Key words
ovarian cancer - breast - biomarker - cell culture
Schlüsselwörter
Biomarker - Mammakarzinom - Ovarmalignom - Zellkultur
Introduction
Every year, around 9600 women in Germany develop ovarian cancer. This makes it the
fifth most common type of cancer in women. Because of its rare symptoms, 65 % of the
cases are diagnosed at a very late stage (FIGO III–IV) [1]. Despite advanced surgical techniques and modern systemic therapies (chemotherapy,
targeted biological therapies), the 5-year probability of survival (around 30 %) has
barely improved at all over recent decades [2], [3]. New therapeutic approaches are therefore urgently needed.
The treatment of ovarian cancers using oncolytic viruses offers a very promising approach
[4]. These are “living” agents which specifically infect and kill tumour cells as part
of the virus replication process. Huge numbers of progeny virions are released, which
in turn attack further tumour cells. The capability of constant, tumour-specific replication
is a property that sets virotherapy apart from classical gene therapy, in which viral
vectors that are not able to replicate are used to insert foreign genetic material into cells. Moreover, oncolytic
viruses can also be used as “gene carriers” to enhance their antineoplastic effects.
In contrast to classic gene therapy, the therapeutic transgene, coupled with the viral
vector from which it is coded, spreads out within the tumour. This overcomes the hitherto
primary transduction inefficiency of tumour cells, a significant limitation in gene
therapy for cancer [5].
The use of oncolytic viruses to treat tumours is not a new idea. Interestingly, viruses
with natural oncolytic properties were first described at the start of the last century;
a retrospective of the history of virotherapy can be found in Kelly et al. [6]. In the mid-20th century, cases of spontaneous tumour remission were reported following
natural infection with measles virus [7], [8]. Clinical trials and case studies followed in which adenoviruses or the Newcastle
Disease Virus (NDV) were used, among others, to treat tumours [9], [10]. However, the inadequate effectiveness, a lack of tumour specificity and dose-limiting
side effects made it clear, that a comprehensive understanding of how oncolytic viruses
work would be essential if they were to be used in clinical practice. Since the capability
for the genetic characterisation and
manipulation of viral vectors did not exist in those early days, virotherapy has only
experienced a renaissance since the start of the rapid developments in the field of
gene- and biotechnology in the 1990s. Now, both the tumour selectivity and the anti-neoplastic
properties of oncolytic viruses can be specifically manipulated and optimised. As
a consequence hundreds of patients are able to take part in prospective clinical virotherapy
studies (including phase III), today [11].
This paper offers an overview of oncolytic viruses that are used in clinical studies
to treat patients with ovarian cancer. The basic principles of virotherapy and its
particular characteristics are also explained. Future challenges and the potential
that oncolytic viruses offer will then be discussed.
Mechanisms of Tumour Selectivity
Mechanisms of Tumour Selectivity
Throughout evolution, viruses have excelled at specialising in penetrating host cells
and appropriating their biosynthetic apparatus. Thereby, they manipulate essential
cell functions such as cell division, differentiation and cell death.
These cellular changes are frequently very similar to the changes that a cell experiences
during carcinogenesis (e.g. inactivation of the tumour suppressor gene p53, manipulation
of the interferon system, stimulation of the cell cycle, suppression of apoptosis)
[12]. This is one of the reasons why various viruses prefer to grow in tumour cells.
Viruses with natural oncolytic properties include Newcastle Disease viruses (NDV)
[13], Vaccinia viruses VV [14], vesicular stomatitis viruses (VSV) [15], parvovirus H1 (H-1PV) [16], measles vaccine viruses (MeV) [17] and reoviruses (RV) [18]. Viruses can also be genetically engineered so that they are dependent on neoplastic
host cells to reproduce. This is achieved by (1) modifying the viral envelope to allow
selective uptake into tumour
cells, (2) disabling a gene needed for efficient replication in normal cells but which
neoplastic cells can do without, and (3) creating tumour or tissue-specific promoters
that regulate the expression of viral genes [12]. It is also possible to combine these approaches [19]. [Table 1] provides an overview of oncolytic viruses that are already used in clinical studies
to treat patients with ovarian cancer.
Tab. 1 Oncolytic viruses that have been used in clinical phase 1 studies on the treatment
of patients with ovarian cancer.
|
Virus
|
Name
|
Mechanism of tumour selectivity
|
Result
|
Reference
|
|
Measles vaccine virus
|
MeV-CEA
|
Natural tumour selectivity
|
Good tolerance. Dose-dependent stabilisation of the progress of the disease in 14
out of 21 patients with an average duration of 93 days.
|
[24]
|
|
Adenovirus
|
Onyx-015
|
Deletion in the E1B and E3B gene (tumour selectivity for cells with defective p53
signal transduction pathway and defective RNA transport)
|
Good tolerance. No clear radiological or clinical tumour response.
|
[39]
|
|
H101
|
Deletion in the E1B and E3B gene (tumour selectivity for cells with defective p53
signal transduction pathway and defective RNA transport)
|
Good tolerance. 3/9 patients with complete remission, 2/9 with partial remission and
4/9 with no tumour response.
|
[40]
|
|
Ad5-delta24-RGD
|
Binds to αvβ3 and αvβ5 integrins; deletion in the E1A gene (tumour selectivity for
cells with defective retinoblastoma protein-dependent cell cycle control)
|
Good tolerance. 15/21 patients with stable disease, 6/21 with progressive disease
and 7/21 with decreasing CA125.
|
[42]
|
Viruses with Natural Tumour Selectivity
Viruses with Natural Tumour Selectivity
Living viruses capable of replication have already been used millions of times in
the context of vaccination and are known to be extremely safe therapeutic agents with
low side effects [20]. The use of “live” vaccine viruses for oncolytic virotherapy therefore would seem
to be an elegant approach. Interestingly, some vaccine strains replicate better in
neoplastic cells than the corresponding wild type viruses. Measles vaccine viruses,
for example, have natural oncolytic properties. In contrast to wild type measles virus
they predominately enter cells via the CD46 receptor which is over-expressed by malignant
cells including ovarian cancer [21], [22]. An innovative approach was described by Peng et al at the Mayo Clinic in Rochester,
USA: they generated a measles vaccine virus encoding for the human carcino-embryonic
antigen (CEA) (MeV-CEA) [23]. During virotherapy
with MeV-CEA, a simple blood test can be taken to determine the CEA level, thereby
allowing viral replication to be monitored in real time. Galanis et al. recently published
the results of a phase I trial on the intraperitoneal use of MeV-CEA in patients with
advanced ovarian cancer [24]. The virus application was well tolerated, could easily be monitored by determining
serum CEA levels and demonstrated promising clinical activity.
Another vaccine virus, the Vaccinia virus (VV), has successfully been used to treat
smallpox. Numerous clinical studies have also demonstrated that VV has natural oncolytic
properties [25], [26], [27], [28]. The use of VV for oncolytic virotherapy is regarded as very safe and generally
only causes mild, flu-like symptoms. Disabling two viral genes enhances tumor selectivity:
thymidine kinase (TK) enables the virus to replicate independently of the host cellʼs
cell cycle, and the Vaccinia growth factor (VGF, similar to the epidermal growth factor
EGF) makes it easier for the virus to infect neighbouring cells [29]. Both TK and EGF are over-expressed by many tumour cells, which is why their deletion
within the VV genome makes virus replication more difficult in non-neoplastic cells,
while neoplastic cells are able to produce large volumes
of progeny viruses. Pre-clinical studies using VV to treat ovarian cancer demonstrated
an excellent anti-tumour activity [30]. In view of the large virus genome, VV is also an excellent vector for additional
therapeutic transgenes. Chalikonda et al. generated a VV encoding for the suicide
gene cytosine deaminase (CD) (vvDD-CD). This converts the non-toxic prodrug 5-FC into
cytotoxic 5-FU. In an animal model to treat ovarian cancer, the addition of the prodrug
increased the oncolytic activity of vvDD-CD in a tumour-specific and highly significant
manner [31].
Multiple phase I/II clinical trials using VV are currently being carried out on the
treatment of ovarian cancer (http://www.jenerex.com). Currently, the first german virotherapy phase I trial to treat therapy resistant
peritoneal carcinosis is initiated, which includes a large proportion of ovarian cancer
patients with peritoneal recurrence (http://www.clinicaltrials.gov/ct2/show/NCT01443260?term=GL-ONC1&ra=1).
One of the first virotherapy approaches for the treatment of ovarian cancer was the
use of oncolytic reoviruses. These double-stranded RNA viruses replicate highly selectively
in tumour cells with an activated Ras signal transduction pathway. Hirasawa et al.
demonstrated in animal models that reoviruses are able to shrink ovarian cancer, reduce
the formation of ascites and significantly prolong the survival of animals given this
treatment [32]. Reoviruses of serotype 3 (Reolysin®, Oncolytics Biotech) are currently being used
in numerous clinical phase I and II trials that include to treat advanced ovarian
cancer (http://www.clinicaltrials.gov/ct2/results?term=Reolysin) [33]. Following both, intraperitoneal and intravenous virus application, there was excellent
tolerance, tumour-specific viral replication and oncolytic activity [34], [35].
Viruses with Genetically-Engineered Tumour Selectivity
Viruses with Genetically-Engineered Tumour Selectivity
In many cases, viral gene products require the proliferation of the host cell or inhibit
anti-viral defence mechanisms. Since tumour cells proliferate actively and frequently
have limited viral defences, the disabling of certain viral genes brings about artificial
tumour selectivity. Consequently, the adenoviral protein E1B binds to and inactivates
tumour suppressor p53, thereby promoting continuous viral replication [36]. Disabling E1B accordingly leads to the targeted infection of cells with defective
p53 signal transduction pathway. Both adenoviruses Onyx-015 and H101 (Sunway Biotech,
Shanghai, China) have corresponding deletions in the E1B gene [37], [38]. Onyx-015 was the first genetically modified oncolytic virus to be used in clinical
studies. Although the virus demonstrated promising oncolytic activities in pre-clinical
studies, a phase I trial on the treatment of patients with
ovarian carcinoma showed no clear clinical or radiological tumour response [39]. H101 is the first oncolytic virus to receive market approval (in China, not in
western countries) based on phase III trials. A phase I trial on the treatment of
malignant ascites in ovarian cancer patients led to a significant reduction in the
frequency of paracentesis, which markedly improves quality of life [40].
The primary point of attack for the adenoviruses mentioned is the Coxsackie adenovirus
receptor (CAR). The reason for the inadequate clinical effectiveness of Onyx-015 in
the treatment of ovarian cancer may be the highly variable expression of CAR and a
resulting inadequate transduction efficiency of the addressed tumour cells. Genetic
modifications of the viral envelope may accordingly lead to an increased binding affinity
towards ovarian cancer cells. The adenovirus Ad-delta24-RGD, for example, binds to
integrins in the cell surface, including those of ovarian carcinoma cells [41]. The adenoviral E1A protein also lacks the binding point for the cell cycle-regulating
retinoblastoma (Rb) protein. Consequently, Ad-delta24-RGD replicates selectively in
cells with an inactive Rb signal transduction path and accordingly in many neoplastic
cells, including ovarian carcinomas. In a phase I trial on the treatment of patients
with gynaecological
cancers, the intraperitoneal administration of Ad-delta24-RGD was well tolerated [42]. Replication of Ad-delta24-RGD in the patientsʼ ascites and promising clinical activity
was also demonstrated.
Challenges and Requirements of Oncolytic Therapy
Challenges and Requirements of Oncolytic Therapy
Genetic stability is important both for production technology and safety-related reasons.
Ultimately, it must be possible to produce the virus easily and efficiently (i.e.
with a high titre). Vaccine viruses in particular (live vaccines) satisfy these requirements.
In light of the many yearsʼ experience involving enormous patient numbers, there is
plenty of experience available regarding safety and side effects. Technology is available
for efficient virus production with high quality requirements of the production processes,
which also contribute towards a high degree of genetic stability.
One disadvantage of using vaccine viruses, however, is the high seroprevalence for
the agent. With systemic application in particular, which appears to be the medium
of choice for advanced cancer, oncolytic viruses are not only subjected to the innate
immune response, but also to acquired defence mechanisms [43]. When treating ovarian cancer, the frequent loco-regional disease spread lends itself
to intraperitoneal application. Although anti-viral antibodies may be present in malignant
ascites, a phase I study shows that the intraperitoneal use of measles vaccine viruses
does not cause a rise in the antibody titre and that the tumour response does not
correlate with the pre-therapeutic presence of anti-measles antibodies [24], [44]. Various approaches to circumvent anti-viral immune responses have also been described.
On the one hand, there are approaches which eliminate viruses by
modulating the immune response, for example through the simultaneous application of
immuno-suppressive substances [45], [46]. Another approach is one taken naturally by many viruses: by infecting endogenous,
circulating cells, they mask themselves from the immune system. In an analogy to this,
oncolytic viruses can be administered in carrier cells and delivered to the primary
tumour concealed (“Trojan Horses”) [47]. This will ensure that the agent is no longer recognised by the immune system. Viral
replication can also take place within the “Trojan”, and the carrier cells can contribute
towards the tumour selectivity by selecting cells with inherent tumour tropism [48].
The consequences of the immune response, however, do not all have a negative effect
on the effectiveness of virotherapy. The interaction of the immune system with virus-infected
cells appears to contribute to the oncolytic activity in vivo and in particular induce
positive long-term effects by stimulating the anti-tumour immune defence. These effects
can be amplified by cloning transgenic immuno-modulators into the viral genome. The
problem when investigating interactions of oncolytic viruses with the immune system,
however, is that immune-compromised Xenograft mice are frequently used as the tumour
model. Extensive translational research in the context of clinical trials to characterize
immuno-virotherapeutic effects is therefore essential and is one main interest of
the German Consortium for Translational Cancer Research (DKTK), which is currently
being set up. In an innovative clinical approach led by A. Hemminki (Advanced Therapy
Access Program), patients with
advanced, solid tumours that are refractive to treatment (including patients with
ovarian cancer) are treated with adenoviruses that express GMCSF (Granulate Macrophage
Colony Stimulating Factor) [49], [50]. GMCSF stimulates the anti-tumour immune response by activating CD8+ T lymphocytes and natural killer cells. The treatment is tolerated well and has positive
effects in the majority of the patients treated. There is also an anti-tumour as well
as an anti-viral immune response. This in particular indicates that the immunological
tolerance to tumour tissue can be broken through by oncolytic viruses.
Summary and Outlook
Virotherapy is a highly promising approach to treat ovarian cancer. Several clinical
trials have demonstrated the therapyʼs clinical effectiveness. Unlike intraperitoneally
administered chemotherapy, intraperitoneal virus administration is tolerated very
well [51]. The wholly different method of action compared to that of classic cytostatics means
on the one hand that tumours resistant to chemotherapy could be sensitive to oncolytic
viruses [52], [53]. On the other hand, the occurrence of negative side effects is not anticipated when
combined treatment involving oncolytic viruses and classical forms of treatment is
given. Oncolytic viruses are also of interest as a vehicle for therapeutic transgenes
in relation to a whole variety of genetic therapy constructs. As well as generating
oncolytic viruses that are optimised for the treatment of ovarian cancer, future studies
should also analyse
the ideal form of virus administration, the identification of potential therapeutic
combination partners and the interaction of virotherapy with the immune system of
affected patients.
