Available online at www.sciencedirect.com
Cancer Letters 267 (2008) 133–164
www.elsevier.com/locate/canlet
Curcumin and cancer: An ‘‘old-age” disease
with an ‘‘age-old” solution
Preetha Anand, Chitra Sundaram, Sonia Jhurani, Ajaikumar B. Kunnumakkara,
Bharat B. Aggarwal *
Cytokine Research Laboratory, Department of Experimental Therapeutics, The University of Texas M.D. Anderson Cancer Center,
Houston, TX, USA
Received 11 March 2008; received in revised form 11 March 2008; accepted 12 March 2008
Abstract
Cancer is primarily a disease of old age, and that life style plays a major role in the development of most cancers is now
well recognized. While plant-based formulations have been used to treat cancer for centuries, current treatments usually
involve poisonous mustard gas, chemotherapy, radiation, and targeted therapies. While traditional plant-derived medi-
cines are safe, what are the active principles in them and how do they mediate their e?ects against cancer is perhaps best
illustrated by curcumin, a derivative of turmeric used for centuries to treat a wide variety of in?ammatory conditions. Cur-
cumin is a diferuloylmethane derived from the Indian spice, turmeric (popularly called ‘‘curry powder”) that has been
shown to interfere with multiple cell signaling pathways, including cell cycle (cyclin D1 and cyclin E), apoptosis (activation
of caspases and down-regulation of antiapoptotic gene products), proliferation (HER-2, EGFR, and AP-1), survival
(PI3K/AKT pathway), invasion (MMP-9 and adhesion molecules), angiogenesis (VEGF), metastasis (CXCR-4) and
in?ammation (NF-jB, TNF, IL-6, IL-1, COX-2, and 5-LOX). The activity of curcumin reported against leukemia and
lymphoma, gastrointestinal cancers, genitourinary cancers, breast cancer, ovarian cancer, head and neck squamous cell
carcinoma, lung cancer, melanoma, neurological cancers, and sarcoma re?ects its ability to a?ect multiple targets. Thus
an ‘‘old-age” disease such as cancer requires an ‘‘age-old” treatment.
Ó 2008 Elsevier Ireland Ltd. All rights reserved.
Keywords: Curcumin; Cancer; In?ammation; Anticancer activity; Chemoprevention; Chemosensitization; Radiosensitization
1. Introduction
for cancer, it has involved the use of harmful sub-
stances, such as poisonous mustargen introduced
Studies have estimated that genetic factors cause
in 1941; chemotherapy, introduced in 1971; and
only 5–10% of all human cancers, while the remain-
then now targeted therapies, introduced in 1991.
ing percentage is caused by lifestyle. In spite of an
The progress in cancer research is determined by
extensive search for safe and e?cacious treatments
the number of approvals from the U.S. Food and
Drug Administration (FDA), as indicated by very
*
few in 1970; seven in 1987; 16 in 1996; 21 in 1998,
Corresponding author. Tel.: +1 713 792 3503; fax: +1 713 794
and 28 in 2006 [1]. More than 70% of the FDA
1613.
E-mail address: aggarwal@mdanderson.org (B.B. Aggarwal).
approved anticancer drugs can be traced back to
0304-3835/$ - see front matter Ó 2008 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.canlet.2008.03.025

---

134
P. Anand et al. / Cancer Letters 267 (2008) 133–164
their origin in plant-derived natural products, which
Curcumin is a hydrophobic polyphenol derived
were traditionally used as ancient remedies for var-
from turmeric: the rhizome of the herb Curcuma
ious ailments. Vinblastine from Vinca rosea is one of
longa. Chemically, it is a bis-a,b-unsaturated b-
the earliest example that originated from an Ayurv-
diketone
(commonly
called
diferuloylmethane)
edic medicine described for cancer and paclitaxel is
that exhibits keto-enol tautomerism, having a pre-
perhaps one of the most recent example that origi-
dominant keto form in acidic and neutral solu-
nated from Chinese paci?c yew plant.
tions and a stable enol form in alkaline media.
Cancer is well recognized as a disease of old age
Commercial curcumin is a mixture of curcumi-
(Fig. 1). It is estimated that the process of tumori-
noids,
containing
approximately
77%
difer-
genesis starts at around the age of 20 and detection
uloylmethane, 18% demethoxycurcumin, and 5%
of cancer is normally around the age of 50 or later
bisdemethoxycurcumin.
Traditionally,
turmeric
(Table 1); thus with an estimated incubation time
and other curcuminoids have been used in thera-
of around 20–30 years. Recent studies indicate that
peutic preparations for various ailments in di?er-
in any given type of cancer 300–500 normal genes
ent parts of the world. Numerous therapeutic
have been modi?ed somehow to result in the cancer-
e?ects of curcumin/turmeric have been con?rmed
ous phenotype. Although cancers are characterized
by modern scienti?c research. Herein, we present
by the dysregulation of cell signaling pathways at
a systematic review of the clinical and experimen-
multiple steps, most current anticancer therapies
tal data on the use of curcumin in the treatment
involve the modulation of a single target. The inef-
of cancer.
fectiveness, lack of safety, and high cost of monotar-
geted therapies have led to a lack of faith in these
2. Molecular targets of curcumin
approaches. As a result, many pharmaceutical com-
panies are increasingly interested in developing
Extensive research conducted within the past
multitargeted therapies. Many plant-based prod-
two decades has revealed that cancer is a result
ucts, however, accomplish multitargeting naturally
of the dysregulation of multiple cell signaling
and, in addition, are inexpensive and safe compared
pathways. Curcumin is a highly pleiotropic mole-
to synthetic agents. However, because pharmaceuti-
cule that modulates numerous targets (Fig. 2),
cal companies are not usually able to secure intellec-
including the activation of transcription factors
tual property rights to plant-based products, the
(e.g., NF-jB, STAT3, AP-1, NRF-2, PPAR-c,
development of plant-based anticancer therapies
and HIF-1), receptors (e.g., HER-2, IL-8, and
has not been prioritized. Nonetheless, curcumin, a
CXCR-4), kinases (e.g., EGFR, ERK, JAK,
plant-based product, has shown signi?cant promise
and AAPK), cytokines (e.g., TNF, IL, MIP,
against cancer and other in?ammatory diseases.
and MCP), enzymes (e.g., MMP, iNOS, GST,
40
Prostate cancer
30
Lung cancer
Urinary bladder cancer
20
ncidence of cancer
Lymphoma
10
% I
0
<20
>85
20-34
35-44
45-54
55-64
65-74
75-84
Age (Years)
Fig. 1. Age dependency of cancer incidence. Data presented in the ?gure is based on the cancer statistics published in 2007 [3].

---

P. Anand et al. / Cancer Letters 267 (2008) 133–164
135
Table 1
3. Anticancer potential
Median age at which most cancers are diagnosed in American
population
Curcumin has been shown to exhibit therapeutic
Cancer site
Median age at diagnosis (years)
potential against variety of di?erent cancers includ-
Breast cancer
61
ing leukemia and lymphoma; gastrointestinal can-
Gastrointestinal cancers
cers, genitourinary cancers, breast cancer, ovarian
Esophagus cancer
69
cancer, head and neck squamous cell carcinoma,
Stomach cancer
71
lung cancer, melanoma, neurological cancers and
Intestine cancer
67
Liver cancer
65
sarcoma (Fig. 3). The current status of curcumin’s
Pancreatic cancer
72
anticancer potential against various cancers is sys-
Colorectal cancer
71
tematically analyzed and presented below under dif-
Genitourinary cancers
ferent headings.
Bladder cancer
73
Kidney cancer
65
3.1. Breast cancer
Prostate cancer
68
Gynecologic cancers
Breast cancer is the most common and frequently
Cervical cancer
48
diagnosed cancer at a median age of 61 years in
Ovarian cancer
63
women [3]. In the United States, breast cancer
Uterine cancer
67
accounts for about 26% of all newly diagnosed neo-
Thoracic/Head and neck cancer
plasms [4]. Even though substantial advances in
Lung cancer
70
Oral cancer
62
therapy and diagnosis have enhanced the survival
Thymus cancer
50
rate of patients with breast cancer, late recurrences
of the disease account for more than 60% of deaths
Hematologic cancers
Leukemia
67
from breast cancer [5]; the survival rate among
Lymphoma
64
patients with metastatic disease does not seem to
Multiple Myeloma
70
be signi?cantly a?ected by the current treatment
Melanoma
59
modalities [6]. Indeed, further studies are needed
Bone cancer
39
to optimize therapeutic interventions in patients
Brain tumor
56
with metastatic breast cancer.
Data presented in the table is based on the cancer statistics
Several reports have described the anticarcino-
published in 2007 [3].
genic activity of curcumin in a variety of breast
cancer cell lines. One of our early studies estab-
lished that the antiproliferative e?ect of curcumin
and ATPase), and growth factors (e.g., EGF,
in human breast cancer cell lines, including hor-
NGF, HGF, and PDGF). Because of its ability
mone-dependent, hormone-independent, and mul-
to interact with a diverse range of molecular tar-
tidrug-resistant
cells,
was
time-
and
dose-
gets, curcumin can a?ect numerous molecular and
dependent and correlated with curcumin’s inhibi-
biochemical cascades. One of our recently pub-
tion of ornithine decarboxylase activity [7]. Several
lished reviews presents a more detailed descrip-
mechanisms have been proposed to account for
tion of the molecular targets of curcumin [2].
the action of curcumin in breast cancer cells.
Extensive research conducted during the past cen-
For example, curcumin was found to inhibit the
tury has established the complexity and involve-
aryl hydrocarbon receptor and cytochrome P450
ment
of
multiple
signaling
pathways
in
the
1A1 [7]; the tyrosine kinase activity of p185neu;
cancer growth and progression, which in turn
the expression of Ki-67, PCNA, p53 mRNAs;
suggests that a drug, which can interact with
COX-I and COX-II enzymes. Curcumin also
multiple target molecules, will be more e?cacious
induced p53-dependent Bax expression, inhibited
than the current monotargeted anticancer drugs.
vascular endothelial growth factor (VEGF), basic
Curcumin’s multitargeting ability may be the
?broblast growth factor (b-FGF) [8,9], disrupted
key to its therapeutic potential against cancer.
mitotic spindle structure and induced micronucle-
In the next section of this review, we analyze
ation [10]. It has been shown to inhibit telomerase
the current status of curcumin’s potential against
activity through human telomerase reverse trans-
various cancers.
criptase
[11], downregulate the expression of

---

136
P. Anand et al. / Cancer Letters 267 (2008) 133–164
Fig. 2. Molecular targets of curcumin. These include, NF-jB, nuclear factor-kappa B; AP-1, activating protein1; STAT, signal
transducers and activators of transcription; Nrf-2, nuclear factor 2-related factor; Egr-1, early growth response gene-1; PPAR-c,
peroxisome proliferator-activated receptor-gamma; CBP, CREB-binding protein; EpRE; CTGF, connective tissue growth factor; EGF,
epidermal growth factor; EGFRK, epidermal growth factor receptor-kinase; FGF, ?broblast growth factor; HGF, hepatocyte growth
factor; NGF, nerve growth factor; PDGF, platelet-derived growth factor; TGF-b1, transforming growth factor-b1; VEGF, vascular
endothelial growth factor; AR, androgen receptor; Arh-R, aryl hydrocarbon receptor; DR-5, death receptor-5; EGF-R, epidermal growth
factor-receptor; EPC-R, endothelial protein C-receptor; ER-a, estrogen receptor-alpha; Fas-R, Fas receptor; H2-R, histamine (2)-
receptor; InsP3-R, inositol 1,4,5-triphosphate receptor; IR, integrin receptor; IL-8-R, interleukin 8-receptor; LDL-R, low density
lipoprotein–receptor; MMP, matrix metalloproteinase; TIMP, tissue inhibitor of metalloproteinase-3; iNOS, inducible nitric oxide
oxidase; COX-2, cyclooxygenase-2; LOX, lipoxygenase; Gcl, glutamate-cysteine ligase; NAT, arylamine N-acetyltransferases; IAP,
inhibitory apoptosis protein; HSP-70, heat-shock protein 70; TNF-a, tumor necrosis factor alpha; IL, interleukin; MCP, monocyte
chemoattractant protein; MIF, migration inhibition protein; MIP, macrophage in?ammatory protein; ERK, extracellular receptor kinase;
IARK, IL-1 receptor-associated kinase; cAK, autophosphorylation-activated protein kinase; CDPK, Ca2+-dependent protein kinase;
cPK, protamine kinase; JAK, janus kinase; JNK, c-jun N-terminal kinase; MAPK, mitogen-activated protein kinase; TK, protein tyrosine
kinase; FAK, focal adhesion kinase; PhK, phosphorylase kinase; pp60c-src, pp60c-src tyrosine kinase; PKA, protein kinase A; PKB,
protein kinase B; PKC, protein kinase C; FPTase, farnesyl protein transferase; GST, glutathione S-transferase; HO, hemeoxygenase;
ICAM-1, intracellular adhesion molecule-1; VCAM-1, vascular cell adhesion molecule-1; ELAM-1, endothelial leukocyte adhesion
molecule-1; SHP-2, Src homology 2 domain-containing tyrosine phosphatase 2, uPA, urokinase-type plasminogen activator.
matrix metalloproteinase-2 (MMP-2), upregulate
LOX pathways [17], induce the degradation of
tissue inhibitor of metalloproteinase-1 (TIMP-1)
cyclin E expression through a ubiquitin-dependent
[12], and block NF-jB and AP-1 activation [13–
pathway,
upregulate
cyclin-dependent
kinase
16]. Studies have also shown curcumin to inhibit
inhibitors p21 and p27 [18] and downregulate

---

P. Anand et al. / Cancer Letters 267 (2008) 133–164
137
Fig. 3. Various cancers against which curcumin has potential for prevention and treatment.
the insulin-like growth factor-1 (IGF-1) [19] in
of mammary DMBA–DNA adducts in the female
breast cancer cell lines.
rat. Administration (i.p.) of curcumin at 100 and
In a study by Zhang et al. [20], exposure of
200 mg/kg doses prevented the development of the
mouse breast tumor cells to curcumin caused a
number of palpable mammary tumors and mam-
dose-dependent increase in ubiquitinated exosomal
mary adenocarcinomas signi?cantly. The in vivo
proteins compared to those in untreated cells. The
formation of mammary DMBA–DNA adducts also
exosomes isolated from tumor cells pretreated with
was depressed in animals administered with curcu-
curcumin have a much attenuated inhibition of
min and there was no signi?cant enhancement of
IL-2-stimulated-NK cell activation. The tumor exo-
liver GST activity following curcumin administra-
somes isolated from curcumin-pretreated tumor
tion. However, it was also showed that animals
cells had lower potency for inhibition of IL-2-stim-
fed with diets containing 1.0% curcumin had no
ulated NK cell cytotoxicity compared to those from
e?ect on DMBA-induced mammary tumor. In
non-treated cells, suggesting that the partial reversal
1996, Pereira et al. showed that curcumin (8 and
of tumor exosome-mediated inhibition of NK cell
16 g/kg in diet) was weakly e?ective in DMBA
tumor cytotoxicity may account for the anticancer
induced mammary carcinogenesis. Another study
properties of curcumin. The antitumor activities of
evaluated the modulating e?ects of turmeric (T),
curcumin and its isoxazole analog were not a?ected
ethanolic turmeric extract (ETE) and curcumin-free
by multiple gene expression changes in a multidrug-
aqueous turmeric extract (CFATE) on the initiation
resistant (MDR) model of the MCF-7 breast cancer
or post-initiation phases of DMBA-induced mam-
cell line [21]. Treatment of breast cancer cells, hav-
mary tumorigenesis in female Sprague–Dawley rats.
ing up-regulated expression of nicotinamide N-
Dietary administration of 1% turmeric/0.05% etha-
methyltransferase (NNMT), with curcumin resulted
nolic turmeric extract 2 weeks before, on the day
in reduction of the Nicotinamide N-methyltransfer-
of DMBA treatment (day 55) and 2 weeks after
ase (NNMT) level [22]. In addition to curcumin,
the single dose (15 mg/animal) of DMBA (during
several derivatives [7,23–25] and analogs [7,21,26]
the initiation period) resulted in signi?cant suppres-
of curcumin were also found to have anticarcino-
sion of DMBA-induced mammary tumorigenesis as
genic property against various breast cancer cell
seen by a reduction in tumor multiplicity, tumor
lines.
burden and tumor incidence. In another study it
Several in vivo studies have established the che-
was showed that feeding 1% dibenzoylmethane
mopreventive e?ect of curcumin against breast can-
(DBM), a derivative of curcumin in AIN 76A diet,
cer. In 1998 a group studied curcumin’s capacity to
inhibited both the multiplicity and incidence of
inhibit
7,12-dimethylbenzanthracene
(DMBA)
DMBA-induced mammary tumor by 97%. In
induced mammary tumor and the in vivo formation
2001, it was also showed that feeding 1% DBM diet

---

138
P. Anand et al. / Cancer Letters 267 (2008) 133–164
inhibited formation of DMBA–DNA adducts in
normally suitable for the xenograft model studies.
mammary glands and the development of mammary
Even though it is the only study reporting the inhi-
tumors in Sencar mice. The chemopreventive e?ect
bition of tumor regression, further studies are
of curcumin on diethylstilbestrol (DES)-induced
needed to resolve the contradictions about the e?ec-
tumor promotion of rat mammary glands initiated
tiveness of curcumin against breast cancer in vivo.
with radiation was evaluated in a study. The admin-
An early clinical trial, evaluated the e?ectiveness
istration of dietary curcumin signi?cantly reduced
of topical application of a curcumin ointment in
the incidence (28.0%) of mammary tumors. Multi-
seven patients with breast cancer. In this study,
plicity and Iball’s index of mammary tumors were
71% of the patients showed a positive response mea-
also decreased by curcumin. Rats fed with the cur-
sured as reduction in lesion size, pain, itching and
cumin diet showed a reduced incidence of the devel-
exudates [7].
opment of both mammary adenocarcinoma and
ER(+)PgR(+) tumors in comparison with the con-
3.2. Gastrointestinal cancers
trol group. Whole mounts of the mammary glands
showed that curcumin yielded morphologically
3.2.1. Oesophageal cancer
indistinguishable proliferation and di?erentiation
Oesophageal cancer is the seventh leading cause
from the glands of the control rats. The e?ect of cur-
of death from cancer in men, with a mean 5-year
cumin on gamma-radiation induced mammary
survival rate in the United States of 15.6%. In the
tumors was also demonstrated in rats [2].
United States, there were an estimated 15,560 new
In addition to the chemopreventive e?ects, anti-
cases of and 13,940 deaths from oesophageal cancer
metastatic e?ect of curcumin was also established
in 2007 [4]. The standard treatment for surgically
by the in vivo model studies. In a xenograft model
resectable tumors is esophagectomy; radiochemo-
study (nude mice) conducted in our own laboratory,
therapy is used for locally advanced, unresectable
the primary tumor was surgically removed after 58–
tumors. Even with these therapies, however, both
60 days of tumor cell inoculation and dietary curcu-
local regional tumor control and the overall survival
min (2%) was given to the animals starting from
of patients with oesophageal cancer remain poor,
?fth day to 5 week of primary tumor removal. We
and treatments are associated with signi?cant
observed that administration of curcumin signi?-
adverse e?ects, including treatment-related pneumo-
cantly decreased the incidence of breast cancer
nitis,
postoperative
pulmonary
complications,
metastasis to the lung and suppressed the expression
oesophagitis, and pericarditis [28]. Innovative treat-
of NF-jB, COX-2, and MMP-9. Another group
ment strategies are needed to improve the outcome
also evaluated the e?ect of curcumin on lung metas-
of patients with oesophageal cancer.
tasis of breast cancer. In this study, intercardiac
Curcumin could be a potential candidate for use
inoculation of breast cancer cells was done in the
in the treatment of esophageal cancer, few studies
nude mice and the animals were fed with diet con-
have examined it in this disease and no in vitro eval-
taining 1% curcumin. Thirty-?ve days after tumor
uations of its anticancer e?ects in oesophageal can-
implantation the animals were sacri?ced and enu-
cer cells have been reported. However, curcumin
merated the lung metastases. It was observed that
was found to inhibit the cytokine-induced activation
all the animals in the untreated group had lung
of iNOS, JNK, VCAM, and NF-jB in human
metastasis whereas 21% animals in the treated
oesophageal microvascular endothelial cells isolated
group were metastases free. In the control group
from normal human oesophageal tissues [29]. Since
only 17% animals were having few metastatic nod-
in?ammatory molecules-like NF-jB are overexpres-
ules (metastatic score <3) whereas in curcumin-trea-
sed in several tumor tissues, these results may be
ted group 68% animals had few metastatic nodules
indirect evidence that curcumin may be e?ective
[2]. In contrast to the above in vivo studies, Somas-
against oesophageal cancer. Two in vivo studies
undaram et al., [27] reported a signi?cant inhibition
have been reported with curcumin in oesophageal
of tumor regression in a xenograft mouse model of
cancer. In one, dietary curcumin (500 ppm) fed dur-
human breast cancer. These contradictory ?ndings
ing initiation and post-initiation stages inhibited the
could have been caused by the di?erence in admin-
incidence of oesophageal carcinogenesis by 27% and
istered doses as well as the time of treatment. For
33%, respectively, in rats [2]. In the other study, the
example, the authors studied the e?ect of curcumin
e?cacy of curcumin as a chemopreventive agent
in a breast xenograft model for 3 days, which is not
was assessed by measuring the modulation in the

---

P. Anand et al. / Cancer Letters 267 (2008) 133–164
139
incidence of neoplastic change in rat oesophagus
min modi?es apoptosis resistance, leading to the
[30].
inhibition of tumor formation and the prevention
of adenoma development in the intestinal tract.
3.2.2. Gastric cancer
The chemopreventive e?ect of curcumin for intesti-
In the United States, in 2007, there were an esti-
nal tumors in Min/+ mice was investigated. A die-
mated 21,260 new cases of and 11,210 deaths from
tary level of 0.15% curcumin decreased tumor
gastric cancer [4]. Current major modalities for the
formation in MinÀ/À mice by 63%. Examination
treatment of gastric cancer include surgery and che-
of intestinal tissue from the treated animals showed
motherapy, but local recurrence and distant metas-
the tumor prevention by curcumin was associated
tases, which lead to poor survival rates, are still
with increased enterocyte apoptosis and prolifera-
unresolved issues in this disease [31], indicating that
tion. Curcumin also decreased expression of the
modi?ed treatment strategies are needed. The cyto-
oncoprotein b-catenin in the erythrocytes of the
toxic e?ect of curcumin on gastric carcinoma cell
Min/+ mouse, an observation previously associated
lines has been established. In a study curcumin
with an antitumor e?ect. Curcumin enhanced PhIP-
and 5-?uorouracil (5-FU) synergistically inhibited
induced apoptosis and inhibited PhIP-induced
the growth of gastric carcinoma cells. In another
tumorigenesis in the proximal small intestine of
study, curcumin reversed the MDR of a human gas-
Apc (min) mice. Evaluation of the preventive e?ect
tric carcinoma cell line in correlation with a decrease
of curcumin on the development of adenomas in
in P-gp function and a promotion of caspase-3 acti-
the intestinal tract using a Min/+mouse model
vation [7].
showed promising chemopreventive e?ect. Mice
Several in vivo chemoprevention studies have
received dietary curcumin for 15 weeks and curcu-
been reported with curcumin in gastric cancers. In
min at 0.1% in the diet was without e?ect whereas
some of the chemoprevention studies, curcumin
at 0.2% and 0.5% it reduced adenoma multiplicity
fed as dietary turmeric (2% or 5%) to mice and Syr-
by 39% and 40%, respectively. How curcumin is
ian golden hamsters signi?cantly inhibited the ben-
metabolized in intact rat intestinal sacs in situ was
zopyrene-induced
forestomach
tumors.
evaluated and showed that curcumin undergoes
Furthermore, the incidence and multiplicity of fore-
extensive metabolic conjugation and reduction in
stomach tumors induced by benzopyrene in female
the gastrointestinal tract and that the process of
Swiss mice were signi?cantly inhibited by pure cur-
metabolism is more complex in human than in rat
cumin given 2 weeks before, during and after the
intestinal tissue [7]. Experiments performed on
carcinogen treatment. Other studies also revealed
intestinal tumors in C57BL/6J-Min/+ (Min/+) mice
the chemopreventive e?ect of curcumin on benzopy-
demonstrated that curcumin has a regulatory role in
rene-induced forestomach cancer. A signi?cant
lymphocyte-mediated immune function [33]. Fur-
reduction in benzopyrene-induced forestomach pap-
ther, levels of COX-2 protein expression have been
illomas in mice due to treatment with dietary tur-
found to re?ect the retardation of adenoma devel-
meric
extract
containing
curcumin
was
also
opment in mouse intestines after treatment with cur-
reported. It was also showed that curcumin inhib-
cumin [34].
ited MNNG-induced duodenal tumor in mice and
In a phase I clinical trial six patients with intesti-
gastric cancer in rats [7].
nal metaplasia of the stomach was treated with 0.5–
12 g/day of curcumin for 3 months. In this study
3.2.3. Intestinal cancer
one out of the six patients showed histologic
According to the estimates of American Cancer
improvement in precancerous lesions after the treat-
Society, 5640 new intestinal cancers will have been
ment [2].
diagnosed and 1090 patients will have died from
intestinal cancer in 2007 [4]. Recent advances in
neoadjuvant
therapies
have
contributed
to
3.2.4. Hepatic cancer
improved survival for patients with intestinal cancer
Hepatocellular carcinoma (HCC) is an aggressive
[32] and various adjuvant treatment modalities are
cancer, and its incidence is increasing in the United
now being explored.
States and worldwide. In 2007, an estimated 19,160
So far, the e?cacy of curcumin in intestinal can-
new cases of HCC will have been diagnosed and
cer has been shown in a few animal studies. In vivo
16,780 patients will have died from HCC in the Uni-
studies using mouse models have proved that curcu-
ted States [4]. Novel neoadjuvant treatments are

---

140
P. Anand et al. / Cancer Letters 267 (2008) 133–164
being investigated for the improvement of the cur-
esis
model,
5-week-old
C3H/HeN
mice
were
rent treatment strategies [35].
injected intraperitoneally with DENA. One group
Several studies have examined the anticarcino-
of the mice were fed with 0.2% curcumin-containing
genic activity of curcumin in hepatic cancer cells
diet, starting 4 days before DENA injection and
in vitro. In one of these studies, conducted in curcu-
until termination of the experiment. At the age of
min-treated human hepatoblastoma cells, several
42 weeks, the curcumin group had 81% less multi-
hallmarks of apoptosis, including DNA laddering,
plicity and 62% fewer hepatocarcinomas than the
chromatin condensation, fragmentation, and apop-
non-treated group. It also suppressed liver in?am-
tosis-speci?c cleavage of 28S and 18S ribosomal
mation in rats. Liver was identi?ed as the major site
RNA were observed. Curcumin has also exhibited
for the metabolism of curcumin, and the major
signi?cant antiinvasion activity in human HCC
metabolites in suspensions of human or rat hepato-
SK-Hep-1 cells, an e?ect that is associated with cur-
cytes were identi?ed as hexahydrocurcumin and
cumin’s-inhibitory action on MMP-9 secretion.
hexahydrocurcuminol. In rats, in vivo, curcumin
Curcumin undergoes metabolic conjugation and
glucuronide and curcumin sulfate were identi?ed
reduction in subcellular fractions of human and
as the major products of curcumin biotransforma-
rat hepatic tissues [7]. It has also been established
tion, whereas hexahydrocurcumin, hexahydrocurcu-
that the elevation of GSH levels mediates the e?ect
minol, and hexahydrocurcumin glucuronide were
of curcumin in hepatocytes [36].
present only in small amounts. Another in vivo
Curcumin has also been found to interrupt the
study showed that curcumin mixed into a diet could
cell cycle, to have cytotoxic e?ects, and to have a
achieve levels of the drug in the liver su?cient to
role in antiproliferation and the induction of apop-
explain its pharmacological e?ects. Dietary curcu-
tosis in a hepatocarcinoma cell line. Curcumin is a
min increased the activity of hepatic UGT enzymes,
potent
inhibitor
of
phenol
sulfotransferase
which can detoxify carcinogens, in male Wistar rats.
(SULT1A1) in human liver and extrahepatic tissues
In an orthotopic implantation model, curcumin
[37]. Curcumin inhibited the IL-6 production, his-
suppressed both intrahepatic metastases and the
tone acetyltransferase (HAT) activity, and AP-1
development of altered hepatic foci (AHF) in rat liv-
activation [38] and prevented cell death and apopto-
ers. Inhibition of tumor growth by systemic admin-
tic biochemical changes, such as the mitochondrial
istration of 20 lg/kg curcumin for 6 consecutive
release of cytochrome c, the activation of caspase-
days to rats bearing the highly cachectic Yoshida
3, and the cleavage of PARP in human hepatoma
AH-130 ascites hepatoma was also reported. In
cells [7,39]. Another proposed mechanism for curcu-
one of the studies, hepatocellular carcinoma cells
min’s inhibition of tumor growth in HCC is through
were injected subcutaneously in mice and 3 weeks
the inhibition of hypoxia-inducible factor-1 by
after cell injection, a tumor fragment from the injec-
degrading the aryl hydrocarbon receptor nuclear
tion site was implanted to liver. Curcumin (100–
translocator [40,41]. Further, it has been shown that
200 mg/kg) was administered after the implantation
mitochondrial hyperpolarization is a prerequisite
for 20 days and then the e?ect of curcumin treat-
for curcumin-induced apoptosis and that mtDNA
ment was evaluated. Although the growth of tumors
damage is the initial event in a chain leading to
at the implanted site was not a?ected by the curcu-
apoptosis in HepG2 cells [42]. In an in vitro study
min treatment there was a signi?cant and dose
using hepatic cancer cells, a combination of curcu-
dependant decrease in number of intrahepatic
min and cisplatin had synergistic antitumor e?ects,
metastases [43].
and that with doxorubicin additivity or sub-additiv-
Curcumin also prevented the induction of hepatic
ity [7].
hyper plastic nodules, body weight loss, and hypo-
A considerable number of reports have also
proteinemia in carcinogen induced as well as xeno-
described curcumin in HCC in vivo. In one of these
graft hepatic cancer models. Both curcumin and
studies, curcumin signi?cantly reduced the number
curcumin complexed with manganese prevented
of gammaglutamyl transpeptidase-positive foci, a
the increase of hepatic lipid peroxidation expressed
characteristic considered to be the precursor of
as MDA level in mice. The antiangiogenic activity
hepatocellular neoplasm, in rats. Curcumin also
of curcumin in hepatocarcinoma cells implanted in
had anticarcinogenic e?ects mediated through the
nude mice was found to be mediated through the
induction
of
glutathione-linked
detoxi?cation
reduction of biomarkers COX-2 and VEGF [43].
enzymes in rat livers. In a murine hepatocarcinogen-
In a pilot trial with 12 patients with hepatic metas-

---

P. Anand et al. / Cancer Letters 267 (2008) 133–164
141
tases from colorectal cancer the concentrations of
Two in vivo studies were reported showing the
the curcumin in normal and malignant human liver
antitumor activity as well as chemosensitization
tissue after patients received 450–3600 mg of curcu-
e?ect of curcumin against pancreatic cancer. In a
min daily for 1 week prior to surgery were not suf-
xenograft model study, pancreatic cancer cells were
?cient to elicit pharmacologic activity, perhaps
injected subcutaneously on the side of the abdomen
because of the extensive degree to which curcumin
of female nude mice. Once tumor masses became
was metabolized in the intestine [7].
established, animals were randomized to receive
intravenous liposomal curcumin (40 mg/kg, 3 time
3.2.5. Pancreatic cancer
per week) for 20 days. Treatment with liposomal
Pancreatic cancer is one of the most common
curcumin resulted in reduced tumor size and visible
cancers, and the fourth leading cause of cancer-
blanching of tumors showing decreased expression
related mortality, accounting for about 6% of all
of CD31 as well as VEGF and IL-8. These results
cancer-related deaths, in both men and women.
indicate that curcumin suppressed pancreatic carci-
The median age of diagnosis is 72 years [3]. Despite
noma growth in murine xenograft models and
advances in molecular pathogenesis, patients with
inhibited tumor angiogenesis [55]. A recent study
pancreatic cancer have a mean relative 5-year sur-
conducted in our group investigated the chemosen-
vival rate of 5%, and the disease remains a major
sitization e?ect of curcumin using an orthotopic
unsolved health problem [4]. In an attempt to
pancreatic cancer model. After 1 week of implanta-
improve
survival
rates,
recent
therapeutic
tion, mice were randomized into the following treat-
approaches have mostly focused on evaluating che-
ment groups: untreated control (olive oil, 100 lL
motherapy regimens in which gemcitabine is com-
daily), curcumin alone (1 g/kg/day), gemcitabine
bined with a second cytotoxic agent.
alone (25 mg/kg twice weekly by i.p. injection) and
Research over the past decade has indicated that
combination of curcumin and gemcitabine. The ani-
curcumin has an anticarcinogenic e?ect in various
mals were sacri?ced 6 weeks after tumor cell injec-
pancreatic cell lines, with numerous mechanisms
tion and 5 weeks from the date of treatment. The
having been proposed to account for this e?ect. In
tumor volume in the combination of curcumin and
human pancreatic cancer MIA PaCa-2 cells, curcu-
gemcitabine group was signi?cantly lower than the
min was found to inhibit the farnesyl protein trans-
gemcitabine alone or control group indicating the
ferase
[7].
Also,
NF-jB
was
found
to
be
chemosensitizing e?ect of curcumin. Our results
overexpressed in human pancreatic tumor tissues
showed that curcumin in combination with gemcit-
and cell lines; investigators suggested that this over-
abine signi?cantly down-regulated the expression
expression could be inhibited by curcumin because
of cell proliferation marker Ki-67 in tumor tissues
it has the ability to suppress the NF-jB expression
compared with the control group. Further, curcu-
[44–46]. Likewise, curcumin reduces numerous IL-
min alone signi?cantly suppressed the expression
8 bioactivities that contribute to tumor growth
of microvessel density marker CD31 and the pres-
and the cell viability of pancreatic carcinoma cells
ence of gemcitabine further enhanced the down-reg-
[7,47]. Other mechanisms have been proposed to
ulation of CD31 [2].
account for the growth-inhibitory e?ect of curcumin
In a clinical trial, researchers evaluated the e?ect
alone [48] or in combination with celecoxib [49]
of oral curcumin with piperine on the pain, and the
including the down-regulation of COX-2, EGFR,
markers of oxidative stress in patients with tropical
ERK1/2 [50], and Notch-1 [51]. When coupled with
pancreatitis (TP). 20 patients with tropical pancrea-
gemcitabine, curcumin has been observed to have
titis were randomized to receive 500 mg of curcumin
synergistic antiproliferative e?ects in pancreatic can-
with 5 mg of piperine, or placebo for 6 weeks, and
cer cell lines [52,53]. Liposomal curcumin down-reg-
the e?ects on the pattern of pain, and on red blood
ulated NF-jB machinery, suppressed growth and
cell levels of malonyldialdehyde (MDA) and gluta-
induced
apoptosis
of
human
pancreatic
cells
thione (GSH) were assessed. There was a signi?cant
in vitro [2]. A polymeric nanocurcumin formulation
reduction in the erythrocyte MDA levels following
also demonstrated a therapeutic e?cacy comparable
curcumin therapy compared with placebo; with a
to that of free curcumin in a panel of human pancre-
signi?cant increase in GSH levels. There was no cor-
atic cancer cell lines in vitro, and the mechanisms of
responding improvement in pain [2].
action of nanocurcumin in pancreatic cancer cells
The studies from our group [56] showed that cur-
mirrored those of free curcumin[54].
cumin inhibited pancreatic cancer in patients. 25

---

142
P. Anand et al. / Cancer Letters 267 (2008) 133–164
patients were enrolled in this study. Patients
curcumin [58]. Curcumin causes cell shrinkage,
received 8 grams of curcumin by orally every day
chromatin condensation, and DNA fragmentation,
until disease progression, with restaging every 2
by enhancing DNA damage in HT-29 cells and
months. Serum cytokine levels for interleukin IL-
HCT-116 colonocytes; it also increases GADD153
6, IL-8, IL-10, and IL-1 receptor antagonists and
mRNA and protein expression [7,59]. Curcumin
peripheral blood mononuclear cells (PBMC) expres-
upregulates TRAIL-induced apoptosis via ROS-
sion of NF-jB and COX-2 were monitored. Out of
mediated DR5 activation in human renal cancer
25 patients, 21 were evaluable for response. Circu-
cells [7]. Likewise, curcumin enhanced the silencing
lating curcumin was detectable in glucuronide and
of hsp70 expression and may therefore prove to be
sulfate conjugates forms, albeit at low steady-state
a valuable therapeutic agent for cancers whose resis-
levels, suggesting poor oral bioavailability. Two
tance is due to hsp70 expression [60]. EF24, a syn-
patients demonstrated clinical biologic activity.
thetic curcumin analog, induces apoptosis in HT-
One had ongoing stable disease for more than 18
29 cells through a redox-dependent mechanism [7].
months and, interestingly, one additional patient
Similarly, the curcumin derivative HBC disrupts cell
had a brief, but marked, tumor regression (73%),
cycle progression in HCT15 cells by antagonizing
accompanied by signi?cant increases (4- to 35-fold)
Ca2+/CaM function [61].
in serum cytokine levels (IL-6, IL-8, IL-10, and IL-1
The fact that curcumin-induced apoptosis is reg-
receptor antagonists). No toxicities were observed.
ulated by Bax suggests that the targeting of Bcl-xL
Curcumin down-regulated expression of NF-jB,
or Smac can be used to treat Bax-de?cient, chemo-
COX-2 and phosphorylated STAT3 in PBMC from
therapy-resistant cancers [62,63]. Together, curcu-
patients (most of whom had baseline levels consid-
min and either 5-FU or celecoxib downmodulate
erably
higher
than
those
found
in
healthy
COX-2 expression via the inhibition of prostaglan-