- Email: [email protected]
MRI findings of granulomatous prostatitis developing after intravesical Bacillus Calmette–Guérin therapy
Clinical Radiology 68 (2013) 595e599
Contents lists available at SciVerse ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
MRI findings of granulomatous prostatitis developing rin therapy after intravesical Bacillus CalmetteeGue T. Suzuki a, *, M. Takeuchi a, T. Naiki b, N. Kawai b, K. Kohri b, M. Hara a, Y. Shibamoto a a b
Department of Radiology, Graduate School of Medical Sciences, Nagoya City University, Nagoya City, Japan Department of Nephro-Urology, Graduate School of Medical Sciences, Nagoya City University, Nagoya City, Japan
art icl e i nformat ion Article history: Received 5 September 2012 Received in revised form 17 December 2012 Accepted 19 December 2012
AIM: To evaluate magnetic resonance imaging (MRI) findings of granulomatous prostatitis rin (BCG) therapy. (GP) developing after intravesical Bacillus CalmetteeGue MATERIALS AND METHODS: Ten patients with pathologically proven GP underwent prostatic MRI. Lesion shape and signal intensity (SI) were evaluated on T2-weighted (T2WI), T1WI, and diffusion-weighted imaging (DWI). RESULTS: Polygonal nodular lesions with notches, diffuse lesions, and cystic lesions with mural nodules were seen in two, six, and one patients, respectively. The remaining patient had a diffuse and cystic lesion. All diffuse lesions showed higher SI than muscle on T1WI and higher SI than the normal peripheral zone (PZ) on DWI. On T2WI, six of seven diffuse lesions showed a slightly lower SI than bone marrow and the remaining one lesion was iso-intense. All nodular lesions showed a low SI similar to muscle on T2WI and were iso-intense to muscle on T1WI. On DWI, two each of the four nodular lesions showed slightly lower SI and slightly higher SI than the normal PZ, respectively. All contents within the cyst and mural nodules showed markedly high and low SI on T2WI, respectively. On DWI, all fluids within cysts showed markedly high SI. One each of the mural nodules showed slightly higher SI and slightly lower SI than the normal PZ on DWI. CONCLUSION: Three main MRI patterns of GP were identified: diffuse, nodular, and cystic with mural nodule; among them, the diffuse type was the most common. Cystic lesions with mural nodules could accompany the lesion. Ó 2013 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction rin (BCG) therapy is Intravesical Bacillus CalmetteeGue an important adjunct to transurethral resection of stage Tis, Ta, or T1 bladder cancer used to reduce the risk of recurrence and progression.1,2 Several complications of this therapy, such as granulomatous cystitis, renal abscess, * Guarantor and correspondent: T. Suzuki, Department of Radiology, Nagoya City University, Graduate School of Medical Sciences, 1 Kawasumi Mizuho-cho, Mizuho-ku, Nagoya, Aichi 467-8601, Japan. Tel.: þ81 52 853 8276; fax: þ81 52 852 5244. E-mail addresses: [email protected], [email protected] (T. Suzuki).
exanthema, hepatitis, orchitis, epididymitis, and granulomatous prostatitis (GP), have been reported.1,3 The prevalence of GP after BCG therapy has been reported to be 1.3e40%.4,5 Patients with symptomatic GP require treatment with anti-tubercular agents,6,7 so early diagnosis of this disease is desirable. Because of its relative rarity, radiological features of GP have not been described extensively. Ma et al.8 reported radiological features of GP in six patients. Computed tomography (CT) showed non-specific prostatic enlargement and magnetic resonance imaging (MRI) showed diffuse decrease of the signal intensity (SI) or wellcircumscribed low SI nodules in the peripheral zone (PZ) on T2-weighted imaging (T2WI). They stated that these
0009-9260/$ e see front matter Ó 2013 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.crad.2012.12.005
596
T. Suzuki et al. / Clinical Radiology 68 (2013) 595e599
features were similar to those of prostatic cancer.9 Takeuchi et al.10 reported that GP incidentally found during staging of bladder cancer showed high SI on diffusion-weighted imaging (DWI), and stated that it should be differentiated from invasion by the bladder cancer. The purpose of the present study is to assess more detailed MRI findings of GP using T1-weighted imaging (T1WI), DWI, and apparent diffusion coefficient (ADC) value as well as T2WI.
level of PSA before MRI was 7.39 (range 1.66e16.1) ng/mg. All patients underwent prostatic biopsy after MRI. The reasons why the patients underwent prostatic biopsy were as follows: serum PSA elevation in six; suspected GP on MRI in three; and suspected prostatic cancer or invasion of bladder cancer in one. The mean interval between MRI and prostatic biopsy ranged from 1e105 days (mean 48 days). No patient was diagnosed with prostatic cancer by biopsy.
Materials and methods
MRI protocol
This retrospective study was approved by the local ethics committee of Nagoya City University, and informed consent was waived. A search of the radiological reporting system in the hospital for the period between November 2007 and February 2012 identified 10 male patients (mean age 70.2; age range 62e81 years) with pathologically proven GP diagnosed by prostatic biopsy. They had all undergone pelvic MRI. All patients had received intravesical BCG therapy after transurethral resection (TUR) of the bladder cancer before the MRI. Three of the 10 patients received a consecutive 8-week course of BCG therapy with a therapeutic fluid containing 81 mg BCG vaccine. One of the 10 patients received a 6-week course with 81 mg BCG vaccine in the fluid. Four had received an 8-week course with 80 mg vaccine, and one had received an 8-week course with 40 mg vaccine in the fluid. Only one patient had received two series of 8-week courses with the fluid containing 81 and 80 mg BCG vaccine. The interval between the last day of BCG therapy and MRI ranged from 87e913 days (mean 349 days). Six of the 10 patients underwent pelvic MRI for staging of recurrent bladder cancer. Two patients underwent prostatic MRI because of serum prostate-specific antigen (PSA) elevation. The remaining two underwent prostatic MRI because prostatic space-occupying lesions were suspected at CT. Only one patient complained of perineal pain and the others were asymptomatic. The mean
MRI was performed using a 1.5 T MRI machine (Gyroscan Intera; Philips Medical Systems, Best, the Netherlands) with a maximum amplitude of gradient of 33 mT/m and equipped with a radiofrequency coil (Quadrature body coil; Philips Medical Systems) and a phased-array five-channel sensitivity encoding cardiac coil (Philips Medical Systems). Endorectal coils were not used in any case. The imaging parameters were not standardized, because the study was retrospective and not all MRI examinations were performed for the purpose of prostatic imaging. All patients underwent axial T2WI [3363e5525 ms repetition time (TR); 120 ms echo time (TE); 256 189 or 195 matrix; 4 mm section thickness; 0.4 or 0.8 mm gap; 21 or 27 sections; 91e121 s acquisition time; 20 or 23 cm field of view] and DWI (echoplanar imaging with chemical shift selective fatsuppression techniques; b-factor, 0 and 1000 or 1200 s/ mm2; TR/TE, 2664e3598/70e88 ms; 128 100 or 101 matrix; 4 mm section thickness; 0.4 or 0.8 mm gap; 32 or 33 cm field of view; 21 or 27 sections; 4e14 signals acquired; 109e176 s acquisition time). ADC maps were generated using 0 and 1000 s/mm2 or 0 and 1200 s/mm2 bfactors. Five, four, and one of the 10 patients underwent spin-echo T1WI (474.0e551.3/10 ms TR/TE; 90 flip angle; 256 or 352 179 or 245 matrix; 4 or 8 mm section thickness; 20e35 cm field of view; 21e23 sections; 97.2e200.2 s acquisition time), gradient-echo T1WI (212.9e272.2/4.6 ms
Table 1 Characteristics and magnetic resonance imaging (MRI) findings of patients with granulomatous prostatitis. Case
Age (years)
Interval days between BCG and MRI
MRI and biopsy
1 2 3 4 5 6 7 8 9
72 71 62 71 81 68 69 69 74
209 357 87 693 105 368 965 913 154
5 105 1 77 12 63 8 81 63
10
65
497
63
Shape
Extent into TZ
Protrusion into periprostatic tissue
SI of T2WI
SI of T1WI
SI of DWI
ADC value (103 mm2/s)
Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Nodular Nodular Cystic Mural nodule Diffuse Cystic Mural nodule
þ þ , ,
þ, þ, þ þ
Iso-SI to BM Lower than BM Lower than BM Lower than BM Lower than BM Lower than BM Iso-SI Iso-SI Iso-SI to urine Lower than BM Lower than BM Iso-SI to urine Lower than BM
Higher Higher Higher Higher Higher Higher Iso-SI to OM Iso-SI to OM Iso Higher Higher Iso Iso
Higher Higher Higher Higher Higher Higher Low Higher Higher Higher Higher Higher Low
0.97 1.28 1.37 1.17 1.66 1.9 1.02 0.84 1.24 1.07 1.18 1.45 1.19
rin therapy; SI, signal intensity; T2WI, T2-weighted imaging; BM, bone marrow; T1WI, T1-weighted imaging; DWI, BCG, intravesical Bacillus CalmetteeGue diffusion-weighted image; ADC, apparent diffusion coefficient; PZ, peripheral zone; OM, obturator muscle. The SI of the lesion on T2WI was compared with the SI of urine, ischial BM, and the obturator muscle (OM). The SI of the lesion on T1WI was compared with the SI of the OM. The SI of the lesion on DWI was compared with the SI of the normal PZ.
T. Suzuki et al. / Clinical Radiology 68 (2013) 595e599
597
TR/TE; 80 flip angle; 256 179 matrix; 4 mm section thickness; 23 cm field of view; 21e27 sections; 63.4e81.1 s acquisition time) and unenhanced T1WI (5.7/2.7 ms TR/TE; 10 flip angle; 240 191 matrix; 2 mm section thickness; 35 cm field of view; 185 sections; 97.2 s acquisition time), respectively.
Image analysis Qualitative study All MRI image sets were reviewed by two radiologists with 5 and 8 years of experience. On T2WI, the presence of a well-circumscribed nodular lesion and its shape and size were evaluated according to the report of Ma et al.8 The presence of a diffuse lesion seen as a decrease of SI in the PZ according to their report. In addition, the presence of a cystic lesion showing a circular nodule with high SI on T2WI was evaluated. Whether the lesion extended into the transition zone (TZ) or periprostatic tissue was also evaluated using T2WI. The SIs of nodular and diffuse lesions on T2WI were compared with the SI of the ischial bone marrow and obturator muscle. If only the obturator muscle was chosen as the reference point, all lesions would show iso or higher SI, because the SI of the obturator muscle is extremely low on T2WI. As the signal intensity of bone marrow is intermediate, bone marrow was chosen in addition to the obturator muscle as a reference point. The SI of cystic lesions on T2WI was compared to that of urine. On T1WI and DWI, the SI of the lesions showing abnormal SI on T2WI was evaluated. On T1WI, the SI of the lesions was compared with that of the obturator muscle. On DWI, the SI of the lesion was compared with the SI of the normal PZ in 75- and 68-year-old men whose ages nearly matched the average age of the patients in the present study.
Quantitative study A radiologist with 5 years of experience measured the mean ADC values of these lesions. When multiple sametype lesions were present, their mean ADC value was calculated.
Results Qualitative study The radiological features of GP are summarized in Table 1. On T2WI, two nodular lesions were seen in two of the 10 patients (Fig 1); a diffuse lesion was seen in six of the 10 patients (Fig 2); a cystic lesion was seen in one (Fig 3); and a diffuse lesion with a cystic part was seen in one patient. Both of the cystic lesions had mural nodules. All nodular lesions showed low SI similar to that of muscle on T2WI. The size of the nodular lesions ranged from 8e16 mm (mean 12 mm) in diameter. All nodular lesions had a polygonal shape with notches. Three of the four nodular lesions protruded to the periprostatic tissue from the PZ and the remaining lesion was within the PZ. Six of the seven diffuse lesions showed slightly lower SI than that of the bone marrow on T2WI. The remaining one was iso-intense
Figure 1 Nodular lesions of granulomatous prostatitis (case 8). (a) T2WI of two polygonal nodules with notches (arrows) seen in the bilateral PZ. The signal intensities of the nodules are markedly low and iso-intense to the obturator muscle. The lesion in the right PZ extends to the periprostatic tissue (arrowhead). (b) DWI of the nodules in the right (arrowhead) and left (arrow) PZ are iso-intense to the PZ and slightly more hyperintense than the PZ, respectively.
to that of bone marrow. Five and two diffuse lesions were seen in bilateral and unilateral prostatic lobes, respectively. Two of the seven diffuse lesions extended into the TZ from the PZ. No diffuse lesion extended to the periprostatic tissue. There were two cystic lesions in two cases. The fluid in the cystic lesion and mural nodule showed markedly high and low SI on DWI, respectively. One of the two cystic lesions protruded to the periprostatic tissue and the other was within the PZ. On T1WI, all nodular lesions were iso-intense to muscle and all diffuse lesions showed higher SI than muscle. The fluids in the cystic lesion and mural nodule were iso-intense
598
T. Suzuki et al. / Clinical Radiology 68 (2013) 595e599
to or slightly more hyperintense than muscle. On DWI, all seven diffuse lesions showed higher SI than the normal PZ. Two of the four nodular lesions showed slightly lower SI than the normal PZ. Two of the four nodular lesions showed slightly higher SI than the normal PZ. The fluid in the cystic lesion showed markedly high SI on DWI. One of the two mural nodules showed slightly higher SI than the normal PZ. The other showed lower SI than the normal PZ.
Quantitative study The ADC values of nodular and diffuse lesions ranged from 0.84e1.40 103 mm2/s (mean 1.04 103 mm2/s) and from 0.97e1.90 103 mm2/s (mean 1.36 103 mm2/ s), respectively. The ADC values of the fluids within the cystic lesion and mural nodules ranged from 1.24e1.72 103 mm2/s (mean 1.47 103 mm2/s) and from 1.07e1.79 103 mm2/s (mean 1.35 103 mm2/s), respectively.
Discussion GP is one of the complications of intravesical BCG therapy for superficial bladder cancer.1 Radiological features of GP have been reported to be non-specific; therefore, a prostatic biopsy might be necessary to differentiate GP from prostatic cancer or prostatic invasion of bladder cancer.8,10 In the present study, the SI of GP on T1WI, DWI, and T2WI was evaluated. In this study, nodular and diffuse lesions were seen, in agreement with the study of Ma et al.8 The polygonal shape of the nodule with notches and markedly low SI on T2WI were characteristic findings and they were considered to be diagnostic keys to differentiate GP from prostatic cancer or prostatic invasion of bladder cancer. The protrusion of the nodule to the periprostatic tissue, which was seen in three of the four cases in this study, has never been reported. The SI of the nodular lesions were similar to or higher than those of the normal PZ on DWI; the reason for this diversity of SI is unknown, but it is supposed that a proportion of the lymphocytes associated with the acute inflammation and caseous necrosis might be related to the SI of the nodule because DWI reflects the cell density of the lesion.11 The nodular lesions were iso-intense to muscle on T1WI and this appeared to be non-specific, because the SI of benign prostatic hyperplasia and prostate cancer are often similar to that of muscle on T1WI. The SI of the diffuse lesions on T2WI were similar to the SI of bone marrow on T2WI and it was not as low as that of the nodular lesion but lower than the SI of the normal PZ. The diffuse lesions were seen across the TZ and PZ, and the SI of the lesion on DWI was high. It was thought that the findings might reflect acute prostatic inflammation. Lim et al.12 reported that mass-forming splenic tuberculosis showed slightly high SI
Figure 2 Diffuse lesions of granulomatous prostatitis (case 4). (a) T2WI. The SI of the left PZ is diffusely decreased (arrow). (b) DWI. The SI of the lesion is diffusely increased (arrows). (c) T1WI. The SI of the PZ (arrow) is diffusely increased.
T. Suzuki et al. / Clinical Radiology 68 (2013) 595e599
599
showed markedly high SI on DWI. Mural nodules showed markedly low SI on T2WI. Although the cause of this condition is unknown, it might involve fibrosis, caseous necrosis, clots, or calcifications, judging from the low SI on T2WI. However, the presence of calcification was not confirmed at CT. The present study had several limitations. The retrospective nature of the study might have introduced some form of patient selection bias. The number of GP was small, and further studies are needed to confirm the characteristic MRI findings of GP. In addition, differing MRI protocols were used and the majority of the patients did not have a dedicated prostate MRI. MRI findings could not be correlated with pathological findings because all specimens were obtained by random biopsy. In conclusion, the three main patterns of GP identified were diffuse, nodular, and cystic with mural nodule. The diffuse type was the most common.
References
Figure 3 Cystic lesion of granulomatous prostatitis (case 9). (a) T2WI. There is a cystic lesion showing markedly high signal intensity (arrow) and a mural nodule showing low SI (arrowhead) within the cystic lesion. The lesion extends to the periprostatic tissue (arrowhead). (b) DWI. Both the fluid within the cyst and the mural nodule show high signal intensity.
on fat-suppressed unenhanced T1WI. In tuberculomas, paramagnetic substances, such as macrophage-laden oxygen-free radicals, were thought to be the factor that caused the reduction of the T1 and T2 values.13,14 There have been no reports describing that GP was accompanied by a cystic lesion. It was thought that the content within the cystic lesion might be an abscess or protein-rich fluid produced by inflammation because it
1. Lamm DL, Stogdill VD, Stogdill BJ, et al. Complications of bacillus CalmetteeGuerin immunotherapy in 1,278 patients with bladder cancer. J Urol 1986;135:272e4. 2. Morales A, Eidinger D, Bruce AW. Intracavitary Bacillus CalmetteeGuerin in the treatment of superficial bladder tumors. J Urol 1976;116:180e3. rin vaccine-induced 3. Torres GM, Kaude JV, Drylie D. Bacille CalmetteeGue granulomatous prostatitis: another hypoechoic nonneoplastic lesion. AJR Am J Roentgenol 1990;155:195e6. 4. Stilmant M, Siroky MB, Johnson KB. Fine needle aspiration cytology of granulomatous prostatitis induced by BCG immunotherapy of bladder cancer. Acta Cytol 1985;29:961e6. 5. Oates RD, Stilmant MM, Freedlund MC, et al. Granulomatous prostatitis following Bacillus CalmetteeGuerin immunotherapy of bladder cancer. J Urol 1988;140:751e4. 6. Uzoh CC, Uff JS, Okeke AA. Granulomatous prostatitis. BJU Int 2007;99:510e2. 7. Noor MA, Biyabani SR, Burney IA, et al. Symptomatic granulomatous prostatitis following Bacillus CalmetteeGuerin immunotherapy for bladder cancer. J Pak Med Assoc 2002;52:578e80. 8. Ma W, Kang SK, Hricak H, et al. Imaging appearance of granulomatous disease after intravesical Bacille CalmetteeGuerin (BCG) treatment of bladder carcinoma. AJR Am J Roentgenol 2009;192:1494e500. 9. Naik KS, Carey BM. The transrectal ultrasound and MRI appearances of granulomatous prostatitis and its differentiation from carcinoma. Clin Radiol 1999;54:173e5. 10. Takeuchi M, Suzuki T, Sasaki S, et al. Clinicopathologic significance of high signal intensity on diffusion-weighted MR imaging in the ureter, urethra, prostate and bone of patients with bladder cancer. Acad Radiol 2012;19:827e33. 11. Qayyum A. Diffusion-weighted imaging in the abdomen and pelvis: concepts and applications. RadioGraphics 2009;29:1797e810. 12. Lim J, Yu JS, Hong SW, et al. A case of mass-forming splenic tuberculosis: MRI findings with emphasis of diffusion-weighted imaging characteristics. J Korean Med Sci 2011;26:457e60. 13. Chung MH, Lee HG, Kwon SS, et al. MR imaging of solitary pulmonary lesion: emphasis on tuberculomas and comparison with tumors. J Magn Reson Imaging 2000;11:629e37. 14. Sze G, Zimmerman RD. The magnetic resonance imaging of infections and inflammatory diseases. Radiol Clin North Am 1988;26:839e59.
Contents lists available at SciVerse ScienceDirect
Clinical Radiology journal homepage: www.clinicalradiologyonline.net
MRI findings of granulomatous prostatitis developing rin therapy after intravesical Bacillus CalmetteeGue T. Suzuki a, *, M. Takeuchi a, T. Naiki b, N. Kawai b, K. Kohri b, M. Hara a, Y. Shibamoto a a b
Department of Radiology, Graduate School of Medical Sciences, Nagoya City University, Nagoya City, Japan Department of Nephro-Urology, Graduate School of Medical Sciences, Nagoya City University, Nagoya City, Japan
art icl e i nformat ion Article history: Received 5 September 2012 Received in revised form 17 December 2012 Accepted 19 December 2012
AIM: To evaluate magnetic resonance imaging (MRI) findings of granulomatous prostatitis rin (BCG) therapy. (GP) developing after intravesical Bacillus CalmetteeGue MATERIALS AND METHODS: Ten patients with pathologically proven GP underwent prostatic MRI. Lesion shape and signal intensity (SI) were evaluated on T2-weighted (T2WI), T1WI, and diffusion-weighted imaging (DWI). RESULTS: Polygonal nodular lesions with notches, diffuse lesions, and cystic lesions with mural nodules were seen in two, six, and one patients, respectively. The remaining patient had a diffuse and cystic lesion. All diffuse lesions showed higher SI than muscle on T1WI and higher SI than the normal peripheral zone (PZ) on DWI. On T2WI, six of seven diffuse lesions showed a slightly lower SI than bone marrow and the remaining one lesion was iso-intense. All nodular lesions showed a low SI similar to muscle on T2WI and were iso-intense to muscle on T1WI. On DWI, two each of the four nodular lesions showed slightly lower SI and slightly higher SI than the normal PZ, respectively. All contents within the cyst and mural nodules showed markedly high and low SI on T2WI, respectively. On DWI, all fluids within cysts showed markedly high SI. One each of the mural nodules showed slightly higher SI and slightly lower SI than the normal PZ on DWI. CONCLUSION: Three main MRI patterns of GP were identified: diffuse, nodular, and cystic with mural nodule; among them, the diffuse type was the most common. Cystic lesions with mural nodules could accompany the lesion. Ó 2013 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved.
Introduction rin (BCG) therapy is Intravesical Bacillus CalmetteeGue an important adjunct to transurethral resection of stage Tis, Ta, or T1 bladder cancer used to reduce the risk of recurrence and progression.1,2 Several complications of this therapy, such as granulomatous cystitis, renal abscess, * Guarantor and correspondent: T. Suzuki, Department of Radiology, Nagoya City University, Graduate School of Medical Sciences, 1 Kawasumi Mizuho-cho, Mizuho-ku, Nagoya, Aichi 467-8601, Japan. Tel.: þ81 52 853 8276; fax: þ81 52 852 5244. E-mail addresses: [email protected], [email protected] (T. Suzuki).
exanthema, hepatitis, orchitis, epididymitis, and granulomatous prostatitis (GP), have been reported.1,3 The prevalence of GP after BCG therapy has been reported to be 1.3e40%.4,5 Patients with symptomatic GP require treatment with anti-tubercular agents,6,7 so early diagnosis of this disease is desirable. Because of its relative rarity, radiological features of GP have not been described extensively. Ma et al.8 reported radiological features of GP in six patients. Computed tomography (CT) showed non-specific prostatic enlargement and magnetic resonance imaging (MRI) showed diffuse decrease of the signal intensity (SI) or wellcircumscribed low SI nodules in the peripheral zone (PZ) on T2-weighted imaging (T2WI). They stated that these
0009-9260/$ e see front matter Ó 2013 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.crad.2012.12.005
596
T. Suzuki et al. / Clinical Radiology 68 (2013) 595e599
features were similar to those of prostatic cancer.9 Takeuchi et al.10 reported that GP incidentally found during staging of bladder cancer showed high SI on diffusion-weighted imaging (DWI), and stated that it should be differentiated from invasion by the bladder cancer. The purpose of the present study is to assess more detailed MRI findings of GP using T1-weighted imaging (T1WI), DWI, and apparent diffusion coefficient (ADC) value as well as T2WI.
level of PSA before MRI was 7.39 (range 1.66e16.1) ng/mg. All patients underwent prostatic biopsy after MRI. The reasons why the patients underwent prostatic biopsy were as follows: serum PSA elevation in six; suspected GP on MRI in three; and suspected prostatic cancer or invasion of bladder cancer in one. The mean interval between MRI and prostatic biopsy ranged from 1e105 days (mean 48 days). No patient was diagnosed with prostatic cancer by biopsy.
Materials and methods
MRI protocol
This retrospective study was approved by the local ethics committee of Nagoya City University, and informed consent was waived. A search of the radiological reporting system in the hospital for the period between November 2007 and February 2012 identified 10 male patients (mean age 70.2; age range 62e81 years) with pathologically proven GP diagnosed by prostatic biopsy. They had all undergone pelvic MRI. All patients had received intravesical BCG therapy after transurethral resection (TUR) of the bladder cancer before the MRI. Three of the 10 patients received a consecutive 8-week course of BCG therapy with a therapeutic fluid containing 81 mg BCG vaccine. One of the 10 patients received a 6-week course with 81 mg BCG vaccine in the fluid. Four had received an 8-week course with 80 mg vaccine, and one had received an 8-week course with 40 mg vaccine in the fluid. Only one patient had received two series of 8-week courses with the fluid containing 81 and 80 mg BCG vaccine. The interval between the last day of BCG therapy and MRI ranged from 87e913 days (mean 349 days). Six of the 10 patients underwent pelvic MRI for staging of recurrent bladder cancer. Two patients underwent prostatic MRI because of serum prostate-specific antigen (PSA) elevation. The remaining two underwent prostatic MRI because prostatic space-occupying lesions were suspected at CT. Only one patient complained of perineal pain and the others were asymptomatic. The mean
MRI was performed using a 1.5 T MRI machine (Gyroscan Intera; Philips Medical Systems, Best, the Netherlands) with a maximum amplitude of gradient of 33 mT/m and equipped with a radiofrequency coil (Quadrature body coil; Philips Medical Systems) and a phased-array five-channel sensitivity encoding cardiac coil (Philips Medical Systems). Endorectal coils were not used in any case. The imaging parameters were not standardized, because the study was retrospective and not all MRI examinations were performed for the purpose of prostatic imaging. All patients underwent axial T2WI [3363e5525 ms repetition time (TR); 120 ms echo time (TE); 256 189 or 195 matrix; 4 mm section thickness; 0.4 or 0.8 mm gap; 21 or 27 sections; 91e121 s acquisition time; 20 or 23 cm field of view] and DWI (echoplanar imaging with chemical shift selective fatsuppression techniques; b-factor, 0 and 1000 or 1200 s/ mm2; TR/TE, 2664e3598/70e88 ms; 128 100 or 101 matrix; 4 mm section thickness; 0.4 or 0.8 mm gap; 32 or 33 cm field of view; 21 or 27 sections; 4e14 signals acquired; 109e176 s acquisition time). ADC maps were generated using 0 and 1000 s/mm2 or 0 and 1200 s/mm2 bfactors. Five, four, and one of the 10 patients underwent spin-echo T1WI (474.0e551.3/10 ms TR/TE; 90 flip angle; 256 or 352 179 or 245 matrix; 4 or 8 mm section thickness; 20e35 cm field of view; 21e23 sections; 97.2e200.2 s acquisition time), gradient-echo T1WI (212.9e272.2/4.6 ms
Table 1 Characteristics and magnetic resonance imaging (MRI) findings of patients with granulomatous prostatitis. Case
Age (years)
Interval days between BCG and MRI
MRI and biopsy
1 2 3 4 5 6 7 8 9
72 71 62 71 81 68 69 69 74
209 357 87 693 105 368 965 913 154
5 105 1 77 12 63 8 81 63
10
65
497
63
Shape
Extent into TZ
Protrusion into periprostatic tissue
SI of T2WI
SI of T1WI
SI of DWI
ADC value (103 mm2/s)
Diffuse Diffuse Diffuse Diffuse Diffuse Diffuse Nodular Nodular Cystic Mural nodule Diffuse Cystic Mural nodule
þ þ , ,
þ, þ, þ þ
Iso-SI to BM Lower than BM Lower than BM Lower than BM Lower than BM Lower than BM Iso-SI Iso-SI Iso-SI to urine Lower than BM Lower than BM Iso-SI to urine Lower than BM
Higher Higher Higher Higher Higher Higher Iso-SI to OM Iso-SI to OM Iso Higher Higher Iso Iso
Higher Higher Higher Higher Higher Higher Low Higher Higher Higher Higher Higher Low
0.97 1.28 1.37 1.17 1.66 1.9 1.02 0.84 1.24 1.07 1.18 1.45 1.19
rin therapy; SI, signal intensity; T2WI, T2-weighted imaging; BM, bone marrow; T1WI, T1-weighted imaging; DWI, BCG, intravesical Bacillus CalmetteeGue diffusion-weighted image; ADC, apparent diffusion coefficient; PZ, peripheral zone; OM, obturator muscle. The SI of the lesion on T2WI was compared with the SI of urine, ischial BM, and the obturator muscle (OM). The SI of the lesion on T1WI was compared with the SI of the OM. The SI of the lesion on DWI was compared with the SI of the normal PZ.
T. Suzuki et al. / Clinical Radiology 68 (2013) 595e599
597
TR/TE; 80 flip angle; 256 179 matrix; 4 mm section thickness; 23 cm field of view; 21e27 sections; 63.4e81.1 s acquisition time) and unenhanced T1WI (5.7/2.7 ms TR/TE; 10 flip angle; 240 191 matrix; 2 mm section thickness; 35 cm field of view; 185 sections; 97.2 s acquisition time), respectively.
Image analysis Qualitative study All MRI image sets were reviewed by two radiologists with 5 and 8 years of experience. On T2WI, the presence of a well-circumscribed nodular lesion and its shape and size were evaluated according to the report of Ma et al.8 The presence of a diffuse lesion seen as a decrease of SI in the PZ according to their report. In addition, the presence of a cystic lesion showing a circular nodule with high SI on T2WI was evaluated. Whether the lesion extended into the transition zone (TZ) or periprostatic tissue was also evaluated using T2WI. The SIs of nodular and diffuse lesions on T2WI were compared with the SI of the ischial bone marrow and obturator muscle. If only the obturator muscle was chosen as the reference point, all lesions would show iso or higher SI, because the SI of the obturator muscle is extremely low on T2WI. As the signal intensity of bone marrow is intermediate, bone marrow was chosen in addition to the obturator muscle as a reference point. The SI of cystic lesions on T2WI was compared to that of urine. On T1WI and DWI, the SI of the lesions showing abnormal SI on T2WI was evaluated. On T1WI, the SI of the lesions was compared with that of the obturator muscle. On DWI, the SI of the lesion was compared with the SI of the normal PZ in 75- and 68-year-old men whose ages nearly matched the average age of the patients in the present study.
Quantitative study A radiologist with 5 years of experience measured the mean ADC values of these lesions. When multiple sametype lesions were present, their mean ADC value was calculated.
Results Qualitative study The radiological features of GP are summarized in Table 1. On T2WI, two nodular lesions were seen in two of the 10 patients (Fig 1); a diffuse lesion was seen in six of the 10 patients (Fig 2); a cystic lesion was seen in one (Fig 3); and a diffuse lesion with a cystic part was seen in one patient. Both of the cystic lesions had mural nodules. All nodular lesions showed low SI similar to that of muscle on T2WI. The size of the nodular lesions ranged from 8e16 mm (mean 12 mm) in diameter. All nodular lesions had a polygonal shape with notches. Three of the four nodular lesions protruded to the periprostatic tissue from the PZ and the remaining lesion was within the PZ. Six of the seven diffuse lesions showed slightly lower SI than that of the bone marrow on T2WI. The remaining one was iso-intense
Figure 1 Nodular lesions of granulomatous prostatitis (case 8). (a) T2WI of two polygonal nodules with notches (arrows) seen in the bilateral PZ. The signal intensities of the nodules are markedly low and iso-intense to the obturator muscle. The lesion in the right PZ extends to the periprostatic tissue (arrowhead). (b) DWI of the nodules in the right (arrowhead) and left (arrow) PZ are iso-intense to the PZ and slightly more hyperintense than the PZ, respectively.
to that of bone marrow. Five and two diffuse lesions were seen in bilateral and unilateral prostatic lobes, respectively. Two of the seven diffuse lesions extended into the TZ from the PZ. No diffuse lesion extended to the periprostatic tissue. There were two cystic lesions in two cases. The fluid in the cystic lesion and mural nodule showed markedly high and low SI on DWI, respectively. One of the two cystic lesions protruded to the periprostatic tissue and the other was within the PZ. On T1WI, all nodular lesions were iso-intense to muscle and all diffuse lesions showed higher SI than muscle. The fluids in the cystic lesion and mural nodule were iso-intense
598
T. Suzuki et al. / Clinical Radiology 68 (2013) 595e599
to or slightly more hyperintense than muscle. On DWI, all seven diffuse lesions showed higher SI than the normal PZ. Two of the four nodular lesions showed slightly lower SI than the normal PZ. Two of the four nodular lesions showed slightly higher SI than the normal PZ. The fluid in the cystic lesion showed markedly high SI on DWI. One of the two mural nodules showed slightly higher SI than the normal PZ. The other showed lower SI than the normal PZ.
Quantitative study The ADC values of nodular and diffuse lesions ranged from 0.84e1.40 103 mm2/s (mean 1.04 103 mm2/s) and from 0.97e1.90 103 mm2/s (mean 1.36 103 mm2/ s), respectively. The ADC values of the fluids within the cystic lesion and mural nodules ranged from 1.24e1.72 103 mm2/s (mean 1.47 103 mm2/s) and from 1.07e1.79 103 mm2/s (mean 1.35 103 mm2/s), respectively.
Discussion GP is one of the complications of intravesical BCG therapy for superficial bladder cancer.1 Radiological features of GP have been reported to be non-specific; therefore, a prostatic biopsy might be necessary to differentiate GP from prostatic cancer or prostatic invasion of bladder cancer.8,10 In the present study, the SI of GP on T1WI, DWI, and T2WI was evaluated. In this study, nodular and diffuse lesions were seen, in agreement with the study of Ma et al.8 The polygonal shape of the nodule with notches and markedly low SI on T2WI were characteristic findings and they were considered to be diagnostic keys to differentiate GP from prostatic cancer or prostatic invasion of bladder cancer. The protrusion of the nodule to the periprostatic tissue, which was seen in three of the four cases in this study, has never been reported. The SI of the nodular lesions were similar to or higher than those of the normal PZ on DWI; the reason for this diversity of SI is unknown, but it is supposed that a proportion of the lymphocytes associated with the acute inflammation and caseous necrosis might be related to the SI of the nodule because DWI reflects the cell density of the lesion.11 The nodular lesions were iso-intense to muscle on T1WI and this appeared to be non-specific, because the SI of benign prostatic hyperplasia and prostate cancer are often similar to that of muscle on T1WI. The SI of the diffuse lesions on T2WI were similar to the SI of bone marrow on T2WI and it was not as low as that of the nodular lesion but lower than the SI of the normal PZ. The diffuse lesions were seen across the TZ and PZ, and the SI of the lesion on DWI was high. It was thought that the findings might reflect acute prostatic inflammation. Lim et al.12 reported that mass-forming splenic tuberculosis showed slightly high SI
Figure 2 Diffuse lesions of granulomatous prostatitis (case 4). (a) T2WI. The SI of the left PZ is diffusely decreased (arrow). (b) DWI. The SI of the lesion is diffusely increased (arrows). (c) T1WI. The SI of the PZ (arrow) is diffusely increased.
T. Suzuki et al. / Clinical Radiology 68 (2013) 595e599
599
showed markedly high SI on DWI. Mural nodules showed markedly low SI on T2WI. Although the cause of this condition is unknown, it might involve fibrosis, caseous necrosis, clots, or calcifications, judging from the low SI on T2WI. However, the presence of calcification was not confirmed at CT. The present study had several limitations. The retrospective nature of the study might have introduced some form of patient selection bias. The number of GP was small, and further studies are needed to confirm the characteristic MRI findings of GP. In addition, differing MRI protocols were used and the majority of the patients did not have a dedicated prostate MRI. MRI findings could not be correlated with pathological findings because all specimens were obtained by random biopsy. In conclusion, the three main patterns of GP identified were diffuse, nodular, and cystic with mural nodule. The diffuse type was the most common.
References
Figure 3 Cystic lesion of granulomatous prostatitis (case 9). (a) T2WI. There is a cystic lesion showing markedly high signal intensity (arrow) and a mural nodule showing low SI (arrowhead) within the cystic lesion. The lesion extends to the periprostatic tissue (arrowhead). (b) DWI. Both the fluid within the cyst and the mural nodule show high signal intensity.
on fat-suppressed unenhanced T1WI. In tuberculomas, paramagnetic substances, such as macrophage-laden oxygen-free radicals, were thought to be the factor that caused the reduction of the T1 and T2 values.13,14 There have been no reports describing that GP was accompanied by a cystic lesion. It was thought that the content within the cystic lesion might be an abscess or protein-rich fluid produced by inflammation because it
1. Lamm DL, Stogdill VD, Stogdill BJ, et al. Complications of bacillus CalmetteeGuerin immunotherapy in 1,278 patients with bladder cancer. J Urol 1986;135:272e4. 2. Morales A, Eidinger D, Bruce AW. Intracavitary Bacillus CalmetteeGuerin in the treatment of superficial bladder tumors. J Urol 1976;116:180e3. rin vaccine-induced 3. Torres GM, Kaude JV, Drylie D. Bacille CalmetteeGue granulomatous prostatitis: another hypoechoic nonneoplastic lesion. AJR Am J Roentgenol 1990;155:195e6. 4. Stilmant M, Siroky MB, Johnson KB. Fine needle aspiration cytology of granulomatous prostatitis induced by BCG immunotherapy of bladder cancer. Acta Cytol 1985;29:961e6. 5. Oates RD, Stilmant MM, Freedlund MC, et al. Granulomatous prostatitis following Bacillus CalmetteeGuerin immunotherapy of bladder cancer. J Urol 1988;140:751e4. 6. Uzoh CC, Uff JS, Okeke AA. Granulomatous prostatitis. BJU Int 2007;99:510e2. 7. Noor MA, Biyabani SR, Burney IA, et al. Symptomatic granulomatous prostatitis following Bacillus CalmetteeGuerin immunotherapy for bladder cancer. J Pak Med Assoc 2002;52:578e80. 8. Ma W, Kang SK, Hricak H, et al. Imaging appearance of granulomatous disease after intravesical Bacille CalmetteeGuerin (BCG) treatment of bladder carcinoma. AJR Am J Roentgenol 2009;192:1494e500. 9. Naik KS, Carey BM. The transrectal ultrasound and MRI appearances of granulomatous prostatitis and its differentiation from carcinoma. Clin Radiol 1999;54:173e5. 10. Takeuchi M, Suzuki T, Sasaki S, et al. Clinicopathologic significance of high signal intensity on diffusion-weighted MR imaging in the ureter, urethra, prostate and bone of patients with bladder cancer. Acad Radiol 2012;19:827e33. 11. Qayyum A. Diffusion-weighted imaging in the abdomen and pelvis: concepts and applications. RadioGraphics 2009;29:1797e810. 12. Lim J, Yu JS, Hong SW, et al. A case of mass-forming splenic tuberculosis: MRI findings with emphasis of diffusion-weighted imaging characteristics. J Korean Med Sci 2011;26:457e60. 13. Chung MH, Lee HG, Kwon SS, et al. MR imaging of solitary pulmonary lesion: emphasis on tuberculomas and comparison with tumors. J Magn Reson Imaging 2000;11:629e37. 14. Sze G, Zimmerman RD. The magnetic resonance imaging of infections and inflammatory diseases. Radiol Clin North Am 1988;26:839e59.