Uterine Uptake of Estrogen and Progestogen-Based Radiotracers in Rhesus Macaques with Endometriosis

The Institutional Animal Care and Use Committee (IACUC) at the Oregon National Primate Research Center (ONPRC) and Oregon Health & Science University authorized all experiments and procedures following the U.S. Public Health Service Policy on Humane Care and Use of Laboratory Animals ( 22 ). The division of Comparative Medicine/Animal Resources and Research Support at ONPRC oversaw all care of rhesus macaques and identified individuals with possible spontaneous endometriosis by cage-side monitoring of macaques that presented with painful or heavy menstrual bleeding and subsequently performing ultrasonography. We did not perform laparoscopic examination of individuals with suspected endometriosis because an aim of this study was to determine if PET imaging could identify lesions without a priori knowledge. Laparoscopic examination ensured no lesions were present in controls. The age of control macaques ranged between 9–17 years with an average of 13, whereas macaques with endometriosis ranged from 11–18 years old with an average of 15. Radiotracers were prepared by the OHSU Center for Radiochemical Research following previously published procedures for FES ( 23 , 24 ) and FFNP ( 25 , 26 ). Synthesis of FES was performed using a routine “large format” using a cyclotron. In contrast synthesis of FFNP was less reliable resulting in the need for small volume radiosynthesis. Troubleshooting and optimizing procedures for FFNP resulted in failed opportunities to scan monkeys particularly for control animals in the proliferative phase. Immediately after synthesis, high-performance liquid chromatography and mass spectroscopy analyses were conducted to evaluate radiochemical purity and measure molar concentrations. Radiochemical purity was 98% or greater for all FES syntheses, and 92% or greater for all FFNP syntheses with the exception of one (78.9%; Figure 1 ). The standard uptake values (SUVs) for the synthesis with lower purity was within the range of SUVs for all other scans. Radioactivity of tracers at injection ranged from 57.0 to 141.0 MBq with an average of 122.0 ± 3.6 MBq. The average concentration (μg/mL ± standard deviation) for FES was 1.42 ± 1.03 and for FFNP was 1.42 ± 1.01. Macaques with and without endometriosis (n = 6 and n = 4, respectively) were fasted overnight, sedated with ketamine, intubated, and anesthetized with isofluorane. We placed arterial, venous, and urinary catheters for the collection of serial blood samples, injection of radiotracers, and to promote the passive emptying of the urinary bladder. Animals were placed in dorsal recumbency after anesthetization. The CT was performed prior to the PET scan using a Discovery 610 Scanner (GE Healthcare, Boston, MA). The PET scan was started 2.5 minutes prior to injection of radiotracers. Images from PET scans were obtained for 1.5 hours. Acquisition rates were as follows: 60 frames at 5 sec/frame, 8 frames at 15 sec/frame, 4 frames at 30sec/frame, 3 frames at 60 sec/frame, and 40 frames at 120 sec/frame. DICOM images were analyzed using the imaging software MIM (version 7.0.6, MIM Software, Inc., Cleveland Ohio). We identified regions of interest (ROIs) across multiple sections using the 2D contouring tool to obtain SUVs accounting for body weight. To gain an idea of how uterine uptake compares to other organs, we also identified intestinal and muscle (i.e., psoas major) ROIs to quantify background radiotracer uptake. Evaluating uptake in two different ROIs helped to account for the excretion of FES and FFNP metabolites through the intestine ( 27 ). Maximum SUVs were exported for background and uterine ROIs. We also obtained average SUV and ROI volumes for the uterus. All statistical analyses were performed in SPSS (version 29, IBM SPSS Statistics, Chicago, IL). Any data not adhering to assumptions of statistical tests were log and/or square root transformed. We initially utilized repeated measure analysis of covariance (RMANCOVA) tests with ligand concentration as a covariate to inform how FES and FFNP uptake may have differed across the menstrual cycle. However, we removed this covariate after finding no significant relationship between radioligand concentration and SUVbw ( Table 1 ). Therefore, we performed repeated measure analysis of variance tests to examine how maximal uterine and background SUVbw differ with radiotracer type and phase of the menstrual cycle (between-subjects factors) across the duration of the scan (within-subjects factor). To account for the likely delay between radiotracer injection and start of scan in a clinical setting we analyzed data from a subset of the scan: 10–90 minutes. To determine if the presence of endometriosis was associated with differences in altered sex-steroid signaling in the uterus we again used RMANCOVA tests. We compared individuals with and without endometriosis and radiotracer type (between-subjects factors) to determine if uterine uptake levels differed across the duration of the scan (within-subjects factor). We included uterine volume, and initially ligand concentration as covariates. However, similar to maximal uptake levels, we subsequently removed ligand concentration as a covariate because no significant relationships were found ( Table 1 ). We focused our analyses on the secretory phase because of increased variation and lower sample sizes particularly for controls injected with FFNP during the proliferative phase ( Figures 4 & 5 ).

For macaques with endometriosis, we were unable to identify smaller endometriotic lesions because post-mortem examination of the reproductive tract revealed that the majority of lesions were present on the serosa of the uterus, and uterine uptake occluded lesion uptake. However, we were able to observe FES and FFNP uptake in atypical soft tissue surrounding the uterus ( Figure 2 ). We subsequently found endometriotic lesions in adipose tissue adhered to the uterus and ovary ( Figure 3 ). Anecdotally, it was difficult to distinguish the myometrium from the endometrium using just CT. With overlaying the PET scan, the highest uptake levels of radiotracers in the field of view of the scan can be observed in the uterus ( Figure 2 ). Within the uterus there is more uptake of radiotracers in the myometrium compared to the endometrium. Other reproductive structures such as the ovary and cervix also display uptake of radiotracers. Accumulation of excreted radiotracers both FES and FFNP was observed in the urinary system: bladder and ureters ( Figure 2 ). For FFNP only, there was skeletal uptake ( Figure 2 ). Maximum uterine uptake levels significantly varied with time*radiotracer and time*menstrual cycle phase ( Table 2 ; Figure 4A & B ; Figure 5A & B ). These relationships are likely driven by the increase in FFNP uptake between 70 and 90 minutes of scan time, and only during the secretory phase of the menstrual cycle. No between-subject factors were significant ( Table 2 ). Maximum intestinal uptake levels significantly varied with time*radiotracer, time*menstrual cycle phase, and time*radiotracer*menstrual cycle phase ( Table 2 ; Figure 4C & D ; Figure 5C & D ). The significant interaction of time*radiotracer*menstrual cycle is likely driven by 1) the increase in FES at 60 minutes during secretion ( Figure 4C & D ), and 2) the increase of FFNP with scan time in the proliferative phase ( Figure 5C & D ). We found a significant between-subjects factor of menstrual cycle phase on maximum intestinal uptake and a near significant effect of radiotracer ( Table 2 ) with increased uptake during the proliferative phase ( Figure 4C & D and Figure 5C & D ). Maximum muscle uptake did not significantly differ with any within- or between-subjects factors ( Table 2 ; Figure 4E & F ; Figure 5E & F ). During the secretory phase we found the within-subjects effects of time, time*uterine volume, time*radiotracer, and time*presence of endometriosis*radiotracer significantly affected average uterine uptake levels ( Table 2 , Figures 2 & 6 ). We also found a significant between-subjects effect of radiotracer on uterine uptake with FFNP displaying significantly higher uptake levels compared to FES ( Table 2 ; Figures 2 & 6 ).

Our data show that standard uptake volume in rhesus macaques is influenced by radiotracer type, phase of the menstrual cycle, and the presence of endometriosis. Our data support the premise that ER and PR binding is altered in individuals with endometriosis with the finding that average uterine uptake is elevated compared to controls. As such investigations concerning the dynamic signaling of ER and PR may reveal potential etiologies for endometriosis. Focusing on the difference in uptake due to radiotracer and menstrual cycle phase, we suggest that scanning individuals during the secretory phase would provide a higher probability of detecting endometriotic lesions because during the proliferative phase, intestinal background uptake is higher and more variable. Warranting further investigation is that uterine uptake is significantly altered with scan time, radiotracer type, and menstrual cycle. Maximum uterine uptake of both FES and FFNP are relatively constant regardless of menstrual cycle phase up until 70 minutes of scan time. At 70 minutes we observe a substantial increase in uterine FFNP uptake and only during the secretory phase. While we did not statistically account for the presence of endometriotic lesions in the maximum uptake analyses, the increase in FFNP after 70 minutes is only observed in individuals without endometriosis. This increase is transient and returns back to ‘baseline’ by the end of the scan. The dynamic change in FFNP at the end of secretory phase scans and the increase in background uptake of FFNP after 50 minutes of scan time during proliferation could be explained by either a change in uterine vascular flow and/or the uptake of FFNP metabolites ( 28 ). Determining the physiological and/or cellular basis of FFNP uptake may elucidate potential underlying mechanisms that differ with the disease state of endometriosis. In addition to alterations in maximum uterine uptake with radiotracer target and scan time, we found changes in average uterine uptake. Uterine uptake levels were altered with the presence of endometriotic lesions during the secretory phase with elevated uptake in individuals with endometriosis. Macaques with endometriosis display increased uterine uptake of both FFNP and FES compared to macaques without endometriosis. However, the timing and magnitude of uptake differs with radioligand. Levels of FES and FFNP are relatively constant until 30 minutes into the scan. At this point the uptake of FES in macaques with endometriosis becomes more disparate from those without endometriosis. For FFNP this difference is less prominent than FES and doesn’t occur until around 60 minutes. The observation that uterine uptake levels are different with disease state may represent an alternative approach to a diagnosis for endometriosis using PET imaging, and also supports a pathophysiological difference in estrogen and progesterone signaling. The ability to identify small endometriotic lesions using this PET/CT procedure remains equivocal as we were unable to detect endometriotic lesions in animals that were later verified to have endometriosis at necropsy. The locality of the lesions made it difficult to delineate endometriotic tissue from reproductive organs. For the vast majority of macaques clinically diagnosed with endometriosis we found small cystic lesions and ectopic endometrium on the serosa of the uterus or in adipose tissue adhered to reproductive structures. The high uptake of FES and FFNP in the myometrium occluded the potential to observe these small lesions. Ongoing studies in humans and macaques will help determine the efficacy in identifying smaller lesions. Our human study counterpart utilizes magnetic resonance imaging, which may allow for the identification of lesions located proximal to the uterus because of the ability of MRI to anatomically distinguish the various layers of the uterus. And our induced macaque model of endometriosis, like other induced primate models ( 29 ), will help determine if we can detect lesions more distally located to the uterus ( 30 ). The results presented here will likely inform optimal imaging procedures for future studies in primates both human and nonhuman. We posit that the following conditions will likely provide the optimal chances of identifying ectopic endometrium. These parameters culminate from identifying times during the scans where variation of uptake is least for maximum SUVs. Because intestinal uptake is higher during proliferation, we suggest that scans be performed during the secretory phase of the menstrual cycle. The various association among the menstrual cycle, the gastrointestinal tract, and estradiol supports our findings of menstrual cycle phase influencing intestinal uptake ( 31 – 33 ). Second, if using FES the scanning window should include 25–50 minutes post-injection. Or if using FFNP the scanning window should include 70–90 minutes post-injection. Although we were unable to identify small endometriotic lesions using this PET procedure, we were able to detect alterations in sex-steroid signaling that differed between individuals with and without endometriotic disease. Our data indicate that performing PET/CT with radiotracers that mimic sex-steroid hormones represents a potential alternative approach to a diagnosis of endometriosis. Ongoing studies in humans and in an induced endometriosis model using macaques will further inform the efficiency in using PET as a diagnostic tool. Because disrupted estrogen and progesterone signaling is associated with infertility and other gynecological disorders [e.g., adenomyosis, leiomyoma, polycystic ovary syndrome ( 2 )], we also posit that performing PET scans with FES and FFNP tracers will likely reveal similar and disparate estrogen and progesterone signaling pathways that may allow for targeted treatment.

Positron emission tomography (PET) imaging with with 16α-[ 18 F]fluoroestradiol (FES) and 12-[ 18 F]fluoro-furanyl-nor-progesterone (FFNP) have been validated for imaging of estrogen receptor (ER) and progesterone receptor (PR) responsive diseases, specifically cancer ( 1 ). Cancers are not the only disease in which ER and PR signaling can be disrupted. Other diseases, especially reproductive disorders (e.g., endometriosis, polycystic ovarian syndrome, leiomyomas, etc.) are associated with altered ER and PR action ( 2 , 3 ). Therefore, examining the in vivo presence, absence, abundance, and activity of ER and PR in the reproductive tract using PET imaging and FES/FFNP radiotracers represents an opportunity to obtain a more dynamic understanding of how certain disease states may alter ER and PR signaling. As nuclear steroid receptors, ER and PR are transcription factors that mediate actions of steroid hormones in the reproductive tract and many other organ systems ( 4 – 6 ). In the reproductive tracts of human and nonhuman primates, estrogen receptors (ER) and progesterone receptors (PR) undergo hormone-dependent regulation during the menstrual cycle ( 7 ). For example, in the uterus, estrogen stimulates expression of endometrial ER and PR while progesterone reduces expression of ER and PR with differential regulation of specific PR isoforms (PRA vs PRB). In certain diseases, such as endometriosis, progesterone action is attenuated resulting in altered hormone responsiveness in the reproductive tract and in endometriotic tissues ( 8 , 9 ). This is often referred to as progesterone resistance ( 10 ). Endometriosis is also an estrogen-dependent disorder in which the ectopic endometrium is present outside of the uterus; ectopic as well as eutopic endometrium responds to typical changes in estrogen throughout the menstrual cycle thereby altering ER and PR abundance ( 11 ). Therefore, clinical and research tools developed for hormone dependent cancers such as PET imaging can be applied to endometriosis ( 12 , 13 ). Our broad premise is that PET imaging of ER and PR with FES and FFNP tracers can provide an invaluable tool for assessing individuals with endometriosis. Although safety profiles for PET imaging with FES and FFNP of patients with hormone-dependent cancer have been well documented ( 14 ), this approach has not been assessed in patients with endometriosis and is not the current “standard of care”. Moreover, little is known regarding the uptake of these tracers in the reproductive tract during the menstrual cycle, or the impact of endometriosis-induced progesterone resistance on tracer uptake. In this study we conducted multiple PET/CT imaging experiments on rhesus macaques ( Macaca mulatta ) using both FES and FFNP. Rhesus macaques are menstruating, nonhuman primates with approximately 28-day menstrual cycles similar to those of humans ( 15 , 16 ). Like humans, macaques develop spontaneous endometriosis ( 17 – 19 ). Additionally, the strengths of the macaque model include: 1) our capability to extensively assess menstrual cycle length and phase by quantifying serum sex-steroid hormones levels and performing daily vaginal swabs ( 9 , 20 ); 2) to carryout extensive PET imaging trials that could expose the animals to repeated radiation risk; and 3) to be capable of collecting and assessing endometriosis and the reproductive tract after PET imaging. Our primary goal was to assess tracer uptake by the uterus across the menstrual cycle. We further sought to determine if either FES or FFNP provides more specificity and/or resolution in the ability to identify potential lesions due to the preferential uptake of radioligand based on hormone receptor ( 21 ). We predicted we would be able to identify adhesions or lesions approximately 1 cm 3 in size as current imaging modalities are able to detect other types of lesions (e.g., pancreatic) of this size. We were unsure whether PET imaging would allow for the identification of adhesions and lesions with large cysts because compared to the larger volume of acellular fluid in the interior of the cysts, few ER and PR are present in the cells surrounding the fluid.