- Email: [email protected]
Testicular ultrasonography after chemotherapy for germ cell neoplasm: Significance of highly hyperechoic lesions
TESTICULAR ULTRASONOGRAPIW AFTER CHEMOTHERAPY FOR GERM CELL NEOPLASM: SIGNIFICANCE OF HIGHLY HYI’ERECHOIC LESIONS KENYON K. KOPECKY, M.D. MICHAEL T. BRENDLE, M.D. JON T. HARMAN, M.D.
RICHARD S. FOSTER, M.D. RICHARD BIHRLE, M.D. PATRICK J. LOEHRER, M.D.
From the Departments of Radiology, Urology, and Medicine, Indiana University Medical Center, Indianapolis, Indiana ABSTRACTThe sonographic appearance of the testis after administration of chemotherapy for metastatic germ cell neoplasm is not well known. Fifty-six patients (60 testes) who were previously treated with chemotherapy for metastatic germ cell neoplasm (originally diagnosed by removal of the contralateral testis or by biopsy of metastatic disease) underwent sonography followed by orchiectomy. The sonographic characteristics found to predict viable intratestitular tumor were: lesion size larger than 5 mm, fewer echoes than adjacent parenchyma (hypoechoic), inhomogeneous echo texture, poor margin definition, cystic areas, or highly hyperechoicfoci within a hypoechoic lesion. Fibrosis was predicted by finding single or multiple small, highly hyperechoic lesions. These results suggest the potential for predicting the pathologic diagnosis in some patients after receiving chemotherapy for germ cell neoplasm.
Approximately 60 percent of patients with testicular cancer present with hematogenous and/ or lymphatic metastases.’ Some of these patients are treated with chemotherapy without orchiectomy based on biopsy results of metastatic lesions or clinical presentation. Some patients experience regression or resolution of metastatic tumor with chemotherapy while harboring viable intratesticular tumor. This has led some investigators to believe that the testis is a “sanctuary” for tumor due to the blood-testis barrier. Consequently, orchiectomy is sometimes performed after administration of chemotherapy for metastatic germ cell cancer.2~5 Also, patients with a tumor of one testis have an increased risk of a synchronous or metachronous tumor of the contralateral testis developing; hence, removal of the remaining testis is sometimes necessary.2 The sensitivity of ultrasonography in the detection of testicular neoplasms is nearly 100 percent. Extratesticular and intratesticular lesions are distinguished with 95 to 99 percent ac-
curacy.4.” Sonography is useful in detecting nonpalpable primary testicular tumors in patients who present with metastatic germ cell tumors.3.46 Sonography has not been accurate, however, in distinguishing benign from malignant lesions within the testis.4,6,7 Because the large majority of intratesticular masses are malignant,4,” and because the ultrasound appearances of testicular lesions are not tissue specific, the identification of any intratesticular abnormality prompts orchiectomy in most patients, including those previously treated with chemotherapy.4.fi.8 Some patients with metastatic germ cell tumors have undergone orchiectomy and have been found to have fibrous tissue within the testis, thought to represent “regressed” or “burned out” germ cell tumor. The testicular ultrasound findings in several patients with “regressed” germ cell tumor have been described,4.Y l1 however, there have been no studies of post-chemotherapy patients which attempted to analyze the sonographic abnormalities in a systematic
490
UROLOGY
/ MAY1992
/ VOLUMEXXXIX,NUMBER5
fashion. This study was undertaken to characterize the testicular ultrasound findings in patients who have received chemotherapy for germ cell neoplasm and correlate them with the physical examination and with histologic findings at orchiectomy. In particular, we were interested in determining if there were any ultrasound characteristics predictive of benign disease which could potentially obviate the need for orchiectomy in some patients. Material
and Methods
All patients who had scrotal ultrasonography followed by orchiectomy between May 1983 and January 1991 were retrospectively evaluated. The medical records, sonograms, and pathologic reports of 69 testes in 65 men were available for review. Two men had non-germ cell neoplasms. Sixty-three men had germ cell neoplasms, and 56 of them, with a mean age of twenty-nine years (range 15-57 years), previously received chemotherapy for metastatic disease. Sixty of their testes were studied sonographically and histologically, and they form the basis of this study. Eleven of these 56 patients had undergone a previous orchiectomy. Physical examinations were performed by a urologist, and the detection of any intratesticular mass was recorded as abnormal. The sonograms were performed with high-resolution real-time equipment (Technicare Autosector with a ;‘.5 MHz transducer [Johnson & Johnson, Cleveland, OH] or Acuson 128 with 5 or 7 MHz linear array transducers [Acuson, Mountain View, CA]). The .sonograms were reviewed without knowledge of the surgical pathologic findings. The median time interval between ultrasonography and orchiectomy was seven days (mean 33, range O-179 days). The sonographic findings in each testis were categorized and recorded as follows: (1) the number and size (longest dimension) of lesions; (2) the echogenicity of each lesion was subjectively categorized as highly hyperechoic, slightly hvpereor hypoechoic relative to isoechoic, choic, normal parenchyma of the ipsilateral or contralateral testis: (3) the echogenicity of lesions was categorized as homogeneous or inhomogeneous: (4) the margins of the lesions were described as sharply defined or poorly defined; (5) the presence or absence of cysts within lesions was noted; (6) in lesions which were predominantly hypoechoic, the presence or absence of highly,
TABI.I
65 88
Positive pred. \.alue Negative pred. value
72 83
100 13 36 100
hyperechoic foci within the lesion was noted. The presence or absence of acoustical shadowing behind highly hyperechoic lesions was not recorded because the degree of shadow+ng appeared to depend more on the size of the lesion, a v:ariable already recorded, and the position of the lesion relative to the focal zone of the transducer, rather than on the acoustic properties of the lesion itself (Figs. 1 and 2). Results Histologic examination of the orchiectomy specimen in the men previously treated with chemotherapy. revealed normal tissue or atrophy in 10 testes (17 % ), fibrosis in 30 testes (50 % ), and tumor in 20 testes (33 4h ) ( 17 pure teratomas and 3 mixed tumors). There was no association between patient age and histologic findings (Wilcoxon ranked sum test, p = 0.24). There was no association between the time interval between u1trasonograph.y and surgery and histologic findings (Wilcoxon ranked sum test, p ==0.13). The sensitivity, specificity. positive predictive and negative predictive values f’or the detection of tumor by physical examination and by sonography are given in Table I. For this table any testicular abnormality was considered suspicious for tumor. The sensitivity of sonography may actually be less than 100 percent because patients do not undergo orchiectomy at our institution if the sonogram is normal unless clinical suspicion of tumor is extremely high (for example, predominantly one-sided retroperitoneal adenopathy). None of the ,3 patients in this series who had normal findings on sonogram had tumor at orchiectomy. The correlations between the ultrasound characteristics of the testicular lesions and the presence of tiimor are shown in Table II. Tumor was predicted by the finding of a lesion (1) with
(A-E)Small, highly hyperechoic intraFIGURE 1. testicular lesions, all representing fibrosis in 5 different patients.
a size larger than 5 mm (odds ratio = 23, p = 0.002), (2) with fewer echoes than normal parenchyma (hypoechoic) (odds ratio = 4.2, p 0.019), (3) with poor margin definition (odds r:tio = 7.6, p = 0.002), (4) with inhomogeneous echo texture (odds ratio = 19.6, p = O.OOOl), (5) with cystic areas (odds ratio = 14.6, p = 0.004), and (6) with hyperechoic foci within a hypoechoic lesion (odds ratio = 8.4, p = 0.004). For this analysis the variable lesion size was collapsed to < 5 mm versus 2 5 mm, and the variable echogenicity was collapsed to hypoechoic versus all others. Of the 30 testes with fibrosis as the only histologic finding, ultrasound demonstrated highly hyperechoic lesions in 16 testes, a hyperechoic lesion in 1 testis, hypoechoic lesions in 11 testes, and normal findings in 2 testes.
492
UROLOGY
/ MAY1992
/ VOLUMEXXXIX,NUMBER5
FIGURE
28. (A and B) Two views of 7-mm highly
calcijication
found
hyperechoic
Highly hyperechoic lesions were found in 18 testes of 16 patients. These lesions were solitary in 7 testes and multiple in 11 testes. Lesions in 13 of these measured less than 5 mm in greatest diameter (Fig. l), and all represented fibrosis histologically. Sixteen of these 18 testes were normal to physical examination, and these 16 had on1.y fibrosis in the orchiectomy specimen. TA ISIX II. Correlation of sonographic characteristics of testicular lesions with presence of tumor in patients previously treated with chemotherapy
Sonoqaphic Characteristics -_-___ Lesion size <5 mm S-10 mm lo-20 mm 220 mm Echogenicitb Hypoechoic Isoechk! Hyperechoic Very hyperechoic Margins Sharply defined Poorly defined Echotexture Homogeneous Inhomogeneous Cysts Absent Present Hyperechoic foci Absent Present
intratesticular
lesion; &mm
teratoma
with
at orchiectomy.
No. of Testes With (%) Tumor/Lesions Oil3 9122 7114 416
(0) (41) (SO) (67)
16133 (48) l/l (100) 113 (33) 2118 (11) 3123 (13) 17132 (53) 2126 (8) 18129 (62) 14148 (29) 617 (86) 7123 (30) 11114 (79) -__
The two palpably abnormal testes contained teratoma with calcifications; one lesion measured 7 mm in greatest diameter (Fig. 2), and the other 20 mm. Fibrosis was present in 16 of 16 testes (100%) with highly hyperechoic, nonpalpable lesions. Comment Testicular cancer is the most common malignancy in men between the ages of twenty and thirty-four years. Ninety-five percent of neoplasms originating from the testis are of germ cell origin. Of those, 35 percent are seminoma, 15 percent are embryonal carcinoma, 5 percent are teratoma, and 4-Opercent show more than one cell type.H The tumor distribution in this study of patients having orchiectomies after receiving chemotherapy for metastatic germ cell tumors was much different from the untreated population: 17 of 120tumors and 3 of 20 (15 % ) (85 % ) were pure teratomas. were mixed tumors. This relatively higher prevalence of intratesticular teratoma persisting after chemotherapy is similar to the prevalence of teratoma found in patients undergoing retroperitoneal lymph node dissection for persistent disease after chemotherapy for nonseminomatous germ cell tumors.r2 Ultrasonographically, most testicular cancers are hypoechoic compared with adjacent normal parenchyma. Seminomas typically, are homogeneous, hypoechoic masses. Nonseminomatous tumors are typically more inhomogeneous, and some, especially teratomas. contain cystic areas or dense highly hyperechoic foci which may be
493
caused by calcification, cartilage, bone, or fibrosis.7 In this study, the numbers of patients were too small to ultrasonically characterize the specific type of tumors. In the patients previously treated with chemotherapy, the presence of tumor (85% of which were pure teratomas) was predicted by a lesion (1) larger than 5 mm in diameter, (2) with fewer echoes than normal parenchyma (hypoechoic), (3) with poor margin definition, (4) with inhomogeneous echo texture, (5) with cystic areas, or (6) with highly hyperechoic foci (within hypoechoic lesions). Thirty of 60 testes in 56 patients in this study who had previously received chemotherapy for metastatic germ cell tumor were found to have fibrosis at orchiectomy, which would be consistent with the diagnosis of “regressed” or “burned out” germ cell tumor. Eleven of these testes had hypoechoic lesions, 3 had isoechoic or mildly hyperechoic lesions, and 16 had highly hyperechoic lesions. These sonographic characteristics are similar to those described in previous case reports,4.gm11 however, the relative prevalence of the sonographic characteristics of regressed tumors has never been investigated. In this experience over half of testes (16 of 30) with regressed tumors (fibrosis) had highly hyperechoic lesions. In this experience, 18 of 60 testes were found to have highly hyperechoic lesions, and 16 of the 18 contained only fibrosis. Thirteen of these testes had lesions that measured less than 5 mm in diameter, and all 13 represented fibrosis. Two of 5 testes with highly hyperechoic lesions measuring more than 5 mm in diameter contained teratoma, and both of these were palpable. In our experience, the positive predictive value of a palpable intratesticular mass in predicting the presence of tumor was 72 percent, a value high enough to justify orchiectomy regardless of the sonographic findings. Sixteen of 60 testes (27 % ) in this series had highly hyperechoic, nonpalpable intratesticular lesions, and none had tumor at orchiectomy. The prevalence of bright testicular benign lesions in the general population is not well known. Leung, Gooding, and Williams13 studied 40 asymptomatic men ranging in age from twenty-four to ninety years with ultrasound and found no hyperechoic foci, and Vick et al. I4 reviewed 506 scrotal lesions examined with ultrasound and found only 5 intratesticular hyperechoic masses (all benign). Therefore, it is likely, that the bright lesions discovered in the
494
present study are a consequence of the prior cancer and/or chemotherapy. In this experience, all 13 testes with small (< 5 mm), highly hyperechoic lesions (100 % ) had only fibrosis at orchiectomy. The nee’d for orchiectomy in this subset of cancer patients must be carefully weighed. The morbidity of orchiectomy in post-chemotherapy patients must be assessed compared with the morbidity of not removing a potential site of incompletely treated tumor which may result in late recurrence of neoplastic disease. Also, the psychologic, hormonal, and fertility effects of orchiectomy must be considered, especially if the contralateral testis already has been removed. If orchiectomy is not performed, follow-up physical examination and ultrasonography may aid in excluding the presence of tumor and confirming the suspected diagnosis of fibrosis. Indiana
University Hospital, Room 0285 926 West Michigan Street Indianapolis, Indiana 46202-5253 (DR. KOPECKY) References
1. Einhorn LII, it al: Cancer of the testes, in DcVita VT (Ed): Cancer: Principles and Practice of Oncology. cd 2. Philadelphia, JB Lippincott, 1989. pp 979-1011. 2. Fowler JE. Vugrin D. Cvitkovic E, and \l:hitmore LVF: Sequential bilateral germ cell tumors of the testis despite interval chemotherapy. J Urol 122: 421 (1979). 3. Glazer HS, Lee JKT, Melson CL. and h4cClcnnan HI,: Sonographic detection of occult testicular neoplasms, AJR 138: 673 (1982). 4. Benson CR: The role of ultrasound in diagnosis and staging of testicular cancer, Semin Urol 6: 189 (1988). 5. Griest A, et al: Pathologic findings at orchiectomy following chemotherapy for disseminated testicular cancer. J Clin Oncol 2: 1025 (1984). 6. Rifkin MD, Kurtz AB, Pasta ME, and Goldberg BB: Diagnostic capabilities of high-resolution scrotal ultrasonographyi prospe&ve evaluation, J Ultrasound Mcd 4: 13 (1985). 7. SchLverk WB. Sch\verk IVN. and Rodeck C: Testicular tumors: prospective analysis of real-time US patterns and ahdominal staging, Radiology 164: 369 (1987). 8. Johnson DE, Swanson DA, and van Eschenbach AC: Tumars of the genitourinary tract, in: General Urolog); ed 11, Los Altos, CA. Lange Medical Publications, 1984, pp 306-404. 9. Shaxrker TH, ct (11: Ultrasonographic detection of “hurncdout” primary testicular germ ccl1 tumors in clinically normal testes, J Ultrasound LMcd 2: 477 (1983). 10. Crantham JC. c’f al: Testicular neoplasms: 29 tumors studied by high-resolution US, RadiolokT 157: 775 (1985). 11. van Dijk R, IIitgc-Boetes C, Debruync FMJ, and Rosenbusch G: Sonographic detection of a nonpalpable regressed germcell tumor in an atrophic testis, J Clin Ultrasound 17: 594 (1989). 12. I.oehrer PJ, cl ul: Teratoma following cisplatin-based combination chemotherapy for nonscminomatous germ ccl1 tumors: a clinicopathological correlation, J Urol 135: 1183 (1986). 13. Leung ML, Gooding CAIl: and Williams RD: IIigh-resolution sonography of scrotal contents in asymptomatic subjects, AJR 143: 161 (1984). 14. Vick 0%‘. ct al: Scrotal masse\ with a uniformly hyperechoic pattern, Radiolok? 148: 209 (1983).
UROLOGY
/
MAY 1992
i
VOLUME
XXXIX, NUMBER
5
RICHARD S. FOSTER, M.D. RICHARD BIHRLE, M.D. PATRICK J. LOEHRER, M.D.
From the Departments of Radiology, Urology, and Medicine, Indiana University Medical Center, Indianapolis, Indiana ABSTRACTThe sonographic appearance of the testis after administration of chemotherapy for metastatic germ cell neoplasm is not well known. Fifty-six patients (60 testes) who were previously treated with chemotherapy for metastatic germ cell neoplasm (originally diagnosed by removal of the contralateral testis or by biopsy of metastatic disease) underwent sonography followed by orchiectomy. The sonographic characteristics found to predict viable intratestitular tumor were: lesion size larger than 5 mm, fewer echoes than adjacent parenchyma (hypoechoic), inhomogeneous echo texture, poor margin definition, cystic areas, or highly hyperechoicfoci within a hypoechoic lesion. Fibrosis was predicted by finding single or multiple small, highly hyperechoic lesions. These results suggest the potential for predicting the pathologic diagnosis in some patients after receiving chemotherapy for germ cell neoplasm.
Approximately 60 percent of patients with testicular cancer present with hematogenous and/ or lymphatic metastases.’ Some of these patients are treated with chemotherapy without orchiectomy based on biopsy results of metastatic lesions or clinical presentation. Some patients experience regression or resolution of metastatic tumor with chemotherapy while harboring viable intratesticular tumor. This has led some investigators to believe that the testis is a “sanctuary” for tumor due to the blood-testis barrier. Consequently, orchiectomy is sometimes performed after administration of chemotherapy for metastatic germ cell cancer.2~5 Also, patients with a tumor of one testis have an increased risk of a synchronous or metachronous tumor of the contralateral testis developing; hence, removal of the remaining testis is sometimes necessary.2 The sensitivity of ultrasonography in the detection of testicular neoplasms is nearly 100 percent. Extratesticular and intratesticular lesions are distinguished with 95 to 99 percent ac-
curacy.4.” Sonography is useful in detecting nonpalpable primary testicular tumors in patients who present with metastatic germ cell tumors.3.46 Sonography has not been accurate, however, in distinguishing benign from malignant lesions within the testis.4,6,7 Because the large majority of intratesticular masses are malignant,4,” and because the ultrasound appearances of testicular lesions are not tissue specific, the identification of any intratesticular abnormality prompts orchiectomy in most patients, including those previously treated with chemotherapy.4.fi.8 Some patients with metastatic germ cell tumors have undergone orchiectomy and have been found to have fibrous tissue within the testis, thought to represent “regressed” or “burned out” germ cell tumor. The testicular ultrasound findings in several patients with “regressed” germ cell tumor have been described,4.Y l1 however, there have been no studies of post-chemotherapy patients which attempted to analyze the sonographic abnormalities in a systematic
490
UROLOGY
/ MAY1992
/ VOLUMEXXXIX,NUMBER5
fashion. This study was undertaken to characterize the testicular ultrasound findings in patients who have received chemotherapy for germ cell neoplasm and correlate them with the physical examination and with histologic findings at orchiectomy. In particular, we were interested in determining if there were any ultrasound characteristics predictive of benign disease which could potentially obviate the need for orchiectomy in some patients. Material
and Methods
All patients who had scrotal ultrasonography followed by orchiectomy between May 1983 and January 1991 were retrospectively evaluated. The medical records, sonograms, and pathologic reports of 69 testes in 65 men were available for review. Two men had non-germ cell neoplasms. Sixty-three men had germ cell neoplasms, and 56 of them, with a mean age of twenty-nine years (range 15-57 years), previously received chemotherapy for metastatic disease. Sixty of their testes were studied sonographically and histologically, and they form the basis of this study. Eleven of these 56 patients had undergone a previous orchiectomy. Physical examinations were performed by a urologist, and the detection of any intratesticular mass was recorded as abnormal. The sonograms were performed with high-resolution real-time equipment (Technicare Autosector with a ;‘.5 MHz transducer [Johnson & Johnson, Cleveland, OH] or Acuson 128 with 5 or 7 MHz linear array transducers [Acuson, Mountain View, CA]). The .sonograms were reviewed without knowledge of the surgical pathologic findings. The median time interval between ultrasonography and orchiectomy was seven days (mean 33, range O-179 days). The sonographic findings in each testis were categorized and recorded as follows: (1) the number and size (longest dimension) of lesions; (2) the echogenicity of each lesion was subjectively categorized as highly hyperechoic, slightly hvpereor hypoechoic relative to isoechoic, choic, normal parenchyma of the ipsilateral or contralateral testis: (3) the echogenicity of lesions was categorized as homogeneous or inhomogeneous: (4) the margins of the lesions were described as sharply defined or poorly defined; (5) the presence or absence of cysts within lesions was noted; (6) in lesions which were predominantly hypoechoic, the presence or absence of highly,
TABI.I
65 88
Positive pred. \.alue Negative pred. value
72 83
100 13 36 100
hyperechoic foci within the lesion was noted. The presence or absence of acoustical shadowing behind highly hyperechoic lesions was not recorded because the degree of shadow+ng appeared to depend more on the size of the lesion, a v:ariable already recorded, and the position of the lesion relative to the focal zone of the transducer, rather than on the acoustic properties of the lesion itself (Figs. 1 and 2). Results Histologic examination of the orchiectomy specimen in the men previously treated with chemotherapy. revealed normal tissue or atrophy in 10 testes (17 % ), fibrosis in 30 testes (50 % ), and tumor in 20 testes (33 4h ) ( 17 pure teratomas and 3 mixed tumors). There was no association between patient age and histologic findings (Wilcoxon ranked sum test, p = 0.24). There was no association between the time interval between u1trasonograph.y and surgery and histologic findings (Wilcoxon ranked sum test, p ==0.13). The sensitivity, specificity. positive predictive and negative predictive values f’or the detection of tumor by physical examination and by sonography are given in Table I. For this table any testicular abnormality was considered suspicious for tumor. The sensitivity of sonography may actually be less than 100 percent because patients do not undergo orchiectomy at our institution if the sonogram is normal unless clinical suspicion of tumor is extremely high (for example, predominantly one-sided retroperitoneal adenopathy). None of the ,3 patients in this series who had normal findings on sonogram had tumor at orchiectomy. The correlations between the ultrasound characteristics of the testicular lesions and the presence of tiimor are shown in Table II. Tumor was predicted by the finding of a lesion (1) with
(A-E)Small, highly hyperechoic intraFIGURE 1. testicular lesions, all representing fibrosis in 5 different patients.
a size larger than 5 mm (odds ratio = 23, p = 0.002), (2) with fewer echoes than normal parenchyma (hypoechoic) (odds ratio = 4.2, p 0.019), (3) with poor margin definition (odds r:tio = 7.6, p = 0.002), (4) with inhomogeneous echo texture (odds ratio = 19.6, p = O.OOOl), (5) with cystic areas (odds ratio = 14.6, p = 0.004), and (6) with hyperechoic foci within a hypoechoic lesion (odds ratio = 8.4, p = 0.004). For this analysis the variable lesion size was collapsed to < 5 mm versus 2 5 mm, and the variable echogenicity was collapsed to hypoechoic versus all others. Of the 30 testes with fibrosis as the only histologic finding, ultrasound demonstrated highly hyperechoic lesions in 16 testes, a hyperechoic lesion in 1 testis, hypoechoic lesions in 11 testes, and normal findings in 2 testes.
492
UROLOGY
/ MAY1992
/ VOLUMEXXXIX,NUMBER5
FIGURE
28. (A and B) Two views of 7-mm highly
calcijication
found
hyperechoic
Highly hyperechoic lesions were found in 18 testes of 16 patients. These lesions were solitary in 7 testes and multiple in 11 testes. Lesions in 13 of these measured less than 5 mm in greatest diameter (Fig. l), and all represented fibrosis histologically. Sixteen of these 18 testes were normal to physical examination, and these 16 had on1.y fibrosis in the orchiectomy specimen. TA ISIX II. Correlation of sonographic characteristics of testicular lesions with presence of tumor in patients previously treated with chemotherapy
Sonoqaphic Characteristics -_-___ Lesion size <5 mm S-10 mm lo-20 mm 220 mm Echogenicitb Hypoechoic Isoechk! Hyperechoic Very hyperechoic Margins Sharply defined Poorly defined Echotexture Homogeneous Inhomogeneous Cysts Absent Present Hyperechoic foci Absent Present
intratesticular
lesion; &mm
teratoma
with
at orchiectomy.
No. of Testes With (%) Tumor/Lesions Oil3 9122 7114 416
(0) (41) (SO) (67)
16133 (48) l/l (100) 113 (33) 2118 (11) 3123 (13) 17132 (53) 2126 (8) 18129 (62) 14148 (29) 617 (86) 7123 (30) 11114 (79) -__
The two palpably abnormal testes contained teratoma with calcifications; one lesion measured 7 mm in greatest diameter (Fig. 2), and the other 20 mm. Fibrosis was present in 16 of 16 testes (100%) with highly hyperechoic, nonpalpable lesions. Comment Testicular cancer is the most common malignancy in men between the ages of twenty and thirty-four years. Ninety-five percent of neoplasms originating from the testis are of germ cell origin. Of those, 35 percent are seminoma, 15 percent are embryonal carcinoma, 5 percent are teratoma, and 4-Opercent show more than one cell type.H The tumor distribution in this study of patients having orchiectomies after receiving chemotherapy for metastatic germ cell tumors was much different from the untreated population: 17 of 120tumors and 3 of 20 (15 % ) (85 % ) were pure teratomas. were mixed tumors. This relatively higher prevalence of intratesticular teratoma persisting after chemotherapy is similar to the prevalence of teratoma found in patients undergoing retroperitoneal lymph node dissection for persistent disease after chemotherapy for nonseminomatous germ cell tumors.r2 Ultrasonographically, most testicular cancers are hypoechoic compared with adjacent normal parenchyma. Seminomas typically, are homogeneous, hypoechoic masses. Nonseminomatous tumors are typically more inhomogeneous, and some, especially teratomas. contain cystic areas or dense highly hyperechoic foci which may be
493
caused by calcification, cartilage, bone, or fibrosis.7 In this study, the numbers of patients were too small to ultrasonically characterize the specific type of tumors. In the patients previously treated with chemotherapy, the presence of tumor (85% of which were pure teratomas) was predicted by a lesion (1) larger than 5 mm in diameter, (2) with fewer echoes than normal parenchyma (hypoechoic), (3) with poor margin definition, (4) with inhomogeneous echo texture, (5) with cystic areas, or (6) with highly hyperechoic foci (within hypoechoic lesions). Thirty of 60 testes in 56 patients in this study who had previously received chemotherapy for metastatic germ cell tumor were found to have fibrosis at orchiectomy, which would be consistent with the diagnosis of “regressed” or “burned out” germ cell tumor. Eleven of these testes had hypoechoic lesions, 3 had isoechoic or mildly hyperechoic lesions, and 16 had highly hyperechoic lesions. These sonographic characteristics are similar to those described in previous case reports,4.gm11 however, the relative prevalence of the sonographic characteristics of regressed tumors has never been investigated. In this experience over half of testes (16 of 30) with regressed tumors (fibrosis) had highly hyperechoic lesions. In this experience, 18 of 60 testes were found to have highly hyperechoic lesions, and 16 of the 18 contained only fibrosis. Thirteen of these testes had lesions that measured less than 5 mm in diameter, and all 13 represented fibrosis. Two of 5 testes with highly hyperechoic lesions measuring more than 5 mm in diameter contained teratoma, and both of these were palpable. In our experience, the positive predictive value of a palpable intratesticular mass in predicting the presence of tumor was 72 percent, a value high enough to justify orchiectomy regardless of the sonographic findings. Sixteen of 60 testes (27 % ) in this series had highly hyperechoic, nonpalpable intratesticular lesions, and none had tumor at orchiectomy. The prevalence of bright testicular benign lesions in the general population is not well known. Leung, Gooding, and Williams13 studied 40 asymptomatic men ranging in age from twenty-four to ninety years with ultrasound and found no hyperechoic foci, and Vick et al. I4 reviewed 506 scrotal lesions examined with ultrasound and found only 5 intratesticular hyperechoic masses (all benign). Therefore, it is likely, that the bright lesions discovered in the
494
present study are a consequence of the prior cancer and/or chemotherapy. In this experience, all 13 testes with small (< 5 mm), highly hyperechoic lesions (100 % ) had only fibrosis at orchiectomy. The nee’d for orchiectomy in this subset of cancer patients must be carefully weighed. The morbidity of orchiectomy in post-chemotherapy patients must be assessed compared with the morbidity of not removing a potential site of incompletely treated tumor which may result in late recurrence of neoplastic disease. Also, the psychologic, hormonal, and fertility effects of orchiectomy must be considered, especially if the contralateral testis already has been removed. If orchiectomy is not performed, follow-up physical examination and ultrasonography may aid in excluding the presence of tumor and confirming the suspected diagnosis of fibrosis. Indiana
University Hospital, Room 0285 926 West Michigan Street Indianapolis, Indiana 46202-5253 (DR. KOPECKY) References
1. Einhorn LII, it al: Cancer of the testes, in DcVita VT (Ed): Cancer: Principles and Practice of Oncology. cd 2. Philadelphia, JB Lippincott, 1989. pp 979-1011. 2. Fowler JE. Vugrin D. Cvitkovic E, and \l:hitmore LVF: Sequential bilateral germ cell tumors of the testis despite interval chemotherapy. J Urol 122: 421 (1979). 3. Glazer HS, Lee JKT, Melson CL. and h4cClcnnan HI,: Sonographic detection of occult testicular neoplasms, AJR 138: 673 (1982). 4. Benson CR: The role of ultrasound in diagnosis and staging of testicular cancer, Semin Urol 6: 189 (1988). 5. Griest A, et al: Pathologic findings at orchiectomy following chemotherapy for disseminated testicular cancer. J Clin Oncol 2: 1025 (1984). 6. Rifkin MD, Kurtz AB, Pasta ME, and Goldberg BB: Diagnostic capabilities of high-resolution scrotal ultrasonographyi prospe&ve evaluation, J Ultrasound Mcd 4: 13 (1985). 7. SchLverk WB. Sch\verk IVN. and Rodeck C: Testicular tumors: prospective analysis of real-time US patterns and ahdominal staging, Radiology 164: 369 (1987). 8. Johnson DE, Swanson DA, and van Eschenbach AC: Tumars of the genitourinary tract, in: General Urolog); ed 11, Los Altos, CA. Lange Medical Publications, 1984, pp 306-404. 9. Shaxrker TH, ct (11: Ultrasonographic detection of “hurncdout” primary testicular germ ccl1 tumors in clinically normal testes, J Ultrasound LMcd 2: 477 (1983). 10. Crantham JC. c’f al: Testicular neoplasms: 29 tumors studied by high-resolution US, RadiolokT 157: 775 (1985). 11. van Dijk R, IIitgc-Boetes C, Debruync FMJ, and Rosenbusch G: Sonographic detection of a nonpalpable regressed germcell tumor in an atrophic testis, J Clin Ultrasound 17: 594 (1989). 12. I.oehrer PJ, cl ul: Teratoma following cisplatin-based combination chemotherapy for nonscminomatous germ ccl1 tumors: a clinicopathological correlation, J Urol 135: 1183 (1986). 13. Leung ML, Gooding CAIl: and Williams RD: IIigh-resolution sonography of scrotal contents in asymptomatic subjects, AJR 143: 161 (1984). 14. Vick 0%‘. ct al: Scrotal masse\ with a uniformly hyperechoic pattern, Radiolok? 148: 209 (1983).
UROLOGY
/
MAY 1992
i
VOLUME
XXXIX, NUMBER
5