Bimodal 18F-Choline-PET/mMRI fusion imaging to detect local recurrence after radical prostatectomy and radiation therapy.
Image fusion is the process of combining relevant information from two or more images into a single image,
which is more informative than either of the images separately [18].
To combine anatomical,
functional and metabolic information obtained from different modalities at different times,
the process of spatial co-registration is necessary,
in order to ensure that the pixels from the various imaging datasets represent the same volume with acceptable precision [13,
19].
Registration can be rigid or elastic (deformable) [13,
18].
Only translation and rotation are possible with rigid co-registration,
whereas rotation,
translation and localized stretching are performed with elastic coregistration,
thus improving the matching of anatomical structures.
This latter modality may be particularly useful in the pelvis,
where the anatomical relationships between different structures is affected by the degree of distension of the urinary bladder [9].
The simultaneous acquisition of choline-PET and mpMRI by means of new PET/MRI scanners seems to be the most direct and effective method for obtaining bimodal fused PET/MRI images with optimal spatial co-registration [16].
However,
the scant availability of these expensive integrated systems is a major limitation to their use in routine clinical practice outside research facilities.
The mpMRI and bimodal fused 18F-choline-PET/mpMRI images shown in this essay were generated by means of the new Quanta Prostate (Camelot Biomedical Systems s.r.l.,
Genoa,
Italy) software,
which uses an elastic co-registration technique to integrate morphological and functional information from mpMRI with metabolic information from choline-PET.
Quanta Prostate is able to simultaneously display different MRI datasets,
allowing calculation of color-coded ADC maps from DWI sequences and wash-in/wash-out rate maps from the time-intensity curves of DCE-MRI.
Color-coded ADC and perfusion maps can be overlapped on T2w images and examined at different levels of transparency.
Using Quanta Prostate,
an elastic deformable registration technique (employing nonlinear transformation and spatially varying deformable models) is performed for MRI/PET co-registration,
in order to compensate for changes in patient positioning and local deformations between different imaging datasets (e.g.,
due to varying degrees of filling of the urinary bladder).
Once MRI/PET co-registration is done,
the operator can perform quantitative measurements (ADC and SUVmax values) by drawing a region of interest on the currently selected ADC or SUV map.
Among the different software platforms that are currently available for multiparametric reading and interpretation of prostate MRI,
Quanta Prostate is the only one that allows bimodal fusion imaging of mpMRI with choline-PET (Table 1).
All the mpMRI examinations in the present essay were acquired with a 1.5T MRI scanner (Signa HDxt,
GE Healthcare,
Milwaukee,
WI) equipped with an 8-channel pelvic phased-array surface coil,
according to a standardized protocol,
which has been described in detail in another study [9].
18F-choline-PET/CT was performed in the fasting state (at least 6 h).
An 18F-choline activity of 3MBq/kg (IASOCholine,
IASON Labormedizin Gesmbh&Co.Kg,
Linz,
Austria) was administered intravenously; data were acquired 10' after the injection by means of a dedicated PET/CT system (Discovery ST; General Electric Healthcare Technologies,
Milwaukee,
WI).
PET was acquired in tridimensional mode from the upper neck to the upper thighs,
by means of sequential fields of view,
each covering 12 cm (matrix of 256×256),
over an acquisition time of 3min.
Low-dose CT was acquired for both attenuation correction and topographic localization.
The CT parameters used for acquisition were 140 kV,
80mA,
and 0.5 s per rotation,
and pitch 6:1,
with a slice thickness of 3.25mm,
equal to that of PET.
State-of-the-art imaging and bimodal 18F-choline-PET/mMRI fusion imaging after radical prostatectomy.
Serum PSA measurement and digital rectal examination (DRE) are the first-line tests in the follow–up of patients after RP and RT [1].
After RP,
a serum PSA level of more than 0.2 ng/mL is considered to be associated with residual or recurrent disease [20].
The volume of locally recurrent PCa is estimated to be <1 cm3 for PSA levels <3.5 ng/mL [21].
As a result,
local recurrences are extremely difficult to detect with conventional modalities such as DRE,
transrectal ultrasound (TRUS) or ceCT [4,
5].
TRUS is neither sensitive nor specific in detecting local recurrences after RP; even with TRUS guidance,
the sensitivity of anastomotic biopsies remains low: 40–71% for PSA levels >1 ng/mL and 14–45% for PSA levels <1 ng/mL [22].
Therefore,
according to the most recent guidelines of the European Urology Association [1],
salvage RT (with or without adjuvant androgen deprivation therapy) is usually decided on the basis of the evidence of biochemical recurrence,
without any histological proof of the local recurrence.
Nonetheless,
mpMRI has been proved to be an accurate imaging modality for the detection of local recurrence,
with overall values of sensitivity ranging from 84 to 88% and specificity ranging from 89 to 100% [6,
23–25].
Choline-PET/CT can also detect local relapse of PCa,
but its sensitivity is lower than that of mpMRI [26–30].
In a recent comparative study,
the patient-based sensitivity,
specificity and accuracy of 11C-choline-PET/CT in diagnosing local recurrence were 54.1%,
92.3% and 65.5%,
respectively,
whereas those of mpMRI were 88.5%,
84.6% and 87.4% [6].
In this field,
performing a choline–PET/CT when the PSA value is <1 ng/mL seems to be questionable [1].
Bimodal 18F–Choline–PET/mpMRI fusion imaging allows the metabolic information from PET to be combined with the functional and anatomical details of mpMRI,
thus improving the diagnostic yield of each separate modality (Figure 1).
In common practice,
choline–PET is acquired immediately after a whole-body unenhanced CT for attenuation correction and anatomical localization [11].
However,
the spatial and contrast resolution of combined choline–PET/CT remains insufficient for an accurate assessment of the pelvis.
Combining PET images with high–resolution mpMRI enables the precise anatomical localization of PET–positive lesions to be identified.
On the other hand,
the high specificity of choline–PET may help to reach a conclusive diagnosis in doubtful cases in which T2w images,
DWI and DCE-MRI yield contradictory results.
RP consists of the complete removal of the prostate gland and seminal vesicles,
and the creation of a vesicourethral anastomosis [31,
32].
After RP,
the neck of the urinary bladder descends into the prostatectomy fossa.
The most common sites of local recurrence are the vesicourethral anastomosis and perianastomotic tissues [4] (Figure 1).
Other sites include the anterior and the posterior bladder neck and,
less frequently,
the retrovesical space (posterior to the bladder neck) [33].
The perianastomotic tissue is characterized by homogeneously low signal intensity on both T1 and T2w images due to postoperative scarring,
regardless of the surgical technique performed (i.e.
open suprapubic or transperineal approach performed with laparoscopic or robotic techniques) [2,
31,
32].
In the detection of local relapse,
T2w images have low sensitivity (48%–61%) and specificity (52%–82%) [23,
25],
and most false positive diagnoses occur when postoperative scarring assumes a nodular appearance,
mimicking recurrence.
When interpreted in combination with T2w images,
DCE-MRI is particularly accurate in detecting PCa recurrence after RP,
and sensitivities of 79%–88% and specificities of 89%–100% have been reported [23–25,
34].
DCE–MRI is useful in distinguishing PCa recurrence from fibrosis in the prostatectomy fossa and remnants of normal prostatic tissue.
Early nodular enhancement with early washout (dynamic curve of type 3) in the prostatic fossa,
perianastomotic tissues and seminal vesicle area are considered highly indicative dynamic features of local PCa relapse [34].
In doubtful cases,
the synchronous interpretation of DCE–MRI and dynamic perfusion curves with the metabolic information from choline–PET may be useful in order to reach a correct diagnosis.
Metallic surgical clips cause susceptibility artifacts in the prostatectomy fossa and seminal vesicle area that may hinder the interpretation of DWI sequences and ADC maps [2] (Figure 2).
By contrast,
choline–PET is not affected by artifacts due to ferromagnetic surgical devices.
When seminal vesicles are completely removed,
only linear fibrous streaks with low signal intensity persist.
However,
in approximately 20% of patients,
the seminal vesicles are not entirely removed during RP,
and they appear as nodular structures posterior to the prostatectomy fossa with variable signal intensity on T2w images [33].
Not infrequently,
seminal vesicle remnants may mimic local relapse [35].
On the other hand,
about 22% of PCa recurrences involve a retained seminal vesicle [33] (Figure 3).
In this setting,
additional choline–PET/CT results may help to clarify doubtful mMRI findings.
State-of-the-art imaging and bimodal 18F-choline-PET/mMRI fusion imaging of local recurrence after radiation therapy.
After RT,
the most reliable sign of disease recurrence is a rising PSA level more than 2 ng/mL above the PSA nadir (the lowest post-treatment PSA value),
rather than a specific threshold value [36].
In addition,
the PSA doubling time has been correlated with the site of recurrence: patients with local recurrence show a mean doubling time of about 13 months,
as compared with 3 months in those with distant metastases [1].
TRUS and ceCT are not reliable in defining local recurrences after RT [1,
2,
4].
In contrast,
mpMRI has an excellent diagnostic performance [2,
37] and may be used for biopsy targeting and planning of the salvage treatment.
After RT,
morphological changes in the prostate include inflammation,
glandular atrophy,
fibrosis,
and shrinkage [2,
4].
On T2w images,
these changes result in a diffusely reduced signal intensity of the prostate parenchyma,
more prominent in the peripheral zone,
causing loss of the normal zonal anatomy.
Local PCa recurrences have a low signal intensity on T2w images,
and are often difficult to distinguish from the surrounding irradiated prostate tissue,
which has a similar signal intensity [38,
39].
T2w imaging alone has low sensitivity (26%–44%) and suboptimal specificity (64%–86%) in detecting local recurrence,
and the use of additional functional techniques is mandatory in order to reach a correct diagnosis [40-42].
DCE-MRI seems to be the best functional imaging modality for the early detection and localization of PCa recurrence after RT,
with reported sensitivity of 90% and specificity of 81% [37,
41].
By means of DCE-MRI,
PCa recurrence can be recognized as an early enhancing area that contrasts well with the surrounding tissue,
which enhances less,
presumably because of radiation-induced fibrosis and vascular damage [43].
Recurrent PCa typically has a lower ADC than the surrounding parenchyma.
DWI displays high specificity (91%–93%) and low sensitivity (49%–69%) in the detection of post-RT PCa recurrence [44].
When combined with T2w imaging,
the diagnostic accuracy of DWI increases,
with reported sensitivity and specificity values of 62% and 97%,
respectively [45].
Detection of PCa recurrence is also feasible by means of choline-PET/CT,
with overall detection rates ranging from 81 to 88%,
including local relapse,
lymph-node and skeletal metastases [9,
46,
47].
With regard to local relapse,
DCE-MRI has a higher sensitivity than choline-PET,
but this latter technique displays good specificity [9] (Figure 4).
The limited spatial resolution of choline-PET/CT may hinder the precise anatomical localization of PET-positive foci within the pelvis,
particularly when the lesion is close to the bladder.
To date,
the available studies on the diagnostic performance of combined 18Fcholine-PET/MRI fusion imaging in PCa recurrence have provided encouraging results [9,
14,
15,
17,
48].
Given that the interpretation of mpMRI may be not immediate,
owing to discrepancies between morphological and functional techniques,
18F-choline-PET/CT fusion imaging may be particularly useful in doubtful cases,
providing metabolic characterization of suspect areas seen on MR images (Figures 5 and 6).
Bimodal 18F-Choline-PET/mMRI fusion imaging for the detection of lymph-node and bone metastases.
After both RP and RT,
the main clinical issue is whether biochemical recurrence is produced by a local relapse or metastatic disease.
In the case of biochemical failure,
bone scan and abdominopelvic contrast enhanced CT should be performed only in patients with a PSA level >10 ng/mL,
or with high PSA kinetics (PSA doubling time <6 months or a PSA velocity >0.5 ng/mL/month),
or in patients with symptoms of bone disease [1,
2,
7].
In the restaging of these patients,
newer imaging techniques are more effective even at low PSA levels [1].
Choline-PET has the great advantage of detecting lymph-node metastases when they are not discernible on morphological imaging (ceCT and MRI) [2,
3,
9,
40].
In fact,
only morphological and size criteria are commonly adopted to distinguish between benign and malignant lymph nodes (i.e.,
short axis diameter >10mm for an oval lymph node and diameter >8mm for a round lymph node) on both mpMRI and ceCT [10] (Figures 7 and 8).
However,
up to 80% of metastatic lymph nodes in prostate cancer have a short axis diameter smaller than 7 mm [10],
which produces a lot of false negatives on morphological imaging.
Indeed,
a meta-analysis by Hövels et al.
[49] reported a pooled sensitivity of 42% and 39% and pooled specificity of 82% and 82% for CT and MRI,
respectively.
Therefore,
it is not surprising that 18F-choline-PET/MRI is able to reveal more lymphnode metastases than mMRI alone.
In addition,
bimodal fusion imaging allows the precise anatomical location of PET-positive foci within the pelvis to be identified (Figure 9).
Recent studies on fusion imaging have found both a moderate but significant inverse correlation between the standardized uptake value (SUV) and the apparent diffusion coefficient (ADC) in metastatic lymph nodes [9,
48],
and a significant difference in mean ADC and SUV values between benign and malignant lymph nodes.
However,
a major limitation of DWI remains the lack of a precise cutoff ADC value for the detection of malignant lymph nodes.
With regard to skeletal metastases,
both choline-PET/CT and mpMRI have excellent diagnostic performances [6] (Figure 8).
The main advantage of choline-PET/CT over mMRI lies in its larger field-of-view,
which can cover the whole body.
However,
choline-PET may occasionally yield false negative results in the case of skeletal lesions with dense sclerosis on CT (>825 Hounsfield Units) [50].
Indeed,
choline activity tends to vary inversely with the degree of lesion sclerosis [3,
50].
Sclerosis of lesions is not a limitation of mpMRI by means of DWI and STIR sequences [2,
9].
Trimodal 18F-Choline-PET/MRI/TRUS fusion imaging for targeting locally recurrent prostate cancer.
Magnetic Resonance Imaging/Transrectal Ultrasound (MRI/TRUS) fusion imaging allows the sensitivity and specificity of MRI to be combined with the real-time guidance of transrectal ultrasound (TRUS).
Currently,
bimodal combined MRI/TRUS guidance is being used to improve the diagnostic yield of prostate biopsy in primary PCa detection [51,
52] (Figure 10).
MRI/TRUS fusion imaging guidance can be adopted when repeated systematic sextant biopsy has not detected the tumor in spite of increasing PSA levels [52].
Given the diagnostic yield of mMRI in the detection of local relapse of PCa,
biopsy procedures under MRI/TRUS guidance seem to be a good way to obtain histological proof of recurrence after both RP and RT (Figure 11).
In patients with biochemical failure after RT,
the biopsy status is a major predictor of outcome [22].
Following RT,
the main salvage option is RP,
but radiation-induced changes in the surgical field are associated to higher risks of urinary incontinence and rectal injury than in the primary setting.
Therefore,
given the morbidity of salvage options,
it is necessary to obtain histological proof of the local recurrence before treating the patient [1,
22].
MRI/TRUS fusion is performed by means of various software-based co-registration platforms (Table 2).
The images shown in the present essay were obtained with the ultrasound system MyLabTMTwice (Esaote,
Genoa,
Italy) equipped with the Virtual Navigator,
a fusion imaging platform that was initially released in 2004 for percutaneous interventional procedures and provides co-registration between real-time ultrasound and prior diagnostic CT,
PET-CT or MRI studies [11,
53].
Its application to mpMRI-targeted prostate biopsy has only recently been explored,
and is yielding encouraging results [54,
55].
From a technical viewpoint,
the process of MRI/TRUS coregistration is more immediate and less time-consuming in the untreated prostate gland (Figure 10),
in which the apex and borders of the gland,
or cysts within the prostate parenchyma,
may be used as internal markers to assist the co-registration process.
With the Virtual Navigator,
trimodal 18Fcholine-PET/MRI/TRUS fusion imaging can also be performed.
This can be carried out according to the following two-step procedure.
Rigid co-registration between 18F-choline-PET and mpMRI is initially obtained; real-time TRUS scans are then co-registered with fused 18F-choline-PET/mpMRI by means of the anatomical landmarks shown by the T2w images of mpMRI (i.e.
co-registration process assisted by internal markers).
When co-registration is over,
the screen of the Virtual Navigator displays both the fused image (i.e.,
real-time TRUS juxtaposed to an 18F-choline-PET image of the same size and in the same cut plane) and the images acquired by the separate modalities (i.e.,
TRUS,
mpMRI and 18F-choline-PET).
The degree of transparency can be manually adjusted,
and the operator can choose the intensity of TRUS and 18F-choline-PET images.
In addition,
the operator can decide to display all the potential combinations of co-registered modalities,
including TRUS + 18F-choline-PET,
TRUS + mpMRI,
and TRUS + 18F-choline-PET + mpMRI.
After RP and RT,
trimodal 18F-choline-PET/mpMRI/TRUS guidance may be used to assist the biopsy procedure of suspect areas,
thus enabling a precise spatial correspondence to be found between TRUS,
mpMRI and PET findings (Figures 12 and 13).