| |
INTRODUCTION
As
cancer continues to be a leading cause of death and primary focus
of physicians, biomedical researches, and pharmaceutic industries
throughout the world, interest remains high in the utilization
of diagnostic technologies that provide unique capabilities for
the detection and management of this disease.
After seven years of experience with MRI/MRS, in a magnetic field
of 1.5 Tesla (T), at the Institute of Oncology in Sremska Kamenica,
I can conclude that we have achieved the ultimate possibilities
of this type of magnet in neuroimaging and MR-spectroscopy in
vivo, which enable us not only to reveal the presence of brain
tumor and to distinguish the active part of the tumor from the
necrotic regions, but also, in the majority of patients, to estimate
tumor grading, separate different types of brain tumors, differentiate
primary brain tumors from solitary metastasis, distinguish brain
haemorrhage from intratumorous bleeding, differentiate postradiation
necrosis from recurrent tumor, and even distinguish apoptosis
from necrosis (Diklic, 1995, 1996, 1997, 2000, 2002). For new
methods we need a magnet with high field gradients which will
enable us to perform functional MRI, diffusion MRI/MRS, multivoxel
spectroscopy, 2D-spectroscopy, MRS with different nuclei, etc.
The aim of this article is to summarize the most important new
possibilities in the diagnosis and treatment in neurooncology
as a result of the innovative solutions in superconducting magnet
technology and molecular neuroimaging.
SUPERCONDUCTING
MAGNETS FOR CHALLENGING THE NEW TECHNOLOGY
Traditional NMR-spectrometers
employed vertical bore magnets which held samples in narrow tubes
of 5-25 mm in diameter and were restricted to samples with ascites
tumor cells and tumors in the tails of small laboratory animals,
usually mice and rats. The first NMR imaging technique called
zeugmatography was described by scientist Paul Lauterbur, and
the first crude MR image showing a rat tumor in vivo was produced
by scientist Raymond Damadian in 1971. In the 80-ies, the development
of large horizontal bore magnets has permitted humans to be examined,
but the strength of the magnetic field was limited to 1.5 - 2.0
T. For MRS in vivo the two most important spectroscopic requirements
are localization of the signal to the tissue under study and sensitivity
adequate to produce data in a reasonable (short) amount of time.
The former was achieved by the use of multivoxel or chemical shift
imaging, also known as CSI or MRSI, where spectra are simultaneously
obtained from voxels smaller than 0.125 cm3, which allows more
accurate characterization of tumor histological heterogenity and
measurements of local temperature and pH-values. The latter is
achieved by the use of high magnetic field magnets, which permit
significantly improved MRI and MRS resolution. Today in many MR-centers,
3.0T clinical and in few centers 7.0T research whole body actively
shielded magnet systems are in use. During the 10th Meeting of
the International Society for Magnetic Resonance in Medicine (ISMRM),
held in May 2002 in Honolulu, Hawaii, I have the opportunity to
see the world's largest research magnet of 9.38T, made in Livermore,
in California.
With higher magnetic field strength, the resolution of MR-spectra
in vivo is significantly increased allowing more metabolites to
be detected and the problems of overlapping spectra and signal
to noise ratios (SNR) to be highly improved. If the gold standard
of the 20th century for tissue diagnosis has been histopathology,
with the commencement of the new millennium, MRS is expected to
replace it. It is now understood that disease is often a multi-step
process and that histopathology is not always able to discern
all of the steps involved. MRS is suitable for screening and follow-up
programs. However, even if the MR-spectra are as good as histopathology,
the question of cost efficacy remains. At the present, however,
MRS can not reliably characterize histologic types or subtypes
of all brain tumors in the clinical routine, but if used for guidance
of stereotactic biopsy it can define the most promising target
point which can be reached safely by the stereotactic procedure.
Thereby unrepresentative biopsy specimens can be avoided.
In the near future we can expect even more diverse applications
of MRS, such as in brain temperature monitoring or water diffusion
measurements.
Interstitial
therapies with MR-thermometry
A
variety of ablative techniques exists, including thermal-based
energies, chemical (e.g. ethanol), and chemotherapeutic. Thermal-based
energies include laser, focused ultrasound, radio frequency, cryotherapy
and microwave. While radiofrequency ablation techniques are well
established within the therapeutic spheres of neurosurgery, laser
interstitial thermal therapy (LITT) is also promising. The biological
efficacy of thermal ablation techniques is strongly dependent
on the temperature achieved in all parts of the lesion. A temperature
of approximately 550- 600 C leads to coagulative destruction of
treated tissue. However, healthy tissue damage has been observed
after prolonged (>60 min.) exposure at 420 C, emphasizing the
need for targeted therapies and real-time procedural monitoring
and adjustment. Effective temperature monitoring using proton
chemical shift MR-thermometry has been successfully applied during
laser interstitial thermal therapy (LITT) of a brain tumor (frontal
astrocytoma WHOII)(Kahn et al., 1998).
Diffusion
weighted MRI/MRS
Diffusion weighted MRI/MRS techniques
also have a new perspective for clinical neurooncology. Molecular
diffusion, a random thermal Brownian motion, is expressed by molecular
water displacement. A starting point in the interpretation of
water diffusion at the cellular level in vivo is a two-compartment
model comprising extracellular (ECS) and intracellular space (ICS)
with exchange between the sites. It is generally assumed that
diffusion in ECS is free of Gaussian propagation because of low
concentrations of macromolecules and the absence of membraneous
organelles. Diffusion in the ICS is taken to be strongly restricted
due to physical (i.e. macromolecules; organelles) and chemical
(specific binding; protein transitions and movements) factors.
In a pioneering study of Zhao et al. (1996) it was shown that
apparent diffusion coefficient (ADC) increases in a mouse RIF-1
tumor after treatment with the anticancer drug cyclophosphamide,
prior to reduction of tumor volume. This study also demonstrated
that elevated ADC reversed upon regrowth of the RIF-1 tumor thus
showing diffusion to be an early index of favourable cytotoxic
treatment response. It is particularly interesting that recent
data from human brain tumors indicate that increased ADC is associated
with tumor regression and vice versa (Chenevert et al., 2000;
Mardor et al., 2001). Diffusion weighted MRI data from different
tumor types tend to suggest that density of viable cells might
be the key factor affecting water diffusion in tumors. In human
gliomas and melanomas low diffusion correlates with high cell
density. 9L glioma treated with nitrosourea (BCNU) shows severe
cell loss, widening of the ECS and an inflammatory response as
a sign of a necrotic process, at a time when ADC is increased.
Augmented water diffusion following cytotoxic tumor eradication
by anti-cancer drugs, gene or radiation therapy, both in experimental
tumors and human brain malignancies, is a universal phenomenon.
This is regarded as a key observation for expanded applications
of diffusion MRI/MRS in clinical oncology, and it may be that
this will grow into a major clinical application of this technique
in the near future. Surgical removal of malignancies is the cornerstone
of solid tumor treatment, often complemented by radiation and/or
drug therapies. A challenge for MR-techniques is to differentiate
between cancer recurrence and necrosis/apoptosis, a very significant
issue for clinical decision-making and patient management. Although
this can be studied by diffusion MRI, the change in water diffusion
in dying tumors, even when preceding tumor volume reduction, is
a rather late event. In fact, the biochemical data show that earlier
steps of ongoing apoptotic cell death become amenable to study
by 1H-MRS (Diklic, 1996, 1997; Hakumäki et al., 1999).
MOLECULAR IMAGING
Molecular imaging is a new radiologic
technique, according to its approach and goals, that allows in
vivo tumor visualisation at the molecular and genetic level. The
aim of this new discipline is to monitor all aspects of gene therapy,
from gene delivery to gene expression. Gene therapy is a term
that broadly defines different manipulations of genetic information
for therapeutic purposes. Individual components of gene therapy
have included the introduction of marker genes, the replacement
of defective genes, or the insertion of new transgenes for therapeutic
enzyme/protein production. More than 4,000 human diseases have
been classified as being genetic in origin and more than 250 gene
therapy trials are currently underway in the US.
Molecular imaging of gene delivery
Gene, i.e. a part of DNA molecule, is typically
delivered to target cells by one of three methods:
- enclosing it into a virus ("viral vector"),
- attaching it to a synthetic delivery system ("artificial vector"),
- by physical means such as electroporation ("gene gun").
Whereas synthetic gene delivery constructs are easily labeled,
tracking of viral particles has been more challenging. Most recently,
different techniques have been developed to label herpes simplex,
adenovirus and amplicons. These techniques rely on intraviral
or surface labeling with isotope containing chelates. For example,
Hakumäki et al (1998) used the BT4C rat glioma transfected with
a herpes simplex virus thymidine kinase (HSV-tk) gene, as a simple
model for gene therapy in vivo, to study the effect of gancyclovir
(GCV) treatment of experimental glioma in rats, by diffusion weighted
MRS. Recently, gene delivery was introduced in the therapy of
human brain tumors. Ram et al. (1997) used intratumoral implantation
of retroviral vector-producing cells in the therapy of malignant
brain tumor, and Sandmair et al (2000) applied thymidine kinase
gene therapy for human malignant glioma, using replication-deficient
retroviruses or adenoviruses.
Molecular imaging of gene expression
The topography and level of gene expression has traditionally
been measured by transgenic marker proteins that are normally
not found in mammalian bodies: green fluorescence protein, luciferase,
beta galactosidase, etc. Although the fluorochromic proteins are
detectable by optical imaging, such detection is usually limited
to surface structure and/or in vivo microscopy because of light
absorption and/or scattering. More recently, dedicated "imaging
marker genes" (IMG) have been developed for PET, planar and MR-imaging.
Two fundamentally different IMG strategies have been investigated:
- marker genes encoding intracellular enzymes,
- marker genes encoding cell surface bound receptors or other
ligands.
The first approach is based on the ability of certain enzymes
to modify imaging prodrugs so that tissue accumulation of such
drugs correlates with expression of the gene. One example of this
system is the HSV-tk that can be probed for with radiolabeled
small molecular weight substrates. Alternative system is the tyrosinase/melanin
system for MR-imaging.
The second approach utilizes cell surface expression of a receptor
or a ligand-binding protein which can then be probed for with
novel imaging tracers. The different nuclear and NMR imaging techniques
have been developed to study gene expression and drug delivery.
19F-MRS was effectively used for non-invasive monitoring the tumor-selective
drug activation by monoclonal antibody-cytosine deaminase conjugates.
CLINICAL APPLICATIONS: Gene directed
enzyme prodrug therapy (GDEPT) and neurotrophin transgenes
A major obstacle in the clinical management of cancer patients
is the limited differential toxicity of chemotherapeutic agents
and radiation towards neoplastic versus normal cells. Gene therapy
offers the possibility of overcoming this limitation. Gene directed
enzyme prodrug therapy (GDEPT) is a cancer treatment modality
designed to overcome the systemic toxicity of chemotherapy by
specifically expressing a foreign enzyme in toxic metabolite.
Stegman et al. (1998) have developed a novel GDEPT strategy based
on the production of reactive oxygen species with tumor cells.
This gene therapy approach uses the highly active D-amino acid
oxidase (DAAO) from the yeast Rhodotorula gracilis, which stereoselectively
deaminates D-alanine (D-Ala) generating stoichiometric amounts
of hydrogen peroxide (H2O2). Ectopic expression of DAAO in the
cytoplasm of 9L rat glioma cells, away from endogenous catalase
contained in the peroxisomal matrix allows production of cytotoxic
levels of H2O2 in the presence of D-Ala, which is scarce in mammals.
Expression of the DAAO transgene can be observed on 13C-MRS by
monitoring the metabolism of 13C-labeled substrates. Cells expressing
DAAO will deaminate D-Ala to pyruvate which can subsequently be
converted to lactate by lactate dehydrogenase (LDH). 13C-labeled
lactate is then measured by 13C-MRS. Additionally, H2O2 impairs
cellular energy metabolism by inhibiting glycolysis and activating
poly(ADP-ribose)polymerase resulting in a 31P-MRS-observable decrease
in ATP. In this way, both 13C-MRS and 31P-MRS can be used for
assessment of transgene expression and metabolic response to novel
oxidative stress-inducing cancer gene therapy.
In vivo gene therapy for irradiation or drug induced neuropathy
offers the possibility to deliver a therapeutic neurotrophin transgene
directly to the vulnerable cell population, so that the local
synthesis and continuous release of the neurotrophic factor may
protect those neurons from degeneration, while avoiding the undesirable
complications created by systemic delivery of high doses of the
peptide. Among the gene transfer vectors, herpes simplex virus
(HSV) is ideally suited for the delivery of genes to the peripheral
nervous system. Neurotrophic factors were identified originally
by the ability of these peptides to prevent the programmed cell
death of subpopulations of neurons during development. It is shown
that nerve growth factor (NGF) prevents cisplatin neuropathy (Apfel
et al., 1992); insulin-like growth factor-1 can prevent neuropathy
caused by administration of vincristine (Contreras et al., 1997),
and neurotrophin-3 (NT-3) can reverse neuropathy caused by administration
of cisplatin and prevent the neuropathy caused by pyridoxine overdose
(Helgren et al., 1997). The same paradigm may be used to prevent
a neuropathy caused by taxol. Although the therapeutic use of
these factors in human diseases has been limited by the short
serum half-life and dose-limiting side effects of these potent
peptides, many studies suggest that, even in the course of anticancer
drug-induced neuropathy, neurons that have not yet degenerated
might be rescued by gene delivery of the HSV vector-mediated neurotrophin.
CONCLUSION
The new knowledge of cancer biology that has become available
from molecular and cellular biology and genetic research, and
innovations in MRI/MRS technology, will profoundly affect our
understanding of the brain tumors, acting like a bridge between
different specialists, and connecting imaging of tumors with imaging
of drug-receptor sites inside the tumor.
REFERENCES
1. Apfel SC, Arezzo JC, Brownlee M et al. (1992): Nerve growth
factor prevents experimental cisplatin neuropathy. Ann. Neurol.
31:76-80.
2. Chenevert TL, Stegman LD, Taylor JM et al. (2000): Diffusion
MRI: an early surrogate marker of therapeutic efficacy in brain
tumors. J. Natl. Cancer Inst. 92: 2029-2036.
3. Contreras PC, Vaught JL, Gruner JA et al. (1997): Insulin-like
growth factor-1 prevents development of a vincristine neuropathy
in mice. Brain Res. 774: 20-26.
4. Damadian R (1971): Tumor detection in rat by nuclear magnetic
resonance. Science, 171: 1151-1153.
5. Diklic V, Vucurevic G. (1995): Magnetic resonance spectroscopy
of tumors - myth or reality? Arch. of Oncology, 3(4):23-26.
6. Diklic V, Vucurevic G, Petrovic B et al. (1996 a): Tissue characterization
of the ruptured intracranial dermoid cyst done by 1H-MR spectroscopy
in vivo. Arch. of Oncology, 4(4): 219-221.
7. Diklic V, Vucurevic G. (1996 b): Differentiation of recurrent
or residual brain tumor from radiation necrosis assessed by MR-spectroscopy.
Arch. of Oncology, 4(3): 143-146.
8. Diklic V, Vucurevic G. (1996 c): Diagnostic potential of 1H-MR
spectroscopy in differentiating between tumor regrowth and postradiation
necrosis. Magnetic Resonance Materials in Physics, Biology and
Medicine, 590: 246-247.
9. Diklic V, Baltic V (1996 d): The potential of MR-spectroscopy
in monitoring of the therapeutic effect in the central nervous
system tumors. Arch. of Oncology, 082: 56.
10. Diklic V (1997 a): Methylene/Choline index as a parameter
of choice in distinguishing cerebral metastases from primary CNS
tumors - preliminary results. Proc. Intl. Soc. Magn. Reson. Med.,5:
1130.
11. Diklic V (1997 b): Methylene/Choline index as a possible parameter
of choice in distinguishing cerebral metastases from primary CNS
tumors - preliminary results. Arch. of Oncology, 5(4): 195-196.
12. Diklic V (1997 c): Magnetno rezonantna spektroskopija in vivo
u neurologiji - naša prva iskustva. Klinicka i eksperimentalna
neurologija, 4: 203-208.
13. Diklic V, Prvulovic M, Baltic V (1997 d): MR spektroskopija
in vivo u klinickoj onkologiji: dometi i ogranicenja. Zbornik
radova Studenicke akademije "Dijagnosticki imidžing u onkologiji":
7.
14. Diklic V (1997 e): Magnetno rezonantna spektroskopija in vivo
u neurologiji: rezultati našeg trogodišnjeg iskustva. Aktuelnosti
iz neurologije, psihijatrije i granicnih podrucja, 5(3-4): 1-8.
15. Diklic V (2000 a): Magnetic resonance spectroscopy in vivo
in the investigations of brain tumors: the results of five years
experience. Annals of the Academy of Studenica, 3: 90.
16. Diklic V, Ivanovic V, Koprivšek K et al. (2000 b): Genetically
proven tumors of the central nervous system: MRI/1H-MR spectroscopic
characteristics. Annals of the Academy of Studenica, 3: 53.
17. Diklic V (2002): Magnetno rezonantna spektroskopija in vivo
u ispitivanju tumora mozga. U: "NMR u onkologiji" (Urednik: Baltic
V), izd. Medicinski fakultet u Novom Sadu, u štampi.
18. Hakumäki JM, Poptani H, Puumalainen AM et al. (1998): Quantitative
1H-NMR diffusion spectroscopy of BT4C rat glioma during thymidine
kinase-mediated gene therapy in vivo: identification of apoptotic
response. Cancer Res. 58: 3791-3799.
19. Hakumäki JM, Poptani H, Sandmair AM et al. (1999): 1H-MRS
detects polyunsaturated fatty acid accumulation during gene therapy
of glioma: implications for the in vivo detection of apoptosis.
Nat. Med. 5: 1323-1327.
20. Helgren ME, Cliffer KD, Torrento K et al. (1997): Neurotrophin-3
administration attenuates deficits of pyridoxine-induced large-fiber
sensory neuropathy. J. Neurosci., 17: 372-382.
21. Kahn T, Harth T, Kiwit JCW et al. (1998): In vivo MRI thermometry
using a phase-sensitive sequence: Preliminary experience during
MRI-guided laser induced interstitial thermotherapy of brain tumors.
JMRI, 8: 160-164.
22. Mardor Y, Roth Y, Lidar Z et al. (2001): Monitoring response
to convention enhanced taxol delivery in brain tumor patients
using diffusion weighted MRI. Proc. Intl. Soc. Magn. Reson. Med.,
9: 145.
23. Ram Z, Culver KW, Oshiro EM et al. (1997): Therapy of malignant
brain tumors by intratumoral implantation of retroviral vector-producing
cells. Nat. Med., 3: 1354-1361.
24. Sandmair AM, Loimas S, Puranen P et al. (2000): Thymidine
kinase gene therapy for human malignant glioma, using replication-deficient
retroviruses or adenoviruses. Hum. Gene Ther.,11: 2197-2205.
25. Stegman LD, Freyer JP, Neal ER et al. (1998): Assessment of
transgene expression and metabolic response to a novel oxidative
stress-inducing cancer gene therapy paradigm using 13C and 31P
MRS. Proc. Intl. Soc. Magn. Reson. Med.,6: 611.
26. Zhao M, Pipe JG, Bonnett J et al. (1996): Early detection
of treatment response by diffusion-weighted 1H-MRS in a murine
tumor in vivo. Br. J. Cancer, 73: 61-64.
|
|
