Bernard M. Hitzig

Tumor size dependent changes in a murine fibrosarcoma: Use of in vivo31P NMR for non-invasive evaluation of tumor metabolic status☆

Tumor tissue contains viable hypoxic regions that are radioresistant and often chemoresistant and may therefore be responsible for some treatment failures. A subject of general interest has been the development of non-invasive means of monitoring tissue oxygen. Pulse Fourier transform 31P NMR spectroscopy can be used to estimate intracellular nucleotide triphosphates (NTP), phosphocreatinine (PCr), inorganic phosphate (Pi) and pH. We have obtained 31P NMR spectra as an indirect estimate of tissue oxygen and metabolic status in a C3H mouse fibrosarcoma FSaII. Sequential spectra were studied during tumor growth in a cohort of animals and peak area ratios for several metabolites were computed digitally by computer. During growth, tumors showed a progressive loss of PCr with increasing Pi, and most tumors greater than 250 mm3 in volume had little or no measurable PCr. The smallest tumors (38 mm3 average volume) had PCr/Pi ratios of 1.03 +/- .24, whereas tumors 250 mm3 or more had an average PCr/Pi ratio of 0.15 +/- .04. Similarly derived NTP/Pi ratios decreased with tumor size, but this change was not significant (p = .17). Radiobiologic hypoxic cell fractions were estimated using the radiation dose required to control tumor in 50% of animals (TCD50) or by the lung colony technique. Tumors less than 100 mm3 had a hypoxic cell fraction of 4% (TCD50) while tumors 250 mm3 had a 40% hypoxic cell fraction (lung colony assay). These hypoxic fraction determinations correlated well with the depletion of PCr and decline in NTP/Pi ratios seen at 250 mm3 tumor volumes. Tumor spectral changes with acute ischemia were studied after ligation of the tumor bearing limb and were similar to changes seen with tumor growth. PCr was lost within 7 minutes, with concurrent increase in Pi and loss of NTP. Complete loss of all high energy phosphates occurred by 40 minutes of occlusion. In vivo tumor 31P NMR spectroscopy can be used to estimate tissue metabolic status and may be useful in non-invasive prediction of hypoxic cell fraction, reoxygenation, and radiation treatment response.

Effects of oxygen on the metabolism of murine tumors using in vivo phosphorus-31 NMR.

The effect of 100% inspired oxygen on in vivo tumor metabolism was examined using phosphorus-31 (31P) NMR spectroscopy. Isotransplants of two murine tumor histologies, designated MCaIV (C3H mammary adenocarcinoma) and FSaII (C3H fibrosarcoma), were used in syngeneic mice. Tumor volumes ranged from 30 to 1,800 mm3. Both tumor histologies are known to have a high hypoxic cell fraction when tumor volumes exceed 250 mm3. 31P nuclear magnetic resonance (NMR) spectra were obtained at 145.587 MHz, and the signal was detected using a 1.4 cm diameter, single loop coil designed to localize the signal from only the tumor. Spectral parameters for optimal signal-to-noise ratio (SNR) included a 60° pulse and a 2-second recycle delay. Tumors were implanted in the hindfoot dorsum to assure that all detected mobile phosphates were of tumor origin. Phosphocreatine/inorganic phosphate (PCr/Pi) ratios of large tumors (greater than 250 mm3) were reduced compared with small tumors (less than 250 mm3) of the same histology. The increased PCr/Pi response to 100% inspired oxygen was greater for large tumors and for tumors with lower baseline PCr/Pi ratios. When host animals were given 10% oxygen for respiration, there was an increase in Pi and a decrease in both PCr and ATP. The response to 10% oxygen was observed in both large and small tumors of both tumor histologies studied. Resting skeletal muscle exhibited no alteration in the NMR spectrum during either 100 or 10% oxygen breathing. We conclude that the fractional increase in PCr/Pi ratio that occurs after 100% oxygen breathing is a sensitive, noninvasive method of detecting tumor hypoxia.

Estimation of tumor oxygenation and metabolic rate using 31P MRS: Correlation of longitudinal relaxation with tumor growth rate and dna synthesis

31P MRS longitudinal relaxation times (T1) were determined for C3H murine fibrosarcomas (FSaII), and mammary carcinomas (MCaIV). Tumors were implanted in the foot dorsum, and were 100-300 mm3 in volume. T1s were repeated after the animal was allowed to breathe 100% oxygen for 30 min and then again 36-48 hr following 30 Gy. The spectrum were obtained using an 8.5 T spectrometer with a 8 cm bore and a 1.4 cm single turn antenna coil. The 31P relaxation times for untreated tumors in air breathing animals were: 3.78 sec for phosphomonoesters, 4.37 sec for inorganic phosphate (Pi), 2.73 sec for phosphocreatine, 1.37 sec for gamma ATP, 1.14 sec for alpha ATP, and 1.18 sec for beta ATP. The Pi T1s were 4.37 and 4.70 sec in control and irradiated tumors in air breathing animals. Respiration of oxygen for 30 min reduced the T1s to 3.02 and 2.62 sec in control and irradiated tumors respectively. The Pi T1 of an anoxic tumor, determined on an in situ tumor 60 min after death was 5.93 sec. The oxygen breathing induced decrease in the T1 of Pi is unlikely to have been caused by the paramagnetic properties of oxygen alone, and suggests a component of increased magnetization transfer secondary to the ATPase reaction. Oxygen breathing following 30 Gy, resulted in a decreased growth time (800 mm3 endpoint) and an increased proportion of cells in S-phase. These results support the hypothesis that the decrease in Pi T1 measured with oxygen breathing is a measure of tumor oxygen tension and metabolic rate, and suggests that T1 measurement may indirectly predict tumor growth rate and DNA synthesis.

Effects of hydralazine on in vivo tumor energy metabolism, hematopoietic radiation sensitivity, and cardiovascular parameters☆

Energy metabolism of murine FSaII foot tumors was studied by in vivo 31P-MRS in C3Hf/Sed mice. Spectroscopy was performed following exposure to escalating doses of hydralazine (HYD) ip. At 0.25 mg/kg, HYD caused a 20% increase in PCr/Pi and had no significant effect on mean arterial blood pressure. HYD doses greater than or equal to 2 mg/kg lead to hypotension which was associated with a decrease in PCr, NTP, pH, and an increase in Pi (p less than 0.01 for control vs 10 mg/kg HYD). When mice were given ip injections of HYD (0.25, 1, 2 and 10 mg/kg) 10 min prior to whole body irradiation, spleen stem cell survival after 6 Gy was increased (2.19 colonies in control animals vs 6.74 colonies per spleen in animals treated with greater than or equal to 2 mg/kg HYD), as was the LD50/30 dose (6.49 Gy [control] vs 9.00 Gy [10 mg/kg HYD]). The data provide evidence that PCr/Pi is a useful indicator of perfusion efficiency (and indirectly of hypoxic cell fraction) in FSaII tumors. These observations suggest that HYD may be a useful adjuvant for hyperthermic treatment of tumors and for potentiation of agents specifically toxic to hypoxic or nutrient-deprived cancer cells. HYD should be used with care in patients receiving radiation treatments or other therapies for which hypoxia can unfavorably affect treatment outcome.

Viability of the neonatal rat isolated brainstem preparation by 31P MRS

The isolated brainstem-spinal axis from the neonatal rat is an established model for studying neuronal responses of the ventilatory control system, however, its viability has not been clearly established. We studied the brainstem-spinal axis from newborn rats at 8.5 T with 31P NMR spectroscopy. The relative pattern of high energy phosphates (HEPs) was similar to that reported for the in vivo neonatal brain. The average pHi was 0.2 to 0.4 units less than the pHi for the in vivo neonatal brain. The HEPs and pHi were stable for 6 h, suggesting extended in vitro viability.

Unknown phosphate compounds in tail muscle of intact conscious newts by 31P NMR.

Unknown phosphate resonances at 0 and -21.6 ppm have been identified in 31P NMR spectra of tail muscle of unanesthetized newts which do not correspond to known phosphate-bearing compounds in skeletal muscle cells. The concentrations of both unknowns decrease markedly during muscular activity and severe hypoxia (conditions associated with decreased intracellular pH and increased cellular levels of inorganic phosphate). The unknown at 0 ppm increases in concentration with imposition of moderate hypoxia. Our data suggest that these unknowns may be liable storage compounds for a high energy phosphate bond, and are involved in newt skeletal muscle phosphogen metabolism.

Phosphorus and fluorine NMR examination of the anesthetized NEWT (notopthalamus viridescens)

We have examined newts by 19F-NMR using the anesthetic halothane as a probe and in another set of experiments taken 31P-NMR spectra under similar conditions. The spectra were recorded from the animals tail. The water soluble 31P-NMR signals point to little difference between anesthetized and unanesthetized newts except for the potential disappearance of two pools of inorganic phosphate in the anesthetized animals. The 19F spectra show two anesthetic populations in the tail which the phosphorus spectra suggest arise from populations of halothane in muscle and in lipid.