In vivo relaxation time measurements on a murine tumor model—Prolongation of T1 after photodynamic therapy

Magnetic

Pergamon

Resonance Imaging, Vol. 13, No. 2, pp. 251-258, 1995 Copyright 0 1995 Elsevier Science Ltd Printed in the USA. All rights reserved 0730-725X195 $9.50 + .OO

0730-725x(94)00107-3

l

Original Contribution IN VIVO RELAXATION MODEL-PROLONGATION

TIME

MEASUREMENTS ON A MURINE TUMOR OF Tl AFTER PHOTODYNAMIC THERAPY

Y.H. Lru,* R.M. HAWK,* AND S. RAMAPRASAD~ *Department of Electronics and Instrumentation, University of Arkansas at Little Rock, Little Rock, AR 72204, USA and -iDepartments of Radiology and Pathology, University of Arkansas for Medical Sciences, Little Rock, AR 72205, USA RIF tumors implanted on mice feet were investigated for changes in relaxation times (Tt and T2) after photodynamic therapy (PDT). Photodynamic therapy was performed using Photofrin II as the photosensitizer and laser light at 630 nm. A home-built proton solenoid coil in the balanced configuration was used to accommodate the tumors, and the relaxation times were measured before, immediately after, and up to several hours after therapy. Several control experiments were performed using the untreated tumors, tumors treated with Photofrin II alone, or tumors treated with laser light alone. Significant increases in Tts of water protons were observed after PDT treatment. In all experiments, 31P spectra were recorded before and after the therapy to study the tumor status and to confirm the onset of PDT. These studies show significant prolongation of T,s after the PDT treatment. The spin-spin relaxation measurements, on the other hand, did not show such prolongation in T2values after PDT treatment. Keywords:

Photodynamic therapy; Murine tumor model; Relaxation time: Solenoid coil.

ies have been discussed in the literature.3-6 Two nuclear species were used to investigate PDT effects in this study, namely, phosphorus and hydrogen. We discuss the results of our studies on the changes associated with the water relaxation times following PDT. Our studies focus mainly on the confirmation of the onset of PDT by in vivo 3’P NMR followed by relaxation measurements to assess the short-term effects (a few hours to about a day) on the water proton relaxation times following PDT. The application of magnetic resonance imaging (MRI) or magnetic resonance spectroscopy (MRS) as a noninvasive technique to predict or monitor therapeutic responses is particularly attractive. However, in vivo MR studies related to water relaxation times in tumors after PDT treatment are limited and have been of a preliminary nature using the image-based evaluation of relaxation times.1*2 ‘H imaging studies of PDTdamaged tissue have been performed in murine mammary tumors.’ These studies showed an increase in T, and T2 1 day after the PDT treatment. Not only were the Tl and T2 values higher in the necrotic zone, but

INTRODUCTION Photodynamic therapy (PDT) is a new cancer treatment modality in which a combination of visible light and photosensitizing drugs produces locally cytotoxic chemical moieties. In vivo, PDT is thought to lead to vascular collapse and massive ischemic necrosis of malignant cells. Most predictive assays for tumor treatment response are invasive procedures that require biopsies or surgical specimens. In the current stage of development of PDT, noninvasive techniques to predict PDT responses and to follow the therapy over time are of great value. Another important consideration is that MR spectroscopic and imaging studies may provide information from areas not accessible for biopsy. The PDT-induced damage in the tumors has been visualized in proton images and an increase in T, value has been suggested for PDT-affected tissues.‘,* The purpose of our studies was to observe the changes detectable by NMR that occur in tumor models following PDT. The early response to PDT can be assessed by performing in vivo 31P NMR studies and such studRECENED

2/10/93;

ACCEPTED

9126194.

Address correspondence to S. Ramaprasad. 251

252

Magnetic Resonance Imaging 0 Volume 13, Number 2, 1995

there was an additional image contrast relative to the viable zone. The high image contrast was attributed in part to an increase in water content in the necrotic zone. Similarly, MRI studies to evaluate PDT effects on normal rat brain showed a significant increase in T, and T2 in the lesion at 1 day and 3 days post-PDT.2 As the proton density did not show significant increase until 3 days after PDT, the authors did not attribute Tl and T2 changes to net increase in water content. There are no detailed spectroscopic studies in the current literature which evaluate the relaxation times after the PDT treatment. To evaluate the potential clinical utility of ‘H NMR spectroscopy in PDT, we have performed preliminary relaxation time studies on the RIF-1 tumor grown on the dorsal side of mice feet. In this paper, we discuss our results on in vivo relaxation time studies of several tumors in a murine tumor model undergoing PDT. We made several relaxation measurements on mice tumors under various control conditions to provide a basis for discussing the PDT-induced changes. We chose the foot tumor as a model because the relaxation parameters can be studied using simple inversion recovery techniques, without resorting to complicated localization techniques. A solenoid coil connected in a balanced configuration7 allowed us to study medium-sized murine tumors. The purpose of this study was to measure tumor water proton relaxation times with the improved isolation of tumor from the normal tissue and the homogeneous B, field of the solenoid configuration. In addition, the study was undertaken to explore the possibility of using water proton relaxation times in monitoring tissue response to PDT. In vivo Tls and T2s were measured on mice feet tumors both before and after PDT. Reliable noninvasive methods for prediction of tumor resistance or response to therapy are needed to individualize and optimize clinical PDT. We chose the RIF tumor as a model for our studies. The RIF-1 tumor in C3H mice8,9 is suitable for both in vivo and in vitro studies, and there is a substantial historic database for the RIF tumor. The RIF tumor grown on the mouse foot is a model that satisfies the criteria for an isolated tumor that can be effectively studied by simple in vivo NMR techniques. RIF tumors have been used successfully to study changes induced by radiotherapy” and the effect of PDT on tumor metabolism and blood flow by 3’P and 2H NMR.” METHODS

light throughout the course of the experiments. Care was taken to prevent stray light entering the animal cages by covering them with dark brown paper. Fresh RIF cells were injected into the flank of male mice. Once the tumors grew to a viable size, they were used as donor tissue for further transplantation. Tumors of about 200 mm3 were dissected under sterile conditions into small cubes of about 1 mm on each side and implanted in the feet of mice. Within lo-15 days, the tumors grew to a size that was suitable for PDT and in vivo NMR studies. The desired size of tumors is one which is sufficiently large to produce sufficient signal strength in 31P NMR spectra, but not too large as to make it a physiological burden on the animal. It has been suggested12 in the literature that the tumor size be kept at or around 1Vo of the body weight and it has been further recommended that it not exceed 10%. From the point of view of PDT studies, the tumor should not be too large as laser penetration deep into the tissue becomes less efficient. From a combined viewpoint, we suggest that the tumor volumes be around 300 mm3. The foot dorsum was chosen to reduce spectral contamination below detectability from nearby tissues, specifically the muscle. 8,9This is further demonstrated by recording the 3’P spectra of the mouse foot tumor and the contralateral tumor-free foot as a control spectrum. These spectra are shown in Figs. 1A and 1B. The resonance assignments have been made based on their chemical shifts.10-‘2 An inspection of the spectra in Figs. 1A and 1B indicate that the various sharp resonances are due entirely to tumor tissue, with minimal contribution from the muscle. We emphasize that although this model significantly reduces the contamination from the surrounding muscle tissue, it does not completely eliminate it. The advantage of this tumor model over other models is that it is less prone to produce contamination from the underlying muscle, as compared with a tumor on the flank. The tumor dimensions in the three mutually orthogonal directions were measured using calipers both before and at several time points after the PDT. The tumor volumes were calculated with the ellipsoid approximation using formula V = d, x d2 x d3 x u/6, where d, , d2, and d3 are the diameters along the three mutually orthogonal directions of the ellipsoid. Tumor volumes that were studied here can be grouped into two categories: one in the range of 180 + 50 mm3 and the other around 430 + 10 mm3.

AND MATERIALS

Animal Model Male C3H mice, three to four weeks of age, were used in all the experiments. Food and water were provided ad lib. The animals were maintained in subdued

Laser and Delivery System A dye laser pumped by a 20 W Argon laser (Model CR 599, Coherent Radiation, Palo Alto, California) was used to generate 630 nm laser light. The output was coupled to a quartz optical fiber 2 m in length and

Water T, prolongation after PDT 6

39

Fig. 1. (A) 31PNMR spectrum of a mouse foot tumor. The peak assignments are as follows: (1) /3 phosphates of NTPs (/3-NTP); (2) a! phosphates of NTPs (CX-NTP); (3) y phosphates of NTPs (-y-NTP); (4) phosphocreatine (PCr); (5) Phosphodiesters (PDEs); (6) inorganic phosphate (Pi); and (7) phosphomonoesters (PMEs). (B) 3*P NMR spectrum from a contralateral tumor free foot. The broad hump is from the immobile phosphates of the bone mass.

400 km in diameter. The power from the optical fiber was measured by a calibrated power meter (Coherent, Model 2100) that has a flat response between 400900 nm. The power density was maintained at 150 mW/cm2 and the total light energy fluence was maintained at 100 Joules/cm2. Animals Both the PDT experiments and in vivo NMR experiments were performed under mild anesthesia. Typically for in vivo experiments, mice were anesthetized using Rompun (20 mg/kg) and Ketamine (100 mg/kg) . During the time of data collection, the mice were kept warm by continuously blowing room air through the magnet bore. PDT Experiments Photofrin (PFII) was obtained from Photomedica Inc. (Raritan, New Jersey), and kept in the dark and stored frozen. Prior to use, the compound was thawed and the solution prepared in distilled water in a brown bottle. The solution was then injected IP. The mice were

0 Y.H.

LIU ET AL.

253

secured to a 2.5cm thick wooden slab and positioned around the periphery of a circular radiation field with the tumor-bearing leg extended to the field of irradiation. Tumor irradiation was carried out in vivo using laser light at 630 nm while the mice were under anesthesia. The tumor was irradiated for the same amount of time on either side of the tumor along the dorsoventral direction with 630 nm laser light. The controls and PDT-treated tumors studied can be classified into the following groups: (1) Untreated tumors (n = 2) which were neither treated with Photofrin II nor with the laser light; (2) laser-alone controls (n = 2), which were treated with the same amount of light used in PDT treatments, but no Photofrin II was injected prior to laser treatment; (3) drug-alone controls (n = 2), which were given the same amount of Photofrin II as in all PDT treated cases; and (4) PDT-treated tumors (n = 5). All controls were administered the same amount of anesthesia as the PDTtreated group. The tumors were monitored over the entire course of therapy. When the mice were not healthy and could not be used for further experiments under anesthesia, they were humanely sacrificed. NMR Experiments ‘H NMR spectroscopy. For these experiments, mice were injected with 25 mg/kg of Photofrin II. The home-built solenoid coil (4 turns, 2 cm length, 2 cm diameter) with tune and match capacitors was used. After a 24-h waiting period, the tumor was placed inside a solenoid coil, which was fixed on a plastic plate. The magnetic field homogeneity was optimized for each tumor by shimming on the water proton signal to a line width of -50 Hz. A typical ‘H spectrum is displayed in Fig. 2. The small peaks to the upfield of the water resonance arise from the fat and lipid in the tumor. Note that there is no further discrimination of water signals, because they all overlap. The important thing to be noticed is that this model, by its very nature, minimizes the contamination from other sources as compared with, for example, a flank tumor. T, values in the in vivo condition were measured by the inversion recovery (IR) method using 20-22 evolution points. Similarly, T2s were measured using the Hahn spin-echo technique with 10 evolution points. The total time involved in acquiring the NMR relaxation data was 5-10 min per tumor. For the Tl relaxation measurements, the evolution times ranged from 10 ps to 2 s and the time delay between successive scans (TR) was 5 s. Similarly, for the T2 relaxation times, measured using the Hahn spin-echo technique, the evolution times ranged from 1 to 200 ms and the TR value was 5 s. The free induction decays (FIDs) were subject

Magnetic Resonance Imaging 0 Volume 13, Number 2, 1995

254

ing at 81.10 MHz for the phosphorus nucleus. Typical ‘H line width of the water signal was in the range of 50 Hz.‘~ The 31P parameters were: spectral width of 10 KHz, pulse width of 13 ps, 1 s repetition time, 4 K data points, number of accumulations 900, and a total accumulation time of 15 min. 31P NMR spectra were collected in an identical manner both before and after the PDT experiments. The chemical shifts were referenced to phosphocreatine (PCr) peak at 0 ppm.

1 ’ ” 30

I I i ’ I ” 10

-10

Data Processing and Analysis A line broadening of 15-25 Hz was applied to increase the signal-to- noise ratio (SNR) of each spectrum. The 31P NMR spectra displayed in Fig. 3 were processed using the convolution difference technique” to remove the broad hump underlying the spectra. Line broadenings of 25 Hz and 1000 Hz were used for this purpose. Each set of T, and T2 spectra were analyzed and the data were found to fit to a single exponential decay, using the routine on the SIGMAPLOT program on a personal computer. Separate error estimates of T, and T2 values were done by averaging several consecutive measurements. The curve fitting programs used to fit T, and T2 data gave fitting errors in the range of l-5010 for T,s and about 1% for T2.

‘I -30

PPm Fig. 2. A typical ‘H NMR spectrum of the RIF tumor ona mouse foot. The small peaks that occur upfield of water resonance are from the lipids and fat.

to 20 Hz exponential line broadening prior to Fourier transformation. The typical acquisition parameters chosen for the ‘H relaxation time studies were: Spectrometer frequency of 200.1 MHz, 4 accumulations, pulse width of 11 ps, spectral width of 40 KHz, 4 K data points for FIDs, and a repetition time (TR) of 5 s. The T, and T2 experiments were performed on both the PDT-treated groups and the control groups. In some representative cases, the experiments were repeated several times to evaluate the reproducibihty of T, and T, measured in these tumors. The T, data were found to fit to a single exponential function. The T2 values were estimated from a single exponential fit to the Hahn spin-echo data. 3’P NA4R spectroscopy. In vivo 3’P NMR spectra of the tumors were obtained using the home- built phosphorus-31 coil (5 turns, 2 cm length, 1.5 cm diameter) on the 4.7 T GE Omega spectrometer operat-

Fig. 3. 3’P spectra of mouse foot tumor recorded before PDT is shown in top trace. Drastic reductions in NTP peaks

and a concomitant increase in Pi peaks that occur at the end of 2 h after PDT treatment is shownin the bottom trace. These results indicate successful initiation of PDT.

Water T, prolongation after PDT 0 Y.H. LIU ETAL.

in the foot RIF tumor is illustrated in Fig. 3. The NTP peaks showed drastic reduction in intensities and there is a concomitant increase in the Pi peak intensity. In general, the NTP peaks vanished almost completely beyond detection around 15-20 h after the PDT treatments. Following each 31P NMR study, T1 and T2 relaxation measurements of water protons were performed on each of the tumors. In Fig. 4, we provide a series of

RESULTS The 31P NMR recorded before and after PDT (see Fig. 3) indicate that the high energy phosphates of NTP decrease drastically after PDT, with a corresponding increase in Pi value. These are very similar to the earlier 31P studies on tumor models3-6 before and after PDT. Typical response to PDT measured by 31P NMR

I B

L

ii

.3b

-id

lb’

L

255

C

3.6

s

2.6

s

0.8

s

0.6

s

10

ms

I&

wm

Fig. 4. A series of inversion recovery spectra before (A) and after (B and C) PDT treatment. The relaxation delay indicated on the right- hand side of each spectra (series C) indicates a steady increase in TI values as PDT progresses.

256

Magnetic ResonanceImaging l Volume13,

inversion recovery spectra that were recorded on tumor #3 before and at two time points post-PDT. A visual inspection of the set of T1 spectra at various relaxation delay times suggest different recovery patterns of the Hz0 signal before and after the PDT treatment. The relaxation delay at which the water signal goes through a null gradually increases (C > B > A), and an evaluation of the relaxation times in a more quantitative manner further confirms these qualitative observations. This pattern was seen in all T, studies, without a single exception. This further confirmed that the changes observed are real, significant, and reproducible. In all the relaxation studies, no multiexponential signal was observed, although some earlier studies have suggested multiexponential decay in other tissues.20.21 The T, and T2 values were measured on a number of tumors with volumes varying from 150 to 440 mm3, and the results are recorded in Table 1. The T, s of tumors with volumes in the range of 180 + 50 mm3 (volumes for tumors #l, #2, and #3 were 239, 159, and 144 mm3, respectively) are comparatively smaller than tumors with volumes around 420 mm3 (volumes for tumors #4 and #5 the volumes were 440 and 419 mm3). This observation is consistent with T, values reported on tumors of varying tumor volumes.22-24 Interesting trends and results can be obtained by examining the results reported in Table 1. A close examination of the Tl and T2 data reported in Table 1 indicates that significant changes in T, s occur l-6 h after PDT. In all cases, the T, relaxation times were prolonged after PDT treatment, compared with before treatment. The increase in water proton Tl s are more pronounced, however, several hours after PDT. The T, values were always found to increase in all the tumors that were subject to PDT. It may be noticed that each tumor, by serving as its own control, enhances the confidence in measured changes in T, values. The possible effects on the water relaxation times of tumors while using drug alone, laser light alone, or those that were untreated were studied over

Number

2, 1995

time. In all the control experiments, the relaxation times were reproducible within l-2% of the respective mean values. Typical errors in the T, measurements of a single tumor were determined by performing several consecutive measurements, and the values were reproducible within 2%. The other sources of error in the measured relaxation times involves the curve fitting which, in our case, ranges from 1% to 5 % for Tl s and about 1% for T2s. The regression coefficients were in the range of 0.95-0.99. Analysis of PDT Effects in Individual Tumors The T,s measured post-PDT consistently showed significant increases when compared with the values measured before PDT. For tumor #1 , the T, increased by 6.5% after 0.5 h of PDT and by 16.5% after 20 h of PDT. Similarly, tumor #2 showed increases of 12.7% and 17.6% after 1.5 h and 20 h after PDT. Tumor #3 showed increases of 8.7% and 17.8% after 2 h and 30 h of PDT. Tumors 4 and 5 showed 40% and 33% increases in T,s after 4 h and 6 h of PDT. A maximum combined error of about 7% in T, s are estimated from our studies. All tumors except tumor #l (at 0.5 h postPDT) showed an increase in Tl that was greater than the maximum estimated error of 7%. Tumor #1 showed an increase of 6.5%, which was was marginally lower than the maximum estimated error. It may be of significance to mention here that the T,s measured for tumor #l were within 30 min of PDT treatment, where PDT effects are just beginning. It may be noted, however, that all the PDT-treated tumors always showed increases in T,s and were always found to be significantly higher than the estimated error limits. No such systematic increases were noticed in any of the control tumors. Based on these results, we can conclude that the PDT-induced changes in T,s are significant and measurable by ‘H NMR. The T, values for the tumors before and post-PDT are displayed in Figs. 5A and 5B, respectively. This response is very similar to those

Table 1. Relaxation timesof water protons in mice RIF tumors under PDT treatment T, values(ms) Mouse no. 1 2 3 5 4

Before 898 f 910 * 923 f 969 f 1127k

32 4 5 31 9

T2 values(ms)

After

Hours after PDT

956 f 48 1026f 11 1003f 14 1359f 2 1504f 3

0.5 1.5 2.0 4.0 6.0

Before 45 It 36 f 35 f 41 * 39 +

0.1 0.1 0.1 0.3 0.2

After 49 + 36 k 39 f 42 f 39 *

0.1 0.1 0.1 0.1 0.2

Hours after PDT 0.5 1.5 2.0 4.0 6.0

Water T, prolongation after PDT 0 Y.H. LIU

251

ET AL.

changes in T2s were observed, but they were compar-

atively smaller (mouse #l, #3, #5). The changes observed in T2s for mouse #2 and #4 were relatively small or none. CONCLUSIONS

800 1 0

’

’

I

5

10

15

Hours

I 20

I

I

J

2.5

30

35

After PDT Treatment

03) 1800

I

I

I

I

I

m = Mouse #4 0 = Mouse #5 1600

Our studies on this tumor model reduces the problem of contamination of signal from underlying skin that occurs in other models and discusses the PDT effects on water relaxation times. Our study, although of limited nature, clearly indicates that the changes are significant following PDT and they are detectable by ‘H NMR spectroscopy. Future studies will more fully examine the characteristics of tumors under various conditions. We believe that this presentation of Tl and T2 results will help future planning of the study protocol and in the assessment of response to PDT in tumors and cancerous tissues. In vivo NMR spectroscopy can offer a new approach to cancer therapy planning. The significant increases in water T,s reflect changes in either the tissue, plasma, interstitial, and extracellular water, or a combination thereof following PDT. These findings may have an important implication in the interpretation of ‘H MR imaging studies during

photodynamic therapy. The mechanisms underlying the prolonged T, relaxation process cannot be deduced from this study. Increased tumor water content is a strong possibility, but certainly not the only one. The dominant sources which contribute to T, and T2 relaxation, when identified, can form a further basis to describe biological or physiological changes which occur during therapeutic intervention. We should be able to address this issue

600

0 Hours

!

I

I

t

10

15

20

25

After PDT Treatment

Fig. 5. In vivo spin-latticerelaxation times (T,) of control and irradiated RIF tumorsimplantedon mousefoot at varioustimespost PDT treatment. The T1valuesat time = 0 are from the correspondingcontrolsbeforePDT treatment. The data pointswhich do not indicatethe error barsaretoo small to be representedon the current scale.For clarity, the data are shownin three parts: (A) data from mouse#l, #2, #3; and (B) data from mouse#4 and #5.

reported by Belfi et al.,” in which the tumors were subjected to radiotherapy instead of PDT. They also reported a parallel increase in tumor water content. Such increasesin water content of tumors is a strong possibility in our casealso, although experiments supporting this issue have not been performed. Some

through detailed Tl and T2 studies via localized MRS techniques. Acknowledgment- Wethank Janath D. McKee for assistance in the preparation of this manuscript.

REFERENCES 1. Dodd, N.J.F.; Moore, J.V.; Poppitt, D.G.; Wood, B. In vivo magneticresonanceimagingof the effects of photodynamic therapy. Br. J. Cancer 60:164-167; 1989. 2. Jiang, Q.; Knight, R.A.; Chopp, M.; Helpern, J.A.; Ordidge, R.J.; Qing, Z.X.; Hetzel, F.W. ‘H Magnetic resonanceimaging of normal brain tissueresponseto photodynamic therapy. Neurosurgery 29(4):538-543; 1991. 3. Ceckler,T.L.; Bryant, R.G.; Penney,D.P.;Gibson,S.L.; Hilf, R. 31PNMR spectroscopydemonstrates decreased ATP levelsin vivo asan early response to photodynamic therapy. Biochem.Biophys.Res.Commun.140:273-279; 1986. 4. Hilf, R.; Gibson, S.L.; Penney, D.P.; Ceckler, T.L.;

258

5.

6.

7. 8.

9.

10.

11.

12.

13.

14.

Magnetic ResonanceImaging 0 Volume 13, Number 2, 1995 Bryant, R.G. Early biochemical responses to photodynamic therapy monitored by NMR spectroscopy. Photochem. Photobiol. 46(5):809-817; 1987. Chopp, M.; Helpern, J.A.; Frinak, S.; Hetzel, F.W.; Ewing, J.R.; Welch, K.M.A. In vivo 31P NMR of photoactivated hematoporphyrin derivative in cat brain. Med. Phys. 12:256-258; 1985. Chopp, M.; Evelhoch, J.L.; McCoy, C.; Farmer, H.; Hetzel, F.W. ; Welch, K.M.A.; Schaap, A.P. In vivo 3 1-P NMR spectroscopy: A method to evaluate photodynamic therapy. 1st Intl. Conf. Clinical Applications of Photosensitization for Diagnosis and Treatment. Tokyo, Japan, p. 186, 1986. Murphy-Boesch, J.; Koretsky, A.P. An in vivo NMR probe circuit for improved sensitivity. J. Magn. Reson. 54:526-532; 1983. Brown, J.M.; Twentyman, P.R.; Zamvil, S.S. Response of the RIF-1 tumor in vitro and in C,H/Km mice to X-irradiation (cell survival, regrowth delay, and tumor control), chemotherapeutic agents, and activated macrophages. J. Natl. Cancer Inst. 64:605-611; 1980. Rofstad, E. J.; Howell, R.L.; DeMuth, P.; Ceckler, T.; Sutherland, R.M. 31P NMR spectroscopy in vivo of two murine tumor lines with widely different fractions of radiobiologically hypoxic cells. Int. J. Radiat. Biol. 54:635649; 1988. Rajan, S.S.; Wherle, J.P.; Li, S.J.; Steen, R.G.; Glickson, J.D. 31P NMR spectroscopic study of bioenergetic changes in radiation induced fibrosarcoma-I after radiation therapy. NMR Biomed. 2:165-171; 1989. Mattiello, J.; Evelhoch, J.L.; Brown, E.; Schaap, A.P.; Hetzel, F.W. Effect of photodynamic therapy on RIF-1 tumor metabolism and blood flow examined by 31Pand 2H NMR spectroscopy. NMR Biomed. 3:64-70; 1990. Denekamp, J. The choice of experimental models in cancer research: The key to ultimate success or failure? NMR Biomed. 5:234-237; 1992. Okunieff, P.G.; Koutcher, J.A.; Gerweck, L.; McFarland, E.; Hitzig, B.; Urano, M.; Brady, T.; Neuringer, I.; Suit, H.D. Tumor size dependent changes in a murine fibrosarcoma: Use of in vivo 31P NMR for noninvasive evaluation of tumor metabolic status. Int. J. Radiat. Oncol. Biol. Phys. 12:793-799; 1986. Okunieff, P.G.; McFarland, E.; Rummenney, E.;

15.

16.

17.

18.

19.

20.

21.

22.

23. 24.

Willett, C.; Hitzig, B.M.; Neuringer, L.; Suit, H.D. Effects of oxygen on metabolism of murine tumors using in vivo phosphorus-31 NMR. Am. J. Clin. Oncol. 10: 475-482; 1987. Evanochko, W.T.; Sakai, T.T.; Ng, T.C.; Krishna, R.; Kim, H.D.; Zeidler, R.B.; Ghanta, V.K.; Brockman, R.W.; Schiffer, L.M.; Braunschweiger, P.G.; Glickson, J.D. NMR study of in vivo RIF-1 tumors: Analysis of perchloric acid extracts and identification of ‘H, 31P, and 13C resonances. Biochim. Biophys. Acta 805: 104116; 1984. Hoult, D.I.; Busby, S.J.W.; Gadian, D.G.; Radda, G.K.; Richards, R.E.; Seeley, P.J. Observation of tissue metabolites using 31-P nuclear magnetic resonance. Nature 252:285-289; 1974. Navon, G.; Ogawa, S.; Shulman, R.G.; Yamane, T. 31P nuclear magnetic resonance studies of Ehrlich ascites tumor cells. Proc. Natl. Acad. Sci. USA 74:87-91; 1977. Ackerman, J.J.H.; Gadian, D.G.; Radda, G.K.; Wong, G.G. Observation of ‘H-NMR signals with receiver coils tuned for other nuclides. J. Magn. Reson. 42:498500; 1981. Campbell, I.D.; Dobson, C.M.; Williams, R.J.P.; Xavier, A.V. Resolution enhancement of protein PMR spectra using the difference between a broadened and a normal spectrum. J. Magn. Reson. 11:172-181; 1973. Hazelwood, C.F.; Chang, D.C.; Nichols, B.L. Nuclear magnetic resonance transverse relaxation times of water protons in skeletal muscle. Biophys. J. 14:585-606; 1974. Mardidni, I.A.; McCarter, R.J.M.; Fullerton, G.D. NMR relaxation times of skeletal muscle: Dependence on fiber and diet. Magn. Reson. Imaging 4:393-398; 1986. Belfi, C.A. ; Medendorp, S.V.; Ngo, F.Q.H. The response of the KHT sarcoma to radiotherapy as measured by water proton NMR relaxation times: Relationships with tumor volume and water content. Int. J. Radiat. Oncol. Biol. Phys. 20:497-507; 1991. Akber, S.F. A new perspective to assess relaxation time of tumor. Med. Hypoth. 37:24-26; 1992. Kroker, R.M.; Mcveigh, V.R.; Hardy, P.; Bronskill, M.J.; Henkel R.M. In vivo measurements of NMR relaxation times. Magn. Reson. Med. 2:1-13; 1985.