Örebro University School of Medicine Degree project, 30 ECTS 2018-01-16
Diagnostic performance of
preoperative SPECT/CT in hyperparathyroidism
Version 3
Author: Patrik Bergius, Bachelor of medicine
Supervisor: Håkan Geijer,
MD, PhD, Associate professor
ii
Abstract
Introduction. 99mTc-MIBI-SPECT/CT is considered as a good preoperative examination for hyperparathyroidism patients according to other studies. Factors like Parathyroid hormone (PTH), calcium and lesion size/weight have been suggested to predict the outcome of the exam by the literature.
Aim: To evaluate the performance of 99mTc-MIBI-SPECT/CT, and investigate if some factors
can predict the outcome of the exam.
Materials and Methods: This retrospective study included 71 consecutive patients.
Comparing radiological statements to the surgical findings the sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) could be calculated and then compared to a previous study examining planar/SPECT outcomes. Factors were analysed with ROC-curves and linear-to-linear association tests, examining their associations to the
radiological outcomes.
Results: The mean patient age was 58.5 years (SD±14.65), 50 (70.4%) patients were female and 21 (29.6%) were male. Sensitivity, specificity, PPV and NPV for SPECT/CT was: 76.1%, 100%, 100% and 15.4% respectively. In a previous planar/SPECT study, the same results were: 85.3%, 66.7%, 99.2% and 9.1%. No significant association was found in the two subgroup analyses for PTH (P= 0.140, 0.168), corrected calcium (P=0.144, 0.104) and lesion size (P=0.482, 0.207). Lesion weight was significantly associated to the outcome of the exam (P=0.05, 0.012). The area under the curve (AUC) was 0.608 for PTH, 0.691 for corrected calcium, 0.684 for lesion size and 0.765 for lesion weight.
Conclusion: The 99mTc-MIBI-SPECT/CT results are uncertain, due to a low number of patients. Only lesion weight had a significant association to the radiological statements.
iii
ABBREVIATIONS
99mTc - Technetium 99m
AUC - Area under the curve
CI - Confidence interval
CT - Computed tomography
FN - False negative
FP - False positive
MBq - Megabecquerel
MGD - Multi glandular disease
MIBI - Methoxyisobutylisonitrile
MIP - Minimally invasive parathyroidectomy
mpi - Minutes post injection
NPV - Negative predictive value
PET - Positron emission tomography
PHPT - Primary hyperparathyroidism
PPV - Positive predictive value
PTH - Parathyroid hormone
ROC - Receiver operating characteristics
SHPT - Secondary hyperparathyroidism
SPECT - Single photon emission computed tomography
THPT - Tertiary hyperparathyroidism
TN - True negative
iv
Content
Introduction ... 1
Aim ... 2
Material and Methods ... 3
Technical methods ... 3
Collecting data ... 3
Comparing SPECT/CT with planar/SPECT scintigraphy ... 4
Examining associations between factors and radiological outcome... 4
Statistics ... 5 Ethics ... 5 Results ... 6 Discussion ... 9 Limitations ... 11 Conclusions ... 12 Acknowledgement ... 13 References ... 14 Ethical reflection ... 17 Cover letter ... 18 Utvärdering av ny röntgenmetod ... 19
1
Introduction
Primary hyperparathyroidism (PHPT) is a common endocrine disorder[1,2]. It has a variety of causes and can involve one or several of the parathyroid glands[1,2]. In most patients, only one parathyroid gland is affected as a benign parathyroid adenoma[2]. PHPT affects three-four times as many women than men, and is most common at the age of 50-59 years[2]. To surgically remove the pathological tissue is the only way to permanently cure the disease[2]. A full neck exploration, a bilateral method in which the surgeon manually examines all the parathyroid glands to identify the pathological ones used to be the standard surgical
method[2]. Advances in imaging methods and the use of intraoperative rapid parathyroid hormone (PTH) immunoassay has changed the surgical method of choice into a minimally invasive parathyroidectomy (MIP), a unilateral operation which results in fewer
complications[2,3]. A rapid decline in PTH levels after the removal of a lesion is used to define the surgery as a success. Thus, surgeons can be sure that MIP is as good as the bilateral method[2].
A variety of imaging techniques are used to identify pathological parathyroid glands prior surgery, such as ultrasound, CT, 99mTc-MIBI-planar scintigraphy, 99mTc-MIBI SPECT, 99m Tc-MIBI SPECT/CT-scans, 4D-CT scans, PET-scans etc[2–4]. In 1989 it was discovered that sestamibi (99mTcmethoxyisobutyl isonitrile (99mTc-MIBI)), used in cardiac stress tests also could be used for parathyroid imaging as it accumulates in hyperfunctioning parathyroid tissues[5]. Since then, 99mTc-MIBI has become the radiopharmaceutical of choice[3]. 99m Tc-MIBI, a lipophilic cation, accumulates passively in the mitochondria of metabolically active cells[6]. The result is an increased radioactivity in abnormal parathyroid tissue, but no additional activity from normal parathyroid glands[4]. To differentiate the uptake in the parathyroid from surrounding uptake an early and a late scan is used as 99mTc-MIBI is washed out slower from the parathyroid than the thyroid[3,4,7]. Evolution of imaging techniques from planar scintigraphy (two dimensional) to SPECT-scans (three dimensional) and then to
combined SPECT/CT-scans has increased the sensitivity and specificity of parathyroid imaging[3,4,8,9].
Studies have shown that PTH levels, different calcium levels and adenoma sizes often
correlate with the outcome of parathyroid imaging[10–12]. However these correlations are not found in every study[13]. These factors could be beneficial to consider prior to sending some patients to the exam as they perhaps can predict negative outcomes in some patients. In
2 September 2015 the Department of Radiology at Örebro University Hospital switched the preoperative method, from a planar/SPECT-exam to a SPECT/CT-exam.
Aim
The primary aim of this retrospective quality evaluation was to see how SPECT/CT has performed since the start considering its sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) and to compare these findings to a previous study at the same centre using planar/SPECT scintigraphy. The secondary aim is to see if a cut-off value for PTH, corrected calcium, adenoma weight and adenoma size can be found and used to predict the radiological outcome.
3
Material and Methods
Technical methods
The radiological equipment used in this study was a GE Discovery NM/CT 370Pro with a low energy high resolution (LEHR) collimator. An injection of 750 MBq Tc-MIBI (99m Tc
TechneScan Sammibi) was given at the start of the examination. After 10 minutes post injection (mpi), the first SPECT was performed. This included 120 projections taking 20 seconds per projection, in total 1200 seconds (using two gamma camera detectors). Ninety mpi, a second SPECT-scan was made in the same way, a CT-scan with contrast was also added if possible. The previous planar/SPECT-exam was performed like this: First the patient was injected with 99mTc-MIBI, 750 MBq. Then at 2 mpi, 30 mpi and 120 mpi planar images was taken using a pinhole collimator. Also, at 45 mpi a SPECT exam was performed
[Unpublished data]. Collecting data
Seventy-one consecutive hyperparathyroidism patients who underwent a preoperative
SPECT/CT-scan from September 30, 2015 to September 5, 2017 were included in this study. From the start date, all patients were scanned with SPECT/CT instead of the previously used planar/SPECT scintigraphy. Data was collected from two sources, radiological statements, patient age, sex and amount of contrast used was collected from Sectra RIS. More data was added from Kliniska portalen, including preoperative and postoperative corrected calcium, the first and fourth PTH taken during surgery, lesion location during surgery, diagnosis set by the surgeon, lesion size and weight obtained from the pathological examination and pathological diagnosis. All radiological statements were compared to the surgical findings (the reference standard) and then considered as correct or incorrect. The radiological statements were then categorized as True positive (TP), False positive (FP), False negative (FN) or True negative (TN). This categorization was then used to calculate:
• Sensitivity (TP / (TP + FN)) • PPV (TP / (TP + FP))
• Specificity (TN / (TN + FP)) • NPV (TN / (TN + FN))
4 Statements indicating a positive finding that later showed to be incorrect when the surgeons found a lesion in another location, were considered as False negative since radiology failed to identify the correct (true) lesion.
The preoperative corrected calcium used was the calcium test taken closest in time prior to surgery, the postoperative corrected calcium was taken the morning after surgery. The first PTH shows the PTH-levels at the start of surgery and the fourth PTH-test shows PTH-levels after removal of the suspected lesion.
Comparing SPECT/CT with planar/SPECT scintigraphy Patients who were included had a parathyroid
SPECT/CT-scan for the first time. Exclusion criterions were if the patients had no surgery or if patients had more than one exam, in this case, only the first exam was included. Figure 1 shows how patients were excluded when the PHPT study group was created to make the data in the two studies as similar as possible prior the comparison.
Examining associations between factors and radiological outcome
To be included in this part of the study, patients had to be surgically treated, and information about each factor had to be available. In patients with multi glandular disease (MGD) each lesion was included separately, so they had the same PTH and corrected calcium but different weights/sizes. This included three MGD patients in total, two with two lesions
and one with four lesions. Patients with SHPT/THPT were also included. This resulted in 54 lesions in the PTH-analysis, 53 lesions in the corrected calcium analysis, 52 lesions in the size analysis (the widest diameter measured by the pathologist) and 47 lesions in the weight analysis. The PTH test used in this part of the study was PTH 1 (the start of surgery) and the corrected calcium used was the preoperative sample taken closest in time prior the surgery.
5 Statistics
ROC-curves were created to try to find a cut-off value for each factor, and to see how well the factors predicted the radiological outcomes. A continuous variable (like PTH-values) was combined to a nominal outcome (Success or fail) and the program (SPSS) then could calculate the correlation[14].
The ROC-curves suggested that the factors were not particularly good in predicting the
radiological outcome. Therefore, subgroup information was put into cross-tabs to see if even a significant association could be found, and then chi-squared based test information was given by the linear-to-linear association test (trend test). Each lesion in each analysis was linked to if the radiological statement was correct or not. Then, each factor was split into subgroups for the trend test. Subgroups were created in two different ways to minimize bias. One way of creating subgroups was to have the same number of lesions in each group (+/- one if there was an odd number of lesions). The other way of creating subgroups was by dividing them according to their factor level. Using PTH as an example, group one had PTH-levels between 6-14.9, group two had PTH-levels between 15-25 and group three had a PTH-level > 25. The factor analysis part of this study was a pilot study of descriptive nature, which is why the statistical power was not calculated prior the study.
SPSS 23 (IBM SPSS Statistics for Windows, version 23 (IBM Corp., Armonk, N.Y., USA))
was applied for the statistics in this study. A P-value under 0.05 is considered a significant association in the trend tests.
Ethics
Consents were given by the Head of Department of Radiology and the Head of Department of Surgery, Örebro University Hospital to enter the patient’s medical records. No patients can be identified as analyses were made on the group level. The study was made to evaluate the SPECT/CT-exam, to see if guidelines could be improved and by this give better health care to patients in the future.
6
Results
Seventy-one patients were examined with SPECT/CT during the study period. Fifty patients (70.4%) were female and 21 (29.6%) were male. The mean patient age was 58.5 years (SD 14.65, range 22-86). Diagnoses were: 50 PHPT, five secondary hyperparathyroidism (SHPT), one tertiary hyperparathyroidism (THPT) and 15 had no diagnosis. PAD-diagnosis showed 42 adenomas, eight hyperplasia, one hypertrophy and two normal parathyroid glands. Fifty-three statements were positive and 18 were negative. Considering contrast medium concentration, 40 patients got 300 mg/ml and 21 patients got 350 mg/ml. Contrast medium volume ranged from 56-100 ml, with a mean of 94.9±9.61 ml. See table 1 for further factor information.
Seventeen patients were not operated (twelve had no indications, three were still waiting and two of them declined surgery). Table 2 shows where lesions were found. When categorizing statements after the surgical findings in all operated patients, there were 40 TP, no FP, 12 FN and two TN. This resulted in a sensitivity of
76.9% (95% Confidence interval (CI) 63% - 87%), a specificity of 100% (95% CI 16% - 100%), a PPV of 100% (CI couldn’t be calculated), and a NPV of 14.3% (95% CI 9% - 21%). In the SPECT/CT data for comparison with the previous study, there was 35 TP, no FP, 11 FN and two TN statements. Resulting in a sensitivity 76.1% (95% CI 61% - 87%), a
specificity 100% (95% CI 16% - 100%), PPV of 100% (CI couldn’t be calculated) and a NPV of 15.4% (95% CI 10% - 23%). The same data from the previous study had statements
categorized as: 116 TP, one FP, 20 FN and two TN [Unpublished data]. Resulting in a
Sensitivity of 85.3% (95% CI 78% - 91%), a Specificity of 66.7%, (95% CI 9% - 99%) a PPV Location SPECT/CT Surgery
Right side 25 22
Left side 26 22
Bilateral 3 7
Ectopic 3 1
Negative 14 2
Table 2. Showing the locations where
lesions were found.
Table 1. Showing patient lab values and lesion size/weight information.
Data Pre.op Corr.Ca (mmol/L) Post.op Corr.Ca (mmol/L) S-PTH sample 1 (pmol/L) S-PTH sample 4 (pmol/L) Lesion weight (g) Lesion size (mm) Mean 2.7 2.4 23.26 4.65 1.61 20.5 Max 3.03 2.92 155 18.7 14.5 6 Min 2.32 2.16 6.7 1.7 0.15 45 SD 0.14 0.15 23.02 3.34 2.3 8.84
7 of 99.2% (95% CI 94% - 100%) and a NPV of 9.1% (95% CI 4% - 20%). Figure 2 shows a comparison between the two methods, with 95% CI included in each bar.
PTH (pmol/L) Correct Accuracy PTH. (pmol/L)2 Correct2 Accuracy'
Group 1 6-14.9 14 of 23 61% Group 1 6.7 - 13.7 10 of 18 56%
Group 2 15-25 13 of 16 81% Group 2 14.5-23.2 15 of 18 83%
Group 3 >25 12 of 15 80% Group 3 23.5-155 14 of 18 79%
Corr.Ca (mmol/L) Correct Accuracy Corr.Ca (mmol/L) Correct Accuracy
Group 1 2.32-2.63 9 of 14 64% Group 1 2.32-2.65 12 of 18 67% Group 2 2.64-2.75 15 of 21 71% Group 2 2.65-2.76 13 of 18 72% Group 3 >2.76 16 of 18 89% Group 3 2.76-3.03 15 of 17 88%
Weight (g) Correct Accuracy Weight (g) Correct Accuracy
Group 1 0.16-0.5 8 of 13 62% Group 1 0.16-0.56 9 of 16 56%
Group 2 0.51-1.49 16 of 20 80% Group 2 0.56-1.49 14 of 16 88% Group 3 1.5-14.5 13 of 14 93% Group 3 1.49-14.5 14 of 15 93%
Size (mm) Correct Accuracy Size (mm) Correct Accuracy
Group 1 <15 11 of 15 73% Group 1 6-15 14 of 18 78%
Group 2 15-20 11 of 17 65% Group 2 15-22 11 of 17 65%
Group 3 >20 18 of 20 90% Group 3 22-45 15 of 17 88%
Groups divided by value Same Sized groups +/- 1 Table 3. Showing data from each group analysis when lesions were divided into subgroups,
testing each factor the same way. It also displays numbers of correct statements and accuracy (The number of correct statments / total number of statements) for each subgroup. Group 1 had the low numbers, group 2 had middle range numbers and group 3 contained high numbers for the different factors in each group analysis.
8 Table 3 shows the subgroups used to analyse the factors, the number of correct statements in each subgroup and the accuracy of the subgroups. This information was used to make linear-to-linear association tests. P-values from each of these tests are presented in Table 4.
The ROC-curves that were created showed an area under the curve (AUC) of 0.608 for PTH, 0.691 for corrected calcium, 0.765 for lesion weight and 0.684 for lesion size.
Analysed PTH Corr. Ca Lesion size Lesion Weight
Same sized groups 0.140 0.144 0.482 0.012
Groups divided by value 0.168 0.104 0.207 0.05
Table 4. P-values for each factor testing their association to the radiological
9
Discussion
The primary aim in this study was to see how the SPECT/CT-exam performed in patients with PHPT and to compare the results to the results from a previous study from the same centre when planar/SPECT scintigraphy was used. The SPECT/CT-results are somewhat surprising, considering that other studies have shown that SPECT combined with contrast enhanced CT has an advantage over only SPECT as it allows the nuclear medicine physicians an
opportunity to see the lesion on the CT which is beneficial for smaller lesions and it also gives better anatomical localization[15–18]. This has led to higher sensitivity and PPV numbers for SPECT/CT compared to SPECT and planar scintigraphy which was shown in a meta-analysis by Wei et al. from 2015[8]. It should be noted that in this meta-analysis some of the best SPECT/CT results were achieved in smaller studies[18–20]. Also, in some studies no added benefit regarding sensitivity using SPECT with CT could be found, what the CT added was better topographical localization[13,17].
The patients in this study and the previous one are not the same, which is the case in many other studies where SPECT/CT was the superior exam[18–23]. It would probably benefit SPECT/CT to compare them this way, however the planar/SPECT study was available, and it had good results, so it was considered as a good comparison for the SPECT/CT-results in this study. Also, as the data in both studies was collected retrospectively, a benefit would be that some bias would be removed as all examined patients would be included and the data was already set, it could not be altered to improve the results. Several things could explain the SPECT/CT-results, but most importantly, it should be mentioned that the sensitivity CI for the two different methods in this study overlap which means that no method is significantly better than the other, the results are uncertain. Another thing that probably affects the results of this study is that it only has 48 patients/statements in the PHPT comparison, which means lower power and more uncertainty of the results compared to the SPECT/planar results which were based on 136 patients [Unpublished data]. An additional observation was that in two exams, there were findings, but these findings were deemed as insufficient and resulted in negative statements. The surgeons later performed MIP on the side where the findings were noted and found a lesion there. In these (two cases) the statements had to considered as incorrect although they helped the surgeon perform successful MIPs. It could therefore be discussed if the information that the CT adds always is beneficial to perform good exams. In the two patients mentioned earlier perhaps it made it more complex for the nuclear medicine
10 is that the SPECT/CT is a new method at this centre, and that a learning curve could be
expected. Also with only 71 exams in a period of almost two years one question could be if this is enough to gain and maintain sufficient quality. Specificity is calculated on only two SPECT/CT patients and three planar/SPECT patients, therefore the CI for specificity are huge, which can be seen in Figure 2. Also, the NPV is heavily affected by the decision to categorize all incorrect exams as FN. These are the reasons why the sensitivity and PPV should be in focus. A meta-analysis from 2016 by Treglia et al. examined SPECT/CT results from 23 studies which included 1236 patients in total with PHPT and found a pooled
detection rate (sensitivity) of 88% on both a per patient and per lesion based analysis[4]. These results are slightly better than what the previous study with planar/SPECT showed, and it is what was expected in this study for the SPECT/CT-exams. Another thing Treglia et al. discuss is the superior localization that the CT adds which gives the surgeon a good
preoperative insight of the anatomy and it also gives an advantage in finding ectopic lesions. In this study, the SPECT/CT suggested three ectopic findings, however only one was
accurate, while the other two were in normal locations.
The secondary aim was to see if a cut-off value could be found and used for the different factors to predict the radiological outcomes. The ROC-curves showed a weak correlation to most factors, therefore linear-to-linear association tests were used to see of there even was a significant association between the factors and the radiological outcomes, but this could only be found for lesion weight in this study. What can be seen by looking at Table 3 is however, a trend of increasing accuracy between the subgroups with increasing corrected calcium. This could not be seen between the subgroups for PTH and lesion size. Studies have previously found correlations between some of the different factors and radiological outcomes, but not for others[10–13,25–28]. When calculating these significances most of these studies look at the mean factor value for the positive outcomes and statistically compare it to the mean factor value of the negative outcomes. If this study was performed in the same way, more
significances would likely be found. However, to split the statements into subgroups was considered as better because if any significant associations would be found, it would then motivate a continued search for cut-off values in a later analysis which was the secondary aim of this study, it would also remove some possible bias as subgroups were created in two different ways. The low number of statements for each subgroup and its effect on the power of this study to find significant results probably affected the group analysis for each factor in a negative way, more patients in each subgroup would have been beneficial at least for some
11 results. The ROC-analysis showed a AUC of 0.608 for PTH, 0.691 for corrected calcium and 0.684 for lesion size which says that these factors have a low correlation to the radiological outcomes[14]. Considering this and the results from the trend analysis, it was decided not to analyse them further. The AUC for lesion weight of 0.765 (which is considered as fair) and the fact that the two trend analyses for lesion weight had P-values of 0.05 and 0.012 says that it has a significant association to the radiological outcome[14]. Sandqvist et al. showed that an advantage of the added CT is that it helps the identification in smaller (lower weight) lesions. In this study the subgroups for low weights had an accuracy of only 62% and 56% but
according to the findings of Sandqvist et al. it could have been even lower with only the SPECT part of the exam[15]. The CT is probably also what led to the good accuracy for the small sized lesions in Table 3, it does however not explain why the accuracy was lower for medium sized lesions. It could be a random result or depend on other factors, such as the number of oxyphil cells being lower in the medium sized groups which affects the SPECT part of the exam in a negative way[27,28]. PTH and corrected calcium is taken before the SPECT/CT-exam and lesion size can be seen by the radiologist. These factors could therefore be used to influence the decision to send a patient to the SPECT/CT-exam or even by the nuclear medicine physicians to increase the likelihood of a correct statement when the
SPECT/CT-exam results in an uncertain finding. Lesion weight should be linked to lesion size and was therefore one of the factors in this study, however it can only be measured by the pathologist after surgery, therefore the relevance of this finding is limited, which is why this factor is not analysed further.
Limitations
This study has several limitations. The most important is the low number of patients, which resulted in low power for the SPECT/CT results and in failure to find significance when analysing the factors (except lesion weight), even if a probable trend can be seen for corrected calcium levels. The retrospective design of the study is also a limitation as a prospective design would give more opportunities to optimize comparisons and it would also be possible to collect more data. Also, some patient information could not be found as it was not always collected (like some adenoma weights for example). The choice to look at radiological statements is also a limitation, as some of them were unclear. Another limitation in the comparison between SPECT/CT and planar/SPECT scintigraphy is that it is not the same nuclear medicine physician that has written every statement, also the patients are not the same.
12
Conclusions
The SPECT/CT results in this study are uncertain as the number of patients was limited. Our results indicate that it is possible that it performs as well as what has been found in other studies. SPECT/CT is widely considered as one of the best ways to correctly lateralize
parathyroid lesions prior to surgery which has been shown in several other studies. The added information the CT gives makes it easier to find small or ectopic lesions. However, it perhaps results in more complex decisions for the nuclear medicine physicians when reviewing some exams, as more information is available than needed, it could lead to uncertainty or even incorrect conclusions.
Due to the low number of patients in each subgroup, in this study only lesion weight was significantly associated to the radiological statements. Visible trends could however be seen for corrected calcium. If associations would have been found for the other factors, it could be possible to find cut-off values for them and use these to predict the outcome prior to the exam or as added information when the radiological findings are uncertain.
13
Acknowledgement
I would like to give a special thanks to my supervisor, Håkan Geijer, MD for all the
encouragement during the writing process and for the help given whenever it was needed. I also want to say thank you to Anders Magnuson who took time to explain the statistics when I was confused, and to the nuclear medicine staff at the Department of Radiology at Örebro University Hospital for showing me a SPECT/CT-exam.
14
References
1. Pokhrel B, Levine SN. Hyperparathyroidism, Primary. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2017 [cited 2017 Sep 27]. Available from: http://www.ncbi.nlm.nih.gov/books/NBK441895/
2. Bilezikian JP, Brandi M, Eastell G, et al. Primary Hyperparathyroidism. In: Endotext [Internet]. South Dartmouth (MA): MDText.com, Inc.; 2000 [cited 2017 Oct 12]. Available from: http://www.ncbi.nlm.nih.gov/books/NBK278923/
3. Mohebati A, Shaha AR. Imaging techniques in parathyroid surgery for primary hyperparathyroidism. Am J Otolaryngol. 2012 Aug;33(4):457–468.
4. Treglia G, Sadeghi R, Schalin-Jäntti C, Caldarella C, Ceriani L, Giovanella L, et al. Detection rate of (99m) Tc-MIBI single photon emission computed tomography (SPECT)/CT in preoperative planning for patients with primary hyperparathyroidism: A meta-analysis. Head Neck. 2016 Apr;38 Suppl 1:E2159-2172.
5. Coakley AJ, Kettle AG, Wells CP, O’Doherty MJ, Collins RE. 99Tcm sestamibi--a new agent for parathyroid imaging. Nucl Med Commun. 1989 Nov;10(11):791–794. 6. Taïeb D, Ureña-Torres P, Zanotti-Fregonara P, Rubello D, Ferretti A, Henter I, et al. Parathyroid scintigraphy in renal hyperparathyroidism: the added diagnostic value of SPECT and SPECT/CT. Clin Nucl Med. 2013 Aug;38(8):630–635.
7. Taillefer R. , et. al. Detection and localization of parathyroid adenomas in patients with hyperparathyroidism using a single radionuclide imaging procedure with technetium-99m-sestamibi (double-phase study). J Nucl Med. 1992 Oct;33(10):1801-1807. [Internet]. [cited 2017 Nov 20]. Available from:
https://www-ncbi-nlm-nih-gov.db.ub.oru.se/pubmed/1328564
8. Wei W-J, Shen C-T, Song H-J, Qiu Z-L, Luo Q-Y. Comparison of SPET/CT, SPET and planar imaging using 99mTc-MIBI as independent techniques to support minimally invasive parathyroidectomy in primary hyperparathyroidism: A meta-analysis. Hell J Nucl Med. 2015 Aug;18(2):127–135.
9. Hindié E, Zanotti-Fregonara P, Tabarin A, Rubello D, Morelec I, Wagner T, et al. The role of radionuclide imaging in the surgical management of primary hyperparathyroidism. J Nucl Med. 2015 May;56(5):737–744.
10. Zhou J, Lu D-Y, Xia L, Cheng X-J. Diagnosis performance of (99m)Tc-MIBI and multimodality imaging for hyperparathyroidism. J Huazhong Univ Sci Technol Med Sci. 2017 Aug;37(4):582–586.
11. Koberstein W, Fung C, Romaniuk K, Abele JT. Accuracy of Dual Phase Single-Photon Emission Computed Tomography/Computed Tomography in Primary
Hyperparathyroidism: Correlation With Serum Parathyroid Hormone Levels. Can Assoc Radiol J. 2016 May;67(2):115–121.
12. Ciappuccini R, Morera J, Pascal P, Rame J-P, Heutte N, Aide N, et al. Dual-phase 99mTc sestamibi scintigraphy with neck and thorax SPECT/CT in primary
15 hyperparathyroidism: a single-institution experience. Clin Nucl Med. 2012 Mar;37(3):223– 228.
13. Gayed IW, Kim EE, Broussard WF, Evans D, Lee J, Broemeling LD, et al. The value of 99mTc-sestamibi SPECT/CT over conventional SPECT in the evaluation of parathyroid adenomas or hyperplasia. J Nucl Med. 2005 Feb;46(2):248–252.
14. The Area Under an ROC Curve [Internet]. [cited 2017 Dec 22]. Available from: http://gim.unmc.edu/dxtests/roc3.htm
15. Sandqvist P, Nilsson I-L, Grybäck P, Sanchez-Crespo A, Sundin A. SPECT/CT’s Advantage for Preoperative Localization of Small Parathyroid Adenomas in Primary Hyperparathyroidism. Clin Nucl Med. 2017 Feb;42(2):e109–114.
16. Ozkan ZG, Unal SN, Kuyumcu S, Sanli Y, Gecer MF, Ozcinar B, et al. Clinical Utility of Tc-99m MIBI SPECT/CT for Preoperative Localization of Parathyroid Lesions. Indian J Surg. 2017 Aug;79(4):312–318.
17. Oksüz MO, Dittmann H, Wicke C, Müssig K, Bares R, Pfannenberg C, et al. Accuracy of parathyroid imaging: a comparison of planar scintigraphy, SPECT, SPECT-CT, and C-11 methionine PET for the detection of parathyroid adenomas and glandular hyperplasia. Diagn Interv Radiol. 2011 Dec;17(4):297–307.
18. Bural GG, Muthukrishnan A, Oborski MJ, Mountz JM. Improved Benefit of SPECT/CT Compared to SPECT Alone for the Accurate Localization of Endocrine and Neuroendocrine Tumors. Mol Imaging Radionucl Ther. 2012 Dec;21(3):91–96.
19. Serra A, Bolasco P, Satta L, Nicolosi A, Uccheddu A, Piga M. Role of SPECT/CT in the preoperative assessment of hyperparathyroid patients. Radiol Med. 2006 Oct;111(7):999– 1008.
20. Kim Y-I, Jung YH, Hwang KT, Lee H-Y. Efficacy of 99mTc-sestamibi SPECT/CT for minimally invasive parathyroidectomy: comparative study with 99mTc-sestamibi scintigraphy, SPECT, US and CT. Ann Nucl Med. 2012 Dec;26(10):804–810.
21. Tokmak H, Demirkol MO, Alagöl F, Tezelman S, Terzioglu T. Clinical impact of SPECT-CT in the diagnosis and surgical management of hyper-parathyroidism. Int J Clin Exp Med. 2014;7(4):1028–1034.
22. Prommegger R, Wimmer G, Profanter C, Sauper T, Sieb M, Kovacs P, et al. Virtual neck exploration: a new method for localizing abnormal parathyroid glands. Ann Surg. 2009 Nov;250(5):761–765.
23. Pata G, Casella C, Besuzio S, Mittempergher F, Salerni B. Clinical appraisal of 99m technetium-sestamibi SPECT/CT compared to conventional SPECT in patients with primary hyperparathyroidism and concomitant nodular goiter. Thyroid. 2010 Oct;20(10):1121–1127. 24. Peters E, Klein W, Kaufman A, Meilleur L, Dixon A. More Is Not Always Better: Intuitions About Effective Public Policy Can Lead to Unintended Consequences. Soc Issues Policy Rev [Internet]. 2013 Jan 1 [cited 2017 Dec 27];7(1). Available from:
16 25. Dugonjić S, Šišić M, Radulović M, Ajdinović B. Positive (99m)Tc-MIBI and the subtraction parathyroid scan are related to intact parathyroid hormone but not to total plasma calcium in primary hyperparathyroidism. Hell J Nucl Med. 2017 Apr;20(1):46–50.
26. Qiu Z-L, Wu B, Shen C-T, Zhu R-S, Luo Q-Y. Dual-phase (99m)Tc-MIBI
scintigraphy with delayed neck and thorax SPECT/CT and bone scintigraphy in patients with primary hyperparathyroidism: correlation with clinical or pathological variables. Ann Nucl Med. 2014 Oct;28(8):725–735.
27. Mehta NY, Ruda JM, Kapadia S, Boyer PJ, Hollenbeak CS, Stack BC. Relationship of technetium Tc 99m sestamibi scans to histopathological features of hyperfunctioning
parathyroid tissue. Arch Otolaryngol Head Neck Surg. 2005 Jun;131(6):493–498.
28. Erbil Y, Kapran Y, Işsever H, Barbaros U, Adalet I, Dizdaroğlu F, et al. The positive effect of adenoma weight and oxyphil cell content on preoperative localization with 99mTc-sestamibi scanning for primary hyperparathyroidism. Am J Surg. 2008 Jan;195(1):34–39.
17
Ethical reflection
A consent was given by both the Head of Department of Radiology and the Head of Department of Surgery, Örebro University Hospital to enter the patients’ medical records. Data was stored on a USB stick which was kept at the Department of Radiology, in a locked room which only I and my supervisor had access to. No patients can be identified, and each analysis was made on group level. The main ethical problems are that patient journals were entered and read to collect data without approval from the patients. Also, the author was not involved in the patient care at all. Students are allowed to enter patient journals for student theses even if there is no patient contact, it is however still an ethical problem that patients were not asked if they wanted to participate, especially considering that much more patient information is available in the journals than what was needed for this study.
This study was made to evaluate the performance of the preoperative SPECT/CT-scan to identify possible problems, improve guidelines and by this contribute to better healthcare for future patients. More patients could get a positive outcome from the exam which would help the surgeons to minimize the number of unnecessary bilateral operations. It could also result in fewer unnecessary exams, resulting in reduced costs which would be beneficial for a financially limited healthcare and it would also mean less exposure to unnecessary radiation for patients in the future.
18
Cover letter
The Journal of Nuclear Medicine December 12, 2017 1850 Samuel Morse Drive
Reston, VA 20190-5316 U.S.A.
Dear Editors,
Please consider the enclosed manuscript called “Outcomes of preoperative SPECT/CT for hyperparathyroidism – A single centre experience with 71 consecutive patients” for publication in The Journal of Nuclear Medicine.
The study was made to evaluate how SPECT/CT had performed since its introduction at our centre and to examine if certain factors could predict outcomes of the exam. Results indicate that SPECT/CT has similar results as planar/SPECT for correct lateralization of parathyroid lesions and that some factors probably have a significant association to the outcome of the examination.
There are several reasons that I believe will be of interest to your readers. As a retrospective study including consecutive patients, it gives the reader an insight into factors that are not mentioned in other studies, which could be beneficial for readers at other centres to consider in their work. An example is the possibility of using information such as PTH and calcium levels to predict the outcome of an exam. The study also mentions possible reasons to why a new method, which is considered as an upgrade, sometimes at least at first does not live up to its expectations.
The included manuscript hasn’t been published before, it is not in press, under review or being considered for publication in another journal. As a student project, the author has no conflicts of interest.
Yours sincerely,
Patrik Bergius
19
Utvärdering av ny röntgenmetod
En relativt ny metod som används på röntgenkliniken på Universitetssjukhuset i Örebro har utvärderats i en ny studie. Metoden används framför allt hos patienter med godartade tumörer i bisköldkörtlarna. Den enda definitiva boten för dessa patienter är att operera bort tumören. För att göra detta på ett så säkert sätt som möjligt undersöks
patienten på röntgen så att kirurgen vet var tumören sitter. Den nya metoden är en variant av en så kallad nuklearmedicinsk undersökning vilket innebär att patienten ges en injektion av ett radioaktivt ämne, som sedan tas upp av tumören. Då ämnet sönderfaller kan detektorer registrera varifrån sönderfallet kommer. En dator skapar därefter en tredimensionell bild på sönderfallets ursprung. Sedan utförs även en skiktröntgen av patienten och med hjälp av en dator kan en kombinerad bild av båda undersökningarna skapas. Resultatet blir en bild som den till höger. Efter undersökningen kan en röntgenläkare med hög säkerhet säga var tumören är lokaliserad vilket är till hjälp för kirurgerna.
Resultatet från studien indikerar att undersökningen har fungerat ungefär som väntat, men då relativt få patienter har undersökts kan allt för säkra slutsatser ej dras. Man har även undersökt om olika nivåer av vissa faktorer så som bisköldskörtelhormon och kalcium har ett samband med om röntgenundersökningen lyckas eller inte för att bättre kunna välja vilka patienter som har mest nytta av undersökningen. Studien indikerar att samband är troliga men att resultatet är osäkert då relativt få patienter har undersökts.
