PET/CT and contrast-enhanced CT: making a difference in assessment and staging of patients with lymphoma

The aim of this study is to evaluate the diagnostic performance of contrast-enhanced computed tomography (CECT) and 2-[Fluorine-18] fluoro-2-deoxy-d-glucose positron emission tomography combined to computed tomography (18 F-FDG PET/CT) in assessment of lymphoma. Hundred patients, pathologically proven as lymphoma, were evaluated by CECT and 18F-FDG PET/CT for initial assessment and staging of the disease. The number of lesions and the disease stage detected by each modality was calculated and further analyzed to be compared. 18F-FDG PET/CT diagnosed a total number of 545 lymphoma involved regions with sensitivity 96.6%, specificity 98.8%, and accuracy 99% that was higher than CECT which diagnosed a total number of 439 lymphomatous regions with sensitivity 87.5%, specificity 85.7%, and accuracy 88%. Discordant staging by both modalities was found in 23% of the patients. Lymphoma was upstaged by PET/CT in 17% of the patients; with major changes in 12% of them and downstaged in 6% of the patients. 18F-FDG PET/CT scan has a better diagnostic performance, represented by sensitivity, specificity, and accuracy, than CECT scan in the initial assessment of lymphoma regarding its nodal and extra-nodal lesions that could lead to alteration of disease staging which in turn markedly affecting the decision of treatment regimens.


Background
Lymphomas are heterogeneous group of diseases that arise from the constituent cells of the immune system or from their precursors. They are known to arise from virtually any organ or tissue in the body. Lymphomas are broadly classified into two main groups: non-Hodgkin's lymphoma (NHL) and Hodgkin's lymphoma (HL) [1]. Accurate diagnosis, correct staging, and proper therapy are important for successful outcome [2]. Diagnostic imaging plays a critical role in the initial evaluation, monitoring, and follow-up of lymphoma patients [3].
Computed tomography (CT) used to be the imaging modality of choice for staging lymphoma because of its widespread availability and relatively low cost [1]. In combination with powered injectors for rapid bolus administration of intravenous contrast medium used in contrast-enhanced CT (CECT) scan, lymph nodes of 5 mm or less in diameter and extra-nodal lesions with altered enhancement can be identified [4] At initial staging of lymphoma, determination of nodal involvement by CECT is based on size criteria; if the short-axis diameter is more than 10 mm and/or the long-axis diameter of 15 mm are exceeded. General criteria for extra-nodal involvement are any focal density alterations, abnormal contrast enhancement, or mass lesions involving soft tissues, bones, parenchymal organs, or serosal cavities [5].
Lymphoma may be restricted to the lymphatic system and/or present as extra-nodal disease, so diverse imaging appearance is expected [6]. Extra nodal lymphomas, seen mainly in NHL, can involve any organ. Secondary extension from a disseminated disease is the most frequent, while isolated primary lesions are possible but rare [7]. Ann Arbor's staging system was commonly used for both Hodgkin's and non-Hodgkin's lymphoma but currently, the Lugano classification was developed to simplify and standardize staging and response assessment [2]. Positron emission tomography (PET) using 2-[fluorine-18] fluoro-2-deoxy-D glucose (FDG) has already been validated to assess patients with different types of malignant tumors, including lymphomas. The principle of the imaging test is based on metabolic changes that reflect fundamental differences in the central metabolic pathways in malignant tissue [8]. 18F-FDG is the most commonly used radiotracer in PET imaging. FDG is an analogue of glucose and the uptake is directly proportional to the glucose metabolism of tumor tissue. Malignant tumors with high glucose metabolism show preferential uptake of FDG as compared to surrounding normal cells [2]. The standardized uptake value (SUV) is a semiquantitative parameter that is widely used in PET analysis and is proportional to the rate of glucose metabolism within the normal range of serum glucose concentrations [9]. Although the metabolic information afforded by PET imaging is invaluable, the quality of obtained data is poor/ noisy and limits imaging spatial resolution due to lack of detailed anatomical information, which impedes precise localization of sites with FDG uptake. For this reason, PET scan is often combined with CT imaging, allowing correlation between functional and anatomical imaging [10]. PET is especially helpful in oncology to visualize metabolic processes in malignant tissues. Combined with X-ray computed tomography (CT), it substantially enhances the value of diagnostic results [11]. Treatment is based on a variable combination of chemotherapy, radiotherapy, and immunotherapy, adapted to each type [7].

Methods
The research methodology included the following items: -Inclusion and exclusion criteria of the studied patients. -Imaging protocol such as patients preparation, scanning technique and imaging parameters, post scanning instructions, image analysis and data interpretation, and diagnostic criteria of lymphoma by CECT and 18F-FDG PET/CT.

Patients
We retrospectively studied 100 lymphoma patients attended to the radiology department at our institute from December 2017 to July 2019 for initial assessment and staging of the disease by CECT and 18F-FDG PET/ CT scans. The study included adult patients; their ages ranged from 20 up to 70 years distributed as 61 males and 39 females with initial presentation of pathologically proven untreated lymphoma and excluded patients with other malignancies, renal failure, pregnancy, and hypersensitivity to contrast agents.

Imaging method Protocol
All patients in the study were subjected to full history and complete clinical examination including the clinical stage of the disease. The patients were instructed to fast for 6 h before the scan. Creatinine clearance was checked to be > 30 and blood glucose level was ensured to be below 200 mg/dl as non-insulindependent diabetic patients were allowed to take small meal with oral anti-diabetic medications then fasting for 6 h but insulin-dependent diabetic patients were allowed to take early breakfast with their normal insulin dose then fasting for 4 h. On the day of scan, diabetic patients with blood glucose level higher than 200 mg/dl were injected rapid acting subcutaneous insulin and re-evaluated 2 h later. If controlled blood glucose level, they were scheduled late in the list to avoid the high FDG uptake by muscles and heart shortly after insulin injection but if not controlled blood glucose level, the patient was scheduled on another day.

Scanning technique and imaging parameters
Intravenous injection (IV) of 2-[fluorine-18] fluoro-2deoxy-D-glucose 18 F-FDG through IV line with a dose of 0.1 mCi/kg was done. After tracer injection, patient was asked to stay for 60-70 min in a quite dark room covered by warm blankets. No speaking, chewing, or reading was allowed. The patient was taken to the scan room where Siemens Biograph MCT 128 PET/CT exists, as its PET component from Knoxville, USA, and its CT component from Forchheim, Germany. First, a low-dose CT scan at 120 kV, 100 mAs, 0.9 pitch, 5 mm slice thickness, for attenuation correction, was obtained at the range extending from skull base down to upper-thighs. This is immediately followed by a three-dimensional PET acquisition at 3 min/bed position, 6-9 bed positions/patient, field of view 11.4 cm/bed position; obtained at the same scan range. Iterative algorithms, TrueX+TOF ultra HD-PET, 2 iterations, 21 subsets, were used for reconstruction of the PET images. Data were filtered using FWHM 4.0 mm with 3D-Gaussian Kernel and corrected for scatter. After completion of the PET scan, contrast-enhanced CT at 120 kV, 300 mAs, 0.9 pitch, 5 mm slice thickness, and 2.5 mm reconstruction thickness was done at the same scan range after IV injection of 1.5-2 ml/kg of an iodinated contrast material (Optray 300) using automated injector (Medrad Stellant) with flow rate 4 ml/s. A limited breath-hold technique was used to avoid motioninduced artifacts. Additional maximum inspiration CT scan for detection of lung lesion was performed. The radiation dose was 8.19 ± 0.83 mSv and 13.44 ± 5.14 mSv for PET and CT components, respectively, resulting in a total dose of 21.64 ± 5.20 mSv. That was in agreement with Gomez et al. [12] who mentioned that in 2008, in the USA, the diagnose technique that showed the highest dose deposition in the population was CT scans and Yuhao L [13].

Post scanning instructions
Although the level of radiation, emitting out of the patient after PET scan, is low, it is advised to avoid close contact with children and pregnant women and stop breast feeding for 24 h. If traveling abroad within 48 h is planned, it may trigger the sensitive radiation detectors, so it is advised to have a document stating performing PET scan during the last 48 h. Well hydration and frequent voiding are advised to help flush the radioactive material out of the body.

Image analysis and data interpretation
The attenuation-corrected FDG-PET images, CT lowdose images, as well as contrast-enhanced CT images were automatically fused on True D Siemens software or transferred to Philips Intellispace Portal and OSIRIX fusion workstation. Contrast-enhanced CT (CECT) images are evaluated at axial, coronal, and sagittal reconstructed planes. Maximum standardized uptake value (SUVmax) representing the quantitative analysis that measure the degree of FDG uptake at detected lesions was done by applying circular ROI with average diameter 2 cm over the most active part of the lesion. For each patient, 5 lymph node groups including cervical, axillary, mediastinal, abdominopelvic and inguinal groups, spleen, bone marrow, and other body tissues grouped in 9 extranodal sites including lung, liver, rest of GIT, rest of the chest, renoadrenal, muscles, head and neck, breast and skin with subcutaneous tissues were evaluated for lesions at CECT and PET/CT images. So, the total number of examined LN groups was 500 groups/100 patients as the group is considered involved even if a single LN of the group is affected. The spleen is considered involved whatever the number of its lesions, so the total number was 100 spleens/100 patients. The bone marrow was considered involved whatever the number of its lesions, so the total number is 100 bone marrows/100 patients. The other extra nodal lesions examined in 9 main organs as mentioned before, so the total number is 900 other extranodal organs/100 patients. Then staging (according Lugano classification) of lymphoma is done.

Diagnostic criteria of lymphoma by CECT
General criteria for lymph nodal involvement: if the short-axis diameter is more than 10 mm and/or the long-axis diameter of 15 mm are exceeded. General criteria for extra-nodal involvement are any focal density alterations, abnormal contrast enhancement, or mass lesions involving soft tissues, bones, parenchymal organs, or serosal cavities [5]. All lesions detected on CECT images were reevaluated on fused PET/CT images for estimation of their SUV max and correlation with FDG uptake by the mediastinal reference background.

Diagnostic criteria of lymphoma by 18F-FDG PET/CT
Any focus of elevated FDG uptake above mediastinal reference background whether on top of an obvious lesion on CECT or not [14].
Four readers in two teams interpreted the same series of patients, with blinding of the imaging reports and the other reviewers' interpretations. The first team is made up of two radiologists that interpret the CECT scan together. The second team is made up of two radiologists with experience in PET/CT reporting.

Data analysis
Data were statistically described in terms of number of lesions, percentages, mean ±standard deviation (± SD), and significance by P value when appropriate; sensitivity, specificity, and accuracy of each modality were calculated and compared.

Contrast-enhanced computed tomography
CECT detected 439 true positive affected sites and 32/ 471 false positive sites representing a percent of 6.8%, as follows ( Table 2

Regarding the involved groups of lymph nodes by both CECT and 18 F-FDG PET/CT
PET/CT detected more affected LN groups than those detected by CECT with the same order of LN group's involvement by both modalities, as follows: the most common involved lymph node group was the abdomino-pelvic group by a percent of 74% by PET/ CT and 67% by CECT, then the mediastinal group by a percent of 73% by PET/CT and 66% by CECT, followed by the cervical group by percent of 70% by PET/CT and 48% by CECT representing significant P value, then the axillary group by percent of 51% by PET/CT and 44% by CECT and finally the inguinal group by percent of 45% by PET/CT and 40% by CECT (Table 4; Fig. 4).

Regarding the involved other extra nodal organs by 18 F-FDG PET/CT
One hundred true positive extranodal organ involvements were detected. The most common involved extranodal organ was the lung 19%, while the least common involved site was the muscles. The organs showing the highest SUV max by PET/CT were lung 44.0, stomach 33.4, submandibular glands 32.7, and kidney30.8, while the organs showing the lowest SUV max were thymus 2.80, pleura 5.50, peritoneum 6.50, and liver 7.00 (Table 5; Fig. 5).

Discussion
Many lymphomas are potentially curable with their prognosis depends on the stage and histological type [5]. CT scan used to be the cornerstone of imaging in lymphoma and was playing a crucial role in staging. Currently, the advances in molecular imaging with18F-FDG PET/CT scan have facilitated the diagnosis, staging, and response assessment in lymphoma patients [3]. Integrated PET/CT offers the advantage of combining functional and anatomical information and better attenuation correction

Regarding demographic data
In this study, the percent of patients having NHL 66% was higher than those having HL 34%; this was agreed with Raanani P et al. [15] whose study stated that the incidence of NHL 68% was higher than that of HL 35%. Also, Mozaffer R and Sadiqa S's [16] study stated that 81.6% of their patients had NHL compared to 18.4% had HL. The same lymphoma type predilection was resulted by Roman E and Smith A [17]., while the study of Ricard F et al. [18] was carried out on 50.8% of cases with NHL and 49.2% with HL. In our study, NHL patients showed higher percent of males 39% than females 27%, also HL patients showed higher percent of males 22% than females 12% which emphasizes that male affection by both types of lymphoma is higher than female affection. The same sex predilection was resulted by Roman E and Smith A [17]. In our study, the higher percent 27% of cases was at the age group 50-< 60 years distributed as 24% had NHL and 3% had HL, while the lower percent 16% was at the age group of 20-< 30 years distributed as 12% had HL and 4% had NHL. This was agreed with Rodriguez B et al. [19] who stated that mean age of NHL 52± 15 while for HL the mean age was 31 ± 12.

Regarding total lesions
In our study, PET/CT diagnosed a total number of 545 involved regions with sensitivity 96.6%, specificity 98.8%, and accuracy 99% which was higher than those diagnosed by CECT; 439 involved regions with sensitivity 87.5%, specificity 85.7%, and accuracy 88%; la Fougere C et al.'s [20] study showed that PET/CT sensitivity 97% was higher than that of CECT 87.5%, also PET/CT specificity 99% was higher than that of CECT 85.5% as follows.

Regarding lymph nodal involvement
Our study resulted that PET/CT detected 312 true positive nodal group involvement with sensitivity 97.5%, specificity 94%, and accuracy 98% which was higher than CECT that detected 266 true positive nodal group involvement with sensitivity 83.1%, specificity 94%, and accuracy 89.6%, denoting significant difference (P > 0.005). Also, there were 46 false negative lymph node groups on CECT that decreased to only 3 groups by PET/CT. This was in agreement with Ricard F et al. [18] who stated that the sensitivity of PET/CT 99% was higher than the sensitivity of CECT 85% and also detected in their study 32 false negative lymph node groups by CECT that was corrected to 3 groups by PET/CT. Kwee T et al. [21] mentioned that some subtypes of lymphoma, namely small cell lymphocytic lymphoma/chronic lymphocytic leukemia may manifest as an increased number of small LNs, which declares that the size criteria used for the morphological assessment of lymph nodes in oncohematology is insufficient and the metabolic status of lymph nodes by PET scan should be integrated even if their size seems to be normal on CT. The most common involved lymph node group, in this study, was the abdomino-pelvic group, while the least common was inguinal group. In agreement with our study, the most common affected LN group in Mozaffer R and Sadiqa S's [16] study was the abdominopelvic group and the least common was the inguinal group, while the most common affected LN group in Rodriguez B et al.'s [19] study was the mediastinal group and the least common was the abdominal group. Regarding splenic involvement In this study, splenic involvement was truly positive in 79% of the patients by PET/CT with sensitivity 95.2%, specificity 98%, and accuracy 99% which was higher than CECT that detected true positive lesions in 47% of the patients with sensitivity 87.6%, specificity 86.6%, and accuracy76%. There were 22% of the patients with false negative results when assessed by CECT alone that was corrected to zero% by PET/CT. Close results were concluded by De Jong P et al. [22] who stated that PET/CT sensitivity and specificity in detecting splenic involvement were 100% and 95% versus 91% and 96% for CECT and confirmed the importance of metabolic status shown on PET/CT to discover splenic lesions with density similar to splenic tissue that could not be differentiated by CECT alone.

Regarding bone marrow involvement
In our study, bone marrow involvement was truly positive in 54% of the patients by PET/CT with sensitivity 93.7%, specificity of 96%, and accuracy 99% which was higher than CECT that detected true positive lesions in 41% of the patients with sensitivity 88.6%, specificity 86.2%, and accuracy 75%. There were 13% of the patients with false negative bone marrow involvement when assessed by CECT alone that was corrected to zero% by PET/CT. Also, the axial skeleton was the most commonly involved. In agreement with our results, Kwee T et al. [21] who mentioned that PET/CT is more sensitive 95% than CECT 86%. Othman A et al. [23] detected that bone marrow lymphoma mainly affecting the axial skeleton more than the appendicular skeleton.

Regarding other extranodal organ involvement
In our study, PET/CT detected 100 true positive extranodal organ involvements with sensitivity 94%, specificity 96.2%, and accuracy 99.5% which was higher than CECT that detected 85 true positive extranodal organ involvements with sensitivity 80%, specificity 88.6%, and accuracy 95.9%. There were 25 false negative extranodal organ involvements when assessed by CECT alone that was corrected to zero% by PET/CT. In agreement with our results, Ricard F et al. [18] stated that PET/CT sensitivity in detection of extranodal lymphomatous involvement 88% was higher than that of CECT alone 78% and that 9 false negative extranodal results by CECT were corrected to 5 by PET/CT. The most common involved extranodal organs in our study were lung 19%, GIT (mainly the stomach) 16%, liver 14% and renoadrenal 12%. In Das J et al. [22], the most frequently involved  [23] study resulted that the organs showing the highest SUV max were kidney 73.0, oropharynx 60.0, tonsils 40, and thyroid 37.0, while organs showing the lowest SUV max were pancreas 6.0, peritoneum 8.2, and suprarenal 9.0.

Regarding lymphoma staging according to Lugano classification
Differences in staging by PET/CT and CECT were found in our study. Discordant staging by both modalities was found in 23% of the patients. Lymphoma was upstaged by PET/CT in 17% of patients; with major changes in 12% (i.e., upstaging from stages I or II to stages III or IV) and downstaged in 6%. In agreement with our results, Ricard F et al. [18] who stated changes in the staging of 20% of the patients with upstaging in 17% and downstaging in 3% with major changes in 10%. Also, Othman A et al.'s [23] study mentioned that 10% of the patients were upstaged while 5% were downstaged after PET/CT. Luminari et al. [24] resulted that PET/CT helped in upstaging of 11% and downstaging of 1% of patients, while Raanani P et al.'s [15] study revealed changes in staging of 39% of the patients after PET/CT with 33% upstaging and 6% downstaging. This change in staging was explained by the higher ability of PET/CT over CECT in detection of involved sites either with normal morphology; as normal-sized lymph nodes or hidden and missed by its isodensity in extra nodal sites, by detection of increased activity in the form of high FDG uptake.

Conclusion
FDG-PET/CT scan should replace CECT in the initial assessment and staging of lymphoma, as FDG-PET/CT scan showed higher sensitivity, specificity, and accuracy which led to alteration of disease staging with marked effects on the decision of treatment regimens.

Limitation of this study
For clear practical and ethical reasons, it was not possible to take biopsies from all nodal and extranodal suspected sites in order to determine a gold   b Axial PET/CT showing avid FDG uptake of the three hepatic focal lesions seen at segment VIII (red arrows) and peri hepatic peritoneal nodular thickening (blue arrows) previously noted on CECT, with another FDG avid two hepatic focal lesions seen at segment IV (yellow arrows) and right pleural nodular thickening underlying the pleural effusion (green arrow), not seen on CECT. c Axial CECT showing three hypo dense non enhancing hepatic focal lesions; two of them seen at segment VIII (red long arrows) and one seen at segment VII (red short arrow) with peri hepatic peritoneal nodular thickening (blue arrows). d Axial PET/CT showing avid FDG uptake of the three hepatic focal lesions seen at segment VIII (red long arrows) and at segment VII (red short arrow) with peri hepatic peritoneal nodular thickening (blue arrows) previously noted on CECT, with another FDG avid one hepatic focal lesion seen at segment IV (green arrow) and pancreatic two lymphomatous focal lesions (yellow arrows), not seen on CECT standard. So, the standard reference was based on clinical data and other monitoring factors including clinical history; physical examination; laboratory work-up such as cell blood count, serum creatinine, urea, liver function tests, lactate dehydrogenase, b2 microglobulin, and viral serologies; imaging findings such as ultrasound and MRI when necessary; and iliac crest bone marrow biopsy. This limitation is also inherent in other previously published studies such as Raanani P et al. [15].

Recommendations
Novel tools such as the use of Raman spectroscopy (RS) enhanced by using a portable fiber-optic probe, i.e., Raman-enhanced spectroscopy (RESpect) probe, could be leveraged to provide rapid and real-time assessment of malignant lesions as suggested by Agsalda-Garcia et al. [25] that will help to overcome the limitation of taking biopsies form suspected lesions in order to define a gold standard. A future study using the same PET/CT scanner to compare low-dose unenhanced PET/CT and full-dose contrast-enhanced PET/CT in lymphoma assessment is recommended aiming at reducing patient radiation exposure.