In Vivo Manganese–MR Imaging of Calcium Influx in Spontaneous Rat Pituitary Adenoma

Subjects

A total of 16 rats (14 male, 2–22 months of age; Sprague Dawley, Charles River Laboratories, Wilmington, Mass) were used for the in vivo imaging experiments. Originally obtained as part of an aging study, 4 rats (2 male, 18–22 months of age, “aged breeders”) had spontaneous pituitary tumors that were discovered at the time of initial MR imaging. Four aged rats (15–18 months) with normal pituitary glands were scanned as a comparison. In addition, 8 young rats were scanned in a calcium channel modulation and blocking experiment. All subjects were kept on a normal 12-hour light/dark cycle and had free access to food and water. All procedures were approved by the University of Washington Animal Care Committee.

Injections and MR imaging

Under isoflurane gas anesthesia, rats received a baseline MR imaging before administration of MnCl₂. After the prescanning, in the part of the study involving Mn²⁺ uptake modulation via the calcium channel blocker, young healthy rats (2 months of age, n = 8) received one of the following procedures: 1) high-dose verapamil: (3 μL, 5 mg/2 mL) injection into the right nasal cavity, followed after 20 minutes by 8- to 10-μL injection of 1-mol/L MnCl₂; 2) medium-dose verapamil: (3 μL, 5 mg/8 mL) injection into right nasal cavity followed after 1 hour by 8- to 10-μL injection of MnCl₂; or 3) low-dose verapamil: (3-μL, 5 mg/8 mL) injection into right nasal cavity followed after 4 hours by 8- to 10-μL injection of MnCl₂. For MR imaging of normal aged pituitary glands as well as pituitary adenomas, rats (n = 8) received only an 8- to 10-μL injection of MnCl₂ into the right nasal cavity.⁶ Intranasally administered MnCl₂ was rapidly absorbed through nasal mucosa into plasma and also taken up by primary olfactory neurons. Serial MR imaging was performed at 1, 3, 11, 24, 48, 72, and 240 hours postinjection in all subjects. During imaging, body temperature was maintained by using an MR imaging–compatible heating pad (Deltaphase Isothermal Pad; Braintree Scientific, Braintree, Mass). Rats were allowed to move freely in their cages between scanning sessions.

T1-weighted images using a 3D spoiled gradient-recalled echo (3D-SPGR) pulse sequence were acquired on a standard 1.5T MR scanner (Signa; GE Healthcare, Milwaukee Wis) with a custom rat brain volume coil. The original acquisition matrix of 256 × 160 and 44 sections was reconstructed to a 256 × 256 image matrix. A 0.5-mm section thickness in 88 continuous sections was obtained by using zero-filled interpolation with the following scanning parameters: FOV = 7 × 4 cm; TE = 6.8 ms; TR = 15 ms; flip angle = 45°; NEX = 4. Total scanning time was 7 minutes 54 seconds. Voxel size for the image dataset before image processing was 0.273 × 0.273 × 0.5 mm.

FDG-PET Scanning of Pituitary Adenoma

In addition to MR imaging, 3 of the rats with spontaneous pituitary tumors underwent PET scanning by using FDG for characterization of tumor energy metabolism and for differentiation of active tumor from necrosis. Under isoflurane gas anesthesia, 0.5-mCi FDG was administered via tail vein injection, followed by a 40-minute uptake period. PET scanning was performed by using a high-resolution scanner (SHR-7700; Hamamatsu Photonics, Hamamatsu, Japan). 2D imaging for 45 minutes acquired 93 interleaved sections with 1.2-mm section thickness, and images were reconstructed by using a filtered back-image projection for an approximate in-plane resolution of 3-mm full width at half maximum.

Image Processing

Both PET and MR imaging sets for a single subject were coregistered to the baseline scan (no Mn²⁺) by using rigid-body transformation. The algorithm uses mutual information¹⁸ as a cost function and Powell iterative search algorithm to optimize 6 parameters (x y z translation and rotation).¹⁹

Coregistered image sets from the same subject were transformed to the Rat Brain in Stereotaxic Coordinates²⁰ to allow group-wise, pixel-by-pixel statistical analysis. The coregistered image sets for each subject were matched to an MR stereotactic template corresponding to the atlas.⁶ Resultant MR images had an 80 × 80 image matrix of 140 sections with a uniform voxel size of 0.2 mm.

Pixel intensities of MR image sets were normalized to minimize drift in MR sensitivity and differences in intensity across scans and subjects. To remove influence of local Mn²⁺ intensity in normalization, we performed a fine adjustment between individual scans and the template image set from the same subject by using a stochastic normalization algorithm.^21,22 In addition, PET scans were normalized to the global intensity.

In vitro Assessment of Mn²⁺ Uptake into Cultured Cells

Mn²⁺ uptake was evaluated in prolactin- and growth hormone–secreting rat pituitary tumor GH3 cells (ATCC, Manassas, Va). GH3 cells were cultured in Ham F12K medium (ATCC) supplemented with 15% heat-inactivated horse serum and 2.5% heat-inactivated fetal bovine serum with 20-U/mL penicillin-streptomycin in a 5% CO₂ airhumidified atmosphere at 37°C. Approximately 500,000 cells/well in 1 mL of growth medium were seeded in a 12-well plate and stimulated with 2 nM of growth hormone–releasing factor (GHRF; Sigma Aldrich, St Louis, Mo). Two minutes after the addition of GHRF, MnCl₂ (Sigma Aldrich) was added in a final concentration of 0, 0.25, or 0.5 mM. Cells were incubated for 4 minutes at room temperature, then removed, washed 3 times, and suspended in phosphate-buffered saline. Samples were acid-digested and Mn²⁺ content was assessed by using inductively coupled plasma mass spectrometry to quantify the amount of Mn²⁺ in each cell. All liquid in samples was initially evaporated, and cells were digested in an acid solution of 3:1 perchloric to nitric acid for 3 hours at 60°C. A calibration curve was first established for various known concentrations of Mn²⁺ to convert the relative values obtained for cell samples to Mn²⁺ content per cell.

Tumor Characterization

Following the final scanning, aged rats with spontaneous pituitary adenomas were euthanized by anesthetic overdose, and brains with adenoma were removed. Adenomas were preserved in 4% paraformaldehyde until histopathologic characterization was performed on individual tumors. Only 3 of the tumors (rats 1–3) were available for pathology due to the unexpected death of 1 subject during a holiday weekend. Tumor samples were processed routinely with paraffin embedding, sectioned at 4–5 μm, followed by hematoxylin-eosin staining (H&E). Tumors were examined histologically by a board-certified veterinary pathologist. To provide further in vivo characterization and corresponding diagnostic information, a board-certified neuroradiologist inspected the MR images of rat brains.

Data Analysis

Standardized volumes of interest (VOIs) (spherical r = 0.5 mm) were placed on coregistered, standardized, and normalized images for all time points in the pituitary gland (or pituitary adenoma) for all subjects as well as in the olfactory bulb of the subjects in the calcium channel blocker study. In addition, VOIs at prescanning and 1 hour postinjection scanning were measured in the carotid artery close to the area of tumor to confirm rapid absorption of intranasally administered MnCl₂ into plasma. Changes in MR imaging intensity for these VOIs were calculated by subtracting baseline (preinjection) intensity from postadministration time points. Significant changes in intensity were determined by t test (P ≤ .05). The correlation coefficient was calculated between peak average change in MR imaging intensity, peak FDG uptake on PET scans, and overall tumor volume (as assessed on MR imaging) for the 3 subjects that underwent PET scanning.

Pituitary Adenoma Has a Higher Calcium Influx Than a Normally Aged Pituitary Gland

Dynamic assessment of Mn²⁺ uptake in the normal aged rat pituitary gland versus adenoma indicated that the peak uptake is both faster and greater in adenoma (42 ± 18% at 11 hours and 27 ± 12% at 24 hours for adenoma and normal pituitary gland, respectively) (Fig 1). Changes in uptake, representing altered calcium influx between the 2 groups, were significant at 1, 3, 11, and 48 hours (P ≤ .05). The dynamic curves indicated that clearance may be altered in adenomas; however, there was no significant difference between subjects after 48 hours. Individual tumors can be easily distinguished on MR images when compared with a normal pituitary gland, even as soon as 1 hour postadministration of MnCl₂ (Fig 2). For example, the extent of the tumor in rat 1 can be difficult to distinguish from normal brain tissue before MnCl₂ administration but can be clearly seen at 1 hour. Sequential images demonstrate heterogeneous contrast enhancement in pituitary adenomas. VOI analysis of the nearby carotid artery indicated a significant change in signal intensity at 1 hour compared with prescanning (44.5 ± 8.8%, P ≤ .01), indicating rapid absorption of intranasally administered MnCl₂ into plasma. The degree of contrast enhancement (Mn²⁺ uptake) was variable but higher at 11 hours after contrast injection than at 1 hour after injection, indicating slow tumor uptake of the tracer. This gradual uptake is not well explained by the absorption of intranasal Mn²⁺ into plasma but indicates intracellular uptake through calcium channels.

• Download figure
• Open in new tab
• Download powerpoint

Fig 1.

Calcium influx is faster and increased in adenoma versus normal pituitary gland. The vertical axis represents the mean percentage change in intensity from baseline values calculated as the following: (intensity at time point − intensity at baseline) / (intensity at baseline) × 100. The horizontal axis is time in hours postinjection. Error bars show the SD of mean percentage change. The asterisk indicates that differences between the 2 groups reached statistical significance at 1, 3, 11, and 48 hours (P ≤ .05).

• Download figure
• Open in new tab
• Download powerpoint

Fig 2.

Individual Mn-MR images show enhancement of tumor after 1 hour. First column shows coronal sections of a single representative MR image of normal pituitary enhancement at baseline and 1 and 11 hours postinjection (small black arrow). Other columns show coronal sections of each individual subject with pituitary adenoma. Adenomas of individual subjects show clear enhancement at 1 hour postinjection (large black arrow, middle row), which continues to increase for 11 hours (large black arrow, bottom row) and allows superior delineation of the tumor/brain border over nonenhanced baseline images.

Calcium Channel Blocker Inhibits Mn²⁺ Uptake

We investigated the uptake of Mn²⁺ into normal neuronal cells as well as normal pituitary cells via calcium channels. Using a calcium channel blocker, verapamil, administered at 3 different dose/time intervals, we were able to inhibit the uptake into the olfactory primary neurons in the olfactory bulb (from the olfactory epithelium, site of administration via the olfactory nerve) (Fig 3A). Mn²⁺ uptake into the bulb was significantly higher with low-dose verapamil than with medium and high doses at all time points postadministration (P ≤ .05). However, only at the peak uptake at 24 hours of the medium-dose verapamil was there a significant increase over the high dose. We then looked at the Mn²⁺ uptake in the pituitary gland of these normal rats. Rats receiving a high dose of verapamil intranasally showed the same inhibition of uptake that we saw in the olfactory bulb, which was significant at all time points from medium- and low-dose verapamil (Fig 3B). However, pituitary uptake between medium and low doses was nearly identical.

• Download figure
• Open in new tab
• Download powerpoint

Fig 3.

Verapamil alters Mn²⁺ uptake in the olfactory bulb and pituitary gland. Change in intensity with time in the olfactory bulb was modulated by different doses of verapamil injected intranasally before MnCl₂ administration (A). In the same subjects, VOI in the pituitary indicates that low- and medium-dose verapamil have a nearly identical effect on uptake; however, uptake remains virtually extinct with a high dose (B). The vertical axis represents normalized intensity for VOIs in the right olfactory bulb and in the pituitary gland that have been subtracted from baseline values. The horizontal axis is time in hours postinjection of MnCl₂. Error bars show the SD of mean change in intensity. The asterisk indicates differences that reached statistical significance (P ≤ .05).

Calcium Influx Is Related to Energy Metabolism and Tumor Size

We performed FDG-PET scanning on 3 of the 4 rats with spontaneous adenomas (1 subject died before the PET scanning). When the average percentage change in MR imaging intensity, reflecting tumor calcium influx, was compared with peak tumor uptake of FDG on PET images as well as with the size of the tumor, there was a general correlation with all of these indices (Table). Calcium influx was most closely correlated to tumor size (Pearson correlation coefficient = 0.99) and also well correlated to peak FDG uptake (Pearson correlation coefficient = 0.91). Peak FDG uptake was less well correlated to tumor size (Pearson correlation coefficient = 0.84). An example of coregistered PET- and Mn²⁺-enhanced MR images shows good agreement between areas for enhancement and FDG uptake, suggesting that calcium influx may be related to energy metabolism and thus the functional activity of pituitary adenomas (Fig 4).

• Download figure
• Open in new tab
• Download powerpoint

Fig 4.

FDG-PET versus Mn-MR imaging: single subject. Coronal sections through pituitary tumor of rat 2. Left column shows tumor 1 hour after intranasal administration of the calcium analog Mn²⁺. Right column shows an FDG-PET scan of tumor glucose metabolism. Single arrows (white on MR imaging, black on PET) indicate region of high glucose metabolism correlating with high calcium influx. Double arrows (white on MR imaging, black on PET) indicate low glucose metabolism and calcium influx in the area of necrosis.

Mn²⁺ Uptake May Suggest Differences in Tumor Composition

Tumors from subjects 1–3 were removed and stained for pathologic characterization. With some specific differences in individual cellular composition and degree of hemorrhage or numbers of cystic spaces, all of the tumors were classified as pituitary adenomas, typical of “chromophobe adenoma of the pars distalis.”²³ Although there were foci of moderate atypia including karyomegaly and anisokaryocytosis in this study, malignancy of the tumors was not determined because invasion to adjacent tissues or intraorgan metastasis could not be assessed fully by histologic analysis.

Coronal and sagittal MR images obtained 11 hours following intranasal injection of Mn²⁺ demonstrated a large pituitary adenoma in each rat extending to the suprasellar cistern and the central skull base. Although the adenomas from all 3 rats appeared similar on MR images, we did note some distinct characteristics that may correspond to individual histopathologic features (Fig 5). Marked high signal intensity of the pituitary adenoma in rat 1 indicated an intense tumor uptake of Mn²⁺. Most interesting, the tumor morphology from this rat was more solid, with moderate areas of atypia and fewer dilated vascular areas or hemorrhages compared with rats 2 and 3 (Fig 5, upper right panel, H&E). There was heterogeneous contrast enhancement in rats 2 and 3, and the degree of Mn²⁺ uptake was less intense compared with that in rat 1. This may be due to the relative increase in both vascular/cystic spaces as well as the presence of hemorrhage in these tumors. In support of this, a focal area of marked low signal intensity was noted in the superior right lateral aspect of the pituitary tumor in rat 3, corresponding to a focus of hemorrhage on pathology (Fig 5, number sign).

• Download figure
• Open in new tab
• Download powerpoint

Fig 5.

Mn-MR images of individual subjects show correspondence to H&E stain. Coronal 3D-SPGR images through a pituitary adenoma at 11 hours post-MnCl₂ injection show different patterns of uptake between different subjects. Histologic sections are from paraffin-embedded H&E-stained pituitary tumors. Rat 1: The tumor is composed of principally solid sheets of neoplastic cells (asterisk) and a few small-caliber capillaries (arrow) with minimal hemorrhage. Images for rat 1 indicate intense more uniform uptake of Mn²⁺. Rat 2: Note the large dilated vascular and cystic spaces within the tumor section (arrow) and cords of neoplastic cells (asterisk). Rat 3: Note area of hemorrhage within the tumor parenchyma (number sign) and dilated vascular and cystic spaces (arrows) within cords of neoplastic cells (asterisk). MR images of rats 2 and 3 have less-intense uptake with a more heterogeneous appearance, which may correspond to hemorrhage and cystic spaces on pathology. Purple line indicates location of the coronal section on the sagittal image. All sagittal sections are from the midline. The number beneath each subject indicates the distance of the coronal section in millimeters from the bregma landmark.²⁰ (All histologic sections, original magnification ×10).

GH3 Cells Show Dose-Related Mn²⁺ Uptake In Vitro

In vivo cell culture of GH3 pituitary adenoma cells with increasing concentrations of MnCl₂ indicated a dose-response relationship of Mn²⁺ uptake as assessed by mass spectrophotometry. Mn²⁺ content of cultured cells incubated with 0.25-mM and 0.5-mM MnCl₂ (3.06 ± 1.22 and 10.71 ± 3.41 pg/cell, respectively) was significantly higher than the Mn²⁺ content of control samples that were incubated with vehicle alone (0.108 pg/cell, P ≤ .01) (Fig 6).

• Download figure
• Open in new tab
• Download powerpoint

Fig 6.

Manganese content in cells is increased with higher concentrations of MnCl₂. Graph shows intracellular Mn²⁺ (parts per million [ppm]) for 2 concentrations of MnCl₂ incubated with GH3 cells compared with vehicle alone. Error bars show SD of multiple samples. The asterisk indicates significant increase (P ≤ .01).