`
`EXHIBIT 2009
`
`IPR2016-00111
`
` Cephalon Exhibit 2009
`
`Fresenius v. Cephalon
`
`
`
`
`
`b j h b j h
`
`research paper
`
`Bendamustine compared with chlorambucil in previously
`untreated patients with chronic lymphocytic leukaemia:
`updated results of a randomized phase III trial
`
`Summary
`
`The efficacy of bendamustine versus chlorambucil in a phase III trial of pre-
`viously untreated patients with Binet stage B/C chronic lymphocytic leukae-
`mia (CLL) was re-evaluated after a median observation time of 54 months
`in May 2010. Overall survival (OS) was analysed for the first time. At
`investigator-assessed complete response (CR) rate (21 0% vs
`follow-up,
`10 8%), median progression-free survival (21 2 vs 8 8 months; P < 0 0001;
`hazard ratio 2 83) and time to next
`treatment (31 7 vs 10 1 months;
`P < 0 0001) were improved for bendamustine over chlorambucil. OS was
`not different between groups for all patients or those 65 years, >65 years,
`responders and non-responders. However, patients with objective response
`or a CR experienced a significantly longer OS than non-responders or those
`without a CR. Significantly more patients on chlorambucil progressed to
`second/further lines of treatment compared with those on bendamustine
`(78 3% vs 63 6%; P = 0 004). The benefits of bendamustine over chloram-
`bucil were achieved without reducing quality of life. In conclusion, benda-
`mustine is significantly more effective than chlorambucil
`in previously
`untreated CLL patients, with the achievement of a CR or objective response
`appearing to prolong OS. Bendamustine should be considered as a
`preferred first-line option over chlorambucil for CLL patients ineligible for
`fludarabine, cyclophosphamide and rituximab.
`
`Keywords: bendamustine, chlorambucil, chronic lymphocytic leukaemia,
`complete response, overall survival.
`
`Wolfgang U. Knauf,1 Toshko Lissitchkov,2
`Ali Aldaoud,3 Anna M. Liberati,4 Javier
`Loscertales,5 Raoul Herbrecht,6 Gunnar
`Juliusson,7 Gerhard Postner,8 Liana
`Gercheva,9 Stefan Goranov,10 Martin
`Becker,11 Hans-Joerg Fricke,12 Francoise
`Huguet,13 Ilaria Del Giudice,14 Peter
`Klein,15 Karlheinz Merkle16 and Marco
`Montillo17
`1Onkologische Gemeinschaftspraxis, Frankfurt,
`Germany, 2Haematology and Transfusion Medi-
`cine, National Haematological Centre, Sofia,
`Bulgaria, 3Praxis fu¨r Haematologie and Onkolo-
`gie, Leipzig, Germany, 4Azienda Ospidaliera
`Santa Maria di Terni, Universita degli Studi di
`Perugia, Perugia, Italy, 5Haematology, Hospital
`Universitario de la Princesa, Madrid, Spain,
`6Oncology and Haematology, Hopital de Haut-
`epierre, Strasbourg Cedex, France, 7Department
`of Haematology, Lund Stem Cell Centre, Lund,
`Sweden, 8Franz-Josef-Spital, LBI-ACR and ACR-
`ITR Vienna, Vienna, Austria, 9University Hospi-
`tal for Active Treatment, Varna, Bulgaria,
`10University Hospital for Active Treatment
`“St. George, Plovdiv, Bulgaria, 11Onkologische
`Schwerpunktpraxis, Porta Westphalika,
`Germany, 12Haematology, University Hospital,
`Jena, Germany, 13Department of Haematology,
`Hopital de Purpan, Toulouse, France, 14Division
`of Haematology, Department of Cellular Biotech-
`
`nologies and Haematology, Sapienza University,
`Rome, Italy, 15Dsh Statistical Services, Rohrbach,
`Germany, 16Oncology Consulting, Miesbach,
`Germany and 17Ematologia e centro trapianti
`midollo osseo, Ospedale Niguarda Ca’Granda,
`
`Milano, Italy
`
`Received 27 March 2012; accepted for
`
`publication 15 June 2012
`
`Correspondence: Professor Dr Wolfgang U.
`
`Knauf, Onkologische Gemeinschaftspraxis, Im
`
`Pruefling 17-19, 60389, Frankfurt, Germany.
`
`E-mail: Wolfgang.Knauf@telemed.de
`
`ª 2012 Blackwell Publishing Ltd
`British Journal of Haematology, 2012, 159, 67–77
`
`
`
`British Journal of Haematology British Journal of Haematology
`
`First published online 4 August 2012
`doi:10.1111/bjh.12000
`
`CEPHALON, INC. -- EXHIBIT 2009 0001
`
`
`
`W. U. Knauf et al
`
`Chronic lymphocytic leukaemia (CLL) is the most preva-
`lent adult leukaemia in the Western hemisphere. It is pre-
`dominantly a disease of the elderly with a median age at
`diagnosis of 72 years according to Surveillance Epidemiol-
`ogy and End Results (SEER) cancer statistics for 2004–
`2008 (NCI, 2011). Approximately 70% of individuals newly
`diagnosed with CLL are 65 years of age, with 42 5%
`being 75 years or older. Consistent with their advanced
`age, the majority of individuals with CLL have comorbidi-
`ties. In a study of 1,195 individuals with newly diagnosed
`CLL, 89% had 1 comorbidity and 46% had 1 major
`comorbidity (Thurmes et al, 2008).
`The combination of fludarabine with cyclophosphamide
`and rituximab (FCR) is the current recommended standard
`first-line regimen for the treatment of CLL (Eichhorst et al,
`2010). In a phase III trial (CLL8) conducted by the German
`CLL Study Group (GCLLSG), FCR was associated with a sig-
`nificantly higher complete response (CR) rate, median pro-
`gression-free survival (PFS) and overall survival (OS) rate
`than fludarabine plus cyclophosphamide (FC) (Hallek et al,
`2010). However, due to its toxicity, FCR is only considered
`suitable for a minority of ‘fit’ CLL patients without signifi-
`cant comorbidities (Eichhorst et al, 2010; NCCN, 2011). In
`the CLL8 trial, these eligible patients were defined as having
`a Cumulative Illness Rating Scale (CIRS) score 6 (Fortin
`et al, 2005; Hallek et al, 2010).
`Improved first-line treatment options are required for the
`majority of patients with CLL who are ineligible for FCR.
`The alkylating agent chlorambucil has traditionally been the
`first-line treatment of choice for elderly, comorbid or frail
`patients with CLL (Eichhorst et al, 2010). Chlorambucil
`demonstrated similar efficacy in terms of PFS and OS, and
`significantly reduced haematological toxicity by comparison
`with fludarabine alone in a recent phase III trial in treat-
`ment-naive CLL patients aged 65–80 years (Eichhorst et al,
`2009). On the basis of this trial, chlorambucil has become a
`standard of care for patients not fit enough for fludarabine-
`based regimens. However, chlorambucil
`treatment
`is also
`associated with a low CR rate in first-line CLL – 0% in this
`trial versus 7% with fludarabine (Eichhorst et al, 2009). This
`is important because higher CR rates may be associated with
`prolonged PFS (and perhaps OS).
`Bendamustine is a chemotherapeutic agent with structural
`similarities to alkylating agents and purine analogues. How-
`ever, bendamustine has a distinct mechanism of action,
`which includes the induction of TP53-dependent apoptosis,
`the base excision DNA-repair pathway, and TP53/apoptosis-
`independent mitotic catastrophe (Leoni et al, 2008; Dennie
`& Kolesar, 2009).
`In addition, bendamustine is effective
`against lymphoma cells that are resistant to structurally simi-
`lar chemotherapies like cyclophosphamide, at therapeutically
`relevant concentrations (Leoni et al, 2008).
`the European
`Bendamustine has been approved by
`Medicines Agency (EMA) for the first-line treatment of
`patients with CLL (Binet stage B/C) for whom fludarabine
`
`68
`
`combination chemotherapy is not appropriate, and is cur-
`rently licensed in a number of European countries, includ-
`ing Germany and the UK. The approval of bendamustine
`was based on the results of a randomized phase III trial in
`comparison with chlorambucil
`in previously untreated
`patients with CLL (Binet
`stage B or C)
`(Knauf et al,
`2009a). Bendamustine induced significantly higher objective
`response rates (ORR; 68% vs 31%; P < 0 0001) and CR
`rates (31% vs 2%) than chlorambucil (Knauf et al, 2009a,
`2010). Also, bendamustine demonstrated a significant med-
`ian PFS benefit over chlorambucil (21 6 vs 8 3 months;
`P < 0 0001)
`sustained in patients <65 or
`that was
` 65 years (Knauf et al, 2009a,b). Guidance issued by the
`UK’s National Institute for Health and Clinical Excellence
`(NICE) in February 2011 recommended bendamustine for
`use within the National Health Service (NHS) in England
`and Wales (NICE, 2011). The NICE evidence review group
`reported that overall, the economic model was of high qual-
`ity and contained no logical errors. Bendamustine had more
`quality-adjusted life years (QALYs) than chlorambucil (4 82
`QALYs vs 3 55 QALYs). The estimated cost per QALY
`gained for bendamustine according to a de novo Markov
`model was estimated to be GBP 11 960 per QALY gained,
`but was favourably revised following the NICE appraisal
`and was reported to be GBP 9 400 (NICE, 2011).
`Bendamustine, either alone or as combination therapy, is
`a recommended treatment option in European and American
`guidelines
`for CLL patients,
`including
`those
`eligible
`(<70 years and/or without
`significant comorbidities) and
`ineligible for FCR (comorbid, unfit, and/or 70 years)
`(Eichhorst et al, 2010; NCCN, 2011). Furthermore, benda-
`mustine-based regimens are the most commonly used first-
`line treatments for CLL in Germany, according to results
`from a community centre based patient registry of lymphatic
`neoplasias (Wolfgang U. Knauf, unpublished observations).
`The objective of this report, in fulfilment of an EMA post-
`licensing commitment, is to convey updated efficacy results
`from a randomized phase III trial of bendamustine versus
`chlorambucil
`in patients with previously untreated CLL
`(Knauf et al, 2009a), based on an updated 2009 analysis and
`a final follow-up in May 2010. We also report OS results for
`the first time, including comparisons in subsets of patients
`stratified by ORR and CR as well as evaluating the impact of
`treatment on quality of life (QoL).
`
`Methods
`
`As previously reported in the original published results of
`this phase III, multicentre, randomized, open-label parallel
`group trial, the study protocol was approved by the local
`ethics committees at each of the 45 participating centres in
`eight European countries (Knauf et al, 2009a). The study was
`also conducted in accordance with the International Confer-
`ence on Harmonization Good Clinical Practice guidelines
`and the Declaration of Helsinki.
`
`ª 2012 Blackwell Publishing Ltd
`British Journal of Haematology, 2012, 159, 67–77
`
`CEPHALON, INC. -- EXHIBIT 2009 0002
`
`
`
`Patients
`
`As described before,
`eligible patients were previously
`untreated, aged 75 years with a confirmed diagnosis of
`Binet stage B/C CLL, and requiring treatment (Knauf et al,
`2009a). Eligible patients also had a World Health Organiza-
`tion performance status of 0–2 and a life expectancy of at
`least 3 months. Written and informed consent was obtained
`from all patients prior to study inclusion.
`
`Study design
`
`Treatments. The full study design details have been previ-
`ously described (Knauf et al, 2009a). Briefly, patients were
`randomly assigned 1:1 to receive bendamustine or chloram-
`bucil, and stratified by centre and Binet stage. Bendamustine
`(Ribosepharm, Munich, Germany) was administered intra-
`venously over 30 min at a dose of 100 mg/m2/day on days
`1–2, every 4 weeks. Chlorambucil (GlaxoSmithKline, Uxbridge,
`UK) was administered orally at a dose of 0 8 mg/kg (Broca’s
`normal weight in kg: the body weight for the dose being the
`height of the patient in cm minus 100) on days 1 and 15 (or
`as divided doses on days 1–2 and 15–16 for patient comfort
`in some individual cases) every 4 weeks.
`
`Endpoints. The primary study endpoints were ORR and
`PFS. Secondary endpoints included OS (Knauf et al, 2009a).
`
`Follow-up analyses and statistics
`
`A follow-up analysis of this pivotal phase III trial was con-
`ducted in May 2010 on the intention-to-treat (ITT) popula-
`tion, which included all randomized patients. Each endpoint
`listed below was analysed at the 2010 follow-up. However,
`PFS data from an earlier updated analysis in 2009 (con-
`ducted 12 months after the originally published results) will
`also be reported. In contrast to the original trial results, the
`updated and follow-up analyses were investigator assessed
`and were not reviewed by a blinded independent committee
`for response assessment. All statistical analyses and summa-
`ries were generated using SAS version 9.2 software. P val-
`ues < 0 05 were considered statistically significant.
`
`Progression-free survival. PFS was defined as the time from
`randomization until the day of progression/death. Median
`PFS was determined in an update 12 months after the
`originally published trial results in 2009 and at the 2010 fol-
`low-up. In the 2010 follow up PFS was updated for patients
`without progressive disease in the 2009 assessment if progres-
`sion was documented at follow-up and no second-line treat-
`ment was given prior to the date of progression. Differences
`in median PFS between the two treatment groups were anal-
`ysed by log-rank test, stratified by Binet stage B or C. Hazard
`ratios (HRs) for treatment differences and associated 95%
`confidence intervals (CIs) were adjusted for Binet stage B/C
`
`ª 2012 Blackwell Publishing Ltd
`British Journal of Haematology, 2012, 159, 67–77
`
`Bendamustine versus Chlorambucil: Updated Phase III Analysis
`
`and based on a Cox regression (proportional hazard) model.
`Results were plotted as Kaplan–Meier curves.
`
`Last known disease status. All response assessments were
`conducted in accordance with National Cancer Institute
`Working Group (NCIWG) criteria (Cheson et al, 1996).
`Response categories included CR, partial response (PR), PR
`with nodular involvement (nPR), stable disease (SD), or pro-
`gressive disease (PD). The ORR was defined as the sum of
`the PR+nPR+CR rates. The last known status of disease after
`first-line therapy was compared between groups using the
`Cochrane-Mantel-Haenszel (CMH) test adjusting for Binet
`stage B/C.
`
`Overall survival. This was calculated from the time interval
`from the date of randomization to death, regardless of
`cause, for each patient for which data were available at the
`2010 follow-up. Differences in median OS between the ben-
`damustine and chlorambucil treatment groups were com-
`pared by a log-rank test adjusted for Binet stage. HRs for
`treatment differences and associated 95% CIs were adjusted
`for Binet stage and based on a Cox regression (proportional
`hazard) model. Results were plotted as Kaplan–Meier
`curves. The median OS was compared between treatment
`groups in several patient subsets, including those with Binet
`stage B, Binet stage C disease, aged > 65 years, 65 years,
`those with a response, and those without a response. In
`addition, the median OS was compared for all patients with
`a response (regardless of treatment) versus all patients with-
`out a response, and for all patients with a CR versus those
`without a CR.
`
`Time to next treatment. The time to next treatment (TTNT)
`was defined as the time from the date of termination of first-
`line treatment until the start date of a second-line treatment.
`Differences in the calculated median TTNT between first-line
`treatment groups were analysed using an extended Mantel-
`Haenszel test stratified by Binet stage B/C. HRs for treatment
`differences and associated 95% CIs were adjusted for Binet
`stage B/C and based on a Cox regression (proportional haz-
`ard) model. Results were plotted as Kaplan–Meier curves.
`
`Best response after second-line therapy. The best response
`after second-line therapy (CR, PR, SD or PD) was also deter-
`mined. Differences between treatment groups were analysed
`by the Mantel-Haenszel test (for best response) and Fisher’s
`exact test (for ORR).
`
`life. The QoL was analysed using the EORTC
`Quality of
`questionnaires QLQ C30 and QLQ-CLL25.
`
`Second or further lines of treatment. The types of second or
`further lines of treatment received by patients after first-line
`bendamustine or chlorambucil were recorded. Differences in
`the overall proportion of patients who received a second or
`
`69
`
`CEPHALON, INC. -- EXHIBIT 2009 0003
`
`
`
`W. U. Knauf et al
`
`further line of treatment were analysed by the CMH test,
`adjusted for Binet stage B/C.
`
`Results
`
`The original results of this randomized, phase III multicentre
`trial in previously untreated patients with Binet stage B/C
`CLL who received bendamustine (n = 162) or chlorambucil
`(n = 157) were first published in 2009 after a median obser-
`vation time of 35 months (Knauf et al, 2009a).
`This follow-up efficacy analysis was conducted in May
`2010 on the final ITT population (N = 319) after a med-
`ian observation time of 54 months (range: 0–90 months).
`Follow-up results
`for
`the primary endpoints
`(PFS and
`ORR) are provided, and OS results for the ITT population
`and defined patient subgroups are reported for the first
`time.
`
`Patient characteristics
`In total, 247 patients (bendamustine n = 131; chlorambucil
`n = 116) were alive at the end of the study and follow-up
`documentation was available for 244 of them (bendamustine
`n = 129; chlorambucil n = 115). It was previously reported
`that the baseline demographic characteristics of the benda-
`mustine and chlorambucil
`treatment groups were similar
`(Knauf et al, 2009a). The majority of patients were male
`(bendamustine 63%; chlorambucil 60 5%) and the mean age
`was 63 0 years in the bendamustine group and 63 6 years in
`the chlorambucil group. The majority of patients had Binet
`stage B CLL (bendamustine 71 6%; chlorambucil 70 7%),
`with the remainder having Binet stage C disease (Knauf et al,
`2009a).
`
`Progression-free survival
`
`Median PFS was re-assessed by the investigators in 2009 and
`at the 2010 follow-up (Fig 1). According to the 2009 assess-
`ment, the median PFS was significantly longer in the benda-
`mustine group than in the chlorambucil group (21 2 vs
`8 9 months; P < 0 0001). The chlorambucil/bendamustine
`HR, adjusted for Binet stage, was 3 30 (95% CI: 2 48, 4 41).
`The median PFS at the 2010 follow-up was also significantly
`longer in the bendamustine group by comparison with the
`chlorambucil group (21 2 vs 8 8 months; P < 0 0001). The
`chlorambucil/bendamustine HR, adjusted for Binet stage, was
`2 83 (95% CI: 2 16, 3 71).
`
`Last known status of disease after first-line therapy
`
`The last known status of disease after first-line therapy,
`including any further therapies, was assessed by the study
`investigators at the 2010 follow-up. It was reported for 236
`of 319 patients (74%; Table I). The CR rate was 21 0% with
`first-line bendamustine and 10 8% with first-line chlorambucil,
`
`70
`
`respectively. The PR rate was 13 6% with first-line benda-
`mustine and 19 1% with first-line chlorambucil. Overall best
`response (CR,PR, SD, PD) was statistically not significant
`between groups.
`
`Overall survival
`
`A total of 132 patients had died at the time of the 2010 fol-
`low-up. However, the date of death was unknown for 26
`patients (n = 15 bendamustine group; n = 11 chlorambucil
`group). These 26 patients were censored with the time from
`date of randomization until the date of the last contact upon
`which the patient was still documented to be alive. The med-
`ian OS had not yet been reached in the bendamustine group
`and was 78 8 months for patients in the chlorambucil group
`(Table II, Fig 2). Although the chlorambucil/bendamustine
`HR of 1 30 (95% CI: 0 89, 1 91) slightly favoured bendamus-
`tine, there was no statistically significant difference in median
`OS adjusted for Binet stage between groups in this trial at
`this stage (P = 0 1801).
`
`Overall survival in patients with Binet stage B or C. Median
`OS was similar between the two treatment groups when
`observing the entire treated population irrespective of Binet
`stage of disease. In both Binet stage subgroups of patients
`there was a numerical advantage for both Binet B (HR 1 28
`with 95% CI: 0 80, 2 04) and Binet stage C (HR 1 35 with
`95% CI: 0 68, 2 65) patients treated with Bendamustine com-
`pared to those treated with chlorambucil. There was not,
`however, a statistically significant difference between the
`bendamustine and chlorambucil groups at the 2010 follow-
`up (Table II, Fig 2).
`
`Overall survival in patients aged > 65 years. There was also
`no statistically significant difference between the bendamus-
`tine and chlorambucil groups
`in terms of median OS
`adjusted for Binet
`stage
`for
`the
`subset of patients
`aged > 65 years at the 2010 follow-up (Table II, Fig 2).
`Overall survival in patients aged 65 years. Similarly, there
`was no statistically significant difference in median OS
`adjusted for Binet stage between the two treatment groups
`for patients aged 65 years (Table II). However, according
`to the HR, patients aged 65 years who received benda-
`mustine had a 1 66-times greater probability of survival com-
`pared with those who received chlorambucil.
`
`in patients with an objective response.
`survival
`Overall
`Patients who achieved a response (i.e. CR or PR) with
`bendamustine were 1 63 times more likely to survive than
`those who had achieved a response with chlorambucil
`(Table II). There was no statistically significant difference in
`OS between the two treatment groups (P = 0 1642). Median
`OS had not been reached in either group at the time of the
`2010 follow-up.
`
`ª 2012 Blackwell Publishing Ltd
`British Journal of Haematology, 2012, 159, 67–77
`
`CEPHALON, INC. -- EXHIBIT 2009 0004
`
`
`
`Bendamustine versus Chlorambucil: Updated Phase III Analysis
`
`Bendamustine (B: n = 162)
`Chlorambucil (Clb: n = 157)
`Censored observations
`
`Median PFS: B 21·2 months; Clb 8·8 months
`Hazard ratio: 2·83 95% CI: 2·16 – 3·71
`P < 0·0001
`
`54
`48
`42
`36
`Time (months)
`
`60
`
`66
`
`72
`
`78
`
`84
`
`90
`
`0
`
`6
`
`12
`
`18
`
`24
`
`30
`
`1·0
`
`0·9
`
`0·8
`
`0·7
`
`0·6
`
`0·5
`
`0·4
`
`0·3
`
`0·2
`
`0·1
`
`0·0
`
`Survival distribution function
`
`N° left
`162
`B:
`157
`Clb:
`
`127
`81
`
`101
`37
`
`79
`16
`
`63
`7
`
`51
`4
`
`31
`3
`
`15
`3
`
`11
`2
`
`6
`2
`
`3
`2
`
`2
`1
`
`0
`0
`
`0
`0
`
`0
`0
`
`0
`0
`
`Fig 1. Progression-free survival at the 2010 follow up analysis. B, bendamustine; CI, confidence interval; Clb, chlorambucil; PFS, progression-free
`survival.
`
`in patients without an objective response.
`Overall survival
`There was no statistically significant difference in median OS
`for patients who did not achieve a first-line response with
`bendamustine or chlorambucil (i.e. those with SD or PD).
`
`in patients with or without an objective
`survival
`Overall
`response. When patients were stratified into responders and
`non-responders, regardless of their first-line treatment, median
`OS was significantly longer for patients with an objective
`response compared with those without an objective response
`(Fig 3A; median not reached versus 68 3 months; P < 0 0001).
`
`in patients with or without a complete
`survival
`Overall
`response. Similarly, median OS was significantly longer for
`all patients with a CR than for
`those without a CR
`75 9 months;
`(Fig 3B; median
`not
`reached
`versus
`P = 0 0018).
`
`Time to next treatment
`
`The median TTNT was significantly longer for patients who
`received bendamustine compared with those who received
`chlorambucil (Fig 4; 31 7 vs 10 1 months; P < 0 0001).
`
`Table I. Summary of best responses achieved at the 2010 follow-up, taking into account all lines of treatment.
`
`Last known status:
`
`Unknown, n (%)
`CR, n (%)
`PR, n (%)
`SD, n (%)
`PD, n (%)
`ORR, n (%)
`
`Initial treatment Bendamustine (n = 162)
`37 (22 8)
`34 (21 0)
`22 (13 6)
`20 (12 3)
`49 (30 2)
`56 (34 6)
`
`Initial treatment Chlorambucil (n = 157)
`46 (29 3)
`17 (10 8)
`30 (19 1)
`20 (12 7)
`44 (28 0)
`47 (29 9)
`
`Best response after
`second-line therapy*:
`
`Initial treatment Bendamustine (n = 103)
`32 (31 1)
`Unknown, n (%)
`11 (10 7)
`CR, n (%)
`27 (26 2)
`PR, n (%)
`22 (21 4)
`SD, n (%)
`11 (10 7)
`PD, n (%)
`38 (36 9)
`ORR, n (%)
`*Overall best response (CR, PR, SD, PD): P = 0 010 by Mantel-Haenszel test.
`†P = 0 106 by Fisher’s exact test.
`CR, complete response; ORR, overall response rate; PD, progressive disease; PR, partial response; SD, stable disease. ORR = CR + PR.
`
`Initial treatment Chlorambucil (n = 123)
`46 (37 4)
`18 (14 6)
`41 (33 3)
`11 (8 9)
`7 (5 7)
`59 (48 0)†
`
`ª 2012 Blackwell Publishing Ltd
`British Journal of Haematology, 2012, 159, 67–77
`
`71
`
`CEPHALON, INC. -- EXHIBIT 2009 0005
`
`
`
`W. U. Knauf et al
`
`Table II. Summary of overall survival results for all patients and in defined subgroups* (intention-to-treat cohort; 2010 follow-up).
`
`Outcome OS
`All patients
`
`Median OS
`n
`% survived (number left) at:
`12 months
`24 months
`36 months
`48 months
`60 months
`72 months
`84 months
`
`Median OS by subgroup:
`
`Patients with Binet B
`n
`Patients with Binet C
`n
`Patients > 65 years
`n
`Patients 65 years
`n
`Patients with ORR
`n
`Patients with SD
`n
`Patients with PD
`n
`
`Bendamustine
`
`Chlorambucil
`
`Months
`
`NR
`162
`% (number left):
`92 4 (145)
`89 9 (137)
`83 7 (120)
`76 4 (97)
`72 4 (62)
`68 1 (28)
`51 3 (4)
`
`Months
`76 2
`116
`NR
`46
`72 8
`65
`NR
`97
`NR
`110
`38 3
`19
`48 8
`15
`
`Months
`78 8
`157
`% (number left):
`92 0 (136)
`85 8 (125)
`80 3 (116)
`73 2 (93)
`63 4 (57)
`53 1 (25)
`42 6 (4)
`
`Months
`75 6
`111
`68 3
`46
`65 4
`79
`81 9
`78
`NR
`48
`68 3
`32
`65 4
`53
`
`Hazard ratio
`(95%CI: Clb, B)
`1 30 [0 89; 1 91]
`
`P-value
`0 1801
`
`1 28 (0 80; 2 04)
`1 35 (0 68; 2 65)
`0 92 (0 55, 1 55)
`1 66 (0 93, 2 96)
`1 63 (0 81, 3 26)
`0 49 (0 22, 1 08)
`0 68 (0 30, 1 55)
`
`0 3013
`0 3886
`0 7955
`0 0951
`0 1642
`0 0714
`0 3747
`
`*Intention-to-treat cohort analysed at 2010 follow-up; B, bendamustine; CI, confidence interval; Clb, chlorambucil; NR, not reached; ORR, objec-
`tive response rate; OS, overall survival; PD, progressive disease; SD, stable disease. ORR = complete response + nodular partial response + partial
`response.
`
`Favours bendamustine
`
`
`
`I (cid:9) • I (cid:9) •
`
`
`
`I I
`
`
`
`I (cid:9) • I (cid:9) •
`
`
`
`I I
`
`Patients with PD (Ben n = 15 Clb n = 53)
`
`Patients with SD (Ben n = 19 Clb n = 32)
`
`Patients with OR (Ben n = 110 Clb n = 48)
`
`
`
`I I
`
`
`
`• (cid:9)• (cid:9)
`
`
`
`I I
`
`Patients <=65 years (Ben n = 97 Clb n = 78)
`
`
`
`I (cid:9) • (cid:9)I (cid:9) • (cid:9)
`
`
`
`I I
`
`Patients >65 years (Ben n = 65 Clb n = 79)
`
`Binet C (Ben n = 46 Clb n = 46)
`
`
`
`I I
`
`
`
`• (cid:9)• (cid:9)
`
`
`
`I I
`
`Binet B (Ben n = 116 Clb n = 111)
`
`Best response after second-line therapy
`
`The best response after second-line therapy is shown for each
`treatment group in Table I. The ORR achieved with second-
`line therapy was similar in patients who had previously
`received bendamustine or chlorambucil first-line (36 9% vs
`48 0%; P = 0 106). However, the best response rate following
`second-line therapy was
`significantly different
`favouring
`patients who were treated with chlorambucil
`in first-line
`(P = 0 010).
`
`All patients (Ben n = 162 Clb n = 157)
`
`Quality of life
`
`0
`
`1
`
`2
`Hazard ratio
`
`3
`
`4
`
`Fig 2. Overall survival at the 2010 follow up analysis for all patients
`and in defined subgroups after initial therapy– hazard ratios and
`95% confidence intervals. Black circle, hazard ratio; horizontal line,
`95% confidence interval; Ben, bendamustine, Clb, chlorambucil, PD,
`progressive disease; SD, stable disease; OR, objective response.
`
`72
`
`Baseline scores in QoL parameters were similar in the benda-
`mustine and chlorambucil groups (Knauf et al, 2010). There
`was also no difference in QoL between treatment groups at
`the end of treatment (after a mean of 5 cycles administered),
`with similar scores for physical, social, emotional and cogni-
`tive functioning (Knauf et al, 2010). Global health status was
`also similar following bendamustine or chlorambucil treat-
`ment (Knauf et al, 2010).
`
`ª 2012 Blackwell Publishing Ltd
`British Journal of Haematology, 2012, 159, 67–77
`
`CEPHALON, INC. -- EXHIBIT 2009 0006
`
`(cid:9)
`(cid:9)
`(cid:9)
`(cid:9)
`(cid:9)
`(cid:9)
`(cid:9)
`(cid:9)
`
`
`Bendamustine versus Chlorambucil: Updated Phase III Analysis
`
`Complete response (CR: n = 53)
`Non complete response (non-CT: n = 266)
`Censored observations
`
`Median OS: CR not reached; Non CR 75·9 months
`Hazard ratio: 0·34 95% CI: 0·16 – 0·69
`P < 0·0018
`
`(B)
`
`1·0
`0·9
`0·8
`0·7
`0·6
`0·5
`0·4
`0·3
`0·2
`0·1
`0·0
`
`Survival distribution function
`
`Objective response (OR: n = 158)
`No response (NR: n = 161)
`Censored observations
`III III
`
`Median OS: OR not reached; NR 68·3 months
`Hazard ratio: 0·87 95% CI: 0·82 – 0·92
`P < 0·0001
`
`(A)
`
`1·0
`0·9
`0·8
`0·7
`0·6
`0·5
`0·4
`0·3
`0·2
`0·1
`0·0
`
`Survival distribution function
`
`0
`
`6 12 18 24 30 36 42 48 54 60 66 72 78 84 90
`Time (months)
`
`0
`
`6 12 18 24 30 36 42 48 54 60 66 72 78 84 90
`Time (months)
`
`Fig 3. Overall survival in patients ‘with’ or ‘without’ an objective response (A) and in patients ‘with’ or ‘without’ CR (B) at 2010 follow-up.
`CI, confidence interval; CR, complete response; NR, not reached; OR, overall response; OS, overall survival. The ‘Other’ group comprises nodular
`partial response + partial response + stable disease + progressive disease.
`
`Bendamustine (B: n = 162)
`Chlorambucil (Clb: n = 157)
`Censored observations
`
`Median TTNT: B 31·7 months; Clb 10·1 months
`Hazard ratio: 1·97 95% CI: 1·51 – 2·57
`P < 0·0001
`
`1·0
`
`0·9
`
`0·8
`
`0·7
`
`0·6
`
`0·5
`
`0·4
`
`0·3
`
`0·2
`
`0·1
`
`0·0
`
`Survival distribution function
`
`0
`
`6
`
`12
`
`18
`
`24
`
`30
`
`54
`48
`42
`36
`Time (months)
`
`N° left
`
`60
`
`66
`
`72
`
`78
`
`84
`
`90
`
`B:
`Clb:
`
`162
`157
`
`123
`88
`
`113
`63
`
`103
`47
`
`88
`41
`
`75
`32
`
`58
`27
`
`50
`24
`
`37
`16
`
`27
`15
`
`18
`10
`
`12
`7
`
`3
`5
`
`1
`3
`
`0
`1
`
`0
`0
`
`Fig 4. Time to next treatment, adjusted for Binet B and C stage, at 2010 follow-up. B, bendamustine; CI, confidence interval; Clb, chlorambucil;
`TTNT, time to next treatment.
`
`Most commonly utilized second- and third-line
`regimens
`A total of 103 (63 6%) patients in the bendamustine group
`and 123 (78 3%) of those in the chlorambucil group subse-
`quently received second-line or further lines of treatment
`(P = 0 004). The most
`common second-line
`regimens
`administered after first-line bendamustine were fludarabine
`(18 4%), bendamustine (16 5%) and chlorambucil (16 5%;
`Table III). The most common second-line regimens adminis-
`tered after first-line
`chlorambucil were
`bendamustine
`(22 8%), chlorambucil (15 4%) and fludarabine (10 6%).
`Thirty six patients in the bendamustine group and 53
`patients in the chlorambucil group subsequently received
`
`ª 2012 Blackwell Publishing Ltd
`British Journal of Haematology, 2012, 159, 67–77
`
`third-line or further lines of therapy (Table III). The most
`common regimens administered third-line in the benda-
`mustine group were cyclophosphamide, vincristine and pred-
`(CVP; 9 7%), fludarabine
`(9 7%), alemtuzumab
`nisone
`(4 9%), fludarabine plus cyclophosphamide (4 9%) and ben-
`damustine (4 9%). The most common regimens administered
` third-line in the chlorambucil group were fludarabine
`(13 8%), bendamustine
`(12 2%),
`and fludarabine plus
`cyclophosphamide (9 8%) (Table III).
`
`Discussion
`
`This investigator-assessed efficacy follow-up, conducted in
`May 2010, demonstrated that bendamustine had a statistically
`
`73
`
`CEPHALON, INC. -- EXHIBIT 2009 0007
`
`
`
`W. U. Knauf et al
`Table III. The most commonly administered second- and third-
`line regimens.*
`
`Bendamustine
`n = 103
`
`Chlorambucil
`n = 123
`
`103 (100)
`19 (18 4)
`17 (16 5)
`17 (16 5)
`9 (8 7)
`7 (6 8)
`6 (5 8)
`4 (3 9)
`3 (2 9)
`2 (1 9)
`2 (1 9)
`0 (0 0)
`0 (0 0)
`36 (35 0)
`10 (9 7)
`10 (9 7)
`5 (4 9)
`5 (4 9)
`5 (4 9)
`4 (3 9)
`3 (2 9)
`3 (2 9)
`3 (2 9)
`2 (1 9)
`1 (1 0)
`1 (1 0)
`1 (1 0)
`0 (0)
`0 (0)
`
`Second-line regimen
`Total receiving second-line
`therapy, n (%)
`Fludarabine, n (%)
`Bendamustine, n (%)
`Chlorambucil, n (%)
`FC, n (%)
`CVP, n (%)
`FCR, n (%)
`Prednisone, n (%)
`BR, n (%)
`CHOP, n (%)
`CHOP-R, n (%)
`FR, n (%)
`Rituximab, n (%)
` Third-line regimen
`Total receiving third-line
`therapy, n (%)
`CVP
`Fludarabine
`Bendamustine
`Alemtuzumab
`FC
`Chlorambucil
`FCP
`Prednisone
`Rituximab
`CHOP
`Cyclophosphamide
`CV
`Transplant
`Donor lymphocyte transfusion
`BR
`
`123 (100)
`13 (10 6)
`28 (22 8)
`19 (15 4)
`12 (9 8)
`5 (4 1)
`7 (5 7)
`6 (4 9)
`3 (2 4)
`3 (2 4)
`2 (1 6)
`2 (1 6)
`2 (1 6)
`53 (43 1)
`7 (5 7)
`17 (13 8)
`15 (12 2)
`6 (4 9)
`12 (9 8)
`10 (8 1)
`0 (0 0)
`2 (1 6)
`5 (4 1)
`6 (4 9)
`4 (3 3)
`2 (1 6)
`4 (3 3)
`3 (2 4)
`2 (1 6)
`*Only regimens received by > 1% of patients in at least one group
`are included.
`FC, fludarabine, cyclophosphamide; CVP, cyclophosphamide, vincris-
`tine, prednisone; FCR, fludarabine, cyclophosphamide, rituximab;
`BR, bendamustine, rituximab; CHOP, cyclophosphamide, doxorubi-
`cin, vincristine, predisolone; CHOP-R, CHOP + rituximab; FR,
`fludarabine, rituximab; FCP, fludarabine, cyclophosphamide, predni-
`sone; CV, cyclophosphamide, vincristine.
`
`significant > twofold median PFS and approximately three-
`fold median TTNT benefit over chlorambucil after a median
`follow-up of 54 months
`in previously untreated CLL
`patients. Furthermore, achieving a CR or an objective
`response (CR or nPR or PR), irrespective of the first-line
`treatment type, was associated with a statistically significant
`increase in median OS by comparison with not achieving a
`CR or an objective response. Notably, patients who achieved
`a CR (the majority of whom were in the bendamustine
`group) had a 66% reduced risk of dying during the follow-
`up period compared with those without a CR (HR 0 34
`
`74
`
`[95% CI: 0 16, 0 69]; P = 0 0018). However, there was no
`difference in OS between first-line treatment groups, for all
`ITT patients (bendamustine not yet reached versus chloram-
`bucil 78 8 months; P = 0 1801), or for subsets stratified by
`Binet stage (i.e. Binet B or Binet C), age or response (i.e.
`>65 years; 65 years; responders; or non-responders). The
`additional efficacy benefits of bendamustine over chlorambu-
`cil were achieved without a negative impact on QoL, which
`was simil