Validation of prognostic factors and survival of patients with multiple myeloma in a real-life autologous stem cell transplantation setting : a Swiss single centre experience

List of abbreviations ANC absolute neutrophil count ASCT autologous stem cell transplantation BEAM high-dose carmustine, etoposide, cytarabine and melphalan BMI body mass index CD34 cluster of differentiation 34 CI confidence interval CR complete remission CT computed tomography Dex dexamethasone DS Durie Salmon EFS event free survival HR hazard ratio IMIDs immunomodulatory derivatives ISS international staging system M2 cyclophosphamide, carmustine, melphalan, prednisone MM multiple myeloma MR minor remission nCR near complete remission OS overall survival PD progressive disease PR partial remission SD stable disease Thal thalidomide TRM treatment related mortality VAD vincristine, adriamycin, dexamethasone Vel bortezomib (Velcade) VGPR very good partial remission METHODS: Medical histories of the patients were screened for response, event-free survival (EFS) and overall survival (OS). Pre-transplant variables were analysed to identify possible prognostic risk factors. RESULTS: Overall, 182 ASCT were performed in 120 patients with MM from 2002 to 2007. Treatment-related mortality (TRM) was 0.5%. Median EFS was 23.1 months (95% confidence interval [CI]: 19.4–28.4) and median OS was 49.8 months (95%CI: 43.7–not reached) in the whole patient population. The median OS in patients who received one ASCT was 46.4 months (95%CI: 35.2–not reached), and 63.7 months (95%CI: 48.9–not reached) in patients who underwent double ASCT. Patients who already achieved a complete remission (CR) before ASCT had a longer EFS (p = 0.016) than patients without CR. Additionally, patients who achieved a CR after ASCT had a longer EFS (p = 0.0061) and OS (p = 0.0024) than patients without CR. ISS stage <III at first diagnosis strongly correlated with improved EFS (p = 0.0006) and OS (p <0.0001). CONCLUSIONS: ASCT is a safe and effective treatment mode in eligible patients with MM. TRM was below average at our institution. Achievement of CR after transplantation was the most valuable predictor for improved overall survival.

METHODS: Medical histories of the patients were screened for response, event-free survival (EFS) and overall survival (OS).Pre-transplant variables were analysed to identify possible prognostic risk factors.RESULTS: Overall, 182 ASCT were performed in 120 patients with MM from 2002 to 2007.Treatment-related mortality (TRM) was 0.5%.Median EFS was 23.1 months (95% confidence interval [CI]: 19.4-28.4) and median OS was 49.8 months (95%CI: 43.7-not reached) in the whole patient population.The median OS in patients who received one ASCT was 46.4 months (95%CI: 35.2-not reached), and 63.7 months (95%CI: 48.9-not reached) in patients who underwent double ASCT.Patients who already achieved a complete remission (CR) before ASCT had a longer EFS (p = 0.016) than patients without CR.Additionally, patients who achieved a CR after ASCT had a longer EFS (p = 0.0061) and OS (p = 0.0024) than patients without CR.ISS stage <III at first diagnosis strongly correlated with improved EFS (p = 0.0006) and

Introduction
High-dose chemotherapy with autologous stem cell transplantation (ASCT) has been widely established as consolidation or salvage treatment for patients with multiple myeloma (MM) [1][2][3][4].In general the response after first ASCT is being used to direct the subsequent patient treat-ment; application of two consecutive ASCTs within 2-6 months (so-called double or tandem ASCT) is usually reserved for patients not achieving at least a very good partial remission (VGPR) after the first ASCT, according to results from two randomised prospective trials [5][6].The indication and timing of high-dose chemotherapy with subsequent ASCT have been matter of debate recently in light of new active drugs, e.g., proteasome inhibitors and immunomodulatory derivatives (IMIDs), which are increasingly incorporated into various first line treatment regimens before ASCT [7][8][9][10][11][12]; the question whether high-dose chemotherapy can be safely replaced by a non-high dose consolidation treatment will be answered by a recently activated international trial [13].Furthermore, the impact and composition of maintenance treatment after ASCT and even allogeneic transplantation concepts are currently under discussion and various prospective trials will further define future treatment strategies for younger MM patients [14][15][16][17].Until these data are available and also implemented in daily practice, retrospective analyses -especially of patients who are treated in transplantation centres outside the clinical trials setting -are helpful in demonstrating the current options for patients desirous of knowing their specific risks and benefits at a given institution.We started our autologous stem cell programme in Zürich in 1988 and have performed more than 1000 transplantations since that time, with MM the most frequent indication for ASCT [18].In terms of quality control we continuously evaluate our programme's clinical data.We have therefore analysed the clinical course and outcome of our MM patients receiving ASCT between 2002 and 2007 and identified possible risk factors for survival.Additionally, we have assessed the morbidity of our treatment regimens with regard to length of hospital stay, need for antibiotics and blood product transfusions.

Patient data assessment
Records from patients with MM who received at least one ASCT at our transplantation centre were analysed retrospectively on the basis of a prospective database.This analysis was approved by our local ethics committee.The Zürich transplantation centre comprises two hospitals, Zürich University Hospital and Triemli City Hospital, Zürich.Data regarding myeloma stage according to the international staging system (ISS), response to treatment before and after ASCT, event free survival (EFS) and overall survival (OS) were collected.Also, various parameters such as age, gender, body mass index, time from initial diagnosis to ASCT and number of ASCTs per patient were documented if available.As quality control for our transplantation programme we also collected data on the haematological toxicity of the treatment and the supportive care measures performed during the post-transplant period.

Treatment
Individual eligibility for ASCT was discussed in our multidisciplinary autologous transplantation board and determined in accordance with international recommendations.
Patients received either one or two ASCTs.The second ASCT was usually performed sequentially 2-6 months after the first ASCT (so-called double or tandem ASCT), or, in a few patients who had initially received single ASCT, later in their disease course in the event of progression.All patients received primary induction chemotherapy as chosen by the attending oncologists.For stem cell mobilisation, either cyclophosphamide on day -12 or vinorelbine on day -8 was given, followed by collection of the stem cells on day 0. Additionally, patients received filgrastim at a dose of 10 μg/kg body weight per day from day -4 on.Aphereses were performed at our transplantation centre until at least 4 x 10 6 CD34-positive cells per kilogram body weight were collected.Subsequently stem cells were cryopreserved and thawed immediately before retransfusion.Conditioning chemotherapy consisted of melphalan at a dose of 200 mg/m 2 , given in two doses at days -3 and -2 or as single dose at day -2 before stem cell transplantation.Patients with reduced performance score, organ dysfunctions, e.g.renal insufficiency or age over 65 years received a reduced dose in the range of 100-140 mg/m 2 melphalan.Patients received filgrastim from day +5 on or pegfilgrastim once on day +1 after ASCT to shorten time to engraftment.

Response criteria
Assessment of disease response was chiefly based on serological testing because bone marrow biopsy after transplantation was omitted in most of our patients.Thus, our response assessment varied from the International Myeloma Working Group uniform criteria, which are currently proposed as a new standard for response assessment [19].Complete Remission (CR) was defined as negative immunofixation and disappearance of monoclonal protein in serum and urine.Near Complete Remission (nCR) was defined as disappearance of monoclonal protein in serum and/or urine but persistence of positive immunofixation.Very Good Partial Remission (VGPR) was defined as reduction of monoclonal protein >90% in serum.Partial Remission (PR) was defined as reduction of paraprotein ≥50% in serum.Patients with Minor Response (MR) showed a paraprotein reduction of between 25-49% respectively.Stable Disease (SD) was defined as every response between minor response and progressive disease (PD).The latter was documented in patients who had an increase in monoclonal protein ≥25% or clinical progression after transplantation.Patients with no data available regarding their response were regarded and coded as non-responders (i.e.SD).Event free survival (EFS) was defined as time from ASCT to the time of first recurrence after achievement of CR or nCR, to the time of progression in non-CR patients, as verified by serological assessment, bone marrow biopsy or imaging, or to the time of death by any cause.Overall survival (OS) was defined as time from ASCT to the time of death by any cause, as documented in the patient charts.

Statistics
Descriptive statistics (median and range, or counts) were calculated for all variables.Event free and overall survival

Original article
was calculated from the date of ASCT and censored at the date of last follow-up.Survival curves were computed using the method of Kaplan and Meier, and compared using the logrank test [20].Univariate regression analysis was performed using Cox proportional hazards regression [21].P-values <0.05 were considered statistically significant.All analyses were performed in the R programming language [22].

Patient demographics
Between January 2002 and December 2007 a total of 182 ASCTs were performed in 120 MM patients.Sixty-three (53%) patients received single ASCT, and 57 (47%) patients tandem ASCTs during this period.Five (8%) of the single-transplanted patients received a second ASCT later in the disease course after disease progression.The median time from diagnosis to the first ASCT was 6.24 months (range 3.55-283.04months).Patients predominantly received split dose melphalan for conditioning.Only in six cases (3%) was melphalan given as a single dose at day -2 before ASCT.Three patients (2%) underwent haemo-  dialysis at the time of ASCT, in two patients (2%) underlying amyloidosis of the kidneys had been diagnosed.Cytogenetics were not done routinely at our institution during that time period, pathological findings are reported only in five patients (4%) (three patients with del(13q), one patient with del(13q) and t(4;14), one patient with complex aberrations).Patient characteristics are shown in table 1.

Risk factor analysis
By univariate analysis, achievement of a CR before ASCT correlated with better EFS compared with any other response (p = 0.016), without impacting on OS (p = 0.21).Also, CR after ASCT was associated with improved EFS (p = 0.006) and OS (p = 0.0024) compared with any other response.In the subgroup analysis, the achievement of CR after ASCT was associated with better OS especially compared to PR (p = 0.026) and responses less than PR (p = 0.009).Other parameters significantly associated with a better EFS were paraprotein levels of less than 5 g/dL at diagnosis (p = 0.007), a time of 6-12 months from diagnosis to ASCT compared with earlier transplantation (p = 0.042), and MM stage of less than III according to ISS (p = 0.0006).The latter parameter was also significantly linked with better OS (p <0.0001) (table 4).

Discussion
In the present analysis we observed that ASCTs performed for MM at our centre results in high response rates of over 80% in single and in double transplanted patients.Compared to available data in the literature, the procedure-related mortality during the years 2002 and 2007 was very low with 0.5% [23].A CR after ASCT and ISS stage of less than III were the best predictors for prolonged survival.Age was not associated with a worse outcome in selected patients eligible for ASCT.We were able to document a median survival of 49.8 months from the first ASCT in our patients.Separately analysed by the number of ASCTs, the median survival was 46.4 months for single-transplanted patients, and 63.7 months for double-transplanted patients respectively.These results are in accordance with other published outcome analyses [5,[24][25][26].In general, only patients who do not achieve at least a VGPR after the first ASCT might benefit from a double ASCT [5][6].Following this assumption, chiefly patients with an insufficient response proceeded to a second ASCT at our centre.We showed that patient survival correlates with the response achieved after ASCT.Patients who achieved only a PR or less than PR had a significantly shorter OS compared to patients achieving a CR.On the other hand, CR did not result in a significantly longer OS compared to the patient subgroup achieving nCR or VGPR.This result is remarkable considering that our patient number was rather small and usually only studies with large patient numbers were able to link survival with response status achieved after treatment.A similar association of patient survival with CR after ASCT was reported recently in a subgroup analysis of more than 750 patients treated within the GEM2000 protocol, where the quality of response to induction therapy had indeed a positive impact on the quality of the response to ASCT, but without impacting on survival as impress-

Original article
Swiss Med Wkly.2011;141:w13203 Swiss Medical Weekly • PDF of the online version • www.smw.chively as the post-transplant response status [27].The superiority of VGPR compared with PR after ASCT was demonstrated by an analysis of the IFM-99 trials [28].We also found a significant impact of CR on patient survival, but we cannot confirm this observation for patients with nCR or VGPR after ASCT, most probably because of small Myeloma stage (ISS) ASCT, autologous stem cell transplantation; BMI, body mass index; ASCT, autologous stem cell transplantation; DS, Durie-Salmon; ISS, international staging system; VAD, vincristine, adriamycin, dexamethasone; M2, cyclophosphamide, carmustine, melphalan, prednisone; Dex, dexamethasone; Thal, thalidomide; Vel, bortezomib (velcade©).

Original article
Swiss Med Wkly.2011;141:w13203 Swiss Medical Weekly • PDF of the online version • www.smw.chpatient numbers.However, CR was not superior to nCR or VGPR in our patient cohort, and in our view this finding supports the general recommendation to omit a second ASCT in patients who achieve at least a VGPR after first ASCT.
We transplanted a very heterogeneous patient cohort with MM at our transplantation centre.Notably, a few selected high-risk patients with co-morbidities such as renal insufficiency or light chain amyloidosis were also treated and a substantial subset of patients (16%) were aged over 65 years.Patients were considered transplantable if they were biologically fit and judged able to tolerate high dose therapy by the attending physician and our regular stem cell transplantation board.The dose of melphalan was individually adjusted according to physical status and organ function.This heterogeneity of the patient collective is reflected in the trend to prolonged EFS and OS in patients receiving standard dose melphalan for conditioning compared to patients with presence of risk factors, whose melphalan dose was reduced.One interesting finding of this analysis is that elderly patients did not have, in general, a worse outcome than younger patients.The fact that EFS was positively influenced by a higher melphalan dose may be partially explained by a simple dose-efficacy correlation, as higher drug doses may achieve a better treatment result [29].Furthermore, we observed the well known association of initial ISS stage with patient survival.Our findings are in accordance with other reports showing similar data [3,[30][31][32][33].In contrast, the number of previous chemotherapy regimens before ASCT did not affect patient outcome.This finding may be explained by the fact that the majority of the patients received only one regimen for induction,

Figure 1 Kaplan-
Figure 1Kaplan-Meier analysis of the event free survival and the overall survival of all patients.Median EFS was 23.1 months (95% CI: 19.4-28.4) and median OS was 49.8 months (95% CI: 43.7 -not reached).ASCT, autologous stem cell transplantation.

Figure 2
Figure 2Impact of the response to ASCT treatment on patients' outcome.(A) Event free survival, and (B) Overall survival.ASCT, autologous stem cell transplantation; CR, complete remission; nCR, near complete remission; VGPR, very good partial remission; PR, partial remission.

Figure 3
Figure 3 Impact of ISS myeloma stage (A, B), patient age (C, D), and melphalan dose (E, F) on patients' outcome.(A, C, E) Event free survival, and (B, D, F) Overall survival.ISS, international staging system; Mel, melphalan.

Table 3 :
Response rates before and after ASCT.and that most patients with insufficient response could be "rescued" by a second regimen, either by using combination chemotherapy (M2 protocol) initially or bortezomib and IMIDs during the last years of the observation period.Finally, we did not find a statistically significant impact of time between diagnosis and first ASCT on overall survival.The main limitation of this analysis is the limited patient number.Additionally, patients were treated outside the framework of a controlled clinical trial and differed considerably regarding number and duration of preceding treatment regimens, induction treatment, and risk factor profile.The data presented here represent a "real-life" transplantation experience, a scenario that is applicable to the majority of patients.These results are free from possible selection bias created by entrance criteria of clinical studies, and the conclusions reached by an analysis such as ours have their impact on daily practice and on communication with patients.The main goal of our efforts remains the continuous improvement in patient care we aim to achieve by monitoring and analysing treatment outcome.In conclusion, ASCT is a very safe and effective treatment option for eligible patients with MM.Treatment-related mortality is very low.Older patients with good performance status also benefit from an ASCT.Achievement of a CR before ASCT is a predictor for better EFS, and CR status after ASCT remains the best predictor for a prolonged OS and may spare the patients multiple ASCTs.Panagiotis Samaras and Marcel Blickenstorfer contributed equally to this work.Frank Stenner-Liewen and Christoph Renner share senior-authorship Christoph Renner, MD, Department of Oncology, University Hospital Zürich, Rämistrasse 100, CH-8091 Zürich, Switzerland, christoph.renner@usz.ch ASCT, autologous stem cell transplantation; CR, complete remission; VGPR, very good partial remission; nCR, near complete remission; PR, partial remission.Swiss Medical Weekly • PDF of the online version • www.smw.chmainly VAD, Correspondence:

Table 4 :
Prognostic factors for survival by univariate analysis.