- Email: [email protected]
Lymph Node Growth Rate in Testicular Germ Cell Tumours: Implications for Computed Tomography Surveillance Frequency
Clinical Oncology 23 (2011) 333e338 Contents lists available at ScienceDirect
Clinical Oncology journal homepage: www.elsevier.com/locate/clon
Original Article
Lymph Node Growth Rate in Testicular Germ Cell Tumours: Implications for Computed Tomography Surveillance Frequency W. Mazrani *, M.E. O’Malley *, P.W.M. Chung y, P. Warde y, D. Vesprini z, T. Panzarella x * Joint
Department of Medical Imaging, University of Toronto, Canada Department of Radiation Oncology, Princess Margaret Hospital, University of Toronto, Canada z Department of Radiation Oncology, Sunnybrook Odette Cancer Centre, University of Toronto, Canada x Department of Biostatistics, Princess Margaret Hospital Cancer Program and Dalla Lana School of Public Health, University of Toronto, Canada y
Received 10 May 2010; received in revised form 26 July 2010; accepted 5 October 2010
Abstract Aim: To estimate the growth rate of lymph nodes in patients on surveillance for testicular cancer who developed recurrent disease. Materials and methods: During a 7-year period, 318 patients at our institution were managed by surveillance and 39 relapsed (12.3%). The computed tomography scans of 28 patients (median age 32 years; range 19e51 years) who met our inclusion criteria and who developed recurrent disease in the abdomen/pelvis were retrospectively reviewed. Thirteen patients had non-seminoma and 15 had seminoma. To estimate the lymph node growth rate, the slope of lymph node size over time was calculated. Results: The median length of time from orchiectomy to the recurrence computed tomography was 131 days (range 49e520) or about 4.4 months for non-seminoma patients and 373 days (range 129e675) or about 12.3 months for seminoma patients. The median size of the involved lymph node at final computed tomography for seminoma patients was 12 mm (range 9e31 mm) and for non-seminoma patients was 15 mm (range 10e56 mm). The median lymph node growth rate for patients with seminoma was 1.35 mm/month (range 0.62e4.56) and for patients with non-seminoma 2.99 mm/month (range 0.77e7.06); the difference in growth rates was statistically significant (P ¼ 0.029). Conclusions: There is a statistically significant faster growth rate of lymph nodes in patients with recurrent non-seminoma compared with patients with seminoma. This finding supports a more frequent computed tomography schedule during the first 2 years of surveillance in non-seminoma patients compared with seminoma patients. Ó 2010 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. Key words: Computed tomography; growth rate; lymph node; surveillance; testicular cancer
Introduction Testicular cancer is the most common malignancy in males aged 15e34 years [1]. Most testicular cancers are germ cell tumours (GCT) and are classified according to their histology into two main groups: seminoma and nonseminoma. Most patients with GCT of the testis present with clinical stage I disease, i.e. confined to the testis [2e4]. After orchiectomy, the management of patients with stage I tumours is guided by the histology. For patients with seminoma, options include adjuvant radiotherapy, singleAuthor for correspondence: M.E. O’Malley, Princess Margaret Hospital 3-920, 610 University Avenue, Toronto, Ontario, Canada M5G 2M9. Tel: þ1-416-946-4501x4833; Fax: þ1-416-946-6564. E-mail address: martin.o’[email protected] (M.E. O’Malley).
dose carboplatin or surveillance. For patients with nonseminoma, options include adjuvant chemotherapy, surveillance or, in selected cases, retroperitoneal lymph node dissection. If surveillance is chosen, the relapse rate is about 17% for seminoma and 28% for non-seminoma. Therefore, surveillance allows most patients who are already cured of disease after orchiectomy to avoid the long-term toxicity and other potential side-effects of adjuvant treatments [5]. Regardless of the management chosen, disease-specific survival approaches 100% for patients with stage I tumours [6]. Thus, active surveillance is the approach that is increasingly preferred for patients with stage I tumours. The most common site of relapse with stage I GCT of the testis is in the retroperitoneal lymph nodes followed by a small percentage in the pelvic lymph nodes. The vast majority of relapses occur within 2 years of diagnosis.
0936-6555/$36.00 Ó 2010 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.clon.2010.10.006
334
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
Computed tomography is the standard imaging modality used for assessment of the abdominal and pelvic lymph nodes in surveillance protocols. However, there is no international consensus on the frequency of computed tomography scanning for patients on surveillance [5]. Because of the theoretical risk from low-level radiation associated with computed tomography, more attention has recently been paid to computed tomography schedules for patients on surveillance [7]. The computed tomography surveillance schedule at our institution for seminoma stage I is every 4 months for 3 years, every 6 months for 4 years and annually for 3 years (20 computed tomography studies in 10 years). For non-seminoma stage I, our computed tomography schedule is every 4 months for 2 years (six computed tomography studies in 2 years). In light of the heightened awareness of the potential risks related to low-level radiation from computed tomography [8], we undertook a review of our current surveillance protocols. As part of that review, we felt that knowledge of the growth rate of lymph nodes involved with recurrent disease may influence the computed tomography scanning frequency used in surveillance protocols. To our knowledge, this growth rate has not been previously reported. The aim of this study was to estimate the growth rate of abdominal/ pelvic lymph nodes in patients with stage I GCT of the testis on active surveillance who relapsed.
Materials and Methods This was a retrospective study approved by our research ethics board. Informed consent was not required. Using an institutional database, all patients with testicular GCT managed at our institution between May 2001 and May 2008 were identified. The availability of imaging on our picture archiving and communications system dates back to May 2001. The inclusion criteria for this study were as follows: patients with stage I GCT of the testis who were managed with active surveillance; disease recurrence in the abdomen or pelvis; computed tomography scans available on the picture archiving and communications system, including the computed tomography scan at which recurrence was diagnosed and at least one preceding computed tomography scan within the previous 12 months. In these patients, relapse was diagnosed when enlarged (>10 mm in the short axis) lymph node(s) were present in the expected location for recurrent disease (i.e. aortocaval region for right-sided tumours and left para-aortic region for left-sided tumours) or alternatively when lymph nodes in these regions measured <10 mm in the short axis but grew by at least 2 mm between scans. Two radiologists reviewed the images in consensus from the computed tomography scan at which disease recurrence was identified and all preceding scans. At the time of image review, the radiologists were blind to the histology. The anatomical location and short axis of the single largest (if there was more than one site) lymph node recurrence was measured and recorded. The short axis rather than the volume of each lymph node was used as the metric to reflect
clinical practice. All previous computed tomography scans were reviewed and the short axis of the involved lymph node was measured and recorded. For the purposes of this study, a change in size of 2 mm or more in the short axis from one scan to the next was considered to represent growth. The last computed tomography scan showing no growth in the lymph node was used as the baseline scan. The last computed tomography scan before treatment was instituted was considered the final scan. All imaging studies carried out at the time of recurrence were reviewed to identify other sites of disease recurrence. The type of imaging carried out and the presence or absence of disease outside the abdomen and pelvis was recorded. The type of scanner and the scanning parameters for each computed tomography scan were reviewed and recorded. For each patient, the slope (lymph node size/time) was calculated to estimate lymph node growth rate. The twosample t-test and non-parametric equivalent Wilcoxon rank sum test (exact test) were applied to compare lymph node growth rates.
Results During the period covered by our study, May 2001 to May 2008, of 318 patients with stage I GCT on active surveillance, 39 patients relapsed (12.3%); 18 had seminoma (relapse rate 9.1%) and 21 had non-seminoma (relapse rate 17.4%), of whom 15 and 13 met our inclusion criteria, respectively. These 28 patients (76% of all patients with recurrent disease in the abdomen and pelvis) formed the study group (Fig. 1). Three patients with seminoma and six patients with nonseminoma who relapsed were excluded due to a lack of access to a previous computed tomography scan within the preceding 12 months. One patient with non-seminoma had recurrent disease in the chest only and one patient with nonseminoma had positive tumour markers only; both of these patients were excluded from the study. The median age of patients in the study group at first scan was 32 years (range 19e51 years); seminoma 38 years (range 29e51 years) and non-seminoma 29 years (range 19e36 years). In 24 of 28 patients there was a single site of lymph node recurrence in the abdomen or pelvis. In four patients there was more than one site of lymph node recurrence (Table 1). None of the patients in this study who had disease recurrence in the abdomen and pelvis had simultaneous recurrence in the chest. Because our study covered a span of several years, a variety of computed tomography scanners were used. Sixty-five computed tomography studies were carried out on four-, eight- or 64-slice scanners with a slice thickness of 5 mm, reconstructed every 2.5 mm. Two computed tomography studies were carried out on a single slice scanner with a slice thickness of 7 mm, reconstructed every 3.5 mm. Thirty-three of 67 (49%) computed tomography studies were carried out with intravenous contrast and 34 of 67 (51%) without contrast. Anatomical coverage extended from the domes of the diaphragm to the ischial tuberosities.
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
335
Fig. 1. Patients with clinical stage I germ cell tumour of the testis managed at the Princess Margaret Hospital, Toronto, Canada between May 2001 and May 2008. AP, abdomen and pelvis; GCT, germ cell tumour; RPLND, retroperitoneal lymph node dissection; RT, radiotherapy.
The median length of time from orchiectomy to the recurrence computed tomography scan was 131 days (range 49e520) or about 4.4 months for non-seminoma patients and 373 days (range 129e675) or about 12.3 months for seminoma patients. The median size of the involved lymph node at baseline computed tomography for seminoma patients was 5 mm (range 2e8 mm) and for non-seminoma patients 5 mm (range 2e9 mm). The median size of the involved lymph node at final computed tomography for seminoma patients was 12 mm (range 9e31 mm) and for non-seminoma patients 15 mm (range 10e56 mm). It should be noted that the 56 mm measurement represents the largest lymph node in a patient with massive confluent retroperitoneal lymphadenopathy and was a statistical outlier. The next largest involved lymph node in a patient with non-seminoma was 26 mm. The change in lymph node size (mm) over time (the slope) was used to determine the rate of growth of recurrent lymph node disease for each patient (Fig. 2). The number of computed tomography scans per patient used to calculate growth rates was as follows: two for 18
patients (eight non-seminoma, 10 seminoma); three for seven patients (three non-seminoma, four seminoma) and four for three patients (two non-seminoma, one seminoma). The median slope (lymph node growth) for seminoma was 1.35 mm/month (range 0.62e4.56) and for non-seminoma was 2.99 mm/month (range 0.77e7.06). The mean/ median slope was at least double for patients with nonseminoma compared with patients with seminoma. The two-sample t-test (P ¼ 0.029) and the non-parametric equivalent Wilcoxon rank sum test (exact test) were statistically significant (P ¼ 0.029). There was no statistically significant difference in growth rate among patients with more than one lymph node relapse site in the abdomen and pelvis compared with a single lymph node relapse site (P ¼ 0.10 two-sample t-test; P ¼ 0.36 exact Wilcoxon rank sum). However, there were only four patients who had multiple nodes at relapse. The results were similarly not statistically significant comparing slopes among the six subgroups of relapse location (P ¼ 0.38, analysis of variance; P ¼ 0.26, non-parametric KruskaleWallis).
Table 1 Sites of relapse in the study group Site of relapse
Left seminoma
LPA AC LEI RCI LPA gonadal RCI gonadal
6
Total
9
Right seminoma
Left non-seminoma
Right non-seminoma
8* 4
4
2yz 1 1x
1 1 6
Total 14 8 2 1 2 1
9
4
28
LPA, left para-aortic; AC, aortocaval; LEI, left external iliac; RCI, right common iliac; LPA gonadal, LPA distribution in the region of the left testicular vein; RCI gonadal, RCI lateral to the right testicular vein. * One had a second site ¼ LPA. y One had a second site ¼ LPA. z One had a second site ¼ multiple sites including aortic bifurcation. x Second site ¼ LEI.
336
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
a
b
Fig. 2. Rate of growth of lymph nodes (LN), mm per month, least squares linear regression lines. (a) Non-seminoma patients (n ¼ 13); (b) Seminoma patients (n ¼ 15).
Discussion
lymph node involvement, recognising that false negatives and false positives may occur with this threshold value.
To our knowledge, the growth rate of lymph nodes in recurrent GCT of the testis has not been previously reported. This information may help in determining the frequency for active surveillance with computed tomography. In this study we found that the median growth rate of lymph nodes in patients with recurrent testis cancer was 1.35 mm/ month for seminoma and 2.99 mm/month for non-seminoma. The difference in growth rates was statistically significant. The mean length of time to relapse was 4.4 months for non-seminoma and 12.3 months for seminoma; all relapses occurred within 24 months of diagnosis. For patients with stage I GCT of the testis, surveillance is an attractive option as only about 17% of seminoma and 28% of non-seminoma patients will ultimately relapse and require further therapy after orchiectomy [6]. The most common site for recurrent disease is in the retroperitoneal lymph nodes and computed tomography is considered the standard imaging modality for lymph node surveillance. Recurrent lymph node disease is diagnosed when enlarged or enlarging lymph nodes are present. For practical purposes, a threshold of 10 mm in the short axis is used to diagnose
The ideal computed tomography surveillance schedule has not been determined and there is considerable variation among large cancer centres and in published recommendations. For seminoma, the National Comprehensive Cancer Network practice guidelines recommend abdominal/pelvic computed tomography every 3e4 months for the first year, every 6 months for the second year and annually thereafter [9]. Recent evidence-based guidelines from the UK recommend 6 monthly abdominal computed tomography for the first 2 years and then annually to year 5. The pelvis is only scanned if there has been scrotal interference or previous pelvic surgery [10]. A Canadian publication of evidencebased guidelines for stage I seminoma suggested a frequency of every 4 months for abdominal/pelvic computed tomography during the first 2 years of surveillance when the annualised hazard rate for relapse is >5% [11]. For non-seminoma, National Comprehensive Cancer Network practice guidelines recommend abdominal/pelvic computed tomography every 2e3 months for the first year and every 3e4 months during the second year. The UK guidelines recommend abdominal computed tomography
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
scans (the pelvis is only included if deemed high risk) at 3, 12 and 24 months. These recommendations are based on the results of a trial of a two-scan (at 3 and 12 months) versus a five-scan (at 3, 6, 9, 12 and 24 months) surveillance strategy [7]. In that trial, the primary outcome was the proportion of patients relapsing with intermediate- or poor-prognosis disease with each strategy. There were no patients who relapsed with poor-prognosis disease. Two (0.8%) patients relapsing in the two-scan strategy had intermediate prognosis compared with one (0.6%) patient in the five-scan strategy. Most patients managed by surveillance will already be cured of their disease and will not relapse. Of the patients who relapse, most will have small volume disease (lymph nodes <2 cm or lymph nodes 2e5 cm). Whichever treatment option is chosen at relapse, the results are similar for these patients compared with patients treated for stage IIA or IIB disease at initial diagnosis; the overall survival for seminoma is close to 100% and for non-seminoma almost 98% [12]. In other words, the size of lymph nodes at the time of recurrence does not strongly correlate with survival after treatment. Therefore, although recurrent disease should be detected as early as possible, the intensity of surveillance and any associated potential negative effects of surveillance need to be balanced with the excellent prognosis for testicular GCT. With regards to imaging surveillance, the main risk to the patient is related to ionising radiation. Although there have been no studies showing a direct link between ionising radiation from computed tomography scanning and cancer, and some scientists actually suggest a benefit from low-level radiation, it is prudent to accept the radiation risk estimates, such as those of the Biological Effects of Ionizing Radiation (BEIR) VII, phase 2 report [13]. In that report, the committee adopted a linear no threshold model for radiation-induced cancers. In other words, the report assumes that any amount of radiation has the potential to cause cancer and that there is a direct relationship between the amount of radiation and the risk of inducing cancer. The risk is also inversely related to age. This is particularly important in patients with testis GCT who tend to be young and otherwise healthy. In our study, for example, the median age of patients was 32 years. The BEIR VII phase 2 report estimates the risk of a radiationinduced cancer from 10 mSv (approximately equivalent to one computed tomography scan of the abdomen) to be one in 1000. Therefore, it would be prudent to adopt radiation reduction strategies for patients on surveillance. These strategies could include reducing the frequency of computed tomography surveillance, using low-dose computed tomography protocols [14] or using alternative imaging modalities that do not use ionising radiation, such as magnetic resonance imaging [15]. Extrapolating from our data, and assuming a constant growth rate, to detect recurrent disease in all seminoma patients before the lymph nodes reached 2 cm (the threshold for N2 disease), the frequency of computed tomography would have to be at least every 3.25 months (Figure 2a) during the first 2 years of surveillance. To detect a relapse in every patient before the lymph nodes reached greater than 5 cm (the threshold for N3 disease), computed tomography would
337
have to be carried out at least every 9.8 months. For nonseminoma, the frequency of computed tomography would have to be at least every 2.4 months (Figure 2b) to detect all the patients before the lymph nodes reached 2 cm and at least every 6.6 months to detect all recurrences less than 5 cm. However, the growth rate of lymph nodes in patients who relapse is not constant and a small percentage of patients will have a rapidly progressive relapse, especially with non-seminoma. In our study, one patient with non-seminoma had an estimated lymph node growth rate of 7.06 mm/month and multifocal recurrence. Therefore, there is unlikely to be an ideal computed tomography surveillance schedule, but rather one that will be practical and minimise potential risks to most patients who will not relapse. At our own institution, we have been studying the use of low-dose computed tomography to manage the radiation risk related to surveillance without altering our computed tomography schedule. With the use of low-dose computed tomography, we have achieved a mean dose reduction of 55% compared with standard-dose computed tomography while maintaining a diagnostically acceptable image quality [14]. On the basis of the results of our current study and the existing literature, we are now reviewing our surveillance schedules to determine whether a less frequent computed tomography schedule would be appropriate during the first 2 years of surveillance. Our study had limitations. Our institution manages a large number of patients with testicular cancer and we analysed a fairly high percentage of patients who relapsed. However, the absolute number of patients in this study was relatively small and the rate of relapse during the time period studied was lower than would have been predicted. Ideally, our results could be validated by a study with a larger number of patients with recurrent disease. Another limitation was that some patients only had two computed tomography scans for analysis separated by up to 8 months. Therefore, our results only represent an estimate of the growth rates of lymph nodes involved with recurrent disease. However, this was a retrospective study and our current schedule involves carrying out computed tomography scans on a relatively frequent basis compared with other institutions with large groups of testis cancer patients on surveillance. Finally, this study was focused on only one aspect of surveillance, i.e. computed tomography scanning, and did not take into account other surveillance tools, such as tumour markers, which play an important role in the surveillance of patients, particularly with non-seminoma GCT [7,16].
Conclusion In summary, we have estimated the growth rate for abdominal/pelvic lymph node recurrence in stage I GCT of the testis. There was a statistically significant faster growth rate in patients with non-seminoma compared with patients with seminoma and the median time to recurrence was shorter with non-seminoma than with seminoma. These results support a more frequent computed tomography surveillance schedule for non-seminoma than for seminoma during the first 2 years of surveillance. The computed tomography
338
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
surveillance schedules referenced here fall within a practical range while recognising that steps should be taken to minimise the risks from ionising radiation when more frequent schedules are used.
[8] [9]
Acknowledgement M.E. O’Malley was supported in part by a research and development grant from the Department of Medical Imaging, University of Toronto, Canada.
[10]
[11]
References [1] McGlynn KA, Devesa SS, Graubard BI, et al. Increasing incidence of testicular germ cell tumors among black men in the United States. J Clin Oncol 2005;23:5757e5761. [2] Sant M, Aareleid T, Artioli ME, et al. Ten-year survival and risk of relapse for testicular cancer: a EUROCARE high resolution study. Eur J Cancer 2007;43:585e592. [3] Sternberg CN. The management of stage I testis cancer. Urol Clin North Am 1998;25:435e449. [4] Warde P, Jewett MA. Surveillance for stage I testicular seminoma. Is it a good option? Urol Clin North Am 1998;25:425e433. [5] Huddart RA, Joffe JK. Preferred treatment for stage I seminoma: a survey of Canadian radiation oncologists. Clin Oncol (R Coll Radiol) 2006;18:693e695. [6] Groll RJ, Warde P, Jewett MA. A comprehensive systematic review of testicular germ cell tumor surveillance. Crit Rev Oncol Hematol 2007;64:182e197. [7] Rustin GJ, Mead GM, Stenning SP, et al. Randomized trial of two or five computed tomography scans in the surveillance of patients with stage I nonseminomatous germ cell tumors of the testis: Medical Research Council trial TE08, ISRCTN56475197 e the National Cancer Research Institute
[12]
[13]
[14]
[15]
[16]
Testis Cancer Clinical Studies Group. J Clin Oncol 2007;25: 1310e1315. Brenner DJ, Hall EJ. Computed tomography e an increasing source of radiation exposure. N Engl J Med 2007;357:2277. National Comprehensive Cancer Network. Clinical practice guidelines in oncology. Testicular cancer, version 1, 2009. Available at; www.nccn.org. Accessed 1 March 2009. Van As NJ, Gilbert DC, Money-Kyrle J, et al. Evidence-based pragmatic guidelines for the follow-up of testicular cancer: optimising the detection of relapse. Br J Cancer 2008;98: 1894e1902. Martin JM, Panzarella T, Zwahlen DR, et al. Evidence-based guidelines for following stage 1 seminoma. Cancer 2007;109: 2248e2256. Krege S, Beyer J, Souchon R, et al. European consensus conference on diagnosis and treatment of germ cell cancer: a report of the second meeting of the European Germ Cell Cancer Consensus group (EGCCCG): part I. Eur Urol 2008;53: 478e496. National Research Council Committee to Assess Health Risks from Exposure to Low Levels of Ionizing Radiation. Health risks from exposure to low levels of ionizing radiation: BEIR VII, Phase 2. Washington, DC: National Academies Press; 2006. O’Malley ME, Chung P, Haider M, et al. Comparison of low dose with standard dose abdominal/pelvic multidetector CT in patients with stage 1 testicular cancer under surveillance. Eur Radiol 2010;20:1624e1630. Sohaib SA, Koh DM, Barbachano Y, et al. Prospective assessment of MRI for imaging retroperitoneal metastases from testicular germ cell tumours. Clin Radiol 2009;64:362. Crawford SM, Rustin GJ, Begent RH, Newlands ES, Bagshawe KD. Safety of surveillance in the management of stage I anaplastic germ cell tumours of the testis. Br J Urol 1988; 61:250e253.
Clinical Oncology journal homepage: www.elsevier.com/locate/clon
Original Article
Lymph Node Growth Rate in Testicular Germ Cell Tumours: Implications for Computed Tomography Surveillance Frequency W. Mazrani *, M.E. O’Malley *, P.W.M. Chung y, P. Warde y, D. Vesprini z, T. Panzarella x * Joint
Department of Medical Imaging, University of Toronto, Canada Department of Radiation Oncology, Princess Margaret Hospital, University of Toronto, Canada z Department of Radiation Oncology, Sunnybrook Odette Cancer Centre, University of Toronto, Canada x Department of Biostatistics, Princess Margaret Hospital Cancer Program and Dalla Lana School of Public Health, University of Toronto, Canada y
Received 10 May 2010; received in revised form 26 July 2010; accepted 5 October 2010
Abstract Aim: To estimate the growth rate of lymph nodes in patients on surveillance for testicular cancer who developed recurrent disease. Materials and methods: During a 7-year period, 318 patients at our institution were managed by surveillance and 39 relapsed (12.3%). The computed tomography scans of 28 patients (median age 32 years; range 19e51 years) who met our inclusion criteria and who developed recurrent disease in the abdomen/pelvis were retrospectively reviewed. Thirteen patients had non-seminoma and 15 had seminoma. To estimate the lymph node growth rate, the slope of lymph node size over time was calculated. Results: The median length of time from orchiectomy to the recurrence computed tomography was 131 days (range 49e520) or about 4.4 months for non-seminoma patients and 373 days (range 129e675) or about 12.3 months for seminoma patients. The median size of the involved lymph node at final computed tomography for seminoma patients was 12 mm (range 9e31 mm) and for non-seminoma patients was 15 mm (range 10e56 mm). The median lymph node growth rate for patients with seminoma was 1.35 mm/month (range 0.62e4.56) and for patients with non-seminoma 2.99 mm/month (range 0.77e7.06); the difference in growth rates was statistically significant (P ¼ 0.029). Conclusions: There is a statistically significant faster growth rate of lymph nodes in patients with recurrent non-seminoma compared with patients with seminoma. This finding supports a more frequent computed tomography schedule during the first 2 years of surveillance in non-seminoma patients compared with seminoma patients. Ó 2010 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. Key words: Computed tomography; growth rate; lymph node; surveillance; testicular cancer
Introduction Testicular cancer is the most common malignancy in males aged 15e34 years [1]. Most testicular cancers are germ cell tumours (GCT) and are classified according to their histology into two main groups: seminoma and nonseminoma. Most patients with GCT of the testis present with clinical stage I disease, i.e. confined to the testis [2e4]. After orchiectomy, the management of patients with stage I tumours is guided by the histology. For patients with seminoma, options include adjuvant radiotherapy, singleAuthor for correspondence: M.E. O’Malley, Princess Margaret Hospital 3-920, 610 University Avenue, Toronto, Ontario, Canada M5G 2M9. Tel: þ1-416-946-4501x4833; Fax: þ1-416-946-6564. E-mail address: martin.o’[email protected] (M.E. O’Malley).
dose carboplatin or surveillance. For patients with nonseminoma, options include adjuvant chemotherapy, surveillance or, in selected cases, retroperitoneal lymph node dissection. If surveillance is chosen, the relapse rate is about 17% for seminoma and 28% for non-seminoma. Therefore, surveillance allows most patients who are already cured of disease after orchiectomy to avoid the long-term toxicity and other potential side-effects of adjuvant treatments [5]. Regardless of the management chosen, disease-specific survival approaches 100% for patients with stage I tumours [6]. Thus, active surveillance is the approach that is increasingly preferred for patients with stage I tumours. The most common site of relapse with stage I GCT of the testis is in the retroperitoneal lymph nodes followed by a small percentage in the pelvic lymph nodes. The vast majority of relapses occur within 2 years of diagnosis.
0936-6555/$36.00 Ó 2010 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.clon.2010.10.006
334
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
Computed tomography is the standard imaging modality used for assessment of the abdominal and pelvic lymph nodes in surveillance protocols. However, there is no international consensus on the frequency of computed tomography scanning for patients on surveillance [5]. Because of the theoretical risk from low-level radiation associated with computed tomography, more attention has recently been paid to computed tomography schedules for patients on surveillance [7]. The computed tomography surveillance schedule at our institution for seminoma stage I is every 4 months for 3 years, every 6 months for 4 years and annually for 3 years (20 computed tomography studies in 10 years). For non-seminoma stage I, our computed tomography schedule is every 4 months for 2 years (six computed tomography studies in 2 years). In light of the heightened awareness of the potential risks related to low-level radiation from computed tomography [8], we undertook a review of our current surveillance protocols. As part of that review, we felt that knowledge of the growth rate of lymph nodes involved with recurrent disease may influence the computed tomography scanning frequency used in surveillance protocols. To our knowledge, this growth rate has not been previously reported. The aim of this study was to estimate the growth rate of abdominal/ pelvic lymph nodes in patients with stage I GCT of the testis on active surveillance who relapsed.
Materials and Methods This was a retrospective study approved by our research ethics board. Informed consent was not required. Using an institutional database, all patients with testicular GCT managed at our institution between May 2001 and May 2008 were identified. The availability of imaging on our picture archiving and communications system dates back to May 2001. The inclusion criteria for this study were as follows: patients with stage I GCT of the testis who were managed with active surveillance; disease recurrence in the abdomen or pelvis; computed tomography scans available on the picture archiving and communications system, including the computed tomography scan at which recurrence was diagnosed and at least one preceding computed tomography scan within the previous 12 months. In these patients, relapse was diagnosed when enlarged (>10 mm in the short axis) lymph node(s) were present in the expected location for recurrent disease (i.e. aortocaval region for right-sided tumours and left para-aortic region for left-sided tumours) or alternatively when lymph nodes in these regions measured <10 mm in the short axis but grew by at least 2 mm between scans. Two radiologists reviewed the images in consensus from the computed tomography scan at which disease recurrence was identified and all preceding scans. At the time of image review, the radiologists were blind to the histology. The anatomical location and short axis of the single largest (if there was more than one site) lymph node recurrence was measured and recorded. The short axis rather than the volume of each lymph node was used as the metric to reflect
clinical practice. All previous computed tomography scans were reviewed and the short axis of the involved lymph node was measured and recorded. For the purposes of this study, a change in size of 2 mm or more in the short axis from one scan to the next was considered to represent growth. The last computed tomography scan showing no growth in the lymph node was used as the baseline scan. The last computed tomography scan before treatment was instituted was considered the final scan. All imaging studies carried out at the time of recurrence were reviewed to identify other sites of disease recurrence. The type of imaging carried out and the presence or absence of disease outside the abdomen and pelvis was recorded. The type of scanner and the scanning parameters for each computed tomography scan were reviewed and recorded. For each patient, the slope (lymph node size/time) was calculated to estimate lymph node growth rate. The twosample t-test and non-parametric equivalent Wilcoxon rank sum test (exact test) were applied to compare lymph node growth rates.
Results During the period covered by our study, May 2001 to May 2008, of 318 patients with stage I GCT on active surveillance, 39 patients relapsed (12.3%); 18 had seminoma (relapse rate 9.1%) and 21 had non-seminoma (relapse rate 17.4%), of whom 15 and 13 met our inclusion criteria, respectively. These 28 patients (76% of all patients with recurrent disease in the abdomen and pelvis) formed the study group (Fig. 1). Three patients with seminoma and six patients with nonseminoma who relapsed were excluded due to a lack of access to a previous computed tomography scan within the preceding 12 months. One patient with non-seminoma had recurrent disease in the chest only and one patient with nonseminoma had positive tumour markers only; both of these patients were excluded from the study. The median age of patients in the study group at first scan was 32 years (range 19e51 years); seminoma 38 years (range 29e51 years) and non-seminoma 29 years (range 19e36 years). In 24 of 28 patients there was a single site of lymph node recurrence in the abdomen or pelvis. In four patients there was more than one site of lymph node recurrence (Table 1). None of the patients in this study who had disease recurrence in the abdomen and pelvis had simultaneous recurrence in the chest. Because our study covered a span of several years, a variety of computed tomography scanners were used. Sixty-five computed tomography studies were carried out on four-, eight- or 64-slice scanners with a slice thickness of 5 mm, reconstructed every 2.5 mm. Two computed tomography studies were carried out on a single slice scanner with a slice thickness of 7 mm, reconstructed every 3.5 mm. Thirty-three of 67 (49%) computed tomography studies were carried out with intravenous contrast and 34 of 67 (51%) without contrast. Anatomical coverage extended from the domes of the diaphragm to the ischial tuberosities.
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
335
Fig. 1. Patients with clinical stage I germ cell tumour of the testis managed at the Princess Margaret Hospital, Toronto, Canada between May 2001 and May 2008. AP, abdomen and pelvis; GCT, germ cell tumour; RPLND, retroperitoneal lymph node dissection; RT, radiotherapy.
The median length of time from orchiectomy to the recurrence computed tomography scan was 131 days (range 49e520) or about 4.4 months for non-seminoma patients and 373 days (range 129e675) or about 12.3 months for seminoma patients. The median size of the involved lymph node at baseline computed tomography for seminoma patients was 5 mm (range 2e8 mm) and for non-seminoma patients 5 mm (range 2e9 mm). The median size of the involved lymph node at final computed tomography for seminoma patients was 12 mm (range 9e31 mm) and for non-seminoma patients 15 mm (range 10e56 mm). It should be noted that the 56 mm measurement represents the largest lymph node in a patient with massive confluent retroperitoneal lymphadenopathy and was a statistical outlier. The next largest involved lymph node in a patient with non-seminoma was 26 mm. The change in lymph node size (mm) over time (the slope) was used to determine the rate of growth of recurrent lymph node disease for each patient (Fig. 2). The number of computed tomography scans per patient used to calculate growth rates was as follows: two for 18
patients (eight non-seminoma, 10 seminoma); three for seven patients (three non-seminoma, four seminoma) and four for three patients (two non-seminoma, one seminoma). The median slope (lymph node growth) for seminoma was 1.35 mm/month (range 0.62e4.56) and for non-seminoma was 2.99 mm/month (range 0.77e7.06). The mean/ median slope was at least double for patients with nonseminoma compared with patients with seminoma. The two-sample t-test (P ¼ 0.029) and the non-parametric equivalent Wilcoxon rank sum test (exact test) were statistically significant (P ¼ 0.029). There was no statistically significant difference in growth rate among patients with more than one lymph node relapse site in the abdomen and pelvis compared with a single lymph node relapse site (P ¼ 0.10 two-sample t-test; P ¼ 0.36 exact Wilcoxon rank sum). However, there were only four patients who had multiple nodes at relapse. The results were similarly not statistically significant comparing slopes among the six subgroups of relapse location (P ¼ 0.38, analysis of variance; P ¼ 0.26, non-parametric KruskaleWallis).
Table 1 Sites of relapse in the study group Site of relapse
Left seminoma
LPA AC LEI RCI LPA gonadal RCI gonadal
6
Total
9
Right seminoma
Left non-seminoma
Right non-seminoma
8* 4
4
2yz 1 1x
1 1 6
Total 14 8 2 1 2 1
9
4
28
LPA, left para-aortic; AC, aortocaval; LEI, left external iliac; RCI, right common iliac; LPA gonadal, LPA distribution in the region of the left testicular vein; RCI gonadal, RCI lateral to the right testicular vein. * One had a second site ¼ LPA. y One had a second site ¼ LPA. z One had a second site ¼ multiple sites including aortic bifurcation. x Second site ¼ LEI.
336
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
a
b
Fig. 2. Rate of growth of lymph nodes (LN), mm per month, least squares linear regression lines. (a) Non-seminoma patients (n ¼ 13); (b) Seminoma patients (n ¼ 15).
Discussion
lymph node involvement, recognising that false negatives and false positives may occur with this threshold value.
To our knowledge, the growth rate of lymph nodes in recurrent GCT of the testis has not been previously reported. This information may help in determining the frequency for active surveillance with computed tomography. In this study we found that the median growth rate of lymph nodes in patients with recurrent testis cancer was 1.35 mm/ month for seminoma and 2.99 mm/month for non-seminoma. The difference in growth rates was statistically significant. The mean length of time to relapse was 4.4 months for non-seminoma and 12.3 months for seminoma; all relapses occurred within 24 months of diagnosis. For patients with stage I GCT of the testis, surveillance is an attractive option as only about 17% of seminoma and 28% of non-seminoma patients will ultimately relapse and require further therapy after orchiectomy [6]. The most common site for recurrent disease is in the retroperitoneal lymph nodes and computed tomography is considered the standard imaging modality for lymph node surveillance. Recurrent lymph node disease is diagnosed when enlarged or enlarging lymph nodes are present. For practical purposes, a threshold of 10 mm in the short axis is used to diagnose
The ideal computed tomography surveillance schedule has not been determined and there is considerable variation among large cancer centres and in published recommendations. For seminoma, the National Comprehensive Cancer Network practice guidelines recommend abdominal/pelvic computed tomography every 3e4 months for the first year, every 6 months for the second year and annually thereafter [9]. Recent evidence-based guidelines from the UK recommend 6 monthly abdominal computed tomography for the first 2 years and then annually to year 5. The pelvis is only scanned if there has been scrotal interference or previous pelvic surgery [10]. A Canadian publication of evidencebased guidelines for stage I seminoma suggested a frequency of every 4 months for abdominal/pelvic computed tomography during the first 2 years of surveillance when the annualised hazard rate for relapse is >5% [11]. For non-seminoma, National Comprehensive Cancer Network practice guidelines recommend abdominal/pelvic computed tomography every 2e3 months for the first year and every 3e4 months during the second year. The UK guidelines recommend abdominal computed tomography
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
scans (the pelvis is only included if deemed high risk) at 3, 12 and 24 months. These recommendations are based on the results of a trial of a two-scan (at 3 and 12 months) versus a five-scan (at 3, 6, 9, 12 and 24 months) surveillance strategy [7]. In that trial, the primary outcome was the proportion of patients relapsing with intermediate- or poor-prognosis disease with each strategy. There were no patients who relapsed with poor-prognosis disease. Two (0.8%) patients relapsing in the two-scan strategy had intermediate prognosis compared with one (0.6%) patient in the five-scan strategy. Most patients managed by surveillance will already be cured of their disease and will not relapse. Of the patients who relapse, most will have small volume disease (lymph nodes <2 cm or lymph nodes 2e5 cm). Whichever treatment option is chosen at relapse, the results are similar for these patients compared with patients treated for stage IIA or IIB disease at initial diagnosis; the overall survival for seminoma is close to 100% and for non-seminoma almost 98% [12]. In other words, the size of lymph nodes at the time of recurrence does not strongly correlate with survival after treatment. Therefore, although recurrent disease should be detected as early as possible, the intensity of surveillance and any associated potential negative effects of surveillance need to be balanced with the excellent prognosis for testicular GCT. With regards to imaging surveillance, the main risk to the patient is related to ionising radiation. Although there have been no studies showing a direct link between ionising radiation from computed tomography scanning and cancer, and some scientists actually suggest a benefit from low-level radiation, it is prudent to accept the radiation risk estimates, such as those of the Biological Effects of Ionizing Radiation (BEIR) VII, phase 2 report [13]. In that report, the committee adopted a linear no threshold model for radiation-induced cancers. In other words, the report assumes that any amount of radiation has the potential to cause cancer and that there is a direct relationship between the amount of radiation and the risk of inducing cancer. The risk is also inversely related to age. This is particularly important in patients with testis GCT who tend to be young and otherwise healthy. In our study, for example, the median age of patients was 32 years. The BEIR VII phase 2 report estimates the risk of a radiationinduced cancer from 10 mSv (approximately equivalent to one computed tomography scan of the abdomen) to be one in 1000. Therefore, it would be prudent to adopt radiation reduction strategies for patients on surveillance. These strategies could include reducing the frequency of computed tomography surveillance, using low-dose computed tomography protocols [14] or using alternative imaging modalities that do not use ionising radiation, such as magnetic resonance imaging [15]. Extrapolating from our data, and assuming a constant growth rate, to detect recurrent disease in all seminoma patients before the lymph nodes reached 2 cm (the threshold for N2 disease), the frequency of computed tomography would have to be at least every 3.25 months (Figure 2a) during the first 2 years of surveillance. To detect a relapse in every patient before the lymph nodes reached greater than 5 cm (the threshold for N3 disease), computed tomography would
337
have to be carried out at least every 9.8 months. For nonseminoma, the frequency of computed tomography would have to be at least every 2.4 months (Figure 2b) to detect all the patients before the lymph nodes reached 2 cm and at least every 6.6 months to detect all recurrences less than 5 cm. However, the growth rate of lymph nodes in patients who relapse is not constant and a small percentage of patients will have a rapidly progressive relapse, especially with non-seminoma. In our study, one patient with non-seminoma had an estimated lymph node growth rate of 7.06 mm/month and multifocal recurrence. Therefore, there is unlikely to be an ideal computed tomography surveillance schedule, but rather one that will be practical and minimise potential risks to most patients who will not relapse. At our own institution, we have been studying the use of low-dose computed tomography to manage the radiation risk related to surveillance without altering our computed tomography schedule. With the use of low-dose computed tomography, we have achieved a mean dose reduction of 55% compared with standard-dose computed tomography while maintaining a diagnostically acceptable image quality [14]. On the basis of the results of our current study and the existing literature, we are now reviewing our surveillance schedules to determine whether a less frequent computed tomography schedule would be appropriate during the first 2 years of surveillance. Our study had limitations. Our institution manages a large number of patients with testicular cancer and we analysed a fairly high percentage of patients who relapsed. However, the absolute number of patients in this study was relatively small and the rate of relapse during the time period studied was lower than would have been predicted. Ideally, our results could be validated by a study with a larger number of patients with recurrent disease. Another limitation was that some patients only had two computed tomography scans for analysis separated by up to 8 months. Therefore, our results only represent an estimate of the growth rates of lymph nodes involved with recurrent disease. However, this was a retrospective study and our current schedule involves carrying out computed tomography scans on a relatively frequent basis compared with other institutions with large groups of testis cancer patients on surveillance. Finally, this study was focused on only one aspect of surveillance, i.e. computed tomography scanning, and did not take into account other surveillance tools, such as tumour markers, which play an important role in the surveillance of patients, particularly with non-seminoma GCT [7,16].
Conclusion In summary, we have estimated the growth rate for abdominal/pelvic lymph node recurrence in stage I GCT of the testis. There was a statistically significant faster growth rate in patients with non-seminoma compared with patients with seminoma and the median time to recurrence was shorter with non-seminoma than with seminoma. These results support a more frequent computed tomography surveillance schedule for non-seminoma than for seminoma during the first 2 years of surveillance. The computed tomography
338
W. Mazrani et al. / Clinical Oncology 23 (2011) 333e338
surveillance schedules referenced here fall within a practical range while recognising that steps should be taken to minimise the risks from ionising radiation when more frequent schedules are used.
[8] [9]
Acknowledgement M.E. O’Malley was supported in part by a research and development grant from the Department of Medical Imaging, University of Toronto, Canada.
[10]
[11]
References [1] McGlynn KA, Devesa SS, Graubard BI, et al. Increasing incidence of testicular germ cell tumors among black men in the United States. J Clin Oncol 2005;23:5757e5761. [2] Sant M, Aareleid T, Artioli ME, et al. Ten-year survival and risk of relapse for testicular cancer: a EUROCARE high resolution study. Eur J Cancer 2007;43:585e592. [3] Sternberg CN. The management of stage I testis cancer. Urol Clin North Am 1998;25:435e449. [4] Warde P, Jewett MA. Surveillance for stage I testicular seminoma. Is it a good option? Urol Clin North Am 1998;25:425e433. [5] Huddart RA, Joffe JK. Preferred treatment for stage I seminoma: a survey of Canadian radiation oncologists. Clin Oncol (R Coll Radiol) 2006;18:693e695. [6] Groll RJ, Warde P, Jewett MA. A comprehensive systematic review of testicular germ cell tumor surveillance. Crit Rev Oncol Hematol 2007;64:182e197. [7] Rustin GJ, Mead GM, Stenning SP, et al. Randomized trial of two or five computed tomography scans in the surveillance of patients with stage I nonseminomatous germ cell tumors of the testis: Medical Research Council trial TE08, ISRCTN56475197 e the National Cancer Research Institute
[12]
[13]
[14]
[15]
[16]
Testis Cancer Clinical Studies Group. J Clin Oncol 2007;25: 1310e1315. Brenner DJ, Hall EJ. Computed tomography e an increasing source of radiation exposure. N Engl J Med 2007;357:2277. National Comprehensive Cancer Network. Clinical practice guidelines in oncology. Testicular cancer, version 1, 2009. Available at; www.nccn.org. Accessed 1 March 2009. Van As NJ, Gilbert DC, Money-Kyrle J, et al. Evidence-based pragmatic guidelines for the follow-up of testicular cancer: optimising the detection of relapse. Br J Cancer 2008;98: 1894e1902. Martin JM, Panzarella T, Zwahlen DR, et al. Evidence-based guidelines for following stage 1 seminoma. Cancer 2007;109: 2248e2256. Krege S, Beyer J, Souchon R, et al. European consensus conference on diagnosis and treatment of germ cell cancer: a report of the second meeting of the European Germ Cell Cancer Consensus group (EGCCCG): part I. Eur Urol 2008;53: 478e496. National Research Council Committee to Assess Health Risks from Exposure to Low Levels of Ionizing Radiation. Health risks from exposure to low levels of ionizing radiation: BEIR VII, Phase 2. Washington, DC: National Academies Press; 2006. O’Malley ME, Chung P, Haider M, et al. Comparison of low dose with standard dose abdominal/pelvic multidetector CT in patients with stage 1 testicular cancer under surveillance. Eur Radiol 2010;20:1624e1630. Sohaib SA, Koh DM, Barbachano Y, et al. Prospective assessment of MRI for imaging retroperitoneal metastases from testicular germ cell tumours. Clin Radiol 2009;64:362. Crawford SM, Rustin GJ, Begent RH, Newlands ES, Bagshawe KD. Safety of surveillance in the management of stage I anaplastic germ cell tumours of the testis. Br J Urol 1988; 61:250e253.