day mortality rate was

13.4% (7/52), the overall mortality rate was 28.8%.Overall survival (OS) curve of patients is shown inFigs. 1, 2 and 3. The median estimated OS was 72months (95% CI: 27.2–116.8). Although the meanestimated OS was higher in patients with primaryTTP compared to secondary TTP, there were no sta-tistical diVerences between groups [59.4§ 5.15 (95%CI: 49.3–69.6) vs. 29.5§5.6 (95% CI: 18.4–40.5),pD0.07)]. Additionally, we could not demonstrateany association in terms of the mean estimated OSbetween patients on TPE and TPE + prednisolone[57.6§7.6 (95% CI: 42.7–72.6) vs. 48.5§ 6.9 (95%CI: 35.0–62.0), pD 0.35)]. Although the mean esti-mated OS was higher in patients receiving FFP asthe replacement products, compared to CSP therewere no statistical diVerences between these groups[50.6§5.5 (95% CI: 39.3–61.3) vs. 38.5§ 3.2 (95%CI: 31.9–45.0), pD0.2)].

3.7. Factors related to response and survival

A univariate analyses with logistic regressionwere performed to evaluate whether there was anyassociation between response and age, sex, type ofTTP (primary and secondary), anemia, thrombocy-

Fig. 1. OS in primary and secondary TTP (OS D overall survival).

80.0060.0040.0020.000.00Time (months)

1.0

0.8

0.6

0.4

0.2

0.0C

um

Su

rviv

al2.00-censored

1.00-censored

secondary

primary

F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67 63

topenia, neurological manifestations, renal dysfunc-tion, fever, replacement products (FFP and CSP),

Fig. 2. OS in patients on TPE and TPE + prednisolone(OS D overall survival).

80.0060.0040.0020.000.00

Time (month)

1.0

0.8

0.6

0.4

0.2

0.0

Cu

m S

urv

ival

2.00-censored

1.00-censored

PE+steroid

PE

Fig. 3. OS in patients on TPE with CSP and FFP (OS D overallsurvival).

80.0060.0040.0020.000.00

Time (month)

1.0

0.8

0.6

0.4

0.2

0.0

Cu

m S

urv

ival

1.00-censored

.00-censored

FFP

CSP

therapy (TPE alone or TPE + prednisolone), num-ber of TPE, initial serum Cr, BUN, LDH, Hb leveland platelet count. In multivariate analysis withlogistic regression (backward stepwise procedure),there was nearly a 24.9-fold increased responseprobability in patients without fever compared topatients with fever [odds ratio (OR) 0.040, 95% CI:0.004–0.367, p < 0.01]. As the number of TPE wasincreased, the response probability increased [OR1.794, 95% CI: 1.158–2.778; p < 0.01]. None of theother parameters were associated with the probabil-ity of achieving response.

A univariate analysis with Cox’s regression wasused to determine whether there was any associationbetween survival and age, sex, type of TTP, anemia,thrombocytopenia, neurological manifestations,renal dysfunction, fever, replacement products, ther-apy, number of TPE, initial serum Cr, BUN, LDH,Hb level, and platelet count. Fever, LDH level andresponse status [unresponsive and responsive(PR + CR)] were found to be associated with theprobability of survival. However, in multivariateanalysis with Cox’s regression, there was nearly a32.3-fold increased probability of survival inpatients who were responsive to treatment com-pared to patients who were unresponsive [hazardratio 0.031, 95% CI: 0.007–0.141, p < 0.001]. None ofother parameters had statistical signiWcance regard-ing the probability of survival in multivariate Cox’sregression analysis.

4. Discussion

TTP typically aVects women more than men, witha ratio of 2:1 [1,31–34]. The most commonly aVectedindividuals are between 20 and 60 years of age with areported incidence of 3.7 cases per million adult[22,32,33]. In our study, 65% of the patients werefemales, with an age range of 17–73 years (median 34years). It is reported that the most frequentlyobserved abnormalities are MAHA (100%) andthrombocytopenia (100%), neurological manifesta-tions (60–88%), renal dysfunction (18–76%) andfever (22–86%) [18–20,31–36]. The presence of thecomplete pentad only occurs in a few cases at presen-tation [15]. In our study, the classical pentad picturewas present in only 23% of the cases. Additionally,all patients had anemia and thrombocytopenia onpresentation. Neurological manifestations were thechief presenting symptom in 65.4% of the patients.Renal dysfunction was detected in 52% of thepatients. Fever was present in 30.8% of the cases.

64 F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67

It is reported that an excess of unusually largevon Willebrand factor (vWF) multimers have beenlinked to TTP [9,37]. Recent studies suggest that thepresence of auto-antibodies against VWF-CPasemay be responsible for the presence of these multim-ers in adult TTP patients [5,7,27]. Therefore, TPE isconsidered to be the therapy of choice in adult TTPpatients. It removes antibodies against VWF-CPaseand replaces fresh proteases. Response rates to TPEwere 56–90% of acute TTP [15,19,20,31,34,38]. In thepresent study, OR rate to initial therapy was 77%,with a 60% rate of CR.

Although there are no well-conducted random-ized trials regarding the use of corticosteroids inTTP patients, corticosteroids are often used in com-bination with TPE as a frontline therapy [1,39,40].Some studies report that TTP might be controlledwith high dose corticosteroid therapy [29,41]. How-ever, Gurkan E revealed that there was no overallbeneWt in using high-dose corticosteroids in adjunctto TPE [42]. In our study, there was no advantage ofTPE + prednisolone compared to TPE alone interms of OR (75.8% vs. 79%; pD1) and CR rates(60.6% vs. 57.9%; pD 0.8). Additionally, the mediantime to a CR was longer in patients onTPE + prednisolone compared with TPE alone (11days vs. 9 days); however, there were no statisticaldiVerences between groups (pD 0.3).

On the basis of the possible involvement of largeVWF multimers in the pathogenesis of TTP [9,43],the use of CSP seems to be reasonable due to thefact that it is relatively deWcient of VWF multimers[44,45]. Some studies reported that TPE with CSPwas eVective in unresponsive patients or who hadincomplete response to TPE with FFP [44,46,47].Other retrospective studies revealed that TPE withCSP showed an improved response rate and survivalin newly diagnosed TTP patients [45,48]. However,North American TTP Group’s retrospective ran-domized study revealed that FFP and CSP are bothacceptable replacement products in the treatment ofTTP and provide similar results in the acute man-agement of newly diagnosed primary TTP [49]. Inthe present study, OR rate was higher in patients onTPE with FFP compared to CSP (91.6% vs. 72.5%,respectively). However, there were no statisticaldiVerences between the two groups (pD0.25). Incontrast to OR, TPE with CSP resulted in a higherCR rate compared to TPE with FFP (83.3% vs.52.5%). However, the diVerence was not statisticallysigniWcant (pD0.16). Additionally, there were no

statistical diVerences between patients on TPE withCSP and FFP in terms of median time to a CR (11days vs. 10 days; pD 0.8).

Some patients do not respond well to TPE andwill need alternative and/or adjunct therapy withsteroids, vincristine, cyclophosphamide, rituximabor other medications that can suppress inhibitorproduction [25,27–29,50,51]. If all these treatmentsfail, then the removal of the spleen may be consid-ered as a means to reach sustainable remission or toreduce/prevent future relapses [15,26,52]. BecauseTTP is a rare disorder, demonstration of the eYcacyof these alternate and/or adjunct therapies has beenlimited to small studies. Often, several treatmentapproaches may be required to treat unresponsivecases [15,53]. In this study, OR and CR rates 55.5%and 33.3%, respectively, were achieved with othertreatment regimens. Complete response wasachieved in 100% with rituximab, 50% with vincris-tine and 28.5% with cyclophosphamide. There wasno response to high dose methylprednisolone, IVIGand splenectomy.

Most TTPs are called primary TTP because nounderlying cause is detected. However, when TTP isfound in association with certain diseases like malig-nancy, chemotherapy, BMT, autoimmune disorders,pregnancy, drugs and infections, then it is called sec-ondary TTP [10–18,54,55]. In the present study, sec-ondary TTP is identiWed in 32.7% of the patients.The response to TPE in primary TTP is excellent.When TTP is secondary to some other underlyingdisorder, the responses may be minimal and unsus-tainable, especially if the causative ailment or agentcannot be eradicated. In our study, OR and CRrates to initial treatment were 64.7% and 29.4% inpatients with secondary TTP, respectively. 29.4% ofsecondary TTP had CR to initial treatment and33.3% (3/9) achieved CR with cyclophosphamide forSLE and antibiotic for brucellosis. None of thepatients presented with TTP secondary to CsA,BMT and malignancy were assessed to have anysigniWcant response. This is consistent with the expe-rience of TPE in this group, where the utility of TPEis questionable [13,55,56]. TTP is rarely seen as com-plicating SLE [12,17]. There is scarce data regardingoptimal treatment of TTP patients complicated withSLE [24,57,58]. In our patient population, weobserved that all of the SLE patients were unrespon-sive to TPE + steroid. Complete response could beachieved after cyclophosphamide therapy. Our Wnd-ings suggest that the earlier addition of immunosup-

F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67 65

pressive agents to TPE may improve treatmentoutcome in patients with SLE.

TPE has reduced the overall mortality rate from90% in untreated cases to less than 20%[1,18,19,25,38,59]. Although the 30-day mortalityrate (13.4%) in the present study is comparable tothat in the literature [21,45], the overall mortality of28.8% in this study was relatively higher in compari-son with others in the literature [1,18,19,31,32,38].The relatively high overall mortality rate may beattributed to the increased number of secondaryTTP patients, where the mortality rate was highercompared with primary TTP patients (41.2% vs.22.9%; pD0.3) and the CR rate was lower comparedto primary TTP patients (74.2% vs. 29.4%;pD 0.005).

Although, the mean estimated OS was higher inpatients with primary TTP compared to secondaryTTP, no statistical diVerences were noted betweenthe groups (59.4§ 5.15 vs. 29.5§ 5.6; pD 0.07)).Additionally, we could not demonstrate any associ-ation, in terms of the estimated OS, betweenpatients on TPE and TPE + prednisolone (pD 0.35)and on TPE + FFP and TPE + CSP (pD 0.2). How-ever, there was nearly a 32.3-fold increased proba-bility of survival in patients who were responsive totreatment compared to patients who were unre-sponsive (p < 0.001). Additionally, there was nearlya 24.9-fold increased response probability inpatients with fever compared to patients withoutfever (p < 0.01). In the present study, as the numberof TPE was increased, the response probabilityincreased proportionately (p < 0.01). Therefore,TPE should be performed immediately after TTPdiagnosis; it should then be more slowly tapered inTTP patients with fever compare to patients with-out fever.

We conclude that TPE is extremely eVective forprimary TTP. However, it may be used as anadjunctive therapy for secondary TTP, until theprimary disorder is successfully controlled. Theaddition of steroids to TPE had no advantage interms of response rate, time to response and sur-vival compare to TPE alone. Additionally, CSP asa replacement product is not superior compared toFFP in terms of response rate, time to response andsurvival. However, the roles of steroids and CSPneed to be further investigated by randomized pro-spective clinical trials conducted with a largerseries of patients. Additionally, fever appears to bea poor prognostic indicator; further, prolonged

treatment with TPE may be needed in patients withfever.

References

[1] Bell WR, Braine HG, Ness PM, Kickler TS. Improved sur-vival in thrombotic thrombocytopenic purpura-hemolyticuremic syndrome. Clinical experience in 108 patients. N EnglJ Med 1991;325:398–403.

[2] Moake JL. Thrombotic microangiopathies. N Engl J Med2002;347:589–600.

[3] Drew MJ. Therapeutic plasma exchange in hematologic dis-eases and dysproteinemias. In: McLeod BC, Price TH, Wein-stein R, editors. Apheresis: principles and practice. 2nd ed.Bethasda, MD: AABB Press; 2003. p. 345–73.

[4] Chow TW, Turner NA, Chintagumpala M, McPherson PD,Nolasco LH, Rice L, et al. Increased von Willebrand factorbinding to platelets in single episode and recurrent types ofthrombotic thrombocytopenic purpura. Am J Hematol1998;57:293–302.

[5] Tsai HM, Lian EC. Antibodies to von Willebrand factor –cleaving protease in acute thrombotic thrombocytopenicpurpura. N Engl J Med 1998;339:1585–94.

[6] Furlan M, Robles R, Solenthaler M, Wassmer M, Sandoz P,Lammle B. DeWcient activity of von Willebrand factor-cleav-ing protease in chronic relapsing thrombotic thrombocy-topenic purpura. Blood 1997;89:3097–103.

[7] Furlan M, Robles R, Galbusera M, Remuzzi G, Kyrle PA,Brenner B, et al. Von Willebrand factor-cleaving protease inthrombotic thrombocytopenic purpura and hemolytic-ure-mic syndrome. N Engl J Med 1998;339:1578–84.

[8] Levy GG, Nichols WC, Lian EC, Foroud T, McClintick JN,McGee BM, et al. Mutations in a member of the ADAMTSgene family cause thrombotic thrombocytopenic purpura.Nature 2001;413:488–94.

[9] Moake JL, Turner NA, Stathopulos NA, Nolasco L, Hel-lums JD. Involvement of large plasma von Willebrand factor(vWF) multimers and unusually large vWF forms derivedfrom endothelial cells in shear stress-induced platelet aggre-gation. J Clin Invest 1986;78:1456–61.

[10] Altuntas F, Eser B, Sar I, Yildiz O, Cetin M, Unal A. Severethrombotic microangiopathy associated with brucellosis:successful treatment with plasmapheresis. Clin Appl ThrombHemost 2005;11(1):105–8.

[11] Bennett CL, Connors JM, Carwile JM, Moake JL, Bell WR,Tarantolo SR, et al. Thrombotic thrombocytopenic purpuraassociated with clopidogrel. New Engl J Med 2000;342:1773–7.

[12] Nesher G, Hanna VE, Moore TL, Hersh M, Osborn TG.Thrombotic microangiopathic hemolytic anemia in systemiclupus erythematosus. Semin Arthritis Rheum 1994;24:165–72.

[13] Roy V, Rizvi MA, Vesely SK, George JN. Thromboticthrombocytopenic purpura-like syndromes following bonemarrow transplantation: an analysis of associated conditionsand clinical outcomes. Bone Marrow Transplant2001;27:641–6.

[14] Shamseddine A, Chehal A, Usta I, Salem Z, El Saghir N,Taher A. Thrombotic thrombocytopenic purpura and preg-nancy: report of 4 cases and literature review. J Clin Apher2004;19:5–10.

66 F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67

[15] Thompson CE, Damon LE, Ries CA, Linker CS. Throm-botic microangiopathies in the 1980s: clinical features,response to treatment, and the impact of the human immun-odeWciency virus epidemic. Blood 1992;80:1890–5.

[16] Tsai HM, Rice L, Sarode R, Chow TW, Moake JL. Antibodyinhibitors to von Willebrand factor metalloproteinase andincreased von Willebrand factor-platelet binding inticlopi-dine-associated thrombotic thrombocytopenic purpura. AnnInt Med 2000;132:794–9.

[17] Vaidya S, Abulezz S, Lipsmeyer E. Thrombotic thrombocy-topenic purpura and systemic lupus erythematosus. Scand JRheumatol 2001;30(5):308–10.

[18] Lara PN, Coe TL, Zhou H, Fernando L, Holland PV, WunT. Improved survival with plasma exchange in patients withthrombotic thrombocytopenic purpura-hemolytic uremicsyndrome. Am J Med 1999;107:573–9.

[19] Rock GA, Shumak KH, Buskard NA, Blanchette VS, KeltonJG, Nair RC, et al. Comparison of plasma exchange withplasma infusion in the treatment of thrombotic thrombocy-topenic purpura. N Engl J Med 1991;325:393–7.

[20] Sarode R, Gottschall JL, Aster RH, McFarland JG. Throm-botic thrombocytopenic purpura: early and late responders.Am J Hematol 1997;54:102–7.

[21] Hayward CP, Sutton DM, Carter Jr WH, Campbell ED,Scott JG, Francombe WH, et al. Treatment outcomes inpatients with adult thrombotic thrombocytopenic pur-pura-hemolytic uremic syndrome. Arch Int Med 1994;154:982–7.

[22] Torok TJ, Holman RC, Chorba TL. Increasing mortalityfrom thrombotic thrombocytopenic purpura in the UnitedStates: analysis of national mortality data ,1968–1991. Am JHematol 1995;50:84–90.

[23] Ahmad A, Aggarwal A, Sharma D, Dave HP, Kinsella V,Rick ME, et al. Rituximab for treatment of refractory/relaps-ing thrombotic thrombocytopenic purpura. Am J Hematol2004;77(2):171–6.

[24] Akaogi J, Akasaka N, Yamada H, Hama N, Satoh M,Nichols C, et al. Intravenous cyclophosphamide therapy in acase with refractory thrombotic microangiopathic hemolyticanemia and SLE. Clin Rheumatol 2004;23(6):541–3.

[25] Allan DS, Kovacs MJ, Clark WF. Frequently relapsingthrombotic thrombocytopenic purpura treated with cyto-toxic immunosuppressive therapy. Haematologica2001;86(8):844–50.

[26] Kappers-Klunne MC, Wijermans P, Fijnheer R, CroockewitAJ, van der Holt B, de Wolf JT, et al. Splenectomy for thetreatment of thrombotic thrombocytopenic purpura. Br JHaematol 2005;130(5):768–76.

[27] Bohm M, Betz C, Miesbach W, Krause M, von Auer C, Gei-ger H, et al. The course of ADAMTS-13 activity and inhibi-tor titre in the treatment of thrombotic thrombocytopenicpurpura with plasma exchange and vincristine. Br J Haema-tol 2005;129(5):644–52.

[28] Stein GY, Zeidman A, Fradin Z, Varon M, Cohen A, Mittel-man M. Treatment of resistant thrombotic thrombocy-topenic purpura with rituximab and cyclophosphamide. Int JHematol 2004;80(1):94–6.

[29] Toyoshige M, Zaitsu Y, Okafuji K, Inoue Y, Hiroshige Y,Matsumoto N, et al. Successful treatment of thromboticthrombocytopenic purpura with high-dose corticosteroid.Am J Hematol 1992;41:69.

[30] Dervenoulas J, Tsirigotis P, Bollas G, Koumarianou AA,Pappa V, Mantzios G, et al. EYcacy of intravenous immuno-globulin in the treatment of thrombotic thrombocytopaenicpurpura. A study of 44 cases. Acta Haematol2001;105(4):204–8.

[31] Dervenoulas J, Tsirigotis P, Bollas G, Pappa V, Xiros N,Economopoulos T, et al. Thrombotic thrombocytopenicpurpura/haemolytic uremic syndrome (TTP/HUS): treat-ment outcome, relapses, prognostic factors. A single-centerexperience of 48 cases. Ann Hematol 2000;79:66–72.

[32] Ramanan AS, Thirumala S, Chandrasekaran V. Thromboticthrombocytopenic purpura: a single institution experience. JClin Apher 1999;14:9–13.

[33] Rock G, Kelton JG, Shumak KH, Buskard NA, Sutton DM,Benny WB. Laboratory abnormalities in thrombotic throm-bocytopenic purpura. Br J Haematol 1998;103:1031–6.

[34] Shamseddine A, Saliba T, Aoun E, Chahal A, El-Saghir N,Salem Z, et al. Thrombotic thrombocytopenic purpura: 24years of experience at the American University of BeirutMedical Center. J Clin Apher 2004;19(3):119–24.

[35] Gasparovic V, Radonic R, Mejic S, Pisl Z, Radman I. Possi-bilities and limits of treatment in patients with thromboticthrombocytopenic purpura. Intensive Care Med2000;26:1690–3.

[36] RidolW RL, Bell WR. Thrombotic thrombocytopenic pur-pura. Report of 25 cases and review of the literature. Medi-cine 1981;60:413–28.

[37] Moake JL, Rudy CK, Troll JH, Weinstein MJ, ColanninoNM, Azocar J, et al. Unusually large plasma factor VIII: vonWillebrand factor multimers in chronic relapsing thromboticthrombocytopenic purpura. N Engl J Med 1982;307:1432–5.

[38] Bandarenko N, Brecher ME. United States ThromboticThrombocytopenic Purpura Apheresis Study Group (USTTP ASG): multicenter survey and retrospective analysis ofcurrent eYcacy of therapeutic plasma exchange. J ClinApher 1998;13(3):133–41.

[39] De la Rubia J, Lopez A, Arriaga F, Cid AR, Vicente AI,Marty ML, et al. Response to plasma exchange and steroidsas combined therapy for patients with thrombotic thromb-ocytopenic purpura. Acta Haematol 1999;102:12–6.

[40] Yang CW, Chen YC, Dunn P, Chang MY, Fang JT, HuangCC. Thrombotic thrombocytopenic purpura: initial treat-ment with plasma exchange plus steroids and immunosup-pressive agents for relapsing cases. Renal Fail 2003;25:21–30.

[41] Ozsoylu S. High-dose intravenous methylprednisolone forthrombotic thrombocytopenic purpura. Acta Haematol1990;84:110.

[42] Gurkan E, Baslamisli F, Guvenc B, Kilic NB, Unsal C, Kara-koc E. Thrombotic thrombocytopenic purpura in southernTurkey: a single-center experience of 29 cases. Clin Lab Hae-matol 2005;27(2):121–5.

[43] Moake JL, McPherson PD. Abnormalities of von Wille-brand factor multimers in thrombotic thrombocytopenicpurpura and hemolytic-uremic syndrome. Am J Med1989;87:9–15.

[44] Byrnes JJ, Moake JL, Klug P, Periman P. EVectivenss of thecryosupernatant fraction of plasma in the treatment ofrefractory thrombotic thrombocytopenic purpura. Am JHematol 1990;34:169–74.

[45] Rock G, Shumak KH, Sutton DM, Buskard NA, Nair RC.Cryosupernatant as replacement Xuid for plasma exchange

F. Altuntas et al. / Transfusion and Apheresis Science 36 (2007) 57–67 67

in thrombotic thrombocytopenic purpura. Br J Haematol1996;94:383–6.

[46] Obrador GT, Zeigler ZR, Shadduck RK, Rosenfeld CS,Hanrahan JB. EVectiveness of cryosupernatant therapy inrefractory and chronic relapsing thrombotic thrombocy-topenic purpura. Am J Hematol 1993;42:217–20.

[47] Sgarabotto D, Vianello F, Scano F, Stefani PM, Sartori R,Girolami A. Clinical and laboratoristic remission after cryo-supernatant plasma exchange in thrombotic thrombocy-topenic purpura. Hematologica 1998;83:569–76.

[48] Owens MR, Sweeney JD, Tahhan RH, Fortkolt P. InXuenceof type of exchange Xuid on survival in therapeutic apheresisfor thrombotic thrombocytopenic purpura. J Clin Apher1995;10:178–82.

[49] Zeigler ZR, Shadduck RK, Gryn JF, Rintels PB, George JN,Besa EC, et al. Cryoprecipitate poor plasma does notimprove early response in primary adult thrombotic thromb-ocytopenic purpura. J Clin Apher 2001;16(1):19–22.

[50] Gutterman LA, Kloster B, Tsai HM. Rituximab therapy forrefractory thrombotic thrombocytopenic purpura. BloodCells Mol Dis 2002;28(3):385–91.

[51] Yomtovian R, Niklinski W, Silver B, Sarode R, Tsai HM.Rituximab for chronic recurring thrombotic thrombocy-topenic purpura: a case report and review of the literature.Br J Haematol 2004;124(6):787–95.

[52] Aqui NA, Stein SH, Konkle BA, Abrams CS, Strobl FJ. Role ofsplenectomy in patients with refractory or relapsed thromboticthrombocytopenic purpura. J Clin Apher 2003;18(2):51–4.

[53] George JN. How I treat patients with thrombotic thromb-ocytopenic purpura-hemolytic uremic syndrome. Blood2000;96:1223–9.

[54] Mach-Pascual S, Samii K, Beris P. Microangiopathichemolytic anemia complicating FK506 (tacrolimus) therapy.Am J Hematol 1996;52:310–2.

[55] Moake JL, Byrnes JJ. Thrombotic microangiopathies associ-ated with drugs and bone marrow transplantation. HematolOncol Clin North Am 1996;10:485–97.

[56] Sarode R, McFarland JG, Flomenberg N, Casper JT, CohenEP, Drobyski WR, et al. Therapeutic plasma exchange doesnot appear to be eVective in the management of thromboticthrombocytopenic purpura/hemolytic uremic syndrome fol-lowing bone marrow transplantation. Bone Marrow Trans-plant 1995;16(2):271–5.

[57] Perez-Sanchez I, Anguita J, Pintado T. Use of cyclophospha-mide in the treatment of thrombotic thrombocytopenic pur-pura complicating systemic lupus erythematosus: report oftwo cases. Ann Hematol 1999;78(6):285–7.

[58] Vasoo S, Thumboo J, Fong KY. Thrombotic thrombocy-topenic purpura in systemic lupus erythematosus: diseaseactivity and the use of cytotoxic drugs. Lupus 2002;11:443–50.

[59] Amorosi EL, Ultman JE. Thrombotic thrombocytopenicpurpura: report of 16 cases and review of literature. Medi-cine 1966;45:139–59.