Published Ahead of Print on August 18, 2014 as 10.1200/JCO.2013.53.

4537
The latest version is at http://jco.ascopubs.org/cgi/doi/10.1200/JCO.2013.53.4537

JOURNAL OF CLINICAL ONCOLOGY O R I G I N A L R E P O R T

Vitamin D Deficiency Impairs Rituximab-Mediated Cellular

Jörg Thomas Bittenbring, Frank
Neumann, Marina Achenbach, Jörg
Reichrath, Jürgen Geisel, Evi Regitz,
Gerhard Held, and Michael Pfreund-
schuh, Klinik für Innere Medizin I,
Universitätsklinikum des Saarlandes,
Homburg; and Bettina Altmann and
Marita Ziepert, University Leipzig,
Leipzig, Germany.
Published online ahead of print at
www.jco.org on August 18, 2014.
Supported by a grant from Deutsche
Krebshilfe (a charity
organization).
Terms in blue are defined in the glos-
sary, found at the end of this article
and online at www.jco.org.
Authors’ disclosures of potential con-
flicts of interest and author contribu-
tions are found at the end of this
article.
Corresponding author: Michael Pfreund-
schuh, MD, Klinik für Innere Medizin I,
Universitätsklinikum des Saarlandes,
D-66421 Homburg (Saar) Germany;
e-mail: michael.pfreundschuh@uks.eu.
© 2014 by American Society of Clinical
Oncology
0732-183X/14/3299-1/$20.00
DOI: 10.1200/JCO.2013.53.4537
Cytotoxicity and Outcome of Patients With Diffuse Large
B-Cell Lymphoma Treated With but Not Without Rituximab
Jörg Thomas Bittenbring, Frank Neumann, Bettina Altmann, Marina Achenbach, Jörg Reichrath,
Marita Ziepert, Jürgen Geisel, Evi Regitz, Gerhard Held, and Michael Pfreundschuh
A B S T R A C T
Purpose
To investigate the impact and mechanisms of vitamin D deficiency (VDD) on the outcome of
elderly patients with diffuse large B-cell lymphoma (DLBCL).
Patients and Methods
Three hundred fifty-nine pretreatment 25-hydroxyvitamin D3 (25[OH]D3) serum levels from the
RICOVER-60 study (Six Versus Eight Cycles of Biweekly CHOP-14 With or Without Rituximab in
Elderly Patients With Aggressive CD20⫹ B-Cell Lymphomas) and 63 from the RICOVER-noRTh
study (an amendment to the RICOVER-60 study in which patients received six cycles of
cyclophosphamide, doxorubicin, vincristine, and prednisone administered at an interval of 2 weeks
plus two cycles of rituximab [R-CHOP-14], but without radiotherapy) were determined by
chemoluminescent immunoassay. Rituximab-mediated cellular cytotoxicity (RMCC) was assessed
by lactate dehydrogenase release assay of CD20⫹ Daudi cells.
Results
RICOVER-60 patients with VDD (ⱕ 8 ng/mL) and vitamin D levels more than 8 ng/mL treated with
rituximab had 3-year event-free survival (EFS) of 59% and 79% and 3-year overall survival (OS) of
70% and 82%, respectively. These differences were significant in a multivariable analysis
adjusting for International Prognostic Index risk factors with a hazard ratio (HR) of 2.1 (P ⫽ .008)
for EFS and 1.9 (P ⫽ .040) for OS. EFS was not significantly different in patients with vitamin D
levels ⱕ 8 or more than 8 ng/mL (HR, 1.2; P ⫽ .388) treated without rituximab. This was confirmed
in an independent validation set of 63 RICOVER-noRTh patients. RMCC increased significantly
(P ⬍ .001) in seven of seven individuals with VDD after substitution and normalization of their
vitamin D levels.
Conclusion
VDD is a risk factor for elderly patients with DLBCL treated with R-CHOP. That VDD impairs RMCC
and substitution improves RMCC strongly suggests that vitamin D substitution enhances
rituximab efficacy, which must be confirmed in appropriately designed prospective trials address-
ing VDD and substitution not only in DLBCL, but also in malignancies treated with other antibodies,
of which the major mechanism of action is antibody-dependent cellular cytotoxicity (eg, trastu-
zumab in breast cancer and cetuximab in colorectal cancer).

J Clin Oncol 32. © 2014 by American Society of Clinical Oncology

hydroxyvitamin D3 (25[OH]D3) levels decreased

INTRODUCTION
Vitamin D plays an important role in calcium ho-
meostasis and bone health,1 which forms the basis
for recommendations concerning daily oral intake,
sun exposure, and supplementation.2 Many other
nonclassical functions of vitamin D have been estab-
lished, the most important of which may be the
association of vitamin D deficiency (VDD) with all-
cause mortality.3 Moreover, vitamin D is important
in conquering mycobacterial infections,4 and a ret-
rospective analysis suggested that sufficient 25-
the rate of infections in the upper respiratory tract,5
although two prospective interventional trials failed
to reduce such infections.6,7
Retrospective analyses of patients with cancer
demonstrated a positive impact of 25(OH)D3 on
survival in breast cancer8 and chronic lymphocytic
leukemia.9 A retrospective study by Drake et al10
showed that patients with non-Hodgkin lympho-
mas (diffuse large B-cell lymphoma [DLBCL]as well
as T-cell lymphomas) and VDD had an inferior
event-free survival (EFS) and overall survival (OS).

© 2014 by American Society of Clinical Oncology 1

Downloaded from jco.ascopubs.org on March 28, 2016. For personal use only. No other uses without permission.
Copyright © 2014 American Society of Clinical Oncology. All rights reserved.
Copyright 2014 by American Society of Clinical Oncology
 Bittenbring et al
We therefore investigated the role of 25(OH)D3 in patients with
DLBCL treated within the randomized RICOVER-60 trial (Six Versus
Eight Cycles of Biweekly CHOP-14 With or Without Rituximab in
Elderly Patients With Aggressive CD20⫹ B-Cell Lymphomas), which
was performed by the German High-Grade Non-Hodgkin Lym-
phoma Study Group11 and validated the results in an independent
population study (RICOVER-noRTh; an amendment to the
RICOVER-60 study in which patients received six cycles of cyclophos-
phamide, doxorubicin, vincristine, and prednisone administered at an
interval of 2 weeks plus two cycles of rituximab [R-CHOP-14], but
without radiotherapy).12 Because these results pointed to an interfer-
ence with rituximab effector mechanisms, the impact of VDD on
rituximab-mediated cellular cytotoxicity (RMCC) was investigated.
PATIENTS AND METHODS
The RICOVER-60 trial for elderly patients (61 to 80 years of age) with un-
treated DLBCL was registered as NCT00052936 and was performed in accor-
dance with the Helsinki Declaration. The protocol was approved by the ethical
review committee of each participating center. All patients gave written in-
formed consent.11,12 The randomization phase of the RICOVER-60 trial was
stopped after a second planned interim analysis revealed that the predefined
stopping rules were fulfilled. By then, 612 patients had been randomly assigned
to receive cyclophosphamide, doxorubicin, vincristine, and prednisone ad-
ministered at an interval of 2 weeks (CHOP-14) only and 610 patients had
been randomly assigned to receive CHOP-14 plus eight applications of
rituximab, and an amendment was made. According to this amendment—
designated the RICOVER-noRTh study—patients with the same inclusion
and exclusion criteria received the best of the four treatment arms consisting of
six cycles of rituximab plus CHOP-14 (R-CHOP-14) plus two additional
cycles of rituximab but without radiotherapy (36 Gy) to bulky disease and sites
of extralymphatic involvement. In RICOVER-noRTh, 164 of 165 consecutive
patients were evaluable. Further details are available in the articles on
RICOVER-60 and RICOVER-noRTh.11,12
Serum samples were obtained by mail from participating centers in
Germany. Serum samples taken before the first day of the second chemother-
apy cycle were accepted as pretreatment samples. Blood was drawn from
patients after they provided informed written consent. Serum samples were
stored at ⫺80°C until they were analyzed. 25(OH)D3 levels were determined
by the commercially available Diasorin LIAISON chemiluminescent immu-
noassay (Saluggia, Italy) according to the manufacturer⬘s recommendations.
25(OH)D3 values are given in nanograms per milliliter.
Statistical Analysis
The primary end point, EFS, in the RICOVER-60 and RICOVER-
noRTh trials was defined as the time from random assignment or registration
to disease progression, start of salvage treatment, additional (unplanned)
treatment, relapse, or death as a result of any cause. Progression-free survival
(PFS) was defined as time from random assignment or registration to disease
progression, relapse, or death as a result of any cause. Patients with complete
response or unconfirmed complete response and additional treatment were
censored. OS was defined as the time from random assignment or registration
to death as a result of any cause. EFS, PFS, and OS were estimated by using
Kaplan-Meier methods. For comparison of patient characteristics with vita-
min D (ⱕ 8 v ⬎ 8 ng/mL), ␹2 and, if necessary, Fisher’s exact tests were used.
For comparison of the median age, Mann-Whitney U test was used.
Martingale residual analysis13 was performed to assess the cutoff point of
25(OH)D3 to be used in a Cox proportional hazard model. A Cox regression
model without covariates and another Cox regression model with covariates of
interest (ie, risk factors according to the International Prognostic Index
[IPI]14) were fitted. Then the functional shape to be used in the Cox regression
models was checked by smoothed Martingale residual plots. A cutoff point
should be considered if a remarkable knee of this curve can be shown. The
2 © 2014 by American Society of Clinical Oncology
Downloaded from jco.ascopubs.org on March 28, 2016. For personal use only. No other uses without permission.
Copyright © 2014 American Society of Clinical Oncology. All rights reserved.
Martingale residuals decreased with increasing vitamin D levels. The most
commonly observed pattern had two phases. The first phase was up to 8 to 9
ng/mL. In the second phase, the slope was smaller. In addition, we performed
Cox regression models with different cutoff points for 25(OH)D3 to compare
the hazard ratios (HRs). The results of both methods were similar, and 8
ng/mL was identified as the most appropriate cutoff point. The cutoff point
was optimized for data from the training sample (RICOVER-60). To confirm
this cutoff point, it was used in the independent validation sample of the
RICOVER-noRTh study.
With this cutoff point assessed, univariable outcome analyses were per-
formed, and 3-year rates with 95% CIs were determined. Proportional hazard
models were adjusted for the IPI factors (ie, age older than 60 years, lactate
dehydrogenase [LDH] ⬎ normal, Eastern Cooperative Oncology Group per-
formance score [ECOG PS] ⬎ 1, stages III to IV, and extralymphatic involve-
ment ⬎ 1). A preplanned subgroup analysis was done for patients treated
without and with rituximab. Furthermore, a proportional hazard model for
rituximab, vitamin D, and interaction term (rituximab ⫻ vitamin D) adjusted
for the IPI factors was used. HRs with 95% CIs and P values were calculated. All
analyses were done according to intention-to-treat. The significance level was
P ⫽ .05 (two-sided). Statistical analyses were done with IBM SPSS Statistics 20
software (SPSS, Chicago, IL).
RMCC
Natural killer (NK) cells were isolated from donor peripheral blood
mononuclear cells by magnetic CD56 beads (Miltenyi, Bergisch Gladbach,
Germany) and incubated overnight with interleukin-2 at 10 ng/mL (Miltenyi).
Thereafter NK cells were washed and resuspended in X-VIVO 15 medium
(Luna, Köln, Germany) supplemented with 2% human serum albumin.
Daudi cells (LGC Standards, Wesel, Germany) were obtained, washed twice,
and resuspended in X-VIVO 15 medium (2% human serum albumin) at 5,000
cells per 50 microliters. Cells were divided into five fractions. One fraction
remained untreated; rituximab was added at the given concentrations to the
other four fractions. NK cells (50 microliters per well) and Daudi cells (50
microliters per well) were co-incubated at an effector:target ratio of 3.5:1 and
7:1, respectively, for 5 hours in a 96-well plate (U-bottom; Nunc A/S, Roskilde,
Denmark). Thereafter, the plates were centrifuged, and 50 ␮L of the superna-
tant from each well was transferred to a flat-bottom plate. Cytotoxicity of the
NK cells was analyzed by photometrically determining the LDH release from
Daudi cells by using the Cytotoxicity Detection Kit PLUS (Roche Applied
Science, Mannheim, Germany) according to the manufacturer⬘s manual. Rel-
ative lysis rates (minus spontaneous lysis) were calculated in relation to the
maximum LDH release induced by Triton X.
RESULTS
25(OH)D3 Levels in Elderly Patients With DLBCL
The training cohort included 359 of 1,222 patients from the
RICOVER-60 trial. These 359 patients consisted of all the patients
from whom a prospectively collected serum sample was available for
the determination of the serum level of vitamin D (Data Supplement).
This subgroup was representative for the entire RICOVER-60 popu-
lation (Data Supplement). The median 25(OH)D3 serum level for
these 359 patients was 9.2 ng/mL with a range of less than 4.0 ng/mL
(not detectable) to 61.9 ng/mL. According to current guidelines, 193
patients (54%) were considered to be vitamin D deficient (⬍ 10
ng/mL), although 165 patients (46%) had a vitamin D insufficiency
(10 to 30 ng/mL), and only one patient had a normal vitamin D level
(⬎ 30 to 100 ng/mL).
Because cutoff levels for 25(OH)D3 are defined with reference to
bone metabolism, we used Martingale residuals as described in the
Statistical Analysis subsection of the Patients and Methods section to
JOURNAL OF CLINICAL ONCOLOGY
 Vitamin D Deficiency in DLBCL

determine an appropriate cutoff level for this study. For elderly pa-
tients with aggressive B-cell lymphomas, 8 ng/mL was shown to be the
best discriminator for survival parameters in the study population.
A 25(OH)D3 level of ⱕ 8 ng/mL was significantly more common
among female compared with male patients (female-to-male ratio,
2:1; P ⬍ .001), and a lower 25(OH)D3 level was associated with higher
IPI scores (IPI ⬎ 1: 79% v 65%; P ⬍ .001 for all IPI groups), increased
pretreatment LDH, more than one extranodal disease, impaired per-
formance status (ECOG PS ⬎ 1), bulky disease, and older age (older
than 70 years; see Data Supplement for patients’ characteristics).
25(OH)D3 Levels and Outcome
When treated with R-CHOP, patients with 25(OH)D3 serum
levels ⱕ 8 ng/mL had a 3-year EFS of 59% (95% CI, 48% to 69%)

A 1.0 > 8 ng/ml D3 B 1.0 > 8 ng/ml D3

≤ 8 ng/ml D3 ≤ 8 ng/ml D3
Event-Free Survival

Event-Free Survival
0.8 0.8
(proportion)

(proportion)
0.6 0.6

0.4 0.4

0.2 0.2
HR, 1.2; P = .388 HR, 2.1; P = .008

0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Time (months) Time (months)
No. at risk No. at risk
>8 105 76 55 34 26 10 4 0 0 >8 103 83 70 53 36 11 2 0 0
≤8 70 37 32 24 14 5 3 0 0 ≤8 81 55 47 38 30 9 0 0 0

C 1.0 > 8 ng/ml D3 D 1.0 > 8 ng/ml D3

≤ 8 ng/ml D3 ≤ 8 ng/ml D3
Survival (proportion)

Survival (proportion)

0.8 0.8
Progression-Free

Progression-Free

0.6 0.6

0.4 0.4

0.2 0.2
HR, 1.4; P = .172 HR, 1.8; P = .047

0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Time (months) Time (months)
No. at risk No. at risk
>8 105 82 59 37 29 11 5 0 0 >8 103 85 70 53 36 11 2 0 0
≤8 70 40 33 25 15 6 5 0 0 ≤8 81 62 52 42 31 10 0 0 0

E 1.0 > 8 ng/ml D3 F 1.0 > 8 ng/ml D3

≤ 8 ng/ml D3 ≤ 8 ng/ml D3

0.8 0.8
Overall Survival

Overall Survival
(proportion)

(proportion)

0.6 0.6

0.4 0.4

0.2 0.2
HR, 1.8; P = .025 HR, 1.9; P = .040

0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Time (months) Time (months)
No. at risk No. at risk
>8 105 89 76 48 37 14 7 0 0 >8 103 90 76 55 39 14 2 0 0
≤8 70 50 39 28 16 6 4 0 0 ≤8 81 65 56 46 33 10 0 0 0

Fig 1. Outcomes of patients from the training cohort of the RICOVER-60 trial (Six Versus Eight Cycles of Biweekly CHOP-14 With or Without Rituximab in Elderly
Patients With Aggressive CD20⫹ B-Cell Lymphomas) treated (A, C, and E) without rituximab and (B, D, and F) with rituximab. (A, B) Event-free, (C, D) progression-free,
and (E, F) overall survival. Gold line indicates patients with 25-hydroxyvitamin D3 (25[OH]D3 (D3) levels ⱕ 8 ng/mL (n ⫽ 151); blue line indicates patients with 25(OH)D3
levels more than 8 ng/mL (n ⫽ 208). Shown are the hazard ratios (HRs) and P values from the multivariable analysis adjusting for International Prognostic Index risk factors.

www.jco.org © 2014 by American Society of Clinical Oncology 3

Downloaded from jco.ascopubs.org on March 28, 2016. For personal use only. No other uses without permission.
Copyright © 2014 American Society of Clinical Oncology. All rights reserved.
 Bittenbring et al

Table 1. Multivariable Analysis of Patients From the RICOVER-60 Trial Treated With Rituximab
EFS PFS OS

Variable HR 95% CI P HR 95% CI P HR 95% CI P

Vitamin D ⱕ 8 v ⬎ 8 ng/mL 2.1 1.2 to 3.6 .008 1.8 1.0 to 3.2 .047 1.9 1.0 to 3.6 .040
LDH ⬎ normal 1.5 0.9 to 2.7 .139 1.9 1.0 to 3.4 .044 2.1 1.1 to 4.0 .026
ECOG PS ⬎ 1 0.8 0.4 to 1.7 .617 0.9 0.4 to 1.8 .694 0.9 0.4 to 2.0 .774
Extralymphatic involvement ⬎ 1 1.2 0.7 to 2.3 .487 1.3 0.7 to 2.4 .499 1.2 0.6 to 2.4 .650
Stage III to IV 1.8 0.9 to 3.4 .076 2.1 1.0 to 4.1 .037 1.8 0.9 to 3.8 .104

Abbreviations: ECOG PS, Eastern Cooperative Oncology Group performance status; EFS, event-free survival, HR, hazard ratio; LDH, lactate dehydrogenase; OS,
overall survival; PFS, progression-free survival; RICOVER-60 trial, Six Versus Eight Cycles of Biweekly CHOP-14 With or Without Rituximab in Elderly Patients With
Aggressive CD20⫹ B-Cell Lymphomas.

compared with 79% (95% CI, 71% to 87%) of patients with
25(OH)D3 serum levels more than 8 ng/mL; 3-year PFS was 64%
(95% CI, 54% to 75%) and 81% (95% CI, 73% to 89%), and 3-year OS
was 70% (95% CI, 60% to 80%) and 82% (95% CI, 74% to 90%; Fig 1).
These differences were significant in a multivariable analysis
adjusting for IPI risk factors with an HR of 2.1 (95% CI, 1.2 to
3.6; P ⫽ .008) for EFS, an HR of 1.8 (95% CI, 1.0 to 3.2; P ⫽
.047) for PFS, and an HR of 1.9 (95% CI, 1.0 to 3.6; P ⫽ .040) for
OS (Table 1).
In patients treated without rituximab with ⱕ 8 or more than 8
ng/mL 25(OH)D3 had a similar 3-year EFS (43% [95% CI, 31% to
55%] v 48% [95% CI, 38% to 58%]) and 3-year PFS (46% [95% CI,
33% to 58%] v 53% [95% CI, 43% to 63%]), but 3-year OS was better
in patients with a 25(OH)D3 serum level more than 8 ng/mL: 69%
(95% CI, 59% to 79%) versus 53% (95% CI, 41% to 66%), respec-
tively (Fig 1). The multivariable analysis was significant only with
respect to OS (HR, 1.8; 95% CI, 1.1 to 3.0; P ⫽ .025), but not with
respect to EFS (HR, 1.2; 95% CI, 0.8 to 1.8; P ⫽ .388) or PFS (HR, 1.4;
95% CI, 0.9 to 2.1; P ⫽ .172; Table 2). The improvements in 3-year
survival rates achieved with rituximab were smaller in patients with
25(OH)D3 serum levels ⱕ 8 ng/mL than in those with more than 8
ng/mL for EFS (16% v 31%; Fig 2) and PFS (18% v 28%), but not OS
(17% v 13%). We found a nonsignificant interaction term between
rituximab and 25(OH)D3 for EFS (HR, 0.6; 95% CI, 0.3 to 1.1; P ⫽
.087), for PFS (HR, 0.7; 95% CI, 0.3 to 1.3; P ⫽ .243), and for OS (HR,
0.9; 95% CI, 0.4 to 1.9; P ⫽ .713).
The validation cohort consisted of 63 of 164 elderly patients from
the RICOVER-noRTh study treated with six cycles of R-CHOP-14
plus two cycles of rituximab, but in contrast to the RICOVER-60 study
without radiotherapy to bulky or extranodal disease. The median
25(OH)D3 serum level of patients in the RICOVER-noRTh study was
12.8 ng/mL compared with 9.2 ng/mL in RICOVER-60 (P ⫽ .004).
Similar to the cohort from the RICOVER-60 study, patients in the
RICOVER-noRTh study with a lower 25(OH)D3 level were more
often female, were older, and had a trend for higher IPI scores
(Data Supplement).
That a serum cutoff level of 8 ng/mL 25(OH)D3 separated two
lymphoma populations treated with R-CHOP and a significantly
worse outcome was confirmed in the 63 patients of the validation
cohort. Patients with 25(OH)D3 serum levels more than 8 ng/mL had
a 3-year EFS of 65% (95% CI, 52% to 78%) and a PFS of 76% (95% CI,
65% to 87%). In patients with 25(OH)D3 serum levels ⱕ 8 ng/mL, the
observation time was 32 months. At this time point, the EFS rate was
only 11% (95% CI, 0% to 32%). The 3-year PFS was 33% (95% CI, 3%
to 64%), and 3-year OS was 85% (95% CI, 81% to 90%) and 33%
(95% CI, 3% to 64%), respectively (Fig 3). This was confirmed in a
multivariable analysis adjusting for IPI risk factors (Data Supplement)
for EFS (HR, 4.0; 95% CI, 1.6 to 10.2; P ⫽ .003), for PFS (HR, 2.6; 95%
CI, 1.1 to 7.4; P ⫽ .063), and for OS (HR, 4.1; 95% CI, 1.3 to 12.7; P ⫽
.014), respectively.
Because we had recently reported that elderly females had a
superior outcome compared with males when treated with ritux-
imab,15,16 an exploratory analysis was performed to investigate the
outcome of elderly males and females with vitamin D levels ⱕ 8
ng/mL and more than 8 ng/mL treated with R-CHOP-14 (Data
Supplement). The outcome of both males and females was affected
by vitamin D levels. Although the vitamin D effect appears to be
somewhat more pronounced in elderly women compared with
men, our results should be interpreted with caution because of the
small number of patients in these subgroups of a subgroup. The

Table 2. Multivariable Analysis of Patients From the RICOVER-60 Trial Treated Without Rituximab (n ⫽ 175)
EFS PFS OS

Variable HR 95% CI P HR 95% CI P HR 95% CI P

Vitamin D ⱕ 8 v ⬎ 8 ng/mL 1.2 0.8 to 1.8 .388 1.4 0.9 to 2.1 .172 1.8 1.1 to 3.0 .025
LDH ⬎ normal 1.1 0.7 to 1.8 .593 1.2 0.7 to 1.9 .551 1.5 0.9 to 2.6 .162
ECOG PS ⬎ 1 1.7 0.9 to 3.1 .075 1.9 1.0 to 3.4 .042 1.6 0.8 to 3.1 .182
Extralymphatic involvement ⬎ 1 2.0 1.1 to 3.5 .019 2.0 1.1 to 3.6 .015 1.9 1.0 to 3.8 .053
Stage III to IV 1.7 1.1 to 2.7 .028 1.9 1.2 to 3.1 .011 1.5 0.9 to 2.7 .144

Abbreviations: ECOG PS, Eastern Cooperative Oncology Group performance status; EFS, event-free survival, HR, hazard ratio; LDH, lactate dehydrogenase; OS,
overall survival; PFS, progression-free survival; RICOVER-60 trial, Six Versus Eight Cycles of Biweekly CHOP-14 With or Without Rituximab in Elderly Patients With
Aggressive CD20⫹ B-Cell Lymphomas.

4 © 2014 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

Downloaded from jco.ascopubs.org on March 28, 2016. For personal use only. No other uses without permission.
Copyright © 2014 American Society of Clinical Oncology. All rights reserved.
 Vitamin D Deficiency in DLBCL

A 1.0 B 1.0
Event-Free Survival

Event-Free Survival
0.8 0.8
(proportion)

(proportion)
0.6 0.6

0.4 0.4

0.2 0.2
HR, 0.6; P = .042 HR, 0.4; P = .001

0 10 20 30 40 50 60 70 80 0 10 20 30 40 50 60 70 80
Time (months) Time (months)
Fig 2. Improvement of event-free survival achieved by the addition of rituximab according to vitamin D levels. (A) Patients with vitamin D levels ⱕ 8 ng/mL;
blue line indicates patients treated with six to eight cycles of cyclophosphamide, doxorubicin, vincristine, and prednisone administered at an interval of 2 weeks
(CHOP-14) without rituximab (n ⫽ 70); gold line indicates patients treated with six to eight cycles of CHOP-14 plus eight applications of rituximab (n ⫽ 81). (B)
Patients with vitamin D levels more than 8 ng/mL. Blue line indicates patients treated with six to eight cycles of CHOP-14 (n ⫽ 105) without rituximab; gold line
indicates patients treated with six to eight cycles of CHOP-14 plus eight applications of rituximab (n ⫽ 103). The improvement of 3-year event-free survival was
16% for patients with vitamin D levels ⱕ 8 ng/mL and 31% for patients with vitamin D levels more than 8 ng/mL. In a multivariable analysis adjusting for the
International Prognostic Index risk factors, the hazard ratio (HR) for rituximab was 0.6 (P ⫽ .042) in patients with vitamin D levels ⱕ 8 ng/mL and 0.4 (P ⫽ .001)
in patients with vitamin D levels more than 8 ng/mL.

same should also apply to the observation that elderly females with
vitamin D levels more than 8 ng/mL had a surprising 3-year OS rate
of 94% (95% CI, 86% to 100%). In a multivariable analysis adjust-
ing for IPI risk factors and sex, the HR for vitamin D remained
unchanged (data not shown), demonstrating that the vitamin D
effect was independent of sex.
Impact of 25(OH)D3 Levels on RMCC
The differential effect of VDD in patients treated with and
without rituximab and the relevant interaction term between VDD
and rituximab suggested an interference of VDD with rituximab
effector mechanisms. Because ADCC (antibody-dependent cell-
mediated cytotoxicity) is believed to be the major mechanism of
rituximab action,17 we investigated the impact of VDD in eight
otherwise healthy patients with VDD (8.4 ⫾ 3.0 ng/mL) before and
after vitamin D substitution by using an LDH release assay with
rituximab as the antibody and the CD20⫹ Daudi cell line as the
target. One individual did not achieve normal vitamin D levels and
was excluded from the analysis. The remaining seven probands
achieved normal 25(OH)D3 levels after substitution (40.6 ⫾ 13.6
ng/mL) and all had a significantly increased RMCC with P ⬍ .05
above 0.001 ␮g/mL rituximab (Fig 4).
DISCUSSION
Providing pretreatment serum samples was not mandatory in ei-
ther the training cohort (RICOVER-60) or the validation cohort
(RICOVER-noRTh), and the fact that such sera were available from
only roughly 30% of all patients might represent a limitation of this
study; however, it should be emphasized that the patients with pre-
treatment sera were representative for the entire RICOVER-60 popu-
lation (Data Supplement). We accepted all samples taken before the
first day of cycle 2 as pretreatment samples, because vitamin D serum
half-life is more than 1 month.
VDD in our study population was as prevalent as in a compa-
rable group of patients in a German elderly care rehabilitation
facility,18 and more prevalent than in an average population
age 61 to 80 years old.19 The vitamin D level was higher in the
validation set (RICOVER-noRTh) than in the training set (12.8 v
9.2 ng/mL; P ⫽ .004). Besides the small number of patients in the
validation set, the higher age (71 v 68 years) and the higher preva-
lence of bulky disease (39% v 29%) associated with lower vitamin
D levels (Data Supplement) in the training set might explain this
difference, because age and bulky disease are the only clinical
parameters that were significantly different between the training
and the validation sets.12 Although Drake et al10 used a cutoff point
of 25 ng/mL in their study of VDD in a spectrum of different
lymphoma types, the optimal vitamin D cutoff point in our study
population was 8 ng/mL. Our study was restricted to patients older
than age 60 years who are known to have lower vitamin D serum
levels than younger patients. Aside from the fact that in contrast to
the United States, no 25(OH)D3 is added to milk in Germany, the
lower 25(OH)D3 levels measured might be a result of the chemi-
luminescent immunoassay used in this study, which reports sys-
tematically lower values than mass spectrometry which was used in
the US study.
Although low levels of 25(OH)D3 were associated with negative
IPI prognostic factors,14 the confirmation of VDD as an independent
risk factor for outcome of rituximab-treated patients in a multivari-
able analysis demonstrates that VDD contributes to the worse out-
come of these patients per se to a degree that the addition of rituximab
improved EFS and PFS with a considerably smaller margin in patients
with vitamin D levels ⱕ 8 ng/mL than in patients with vitamin D levels
more than 8 ng/mL in the RICOVER-60 trial (Fig 2). These differences
should not be overinterpreted, because they were observed in the
training cohort by using an optimized cut point for vitamin D levels;
nevertheless, the greater improvement achieved by rituximab in pa-
tients with vitamin D levels more than 8 ng/mL (3-year EFS, 31% v

www.jco.org © 2014 by American Society of Clinical Oncology 5

Downloaded from jco.ascopubs.org on March 28, 2016. For personal use only. No other uses without permission.
Copyright © 2014 American Society of Clinical Oncology. All rights reserved.
 Bittenbring et al

A 1.0 > 8 ng/ml D3 70

≤ 8 ng/ml D3
Event-Free Survival

0.8 60
(proportion)

Relative Lysis (%)

0.6 50

0.4 40

0.2 30
HR, 4.0; P = .003
20
0 10 20 30 40 50 60 70 80
Time (months) 10
No. at risk
>8 54 39 35 31 10 2 0 0 0
≤8 9 3 1 1 0 0 0 0 0
0
0.0 μg/mL 0.0001 μg/mL 0.001 μg/mL 0.01 μg/mL 0.1 μg/mL
P = .0013 P = .0001 P = .0027 P = .0031 P = .0001
B 1.0 > 8 ng/ml D3
≤ 8 ng/ml D3 Rituximab Concentration
Survival (proportion)

0.8
Progression-Free

Fig 4. Rituximab-mediated cellular cytotoxicity in individuals with vitamin D

deficiency before and after substitution. Gold bars indicate cytotoxicity
0.6 against CD20⫹ Daudi cells before substitution (vitamin D level, 8.4 ⫾ 3.0
ng/mL); blue bars indicate cytotoxicity after substitution (40.6 ⫾ 13.6 ng/mL).
Effector:target ratio, 7:1. The two left bars (without rituximab) represent
0.4
natural-killer cell activity.

0.2
HR, 2.6; P = .063
and is prospectively addressed in the ongoing OPTIMAL ⬎ 60 study
0 10 20 30 40 50 60 70 80 (OPTIMAL ⬎ 60, Improvement of Therapy of Elderly Patients With
Time (months) CD20⫹ DLBCL Using Rituximab Optimized and Liposomal
No. at risk Vincristine) of the German High-Grade Non-Hodgkin Lymphoma
>8 54 45 40 36 12 2 0 0 0
≤8 9 6 3 3 0 0 0 0 0 Study Group in elderly patients with DLBCL.
That VDD plays an independent prognostic role is further
C 1.0 > 8 ng/ml D3 supported by the seemingly paradoxical observation that patients
≤ 8 ng/ml D3
randomly assigned to CHOP without rituximab with a 25(OH)D3
0.8
serum level above 8 ng/mL had a benefit compared with vitamin
Overall Survival
(proportion)

D– deficient patients treated without rituximab with respect to OS,

0.6
but not EFS and PFS. This can be explained by the fact that all
0.4
patients for whom CHOP chemotherapy without rituximab failed
received rituximab as part of their salvage treatment and benefited
0.2 from this salvage by being able to exploit the potential of rituximab
HR, 4.1; P = .014 as a result of their sufficient vitamin D levels (Bittenbring et al,
manuscript in preparation). Although the interaction between
0 10 20 30 40 50 60 70 80
VDD and rituximab had an HR of 0.6 for EFS, it did not become
Time (months) significant (P ⫽ .087), and therefore a differential effect on patients
No. at risk
>8 54 49 46 42 14 3 0 0 0
treated with rituximab cannot be proven as statistically conclusive.
≤8 9 6 4 3 0 0 0 0 0 Nevertheless, although not reaching significance, this difference
suggested that VDD might interfere with rituximab effector mech-
Fig 3. Outcome of patients in the validation cohort of the RICOVER-noRTh trial
(an amendment to the RICOVER-60 study in which patients received six cycles anisms. Because ADCC is believed to be the major effector mech-
of rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisone admin- anism of rituximab,17 we tested the impact of VDD in otherwise
istered at an interval of 2 weeks plus two additional cycles of rituximab, but healthy controls by analyzing RMCC against the CD20⫹ Daudi cell
without radiotherapy), all treated with rituximab. (A) Event-free, (B) progression-
free, and (C) overall survival. Gold line indicates patients with 25-hydroxyvitamin line before and after vitamin D substitution, and indeed all seven
D3 (D3) levels ⱕ 8 ng/mL (n ⫽ 9), blue line indicates patients with D3 levels more probands who achieved normal levels of 25(OH)D3 (⬎ 30 ng/mL)
than 8 ng/mL (n ⫽ 54). Shown are the hazard ratios (HRs) and P values from the
multivariable analysis adjusting for International Prognostic Index risk factors.
had a highly significantly stronger ADCC (Fig 4), demonstrat-
ing for the first time (to the best of our knowledge) that ADCC
can be improved by intervention with vitamin D in vitamin
D– deficient individuals.
16%) together with the improvement in RMCC after vitamin D sub- The ADCC response to 25(OH)D3 normalization warrants a
stitution in vivo holds promise that interventional vitamin D might prospective study of vitamin D substitution, not only in patients with
improve the outcome of vitamin D– deficient patients with DLBCL DLBCL treated with rituximab but also in other treatment regimens

6 © 2014 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

Downloaded from jco.ascopubs.org on March 28, 2016. For personal use only. No other uses without permission.
Copyright © 2014 American Society of Clinical Oncology. All rights reserved.
 Vitamin D Deficiency in DLBCL

with monoclonal antibodies in which ADCC is a relevant effector
mechanism, such as trastuzumab in breast cancer20 and cetuximab in
colorectal and head and neck cancer.21 That interventional vitamin D
substitution failed in other clinical settings such as respiratory infec-
tions6,7 should not be an argument against such studies, because in
contrast to our study with rituximab and RDCC, no specific (and
putatively predictive) vitamin D target had been identified in
those studies.
AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS
OF INTEREST
Although all authors completed the disclosure declaration, the following
author(s) and/or an author’s immediate family member(s) indicated a
financial or other interest that is relevant to the subject matter under
consideration in this article. Certain relationships marked with a “U” are
those for which no compensation was received; those relationships marked
with a “C” were compensated. For a detailed description of the disclosure
categories, or for more information about ASCO’s conflict of interest policy,
please refer to the Author Disclosure Declaration and the Disclosures of
Potential Conflicts of Interest section in Information for Contributors.
Employment or Leadership Position: None Consultant or Advisory
Role: Michael Pfreundschuh, Roche (C), Celgene (C), Onyx
Pharmaceuticals (C), Merck (C), Pfizer (C), Boehringer Ingelheim (C)
Stock Ownership: None Honoraria: None Research Funding: Michael
Pfreundschuh, Roche, Amgen Expert Testimony: None Patents,
Royalties, and Licenses: None Other Remuneration: None
AUTHOR CONTRIBUTIONS
Conception and design: Jörg Thomas Bittenbring, Frank Neumann,
Bettina Altmann, Michael Pfreundschuh
Collection and assembly of data: Jörg Thomas Bittenbring, Marita
Ziepert, Jürgen Geisel, Evi Regitz, Gerhard Held, Michael Pfreundschuh
Data analysis and interpretation: Jörg Thomas Bittenbring, Bettina
Altmann, Marina Achenbach, Jörg Reichrath, Marita Ziepert, Jürgen
Geisel, Gerhard Held, Michael Pfreundschuh
Manuscript writing: All authors
Final approval of manuscript: All authors

REFERENCES
1. Holick MF: Vitamin D deficiency. N Engl
J Med 357:266-281, 2007
2. Holick MF, Binkley NC, Bischoff-Ferrari HA, et al:
Evaluation, treatment, and prevention of vitamin D defi-
ciency: An Endocrine Society clinical practice guideline.
J Clin Endocrinol Metab 96:1911-1930, 2011
3. Melamed ML, Michos ED, Post W, et al:
25-hydroxyvitamin D levels and the risk of mortality
in the general population. Arch Intern Med 168:
1629-1637, 2008
4. Liu PT, Stenger S, Li H, et al: Toll-like receptor
triggering of a vitamin D-mediated human antimicro-
bial response. Science 311:1770-1773, 2006
5. Sabetta JR, DePetrillo P, Cipriani RJ, et al:
Serum 25-hydroxyvitamin d and the incidence of
acute viral respiratory tract infections in healthy
adults. PLoS One 5:e11088, 2010
6. Manaseki-Holland S, Maroof Z, Bruce J, et al:
Effect on the incidence of pneumonia of vitamin D
supplementation by quarterly bolus dose to infants
in Kabul: A randomised controlled superiority trial.
Lancet 379:1419-1427, 2012
7. Murdoch DR, Slow S, Chambers ST, et al:
Effect of vitamin D3 supplementation on upper
respiratory tract infections in healthy adults: The
VIDARIS randomized controlled trial. JAMA 308:
1333-1339, 2012
8. Goodwin PJ, Ennis M, Pritchard KI, et al:
Prognostic effects of 25-hydroxyvitamin D levels in
early breast cancer. J Clin Oncol 27:3757-3763,
2009
9. Shanafelt TD, Drake MT, Maurer MJ, et al:
Vitamin D insufficiency and prognosis in chronic
lymphocytic leukemia. Blood 117:1492-1498, 2011
10. Drake MT, Maurer MJ, Link BK, et al: Vitamin
D insufficiency and prognosis in non-Hodgkin’s lym-
phoma. J Clin Oncol 28:4191-4198, 2010
11. Pfreundschuh M, Schubert J, Ziepert M, et al:
Six versus eight cycles of bi-weekly CHOP-14 with or
without rituximab in elderly patients with aggressive
CD20⫹ B-cell lymphomas: A randomised controlled
trial (RICOVER-60). Lancet Oncol 9:105-116, 2008
12. Held G, Murawski N, Ziepert M, et al: Role of
radiotherapy to bulky disease in elderly patients with
aggressive B-cell lymphoma. J Clin Oncol 32:1112-
1118, 2014
13. Klein JP, Moeschberger ML: Survival Analy-
sis: Techniques for Censored and Truncated Data.
New York, NY, Springer Verlag, 1997
14. [No authors listed]: A predictive model for
aggressive non-Hodgkin’s lymphoma: The Interna-
tional Non-Hodgkin’s Lymphoma Prognostic Factors
Project. N Engl J Med 329:987-994, 1993
15. Müller C, Murawski N, Wiesen MH, et al: The
role of sex and weight on rituximab clearance and
serum elimination half-life in elderly patients with
DLBCL. Blood 119:3276-3284, 2012
16. Pfreundschuh M, Müller C, Zeynalova S, et al:
Suboptimal dosing of rituximab in male and female
patients with DLBCL. Blood 123:640-646, 2014
17. Weiner GJ: Rituximab: Mechanism of action.
Semin Hematol 47:115-123, 2010
18. Schilling S: Epidemic vitamin D deficiency
among patients in an elderly care rehabilitation facil-
ity. Dtsch Arztebl Int 109:33-38, 2012
19. Hintzpeter B, Mensink GB, Thierfelder W, et
al: Vitamin D status and health correlates among
German adults. Eur J Clin Nutr 62:1079-1089,
2008
20. Arnould L, Gelly M, Penault-Llorca F, et al:
Trastuzumab-based treatment of HER2-positive breast
cancer: An antibody-dependent cellular cytotoxicity
mechanism? Br J Cancer 94:259-267, 2006
21. Kurai J, Chikumi H, Hashimoto K, et al:
Antibody-dependent cellular cytotoxicity mediated
by cetuximab against lung cancer cell lines. Clin
Cancer Res 13:1552-1561, 2007

■ ■ ■

GLOSSARY TERMS

ADCC (antibody-dependent cell-mediated cyto- rituximab: a monoclonal antibody therapy that is indicated for re-
toxicity): a mechanism of cell-mediated immunity whereby an lapsed or refractory low-grade or follicular, CD20⫹, B-cell non-
effector cell of the immune system actively lyses a target cell that Hodgkin lymphoma.
has been bound by specific antibodies.

www.jco.org © 2014 by American Society of Clinical Oncology 7

Downloaded from jco.ascopubs.org on March 28, 2016. For personal use only. No other uses without permission.
Copyright © 2014 American Society of Clinical Oncology. All rights reserved.
 Bittenbring et al

Acknowledgment
We thank the patients and their families and the participating centers of the RICOVER-60 and RICOVER-noRTh studies.

8 © 2014 by American Society of Clinical Oncology JOURNAL OF CLINICAL ONCOLOGY

Downloaded from jco.ascopubs.org on March 28, 2016. For personal use only. No other uses without permission.
Copyright © 2014 American Society of Clinical Oncology. All rights reserved.