ELSEVIER
Contents lists available at ScienceDirect
The American Journal of Surgery
journal homepage: www.elsevier.com/locate/amjsurg
E
AJS The American Journal of Surgery
COLATINO
Original Research Article
Higher SUVmax on FDG-PET is associated with shorter survival in adrenocortical carcinoma
Sean M. Wrenn a,b, Alessandra L. Moore ª, Hina J. Shaha, Justine A. Barletta ª, Anand Vaidya ª, Kerry L. Kilbridge a,C, Gerard M. Doherty ª, Heather A. Jacene a, Matthew A. Nehsa,”
a .Brigham and Women’s Hospital, Boston, MA, USA
b .Rush University Medical Center, Chicago, IL, USA
c .Dana-Farber Cancer Institute Lank Center for Genitourinary Oncology, USA
Check for updates
| ARTICLE INFO | ABSTRACT |
|---|---|
| Keywords: Adrenal gland Adrenocortical carcinoma Endocrine surgery Diabetes mellitus Hyperglycemia Nuclear medicine | Background: Adrenocortical carcinoma (ACC) is an aggressive, rare malignancy. 2-deoxy-2-[18F]-fluoro-D- glucose positron emission tomography (FDG-PET) assesses tumor metabolism and glucose utilization. We hy- pothesized that higher maximum standard uptake value (SUVmax) is associated with decreased survival. Methods: We performed a retrospective analysis of patients with ACC. Included patients (n = 26) had an FDG-PET scan available with a documentable SUVmax. Patients were dichotomized into "High" (≥8.4, n = 12) and "Low" (<8.4, n = 14) SUVmax. Univariate analysis and survival analysis were performed to compare groups. Results: Demographics between groups were equivalent. The high SUVmax cohort demonstrated lower survival (median 479 days or 15.7 months) compared to the low group (median 1490 days or 48.6 months, p = . 01). Log- Rank curve confirmed differences in survival (p = . 007). Conclusions: Higher SUVmax was associated with significantly worse survival in ACC and may reflect a more aggressive phenotype. FDG-PET may provide clinically useful information to determine prognosis and treatment. Further studies should prospectively evaluate using FDG-PET/CT in ACC. |
1. Introduction
Adrenocortical carcinoma (ACC) is a rare endocrine malignancy which generally has a poor prognosis, however clinical outcomes can vary within a small and heterogenous patient population. Few prog- nostic variables exist, with increased age, tumor stage, and cortisol hy- persecretion all associated with worse outcome.” Previous research in our lab has demonstrated that patients with ACC and diabetes mellitus or hyperglycemia have significantly worse survival.2 One hypothesis for this association is that a state of hyperglycemia provides the tumor added glycolytic inputs for growth and metastasis. The utility of positron emission tomography (PET) using 2-deoxy-2-[18F]-fluoro-D-glucose (FDG) is based on the principle that malignant cells preferentially use glucose and glycolysis compared to normal cells and that these meta- bolic abnormalities can be detected on functional imaging prior to
morphological abnormalities.3,4 PET is a semi-quantitative technique and the level of FDG uptake in a tumor can be quantified using the standardized uptake value (SUV). SUVmax is typically calculated SUVmax = tracer uptake in region of interest/(injected activity/patient weight).5 This creates a quantifiable and comparable value demonstrating the extent of FDG (glucose analog) uptake in that lesion. FDG-PET has proven useful in patients with ACC for staging and treatment response but remains understudied and is utilized less frequently than CT or MRI.6
It has been previously demonstrated that many tumors such as ACC feature altered or reprogrammed metabolic pathways which allow for increased glycolytic flux. In particular, mutations in glucose metabolism and uptake via GLUT transporters can lead to constitutively elevated levels of glucose consumption and lactate hyperproduction. This is central to the Warburg Effect, where tumor cells undergo aerobic glycolysis, which allows tumors to generate cellular building blocks
Abbreviations: Adrenocortical carcinoma, (ACC); maximum standard uptake value, (SUVmax); fluorodeoxyglucose positron emission tomography, (FDG-PET); Brigham and Women’s Hospital, (BWH); Dana Farber Cancer Institute, (DFCI).
* Corresponding author. Brigham and Women’s Hospital, Department of Surgery, 75 Francis St, Boston, MA, 02115, USA.
E-mail addresses: AMOORE2@partners.org (A.L. Moore), jbarletta@bwh.harvard.edu (J.A. Barletta), anandvaidya@bwh.harvard.edu (A. Vaidya), Kerry_ Kilbridge@dfci.harvard.edu (K.L. Kilbridge), gmdoherty@bwh.harvard.edu (G.M. Doherty), hjacene@bwh.harvard.edu (H.A. Jacene), MNEHS@bwh.harvard.edu (M.A. Nehs).
https://doi.org/10.1016/j.amjsurg.2022.08.020
Patients with a diagnosis of adrenocortical carcinoma during the study period (n=68)
Excluded
No FDG-PET available (n=40) Post surgical PET with no evidence of disease (n=2)
Patients included in study (n=26)
Allocation/ dichotomization at median SUV max 8.4
Allocation
High SUV max (n=12)
Low SUV max (n=14)
Follow- Up
9 death events, 3 censored
4 death events, 10 censored
Analysis
through the pentose phosphate pathway.7 Additionally, the Warburg Effect results in increased cellular lactate export, which promotes angiogenesis and HIF (Hypoxia-inducible factors) signaling, which can aid in tumor survival in a hypoxic environment as it is rapidly growing. 8-10
Given that SUVmax is a marker of a tumor’s glucose uptake and uti- lization, we postulated that increased levels of FDG uptake given by SUVmax in ACC would portend a deleterious prognosis for patients with ACC. We further hypothesized that SUVmax would be positively corre- lated with key markers of tumor grade such as Ki67 proliferation index and mitotic count.
2. Methods
A retrospective analysis was performed on patients diagnosed with ACC and treated at Dana-Farber Cancer Institute (DFCI) and/or Brigham and Women’s Hospital (BWH). Institutional Review Board approval was obtained prior to initiating the retrospective chart review.
2.1. Study population
Inclusion criteria included patients with histopathologically- confirmed ACC who were diagnosed or cared for at these institutions (BWH and DFCI) from the years 2000-2019 and then longitudinally followed. Patients were only included if they had at least one FDG-PET/ CT scan with reviewable results that was performed as a part of their initial diagnosis, or after their initial diagnosis until the time of their death. SUVmax values were either abstracted from the radiology report, if available and previously documented, or the images were re-reviewed by radiologists (Authors HJS and HAJ, blinded to the analysis) to quantify SUVmax. Relevant patient variables were collected on patient demographics, disease characteristics, metabolic data, radiological characteristics, and surgical pathology. A consort diagram visualizing the inclusion parameters which ultimately merited the study population is visible in Fig. 1. Average glucose was defined in all patients as the
mean serum glucose (measured in mg/dL) obtained during the time period between their initial diagnosis and their death or censor event.
2.2. Primary and secondary outcomes
The primary outcome of this study was overall survival following diagnosis of ACC, measured in days from the time of initial diagnosis until death from any cause. The secondary outcomes were measures of histopathologic tumor grade, specific Ki67 proliferation index and mi- toses per high powered field.
2.3. Clinical practices within study population
FDG-PET was performed on patients at the discretion of the clinical care team, when believed to be clinically indicated-guidelines have not and do not exist at our institutions (or more broadly, nationally) for the inclusion of FDG-PET in the ACC care pathways.
When multiple PET-avid lesions were present, unless there was sus- picion for an alternative concurrent malignancy, the multiple lesions were presumed to represent metastatic disease.
The SUVmax for patients with multiple lesions was defined as the SUVmax of the most PET avid lesion (highest SUVmax of all lesions). If a post-surgical resection PET scan was performed which demonstrated no evidence of disease (and thus no SUVmax value), the patient was excluded from the study (n = 2 subjects). If a PET scan was performed at an outside institution with no report of SUVmax and no images to retroactively compute an SUVmax, these patients were also excluded. Patients that were unable to tolerate adjuvant mitotane were classified as not taking mitotane.
2.4. Dichotomization of comparator groups
Subjects were dichotomized by the median SUVmax in our dataset into high (≥8.4) and low (<8.4) SUVmax groups and compared via univariate statistical analysis.
Survival based on SUV intensity of FDG-PET, adrenocortical carcinoma
100
Probability of Survival
High SUV (≥8.4)
# Low SUV (<8.4)
50
** p=0.007, Log-rank (Mantel-Cox) test
0
0
2000
4000
6000
8000
Days
2.5. Statistical analysis
Univariate statistical analysis was performed with Mann-Whitney U test (continuous/ordinal variables) and Fisher’s exact test (categorical variables), with statistical significance defined as p < . 05. Statistical software was used to calculate values and to create visual representa- tions of the data (GraphPad Prism 9.0).
Our institutional practices at BWH and DFCI for longitudinal follow- up to monitor for recurrence or progression includes initially routine CT imaging of the chest, abdomen, and pelvis every 3 months for the first two years after treatment. Then, if no evidence of recurrence during this period, imaging may be liberalized to every 6 months for at least 5 years post-treatment, followed by annually for several years, followed by imaging every 1.5-2 years thereafter. FDG-PET imaging was used selectively, at the discretion typically of the medical oncologist when metastatic disease was suspected. Routine use of FDG-PET in this patient population is not yet standardized and universally implemented.
2.6. Survival analysis
Survival analysis was performed via Kaplan-Meier survival curves for overall survival only. Comparison of survival curves was performed with Log-rank (Mantel Cox) test via GraphPad Prism software.
Patients were followed longitudinally in the electronic medical re- cord from the initial timepoint where they were clinically diagnosed with ACC until there was a documented record of death from any cause, or there ceased to be further clinical documentation related to their oncologic (or overall health) care. If death was confirmed patients were recorded as such, and if no further records were obtained patients were listed as censored at the final data of available clinical data.
3. Results
Of 68 patients were evaluated for potential inclusion, 26 had at least one FDG-PET/CT scan available to review and were included for
| Low SUV max | High SUV max | p- Value | |
|---|---|---|---|
| (n =14) | (n =12) | ||
| Patient demographics and metabolic data | |||
| Age at diagnosis (mean) | 54.2 | 54.8 | 0.9 |
| Gender (% Female) | 57.1% | 50% | 0.99 |
| Average Glucose (mg/dL) | 105.0 | 118.0 | 0.34 |
| BMI (median) | 30.0 | 24.1 | 0.08 |
| Chronic Insulin Dependence [n (%)] | 0 (0) | 2 (16.7%) | 0.2 |
| Additional (Known) Metabolic Comorbidities | |||
| Hypertension [n (%)] | 11 (78.6%) | 7 (58.3%) | 0.4 |
| Hyperlipidemia [n (%)] | 6 (42.9%) | 5 (41.7%) | 0.99 |
| Obstructive Sleep Apnea [n (%)] | 0 (0%) | 1 (8.3%) | 0.46 |
| Coronary or Peripheral Vascular Disease | 2 (14.3%) | 1 (8.3%) | 0.99 |
| [n (%)] | |||
| Chronic Steroid Use [n (%)] | 5 (35.7%) | 7 (58.3%) | 0.43 |
| Tumor characteristics, radiology, and pathology data | |||
| Median Ki67 Proliferation Index % (n) | 8.75 (8) | 29.0 (4) | 0.09 |
| Median Mitoses/50 hpf (n) | 23 (10) | 35 (9) | 0.34 |
| Median Index Tumor Size (cm) | 8.5 | 14 | 0.01 |
| Laterality of primary tumor | |||
| Right [n (%)] | 7 (53.8%) | 8 (66.7%) | |
| Left [n (%)] | 6* (46.2%) | 4 (33.3%) | 0.69 |
| Multiple avid lesions on PET [n (%)] | 5 (35.7%) | 7 (58.3%) | 0.43 |
| Hormone hypersecretion from tumor | |||
| Cortisol [n (%)] | 6 (42.9%) | 9 (75%) | 0.13 |
| Androgen/DHEA [n (%)] | 1 (7.1%) | 5 (41.7%) | 0.06 |
| Median Serum Cortisol (mcg/dL) (n) | 14.1 (11) | 25.7 (12) | 0.38 |
| Oncological and prognostic data | |||
| Adjuvant Mitotane Therapy [n (%)] | 5 (35.7%) | 9 (75%) | 0.06 |
| Cancer Stage | |||
| Stage 4 [n(%)] | 9 (64.3%) | 10 (83.3%) | 0.54 |
| Less than Stage 4 [n (%)] | 2 (14.3%) | 1 (8.3%) | |
| Stage Unknown [n (%)] | 3 (21.4%) | 1 (8.3%) | |
| Median Days to death or censor | 1490 | 479 | 0.01 |
analysis. Included subjects consisted of 14 women (53.8%) and 12 men with median age of 53 (range 36-81). Median BMI was 27.8 kg/m2 20 patients underwent surgical resection (76.9%), 5 patients did not un- dergo surgical resection (19.2%), and one patient had uncertain surgical status (3.8%). Median follow up per subject, as defined by median days of follow up until death or censor event, was 1204 days (range: 12-8414 days).
| Days | 0 | 500 | 1000 | 2000 | 4000 | 6000 |
|---|---|---|---|---|---|---|
| High SUV | 12 | 6 | 5 | 3 | 1 | 1 |
| Low SUV | 14 | 13 | 13 | 5 | 4 | 3 |
| Subjects | Events | Censored | Median Overall Survival (days) | |
|---|---|---|---|---|
| High SUV | 12 | 9 | 3 | 479 |
| Low SUV | 14 | 4 | 10 | 1490 |
Days to death or censor
Mitoses/ 50 hpf
Days to death or censor
10000
250
8000
Mitoses/ 50 hpf
200
6000
p =. 01
150
4000
*
100
2000
50
0
0
Low SUV
High SUV
Low SUV
High SUV
Ki67 index
Age at diagnosis
60
100
80
40
Ki67%
Age (years)
60
20
40
20
0
0
Low SUV
High SUV
Low SUV
High SUV
SUVmax
Average serum glucose (mg/dL)
Average Glucose
40
250
30
200
SUVmax
150
20
p <. 0001
100
10
50
0
0
Low SUV
High SUV
Low SUV
High SUV
Survival and SUVmax
40
30
r2 = 0.29
SUVmax
20
10
0
32
64
128
256
512
102 2
048
20
409
96 81
Survival (log scale)
Ki67 and SUVmax
25
20
r2=0.60
SUVmax
15
10
5
0
0.5
1
2
4
8
16
32
64
Ki67 (log scale)
FDG-PET/CT scans were performed between October 2010 and July 2019.11 of the PET scans (42.3%) were performed prior to any surgical resection, and 13 (50%) were performed post initial surgical resection. Of the 13 patients that had a PET scan post-operatively, FDG-PET avid disease was visible in the following locations: adrenal resection bed (n = 3), multiple sites (n = 2), lung (n = 2), liver (n =1), abdomen (n=1), contralateral adrenal (n = 1), caval lymph nodes (n = 1), muscle (Psoas) (n = 1), and unknown (n = 1). Median SUVmax was 8.4 (range 0.5-31.9) for the entire cohort. Patients who had their FDG-PET in a pre-surgical setting (n = 8) had a significantly shorter mean survival time (850 days vs 3028 days, p = . 02) than those who had their PET in the post- operative setting. Average SUVmax between the pre-surgical scans (12.46) and post-surgical scans (7.28) was not statistically significant via t-test (p = . 21).
Median primary tumor size was 11 cm (range 3-18 cm). A total of 15 tumors were found to exhibit cortisol hyperproduction, 9 tumors with DHEA or androgen hyperproduction (see Table 1).
Groups were dichotomized into high (≥8.4) and low (<8.4) SUVmax and compared via univariate analysis. Kaplan Meier survival curves comparing these two populations demonstrated a statistically significant difference in overall survival (p = . 007, Log-Rank Mantel Cox test) as visualized in Fig. 2a and 2b. Full comparison between groups on base- line demographic data and outcomes is shown in Table 1 and Fig. 3. Tumors in the high SUVmax group were significantly larger (median 14 cm, range 7-18) compared to those in the low SUVmax group (median 8.5 cm, range 3-14.5) (p =. 01). To further demonstrate the association between survival, SUVmax within the cohort, scatterplots were created (see Fig. 4a).
Univariate analysis of histopathological and immunohistochemical analysis (mitotic count, Ki67 proliferation index) revealed no statisti- cally significant differences in tumor proliferation between the high and low SUVmax groups. Median Ki67 proliferation index was 8.75 in the low SUV max cohort, and 29.0 in the high SUVmax cohort (p = . 09). Fig. 4b demonstrates the relationship between SUVmax and Ki67 with each available data point in scatterplot.
4. Discussion
These data demonstrate a strong negative association between SUVmax (or FDG uptake intensity) on FDG-PET scan and overall survival in a small ACC cohort. This is significant because ACC, while in general having a poor prognosis, has a significant degree of heterogeneity in outcome and relatively few prognostic indicators.
As far back as 1988, Kern et al. demonstrated a relationship between glucose utilization rate (GUR) on FDG-PET and sarcoma tumor grade. Prior studies have demonstrated clearly negative prognostic relation- ships between SUVmax and survival in colorectal cancer, non-small cell lung, esophageal cancer, and gastric cancer.11-14 To our knowledge, this is the first such study suggesting that a higher SUVmax is associated with worse outcome in patients with ACC. Of note, one previous study demonstrated no negative prognostic relationship between SUVmax and survival in the ACC population. 15
While our initial hypothesis included that there would be differences in tumor grading, as measured by mitoses and Ki67 proliferation index (Ki67), these observed differences were not statistically significant, likely due to underpowering from the low numbers of available Ki67 and mitoses values. Previous studies have demonstrated positive correlation between SUVmax with Ki67 in the setting of non-small cell lung can- cer,16,17 and future studies will be needed to elucidate if that is a robust finding in ACC.
4.1. Limitations
Although our data show a statistically significant difference in overall survival when stratifying ACC into high and low FDG avidity, these data have several important limitations to consider. First, these data result from a retrospective chart review, which can always intro- duce confounding factors. Metabolic and laboratory data (glucose, he- moglobin A1c, BMI, cortisol levels, Ki67, mitoses, etc.) were collected inconsistently and incompletely. Given that the SUVmax value was in some patients collected post-surgical resection, the tissue Ki67 did not always correspond to the recorded radiological SUVmax site. Another additional potential confounding variable was a significant median tumor size difference between the high and low SUVmax groups, which could be a product of the faster proliferation of the high SUV max group. Dichotomization was performed at a value of 8.4 (the median SUVmax of the group) to create 2 essentially equally sized comparator groups, however the authors recognize the inherent limitations in this method. Further, due to a small population size, multivariable analysis to control for potential confounding factors was not feasible due to lack of
statistical power. FDG-PET scans were not all performed at the study institution, creating some heterogeneity in scanning protocols and image acquisition. Post-surgical FDG-PET scans, obtained at heteroge- nous timepoints, may present lead-time bias in their results. Future larger studies will be required to determine if SUVmax has an indepen- dent effect regardless of tumor size.
4.2. Future directions
Future directions for this research include the adoption of FDG-PET/ CT scans into an ACC treatment algorithm in a prospective manner to validate the findings and prognostic value. Beyond offering prognostic information, the use of FDG-PET in the diagnosis of ACC for adrenal nodules merits further study. Larger multi-institutional studies with higher power could potentially demonstrate a positive correlation be- tween SUVmax and markers of tumor grade such as Ki67 and/or mitoses. This could potentially even be useful in preoperative planning when performing an operation for an atypical adrenal nodule prior to the diagnosis of ACC, although the use of FDG-PET in this scenario has not been studied.
5. Conclusion
In conclusion, ACC patients with high SUVmax experience a shorter survival and poorer prognosis than their low SUVmax counterparts. This initial exploratory study suggests a link between SUV avidity and prognosis, and that SUV avidity may be a surrogate for more aggressive tumor biology. Further exploration is warranted with prospective studies on the utility of quantitative parameters from FDG-PET/CT for patients with known or suspected ACC.
Funding
Not applicable.
Declaration of competing interest
The authors report no relevant disclosures related to the submission of this manuscript.
Acknowledgements
The authors would like to thank Dr. Lia Wrenn for providing critical appraisal of the manuscript and for her support.
References
1. Abiven G, Coste J, Groussin L, et al. Clinical and biological features in the prognosis of adrenocortical cancer: poor outcome of cortisol-secreting tumors in a series of 202 consecutive patients. J Clin Endocrinol Metab. 2006 Jul;91(7):2650-2655. https:// doi.org/10.1210/jc.2005-2730. Epub 2006 May 2. PMID: 16670169.
2. Wrenn SM, Pandian TK, Gartland RM, Fong ZV, Nehs MA. Diabetes mellitus and hyperglycemia are associated with inferior oncologic outcomes in adrenocortical carcinoma. Langenbeck’s Arch Surg. 2021 Feb 1. https://doi.org/10.1007/s00423- 020-02061-0. Epub ahead of print. PMID: 33521844.
3. Almuhaideb A, Papathanasiou N, Bomanji J. 18F-FDG PET/CT imaging in oncology. Ann Saudi Med. 2011;31(1):3-13. https://doi.org/10.4103/0256-4947.75771.
4. Czernin J, Allen-Auerbach M, Nathanson D, Herrmann K. PET/CT in oncology: current status and perspectives. Curr Radiol Rep. 2013 May 3;1(3):177-190. https:// doi.org/10.1007/s40134-013-0016-x. PMID: 24883234; PMCID: PMC4034170.
5. Gary A. Ulaner, chapter 2 - FDG PET/CT performance and reporting, editor(s). In: Gary A. Ulaner,Fundamentals of Oncologic PET/CT. Elsevier; 2019:5-8. https://doi.org/ 10.1016/B978-0-323-56869-2.00002-8. ISBN 9780323568692.
6. Deandreis D, Leboulleux S, Caramella C, Schlumberger M, Baudin E. FDG PET in the management of patients with adrenal masses and adrenocortical carcinoma. Horm Cancer. 2011 Dec;2(6):354-362. https://doi.org/10.1007/s12672-011-0091-5. PMID :.
7. Warburg O, Wind F, Negelein E. The metabolism of tumors in the body. J Gen Physiol. 1927 Mar 7;8(6):519-530. https://doi.org/10.1085/jgp.8.6.519. PMID: 19872213, PMCID: PMC2140820.
8. Zare ME, Kansestani AN, Hemmati S, Mansouri K, Vaisi-Raygani A. The rate of aerobic glycolysis is a pivotal regulator of tumor progression. J Diabetes Metab Disord. 2021 Apr 1;20(1):523-531. https://doi.org/10.1007/s40200-021-00774-7. PMID: 34178852; PMCID: PMC8212265.
9. Bao X, Zhang J, Huang G, et al. The crosstalk between HIFs and mitochondrial dysfunctions in cancer development. Cell Death Dis. 2021 Feb 26;12(2):215. https:// doi.org/10.1038/s41419-021-03505-1. PMID: 33637686; PMCID: PMC7910460.
10. Vaupel P, Multhoff G. The Warburg effect: historical dogma versus current rationale, 10.1007/978-3-030-48238-1-27 Adv Exp Med Biol. 2021;1269:169-177. PMID :.
11. Cerfolio RJ, Bryant AS, Ohja B, Bartolucci AA. The maximum standardized uptake values on positron emission tomography of a non-small cell lung cancer predict stage, recurrence, and survival. J Thorac Cardiovasc Surg. 2005 Jul;130(1):151-159. https://doi.org/10.1016/j.jtcvs.2004.11.007. PMID: 15999056.
12. Sokolović E, Cerić T, Š Cerić, et al. The prognostic value of SUVmax of 18F-FDG PET/CT in patients with metastatic colorectal cancer. Acta Med Acad. 2020 Apr;49 (1):1-8. https://doi.org/10.5644/ama2006-124.278. PMID: 32738112.
13. Cerfolio RJ, Bryant AS. Maximum standardized uptake values on positron emission tomography of esophageal cancer predicts stage, tumor biology, and survival. Ann Thorac Surg. 2006;82(2):391-395.
14. Chung HW, Lee EJ, Cho YH, et al. High FDG uptake in PET/CT predicts worse prognosis in patients with metastatic gastric adenocarcinoma. J Cancer Res Clin Oncol. 2010;136(12):1929-1935.
15. Tessonnier L, Ansquer C, Bournaud C, et al. (18)F-FDG uptake at initial staging of the adrenocortical cancers: a diagnostic tool but not of prognostic value. World J Surg. 2013 Jan;37(1):107-112. https://doi.org/10.1007/s00268-012-1802-y. PMID: 23001287.
16. Kernstine KH, Faubert B, Do QN, et al. Does tumor FDG-PET avidity represent enhanced glycolytic metabolism in non-small cell lung cancer? Ann Thorac Surg. 2020 Apr;109(4):1019-1025. https://doi.org/10.1016/j.athoracsur.2019.10.061. Epub 2019 Dec 14. PMID: 31846640; PMCID: PMC7370816.
17. Vesselle H, Salskov A, Turcotte E, et al. Relationship between non-small cell lung cancer FDG uptake at PET, tumor histology, and Ki-67 proliferation index. J Thorac Oncol. 2008 Sep;3(9):971-978. https://doi.org/10.1097/JTO.0b013e31818307a7. PMID: 18758298.
18. Kern KA, Brunetti A, Norton JA, et al. Metabolic imaging of human extremity musculoskeletal tumors by PET. J Nucl Med. 1988 Feb;29(2):181-186. PMID: 2831318.