Supplementary Information:
Treatment Plan
Mobilization of hematopoietic stem cells was performed in 32 patients after the first or the second
course of HyperCHidam and in 2 cases after low-dose cytarabine [0.8 gr/m2 every 12 hours for
three days]. Mobilized stem cells were not used in case of a suitable donor (n=12), PD (n=6), NRM
(n=1), physician decision (n=1).
In alloSCT, G-CSF (300 μg/day sc) was administered starting on day + 7 until PMN > 1000/μl for 2
consecutive days. In autoSCT, G-CSF (300 μg/day sc) was administered starting on day + 3 until
PMN > 1000/μl for 2 consecutive days.
Prophylaxis against herpes simplex, varicella zoster virus and against Pneumocystis Carini was
administered (800 mg acyclovir twice a day; 800/160 mg sulphametoxazole/trimethoprim once a
day on alternate days). CMV reactivation was monitored with antigen pp65 or with virus DNA
detection technique according to the methods used by each concerned Department. In case of
laboratory signs of reactivation, pre-emptive treatment was started (oral valgancyclovir or
intravenous gancyclovir).
Study design and Statistical Methods
Clin A sample size calculation: according to the Simon optimal two-stage study design (Simon R.
Optimal two-stage designs for phase II clinical trials. Controlled Clinical Trials 1989; 10: 1-10), the
study has a type I error rate of 5% if the complete clinical response probability is 20%, and 95%
power for concluding that the experimental treatment is promising if the response probability is
40%. Twenty-eight patients were to be accrued in the first stage, with 7 or more responses
required to continue on to the second stage of 34 additional patients. In the final analysis, at least
18 responses out of the total 62 patients were to be considered evidence of promising activity.
Clin B sample size calculation: according to the Simon optimal two-stage study design (Simon R.
Optimal two-stage designs for phase II clinical trials. Controlled Clinical Trials 1989; 10: 1-10), the
study has a type I error rate of 5% if the complete clinical response probability is 20% (reflecting
the experience with standard treatments), and 90% power for concluding that the experimental
treatment is promising if the response probability is 40%. Nineteen patients were to be accrued in
the first stage, with 5 or more responses required to continue on to the second stage of 35
additional patients. In the final analysis, at least 16 responses of the total 54 patients were to be
considered evidence of promising activity.
The CR probability was estimated as proportion of CR over the total number of intention-to-treat
patients; the corresponding 95% confidence limits were calculated using the exact method, i.e.
taking into account the binomial distribution of proportions.
1
DFS time was calculated as the interval from CR achievement to first disease relapse or death,
regardless of the cause, with censoring at the date of last follow-up assessment. PFS time was
calculated as the interval from beginning of treatment to disease progression or death, regardless
of the cause, with censoring at the date of last follow-up assessment. OS time was calculated as
the interval from beginning of treatment to death for all causes, with censoring at the date of last
follow-up assessment.
In the univariable Kaplan-Meier analyses of PFS and OS involving transplant (auto, allo), time was
calculated from the transplant date.
In the multivariable Cox model patient’s age was modelled as continuous variable by means of a
nonlinear transformation, ie 3-knots restricted cubic splines (RCSs), (Durrleman S, Simon R.
Flexible regression models with cubic splines. Stat Med. 1989;8(5):551 – 561). Concordantly, the
values used in the tables presenting the Cox model results were not cut-off values but exact
values, ie the quartiles of the variable distribution. RCS modelling has the advantage of (i) allowing
the age prognostic effect not to be the same in every part of the range; (ii) taking as reference for
hazard ratio estimation every value within the range of the variable distribution.
The analyses were performed using the software SAS (SAS Institute Inc., Cary, NC) and R (R
Development Core Team (2006). R: A language and environment for statistical computing. R
Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL: http://www.rproject.org/, last access Apr 2nd 2013).
2
Figure S1
A) Progression-free survival and B) Overall survival according to the hystotype (excluding
patients with enteropathy-associated T-cell lymphomas).
100
PTCLNOS
100
ALK-neg
75
OS(%)
PFS(%)
75
50
25
AILD
50
25
p=0.83
p=0.99
0
0
0
20
40
time (months)
60
80
0
20
40
60
80
time (months)
3
Figure S2
A) Progression-free survival by treatment modality calculated from date of transplantation; auto:
autoSCT; allo: alloSCT, B) overall survival by treatment modality calculated from date of
transplantation; auto: autoSCT; allo: alloSCT; patients not transplanted are not included in the
analysis.
1.0
A
0.6
0.4
alloSCT
0.2
Probability
0.8
autoSCT
0.0
p=0.92
0
6
12
18
24
Time (months)
B
0.2
0.4
0.6
alloSCT
p=0.10
0.0
Probability
0.8
1.0
autoSCT
0
6
12
18
24
Time (months)
4