Beruflich Dokumente
Kultur Dokumente
Onkologie 2004;27:304–309
DOI: 10.1159/000077983
Summary Zusammenfassung
The combination of magnetic resonance spectroscopy Die Verknüpfung von Magnetresonanzspektroskopie
(MRS) and imaging (MRI) has led to mapping metabo- (MRS) und -tomographie (MRT) bietet die Möglichkeit,
lites from normal and neoplastic tissue within the time den Stoffwechsel gesunden und neoplastischen Gewe-
limits of a routine study. MRSI (magnetic resonance bes im Rahmen einer einzigen Untersuchung bildlich
spectroscopy imaging) detects metabolites that contain darzustellen. Das Kombinationsverfahren erfasst Stoff-
protons, phosphorus, fluorine, or other nuclei. The wechselprodukte, die Wasserstoff, Phosphor, Fluor oder
uniqueness of the information available in vivo and in a andere resonanzfähige Kerne enthalten. Die Aussicht,
non-invasive manner encouraged radiologists and oncol- derartige Informationen in vivo und nicht-invasiv zu ge-
ogists to apply MRSI in research and clinical practice. winnen, macht die spektroskopische Bildgebung auch für
Both 1H- and 31P-MRS have revealed significant distur- die onkologische Forschung und Praxis zu einer wertvol-
bances in amino acids, lipids, and phosphorus-contain- len Technik. Sowohl die Protonen- als auch die Phos-
ing metabolites within tumors. Phosphocreatine is often phor-Spektroskopie haben bedeutende Veränderungen
diminished in neoplasms compared to their primary host im Aminosäuren-, Lipid- und Phosphorstoffwechsel von
or surrounding tissues. However, the reduction of the Tumoren aufgedeckt. Kreatinphosphat liegt in Neoplas-
compound does not appear to be closely correlated to men häufig in geringerer Konzentration vor als im ge-
the degree of malignancy. Moreover, abnormalities in 31P sunden Parenchym. Eine enge Korrelation zum Grad der
spectra from neoplasms are shared by other disorders. Malignität des Tumors fehlt allerdings. Außerdem wer-
Changes in high-energy phosphate levels almost invari- den ähnliche Normabweichungen im Phosphorstoff-
ably occur with radio- and chemotherapy of tumors. The wechsel auch bei nicht-neoplastischen Erkrankungen be-
spectroscopic alterations are often seen before any vari- obachtet. Radio- und Chemotherapie führen nahezu obli-
ations in tumor size and shape can be detected. Howev- gat zu Änderungen im Metabolismus der energiereichen
er, opposite responses can be associated with the same Phosphate. Die Verschiebungen im Spektrogramm wer-
clinical outcome. 1H-MRS has been successfully used to den oft bereits sichtbar, bevor sich Form und Größe des
quantify the extent of neuronal cell loss imposed on the Tumors ändern. Allerdings kann das klinische Ergebnis
brain during radiotherapy. Recently, MRSI was success- trotz divergierender Trends im Resonanzbild identisch
fully integrated into radiotherapy planning in prostate sein. Die Protonen-Spektroskopie eignet sich zur Quanti-
cancer patients. 19F-MRS opens access to artificially in- fizierung des Verlusts von Nervenzellen während und
duced fluorocompounds such as 5-fluorouracil and its nach Strahlentherapie des Gehirns. Jüngst hat man das
metabolites. Verfahren erfolgreich in die Planung der Strahlentherapie
des Prostatakarzinoms einbezogen. Die Fluor-Spektros-
kopie schafft einen nicht-invasiven Zugang zur Beobach-
tung des Metabolismus fluorhaltiger zytotoxischer Sub-
Siriraj Medical Library, Mahidol University
Magnetic resonance spectroscopy (MRS) offers unique possi- Compound Peak position, Change with
ppm malignancies (organ)
bilities for the non-invasive evaluation of biochemical process-
es in vitro and in vivo. In vitro MRS of body fluids, biopsy Creatine +4.0, +3.1 ↓ (brain)
samples and surgical specimens provides metabolic informa- Myo-inositol +3.6 ↑ (prostate)
tion when other sensitive and specific disease markers are Choline +3.3 ↑ (brain)
lacking. In vivo MRS may detect biochemical alterations in Citrate + 2.6 ↓ (prostate)
Glutamine, glutamate +2.3 ↑ (brain)
the absence of morphologic lesions. MRS obtains signals from
N-acetyl-aspartate +2.0 ↓ (brain)
a wide variety of molecules which are present in tissue at Lactate +1.2 ↑ (brain)
much lower concentrations than water and fat. Each chemical- Lipids +1.0 ↑ (brain)
ly distinct nucleus of a given species resonates at a slightly dif-
ferent frequency, resulting in different MR peaks. The crucial
point is the spatial resolution obtainable by MRS which is in-
ferior compared to that usually achieved by MRI [1]. Table 2. Tissue metabolites identified by 31P-MRS [37]
↔↔ ↔
–3.0 (brain)
higher field strength is superior because of improved sensitivi- α−ATP –8.0 (brain)
ty and chemical shift dispersion, other factors like relaxation β-ATP –17.0 (brain)
time differences, line broadening due to magnetic susceptibili-
ty effects, and radiofrequency coil efficiency potentially affect
the expected improvement. Barker et al. [2] report a 20% im-
provement in sensitivity and a 28% increase in signal-to-noise pled spin. In proton-proton coupling, the effect is advanta-
ratio at 3.0 T compared to 1.5 T. An exceptionally high field geous in aiding the identification of particular species.
strength (4.0 T) may overcome current limitations of quantifi- For deep organs some effort is required to localize the spectra
cation of human proton spectra [3]. Maximal homogeneity of to a specific region of the target organ. The respective tech-
the magnetic field is another absolute prerequisite to satisfac- niques fall into several groups [4]. Those that use pulsed spa-
tory MRS results. tial gradients at audio frequencies (e.g. DRESS, ISIS) have
Chemical shift (CS) is defined as the frequency shift relative proven especially valuable. To acquire long and short TE
to some reference chemical, divided by the resonance fre- (echo time) spectra within the sampling volume, point-re-
quency of the reference chemical. Nuclei in different chemical solved spectroscopy (PRESS) and stimulated-echo acquisition
environments experience different fields and hence produce mode (STEAM) technique are applied.
1H-MRS is used to visualize amino acids, lactate, and lipids.
signals at different frequencies. If one nucleus is more shield-
ed than another, its signal will be shifted to low frequency. By The main problem with 1H-MRS is the large signal from water
convention, it will have a more negative chemical shift and in tissue. The metabolites are quantified by using the target
will appear further towards the right-hand side of the spec- organ water concentration as internal reference. The concen-
trum. The amount of the shift is specified in parts per million tration of a metabolite detected is proportional to the area
(ppm) of the resonance frequency relative to a standard. under the peak and not just the peak height. So a tall, narrow
Another important parameter arising from interactions with peak does not necessarily reflect a higher concentration than a
the bonding electrons is the spin-spin or J coupling. J coupling short, broad peak.
1H-MRS of the brain can detect a number of amino acids as
is an interaction between nuclear spins within a molecule and
provides a direct spectral manifestation of the chemical bond. well as choline, phosphocreatine (PCr), creatine (Cr), and lac-
The effect is transmitted via the electrons in orbit around the tate (table 1). N-acetyl-aspartate (NAA), the marker molecule
nuclei and two spins have measurable J coupling only if they for normally functioning neurons, forms the highest peak. In
are linked together through a small number of chemical disease, signals from lactate and mobile lipids may frequently
bonds, including hydrogen bonds. The behavior of an individ- be observed. The metabolites that are typically seen in 31P-
ual spin is influenced by other coupled nuclei. Thus, what MRS are phosphomonoesters (PME), inorganic phosphate
would otherwise be a single resonance, is split into two or (Pi), phosphodiesters (PDE), PCr and ATP (table 2). Pi is
more distinct frequencies, depending on the state of the cou- composed of mono- and divalent phosphate with a pK of 6.75.
Siriraj Medical Library, Mahidol University
198.143.39.97 - 1/26/2016 7:07:44 PM
MRS, the chemical composition of tumors can be analyzed tral differences between normal and neoplastic tissue in vivo
and compared with the biochemical characteristics of adjacent [16]. Spectra from women with cervical cancer reveal strong
tissues. Cancers display typical metabolic alterations and an resonances for choline. In healthy volunteers the spectra are
intracellular pH more alkaline than the extracellular one [6]. dominated by resonances corresponding to creatine and lipids
1H- and 31P-MRS are of value in observing metabolic path- without detection of a choline peak.
ways of both neoplasms and intact parenchyma [7].
Prostate
Brain 1H-MRS has also proven useful for the non-invasive discrimi-
MRS of the brain is a protean tool. Early studies held out nation between benign prostatic hyperplasia and prostate car-
hope that analysis of metabolite composition might improve cinoma [17]. Apart from the typical citrate, creatine, and
differential diagnosis of intracranial lesions. Recent findings choline signals of prostate spectra, both entities showed a
have cast doubt on the ability of 1H-MRS to play such a role peak centered at 3.6 ppm that was assigned to myo-inositol.
[8]. Nevertheless, MRS may contribute to the characterization The intensity of this compound was found significantly in-
of brain neoplasms, as tumor spectra differ considerably from creased in carcinoma compared to hyperplasia.
those obtained in intact brain parenchyma.
Negendank et al. [9] and Kaminogo et al. [10] applied in vivo Bone and Soft Tissue
1H-MRS to characterize brain tumors of the glial type in Tumors can be distinguished from normal muscle and soft tis-
adults. NAA signal intensities were decreased compared to sue by examining the low-energy phosphate resonance peaks
brain, regardless of the grade of malignancy. Choline signal of the spectrum and inorganic phosphate. Malignant lesions
was highest in astrocytomas, creatine signal was lowest in are distinguishable from benign ones on the basis of higher
glioblastomas. New and recidivant astrocytic tumors had iden- mean peak ratios of PME and PDE to β-nucleoside triphos-
tical mean NAA/creatine and choline/creatine ratios. Mobile phates (NTP), a lower mean peak area ratio of PCr to NTP
lipids turned up in 41% of high-grade tumors with higher and a higher mean pH. Early treatment-induced changes in
mean amounts detectable in glioblastomas [9]. the PME peak predict sensitivity of a sarcoma to continued
Siriraj Medical Library, Mahidol University
198.143.39.97 - 1/26/2016 7:07:44 PM
References
1 Zakian KL, Koutcher JA, Ballon D, Hricak H, Ling 11 Ishimaru H, Morikawa M, Iwanaga S, Kaminogo 20 Prescott DM, Charles HC, Sostman HD, Dodge
CC: Developments in nuclear magnetic resonance M, Ochi M, Hayashi K: Differentiation between RK, Thrall DE, Page RL, Tucker JA, Harrelson
imaging and spectroscopy: Applications to radia- high-grade glioma and metastatic brain tumour LM, Leopold KA, Oleson JR: Therapy monitoring
tion oncology. Semin Radiat Oncol 2001;11:3–15. using single-voxel proton MR spectroscopy. Eur in human and canine soft tissue sarcomas using
2 Barker PB, Hearshen DO, Boska MD: Single-voxel Radiol 2001;11:1784–1791. magnetic resonance imaging and spectroscopy. Int J
proton MRS of the human brain at 1.5 T and 3.0 T. 12 Poptani H, Gupta RK, Roy R, Pandey R, Jain VK, Radiat Oncol Biol Phys 1994;28:415–423.
Magn Reson Med 2001;45:765–769. Chhabra DK: Characterization of intracranial mass 21 Walecki J, Sokol M, Pieniazek P, Maciejewski B,
3 Bartha R, Drost DJ, Menon RS, Williamson PC: lesions with in vivo proton MR spectroscopy. Tarnawski R, Krupska T, Wydmanski J, Brzezinski
Comparison of the quantification precision of AJNR Am J Neuroradiol 1995;16:1593–1603. J, Grieb P: Role of short TE 1H-MR spectroscopy
human short echo time (1)H spectroscopy at 1.5 13 Warren KE, Frank JA, Black JL, Hill RS, Duyn JH, in monitoring of postoperation irradiated patients.
and 4.0 Tesla. Magn Reson Med 2000;44:185–192. Aikin AA, Lewis BK, Adamson PC, Balis FM: Pro- Eur J Radiol 1999;30:154–161.
4 Bottomley PA: The rise of human in vivo NMR ton magnetic resonance spectroscopic imaging in 22 Tarnawski R, Sokol M, Pieniazek P, Maciejewski B,
spectroscopy. Solid State Nucl Magn Reson 1997;9: children with recurrent primary brain tumors. J Walecki J, Miszczyk L, Krupska T: 1H-MRS in vivo
29–40. Clin Oncol 2000;18:1020–1026. predicts the early treatment outcome of postopera-
5 Constantinidis I: MRS methodology. Adv Neurol 14 Prost RW, Mark LP, Rand SD, Kim TA, Haughton tive radiotherapy for malignant gliomas. Int J Radi-
2000;83:235–246. VM: Recent advances in MR spectroscopy expand at Oncol Biol Phys 2002;52:1271–1276.
6 Stubbs M, Veech RL, Griffiths JR: Tumor metabo- its applications in neurologic disease. WMJ 1997;96: 23 Schlemmer HP, Bachert P, Herfarth KK, Zuna I,
lism: The lessons of magnetic resonance spec- 41–45. Debus J, van Kaick G: Proton MR spectroscopic
troscopy. Adv Enzyme Regul 1995;35:101–115. 15 Merchant TE, Kasimos JN, Vroom T, de Bree E, evaluation of suspicious brain lesions after stereo-
7 Leclerc X, Huisman TA, Sorensen AG: The poten- Iwata JL, de Graaf PW, Glonek T: Malignant breast tactic radiotherapy. AJNR Am J Neuroradiol
tial of proton magnetic resonance spectroscopy tumor phospholipid profiles using (31)P magnetic 2001;22:1316–1324.
((1)H-MRS) in the diagnosis and managament of resonance. Cancer Lett 2002;176:159–167. 24 Usenius T, Usenius JP, Tenhunen M, Vainio P, Jo-
patients with brain tumours. Curr Opin Oncol 2002; 16 Allen JR, Prost RW, Griffith OW, Erickson SJ, Er- hannson R, Soimakallio S, Kauppinen R: Radia-
14:292–298. ickson BA: In vivo proton (H1) magnetic reso- tion-induced changes in human brain metabolites
8 Golder W: Wandel und Perspektiven der diagnos- nance spectroscopy for cervical carcinoma. Am J as studied by 1H nuclear magnetic resonance spec-
tischen Onkoradiologie. Onkologie 2002;25(Suppl Clin Oncol 2001;24:522–529. troscopy in vivo. Int J Radiat Oncol Biol Phys 1995;
1):48–53. 17 Garcia-Segura JM, Sanchez-Chapado M, Ibarburen 33:719–724.
9 Negendank WG, Sauter R, Brown TR, Evelhoch C, Viano J, Angulo JC, Gonzalez J, Rodriguez- 25 Movsas B, Li BS, Babb JS, Fowble BL, Nicolaou N,
JL, Falini A, Gotsis ED, Heerschap A, Kamada K, Vallejo JM: In vivo proton resonance spectroscopy Gonen O: Quantifying radiation therapy-induced
Lee BCP, Mengeot MM, Moser E, Padavic-Shaller of diseased prostate: Spectroscopic features of brain injury with whole-brain proton MR spec-
KA, Sanders JA, Spraggins TA, Stillman AE, Ter- malignant versus benign pathology. Magn Reson troscopy: Initial observations. Radiology 2001;221:
wey B, Vogl TJ, Wicklow K, Zimmermann RA: Imaging 1999;17:755–765. 327–331.
Proton magnetic resonance spectroscopy in pa- 18 Negendank WG: MR spectroscopy of muscu- 26 Pirzkall A, McKnight TR, Graves EE, Carol MP,
tients with glial tumours: A multicenter study. J loskeletal soft-tissue tumors. Magn Reson Imaging Sneed PK, Wara WW, Nelson SJ, Verhey LJ, Lar-
Neurosurg 1996;84:449–458. Clin N Am 1995;3:713–725. son DA: MR-spectroscopy guided target delin-
10 Kaminogo M, Ishimaru H, Morikawa M, Ochi M, 19 Sostman HD, Prescott DM, Dewhirst MW, Dodge eation for high-grade gliomas. Int J Radiat Oncol
Ushijima R, Tani M, Matsuo Y, Kawakubo J, Shiba- RK, Thrall DE, Page RL, Tucker JA, Harrelson Biol Phys 2001;50:915–928.
ta S: Diagnostic potential of short echo time MR JM, Reece G, Leopold KA: MR imaging and spec- 27 Kristjansen PE, Pedersen EJ, Quistorff B, Elling F,
spectroscopy of gliomas with single-voxel and troscopy for prognostic evaluation in soft-tissue Spang-Thomsen M: Early effects of radiotherapy in
point-resolved spatially localised proton spec- sarcomas. Radiology 1994;190:269–275. small cell lung cancer xenografts monitored by 31P
troscopy of brain. Neuroradiology 2001;43:353–363. magnetic resonance spectroscopy and biochemical
analysis. Cancer Res 1990;50:4880–4884.
Siriraj Medical Library, Mahidol University
198.143.39.97 - 1/26/2016 7:07:44 PM