BRIEF REPORT

Association of PALB2 Messenger RNA Expression

with Platinum-Docetaxel Efficacy in Advanced
Non–Small Cell Lung Cancer
Niki Karachaliou, MD,a,b Jillian Wilhelmina Paulina Bracht, MSc,b
Manuel Fernandez Bruno, MD,a Ana Drozdowskyj, PhD,c Ana Gimenez Capitan, MSc,b
Teresa Moran, MD,d Enric Carcereny, MD,d Manuel Cobo, MD,e Manuel Domine, MD,f
Imane Chaib, PhD,g Jose Luis Ramirez, MSc,g,h Carlos Camps, MD,i
Mariano Provencio, MD,j Alain Vergnenegre, MD,k Guillermo Lopez-Vivanco, MD,l
Margarita Majem, MD,m Bartomeu Massuti, MD,n Rafael Rosell, MDb,g,h,*
a
Institute of Oncology Rosell, University Hospital Sagrat Cor, QuironSalud Group, Barcelona, Spain
b
Pangaea Oncology, Laboratory of Molecular Biology, Quiron-Dexeus University Institute, Barcelona, Spain
c
PIVOTAL SL, Madrid, Spain
d
Catalan Institute of Oncology, Medical Oncology Service, Hospital Germans Trias i Pujol, Badalona, Spain
e
Medical Oncology Service, Hospital Carlos Haya, Malaga, Spain
f
Medical Oncology Service, Fundacion Jimenez Diaz, Madrid, Spain
g
Institute for Health Science Research Germans Trias i Pujol (IGTP), Badalona, Spain
h
Medical Oncology Service, Hospital General de Valencia, Valencia, Spain
i
Medical Oncology Service, Hospital Puerta de Hierro, Madrid, Spain
j
Chest Department, CHU Limoges, Limoges, France
k
Chest Department, Chu de Limoges, Limoges, France
l
Medical Oncology Service, Hospital Sant Pau, Barcelona, Spain
m
Medical Oncology Service, Hospital General de Alicante, Alicante, Spain
n
Institute of Oncology Rosell, Quirón-Dexeus University Institute, Barcelona, Spain

Received 14 May 2018; revised 24 October 2018; accepted 31 October 2018

Available online - 22 November 2018

ABSTRACT Results: In 177 patients with NSCLC (who had a median

age of 62 years and included 140 men and 91 patients with
Introduction: Partner and localizer of BRCA2 (PALB2) is
adenocarcinoma), only high PALB2 mRNA expression was
essential for homologous recombination repair. We
predictive in the progression-free survival Cox regression
examined mRNA levels of DNA repair genes, including
analysis (hazard ratio ¼ 0.63, 95% confidence interval:
partner and localizer of BRCA2 gene (PALB2), ring finger
0.42–0.83, p ¼ 0.0080). PALB2 was also predictive of overall
protein 8 gene (RNF8), replication timing regulatory fac-
survival (hazard ratio ¼ 0.68, 95% confidence interval:
tor 1 gene (RIF1), ATM serine/threonine kinase gene
0.42–0.90, p ¼ 0.0266). Among the 158 patients evaluable
(ATM), and tumor protein p53 binding protein 1 gene
for response, high PALB2 mRNA expression was predictive
(53BP1) as predictive biomarkers for cisplatin-docetaxel
in the European phase III BRCA1, DNA repair associated
(BRCA1)–receptor-associated protein 80 (RAP80) expres- *Corresponding author.
sion customization (BREC) phase III clinical trial Drs. Karachaliou and Rosell are joint senior authors.
(ClinicalTrials.gov identifier NCT00617656). Disclosure: The authors declare no conflict of interest.
Methods: The study was a prespecified secondary objective Address for correspondence: Rafael Rosell, Cancer Biology and Preci-
sion Medicine Program, Catalan Institute of Oncology, Hospital Germans
of the BREC trial. We assessed mRNA levels of PALB2 and Trias i Pujol, Ctra Canyet, s/n 08916 Badalona, Spain. E-mail: rrosell@
four more DNA repair genes (RNF8, RIF1, ATM and 53BP1) iconcologia.net
as biomarkers in tissue from 177 patients with cisplatin- ª 2018 International Association for the Study of Lung Cancer.
Published by Elsevier Inc. All rights reserved.
docetaxel–treated NSCLC. We examined the relationship of
ISSN: 1556-0864
gene expression levels with progression-free survival,
https://doi.org/10.1016/j.jtho.2018.10.168
overall survival, and response.

Journal of Thoracic Oncology Vol. - No. -: ---

2 Karachaliou et al
of response to cisplatin-docetaxel. Specifically, an objective
response rate of 77% to cisplatin-docetaxel was observed
for patients with high PALB2 mRNA expression compared
with a rate of only 23 % for those with low PALB2 mRNA
expression (p ¼ 0.0448).
Conclusions: High PALB2 mRNA expression identified
patients with NSCLC who significantly benefited from
cisplatin-docetaxel chemotherapy in the European BREC
phase III clinical trial. The combination of chemotherapy
with immunotherapy will become the standard of care, and
a predictive marker of response to chemotherapy may
accurately guide therapeutic decision making.
 2018 International Association for the Study of Lung
Cancer. Published by Elsevier Inc. All rights reserved.
Keywords: PALB2; Non–small cell lung cancer; RNA
expression; Docetaxel
Introduction
Pharmacogenomic strategies for customizing
chemotherapy in NSCLC are not recommended for
treatment selection. We linked BRCA1, DNA repair
associated (BRCA1) and its upstream partner,
receptor-associated protein 80 (RAP80 [an alias for
UIMC1]), to platinum resistance and sensitivity to
antimicrotubule drugs.1 RAP80 is required for
BRCA1 accumulation to DNA break sites. 2 The
BRCA1-RAP80 expression customization (BREC)
studies were two industry-independent biomarker-
directed trials in European (Spanish Lung Cancer
Group in collaboration with the French Lung Cancer
Group) and Chinese patients with advanced-stage
NSCLC.3 In the phase III European study, patients
randomized to the control arm (cisplatin-docetaxel)
had significantly better median overall survival (OS)
than did those who received customized chemo-
therapy based on levels of BRAC1 and receptor-
associated protein 80 gene (RAP80 [an alias for
UIMC1]) mRNA. Accrual of the study was closed
prematurely.3 The BREC studies failed to show
benefit from the BRCA1/RAP80 pharmacogenetic
approach. However, other factors can influence the
predictive significance of BRCA1/RAP80.4
In RAP80-depleted cells, tumor protein p53
binding protein 1 (53BP1 [an alias of TP53BP1]) and
RAD51 recombinase (RAD51) assemble at double-
strand breaks, whereas in BRCA1-depleted cells,
RAP80 and 53BP1 take on this role.5 Replication
timing regulatory factor 1 (RIF1 [also known as
Rap1 interacting factor 1]) is the main factor
downstream of 53BP1 in the control of 5’ end
resection (Fig. 1). Its function is dependent on the
Journal of Thoracic Oncology Vol. - No. -
ATM serine/threonine kinase gene (ATM).6,7 Low
BRCA1 and high 53BP1-RIF1 could be a better
combination to predict resistance to platinum-based
chemotherapy. The inhibition of E3 ubiquitin-
protein ligase RING finger protein 8 (RNF8)
suppresses BRCA1 independent of homologous
recombination in 53BP1-depleted cells and RNF8
connects homologous recombination and nonhomol-
ogous end joining.5 Finally, partner and localizer of
BRCA2 (PALB2) is required for BRCA2, DNA repair
associated (BRCA2) localization in DNA damage
sites8 (see Fig. 1). Any defect in the BRCA1-PALB2-
BRCA2-RAD51 complex can result in defective as-
sembly of RAD51 foci and could be predictive of
response to DNA interstrand cross-linking agents.
Because BRCA1 regulates endogenous microtubule
dynamics and its loss confers resistance to taxanes,9
the BRCA1-PALB2-BRCA2 network may be a critical
determinant of the responsiveness to anti-
microtubule agents.
A prespecified secondary objective of the BREC clin-
ical trial was the study of potential genetic markers of
response or resistance to chemotherapy. In the present
study, we have investigated the association between
mRNA expression of partner and localizer of BRCA2 gene
(PALB2), ring finger protein 8 gene (RNF8), replication
timing regulatory factor 1 gene (RIF1), ATM, and tumor
protein p53 binding protein 1 gene (53BP1 [an alias of
TP53BP1]) and outcome to cisplatin-docetaxel in Euro-
pean patients with NSCLC in the control arm and group 2
of the experimental arm of the BREC study.
Methods
The present study includes all European patients
with NSCLC who were treated with cisplatin-docetaxel
in the BREC trial3 and had sufficient stored tumor
material. RNA was extracted from formalin-fixed
paraffin-embedded tissue samples, and mRNA expres-
sion of PALB2, RNF8, RIF1, ATM, and 53BP1 was
evaluated by quantitative real-time polymerase chain
reaction, as previously described.10 Gene expression
levels were grouped on the basis of tertiles (Q33, Q66)
and divided as high (>Q66 and Q33-Q66) or low
(<Q33). The primary end point of the study was to
examine the effects of gene mRNA expression levels on
survival and responses. Further details are provided in
the Supplementary Material. The BREC study is regis-
tered under ClinicalTrials.gov identifier NCT00617656.
Results
A total of 191 patients with NSCLC were treated
with cisplatin-docetaxel (142 in the control group
and 49 in group 2) in the European BREC phase III
--- 2018 PALB2 mRNA and Platinum-Docetaxel Efficacy 3

Figure 1. Our model. Double-strand breaks induce endogenous DNA repair mechanisms. Double-strand breaks can be
repaired by nonhomologous end joining or homologous recombination. Details are provided throughout the text. ATM, ATM
serine/threonine kinase; RAP80, receptor-associated protein 80 (an alias of ubiquitin interaction motif containing 1 [UIMC1]);
PALB2, partner and localizer of BRCA2; RAD51, RAD51 recombinase; 53BP1, an alias of tumor protein p53 binding protein 1
(TP53BP1); RNF8, ring finger protein 8; RIF1, replication timing regulatory factor 1.

study.3 Among them, 177 had material left for
further molecular analysis. The baseline character-
istics of the patients are shown in Supplementary
Table 1.
Among the five biomarkers explored, PALB2 and
RIF1 were significantly associated with progression-
free survival (PFS). At the time of the interim anal-
ysis of the BREC study,3 patients with high PALB2
mRNA expression had a median PFS of 5.6 months,
compared with 4.1 months for those with low PALB2
mRNA expression (hazard ratio [HR] ¼ 0.56, 95%
confidence interval [CI]: 0.38–0.80, p ¼ 0.0020)
(Fig. 2A). In patients with high and low RIF1 mRNA
expression, the median PFS times were 5.9 and 3.9
months, respectively (HR ¼ 0.60, 95% CI: 0.42–0.87,
p ¼ 0.0069). In the univariate analysis for PFS,
PALB2 and RIF1 emerged as significant factors for
better PFS when highly expressed (Fig. 2B). In the
multivariate analysis, only PALB2 remained a sig-
nificant biomarker to affect PFS (HR ¼ 0.63, 95% CI:
0.42–0.83, p ¼ 0.0018). The Cox regression model
included Eastern Cooperative Oncology Group per-
formance status, sex, age, and histologic type as
additional exploratory variables.
PALB2 was found to be significantly associated
with OS. Patients with high PALB2 mRNA expression
had a median OS of 13.2 months compared with 9.9
months for those with low PALB2 mRNA expression
(HR ¼ 0.64, 95% CI: 0.42–0.97, p ¼ 0.0394) (Fig. 3A).
The univariate analysis established the clinical impor-
tance of PALB2 mRNA expression for OS (HR ¼ 0.63,
95% CI: 0.42–0.83, p ¼ 0.0080) (Fig. 3B). In a
multivariate Cox regression model that was built with
PALB2, RIF1, Eastern Cooperative Oncology Group
performance status, sex, age, and histologic type as
exploratory variables, PALB2 mRNA expression
emerged as the only significant factor to affect OS
(HR ¼ 0.68, 95% CI: 0.42–0.90, p ¼ 0.0104).
Concerning response to treatment, among the 158
patients evaluable for response, cisplatin-docetaxel
induced an objective response rate (ORR) of 77% in
patients with high PALB2 mRNA expression compared
with 23% in those with low PALB2 mRNA expression
(p ¼ 0.0448). RIF1 was also significantly correlated with
ORR. ORR was significantly higher in patients with high
RIF1 mRNA expression (79%) than in patients with low
RIF1 mRNA expression (21%) (p ¼ 0.0018).
Finally, using MTT viability (3-(4,5-dimethylthiazol-
2-yl)-2,5-diphenyltetrazolium bromide) assays, we
determined the influence of PALB2 gene silencing on
NSCLC cell viability. We observed that after lentivirus
infection, the cell proliferation rate of the docetaxel-
treated PALB2 knockdown group was increased in
both EBC1 and H1993 cells compared with the rate in
the docetaxel-treated control group, (Fig. 4).
Discussion
PALB2 is recognized as an essential gene of suscep-
tibility to breast and pancreatic cancer. PALB2 loss-of-
function mutations are associated with sensitivity to
DNA-damaging agents and poly(ADP-ribose) polymerase
1 inhibitors.11 To the best of our knowledge, this is the
first study to explore PALB2 mRNA expression as a
4 Karachaliou et al Journal of Thoracic Oncology Vol. - No. -

Figure 2. Progression-free survival according to biomarker analyses. (A) Progression-free survival for high partner and
localizer of BRCA2 (PALB2) expressers and low PALB2 expressers treated with cisplatin-docetaxel. (B) Forest plots of hazard
ratios (HRs) for progression-free survival by each biomarker. 53BP1, an alias of tumor protein p53 binding protein
1 (TP53BP1); RNF8, ring finger protein 8; RIF1, replication timing regulatory factor 1; ATM, ATM serine/threonine kinase;
CI, confidence interval.

predictive pharmacogenomic marker in NSCLC. We
found that PALB2, when highly expressed, is predictive
of outcome in response to cisplatin-docetaxel. Among the
five DNA repair genes explored, RIF1 was also signifi-
cantly related to PFS and response to cisplatin-docetaxel,
but unlike PALB2, it did not remain significant in the PFS
Cox regression model.
PALB2 has functions that are different from those of
its main interactors, BRCA1 and BRCA2, including its
synergism with p53 to suppress breast tumor forma-
tion.12 Constitutive activation of nuclear factor kappa
light-chain enhancer of activated B cells was found in
PALB2-mutant mice after exposure to radiation.13
Although higher p53 induction was observed after ra-
diation, there were lower apoptosis levels than in wild-
type mice, highlighting a prosurvival and prooncogenic
role of nuclear factor kappa light-chain enhancer of
activated B cells in PALB2-mutant mice.13 On the other
hand, PALB2 interacts with Kelch-like ECH-associated
protein 1 (KEAP1), a negative regulator of the antioxi-
dant transcription factor nuclear factor erythroid 2-like
2 (NFE2L2 [also known by the alias NRF2]). Therefore,
PALB2 promotes NRF2 nuclear accumulation and func-
tion. In NSCLC, high levels of NRF2 protein are associ-
ated with poor outcome and resistance to therapy.14
Intriguingly, 20% of KRAS-mutant NSCLC carry Kelch-
--- 2018 PALB2 mRNA and Platinum-Docetaxel Efficacy 5

Figure 3. Overall survival according to biomarker analyses. (A) Overall survival for high partner and localizer of BRCA2
(PALB2) expressers and low PALB2 expressers treated with cisplatin-docetaxel. (B) Forest plots of hazard ratios (HRs) for
overall survival by each biomarker. 53BP1, an alias of tumor protein p53 binding protein 1 (TP53BP1); RNF8, ring finger
protein 8; RIF1, replication timing regulatory factor 1; ATM, ATM serine/threonine kinase; CI, confidence interval.

like ECH-associated protein 1 gene (KEAP1) mutations.15
Higher PALB2 mRNA expression has been associated
with poorer OS in patients with advanced breast can-
cer.16 In The Cancer Genome Atlas RNA-seq database no
significant prognostic significance is reported for PALB2
mRNA expression in patients with lung adenocarcinoma
or lung squamous cell carcinoma.17
Overall, in the BREC study, the BRCA1/RAP80 model
was not able to customize chemotherapy for patients
with NSCLC.3 In non–oncogene-addicted NSCLC, the
combination of platinum-based chemotherapy with
immunotherapy is emerging as a promising therapeutic
approach.18 Therefore, PALB2 mRNA expression as a
predictor of outcome in response to platinum-based
chemotherapy warrants further investigation.
The findings of this prespecified secondary analysis
of the BREC clinical trial are interesting. Caution should
be taken, however, considering that the cutoff points
for defining high and low mRNA expression were
derived internally from this data set. Although tissue
was available for 92% of the cisplatin-docetaxel–
treated patients and PALB2 mRNA expression was
successfully evaluated in almost 90% of them, the 45
cisplatin-gemcitabine–treated patients of experimental
group 1 and the 43 patients of experimental group 3
who were treated with docetaxel alone were not
included in this analysis.
This study suggests that high PALB2 mRNA
expression is related to a better outcome in response
to cisplatin-docetaxel in advanced patients with
6 Karachaliou et al Journal of Thoracic Oncology Vol. - No. -

Figure 4. Knockdown of partner and localizer of BRCA2 (PALB2) and cell proliferation of NSCLC cells. EBC1 (left) and H1993
(right) cells were transiently transfected with two different siRNAs against PALB2 (Hs_PALB2_2 [siPALB2-1] and Hs_PALB2_4
[siPALB2-2]), or negative control small interfering RNA (siRNA) (siCTRL) (15 pmol/well). Tweny-four hours later, the cells were
treated with serial dilutions of docetaxel. Cell viability was assessed by MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide) assay after 72 hours of treatment. Bars shown are representative of three independent ex-
periments. (Left down) Effects of negative control siRNA or PALB2 siRNA on PALB2 protein levels in EBC1 cells after 24 hours of
transfection. EBC1: *p ¼ 0.0346 and ***p ¼ 0.0004; H1993: *p ¼ 0.0391.ns, not significant.

NSCLC. Revalidation of the findings and refinements of
the mRNA expression cutoff points are required
before clinical application. Nowadays, because com-
binations of chemotherapy with immunotherapy will
become the standard of care, an enrichment strategy
with a chemotherapy-related biomarker may increase
the clinical benefit of such approaches.
Acknowledgments
Work in Dr Rosell’s laboratory is partially supported
by a grant from La Caixa Foundation, an Instituto de
Salud Carlos III grant (RESPONSE, PIE16/00011), a
Marie Skłodowska-Curie Innovative Training Net-
works European Grant (ELBA No 765492), and a
Spanish Association Against Cancer (AECC) grant
(PROYE18012ROSE). The Funders had no role in
design or conduct of the study; collection, manage-
ment, analysis, or interpretation of the data; prepa-
ration, review, or approval of the manuscript; or the
decision to submit the manuscript for publication. Drs.
Karachaliou and Rosell had full access to all of the
data in the study and take responsibility for the
integrity of the data and the accuracy of the data
analysis; they also take responsibility for the study
concept and design, drafting of the manuscript, and
study supervision. Drs. Karachaliou, Bracht, and Rosell
take responsibility for the laboratory experiments.
Drs. Karachaliou and Drozdowskyj take responsibility
for statistical analysis. Dr. Karachaliou, Dr. Rosell, Ms.
Bracht and Ms. Gimenez each take responsibility for
administrative, technical, and/or material support. All
of the authors take responsibility for acquisition,
analysis, or interpretation of data, as well as for crit-
ical revision of the manuscript for important intel-
lectual content.
Supplementary Data
Note: To access the supplementary material accompa-
nying this article, visit the online version of the Journal of
Thoracic Oncology at www.jto.org and at https://doi.
org/10.1016/j.jtho.2018.10.168.
References
1. Rosell R, Perez-Roca L, Sanchez JJ, et al. Customized
treatment in non-small-cell lung cancer based on EGFR
mutations and BRCA1 mRNA expression. PLoS One.
2009;4:e5133.
2. Bonanno L, Costa C, Majem M, Favaretto A, Rugge M,
Rossell R. The predictive value of BRCA1 and RAP80
mRNA expression in advanced non-small-cell lung cancer
patients treated with platinum-based chemotherapy.
Ann Oncol. 2013;24:1130–1132.
3. Moran T, Wei J, Cobo M, et al. Two biomarker-directed
randomized trials in European and Chinese patients
with nonsmall-cell lung cancer: the BRCA1-RAP80
Expression Customization (BREC) studies. Ann Oncol.
2014;25:2147–2155.
4. Karachaliou N, Rosell R, Carcereny E. How to personalize
chemotherapy in stage IV NSCLC? In: Peters S, Besse B,
eds. New Therapeutic Strategies in Lung Cancers. New
York, NY: Springer; 2015:119–129.
5. Nakada S, Yonamine RM, Matsuo K. RNF8 regulates as-
sembly of RAD51 at DNA double-strand breaks in the
--- 2018
absence of BRCA1 and 53BP1. Cancer Res. 2012;72:
4974–4983.
6. Zimmermann M, Lottersberger F, Buonomo SB, Sfeir A,
de Lange T. 53BP1 regulates DSB repair using Rif1 to
control 5’ end resection. Science. 2013;339:700–704.
7. Riabinska A, Daheim M, Herter-Sprie GS, et al. Thera-
peutic targeting of a robust non-oncogene addiction to
PRKDC in ATM-defective tumors. Sci Transl Med. 2013;5:
189ra178.
8. Sy SM, Huen MS, Chen J. PALB2 is an integral component
of the BRCA complex required for homologous recombi-
nation repair. Proc Natl Acad Sci U S A. 2009;106:
7155–7160.
9. Sung M, Giannakakou P. BRCA1 regulates microtubule
dynamics and taxane-induced apoptotic cell signaling.
Oncogene. 2014;33:1418–1428.
10. Rosell R, Molina MA, Costa C, et al. Pretreatment EGFR
T790M mutation and BRCA1 mRNA expression in
erlotinib-treated advanced non-small-cell lung cancer
patients with EGFR mutations. Clin Cancer Res.
2011;17:1160–1168.
11. Nepomuceno TC, De Gregoriis G, de Oliveira FMB,
Suarez-Kurtz G, Monteiro AN, Carvalho MA. The role of
PALB2 in the DNA damage response and cancer predis-
position. Int J Mol Sci. 2017;18:E1886.
PALB2 mRNA and Platinum-Docetaxel Efficacy 7
12. Bowman-Colin C, Xia B, Bunting S, et al. Palb2 synergizes
with Trp53 to suppress mammary tumor formation in a
model of inherited breast cancer. Proc Natl Acad Sci
U S A. 2013;110:8632–8637.
13. Mahdi AH, Huo Y, Tan Y, et al. Evidence of intertissue
differences in the DNA damage response and the
pro-oncogenic role of NF-kappaB in mice with disen-
gaged BRCA1-PALB2 interaction. Cancer Res. 2018;78:
3969–3981.
14. Singh A, Misra V, Thimmulappa RK, et al. Dysfunctional
KEAP1-NRF2 interaction in non-small-cell lung cancer.
PLoS Med. 2006;3:e420.
15. Romero R, Sayin VI, Davidson SM, et al. Keap1 loss pro-
motes Kras-driven lung cancer and results in dependence
on glutaminolysis. Nat Med. 2017;23:1362–1368.
16. Li J, Li M, Chen P, Ba Q. High expression of PALB2 pre-
dicts poor prognosis in patients with advanced breast
cancer. FEBS Open Bio. 2018;8:56–63.
17. The Human Protein Atlas. PALB2. https://www.
proteinatlas.org/ENSG00000083093-PALB2/pathology/
tissue/lungþcancer. Accessed September 10, 2018.
18. Gandhi L, Rodriguez-Abreu D, Gadgeel S, et al.
Pembrolizumab plus Chemotherapy in Metastatic
Non-Small-Cell Lung Cancer. N Engl J Med. 2018;378:
2078–2092.