10.1245/s10434-006-9214-6
Annals of Surgical Oncology 14:899-905 (2007)
© 2007 Society of Surgical Oncology
The Reproducibility in Routine Clinical Practice of Sentinel Lymph Node Identification by Pre-operative Lymphoscintigraphy in Patients with Cutaneous Melanoma
Roger F. Uren, MD1,2,
Robert Howman-Giles, MD1,2,
David K. V. Chung, MB1,2,
Rachael L. Morton, BN, Grad Dip (Prac and Mgmt)4 and
John F. Thompson, MB2,3
1 Nuclear Medicine and Diagnostic Ultrasound, RPAH Medical Centre and Discipline of Medicine, The University of Sydney, Sydney, NSW, Australia
2 The Sydney Melanoma Unit, The University of Sydney, Sydney, NSW, Australia
3 Discipline of Surgery, The University of Sydney, Sydney, NSW, Australia
4 Clinical Trials Business Development Manager, Sydney Melanoma Unit, 1A Eden St, Nth Sydney, NSW 2060, Australia
Correspondence: Address correspondence and reprint requests to: Roger F. Uren, Suite 206, RPAH Medical Centre, 100 Carillon Ave, Newtown NSW 2042, Australia; E-mail: ruren{at}mail.usyd.edu.au
 |
ABSTRACT
|
|---|
Pre-operative lymphoscintigraphy (LS) is an important part of successful sentinel lymph node (SLN) biopsy in most melanoma treatment centers. The test accurately maps lymphatic drainage from cutaneous melanoma sites and has been shown to be reproducible in prospective studies. Its reproducibility has not been tested, however, in routine clinical practice. Occasionally, after LS has been performed to map the location of SLNs, the patient is unable to proceed to SLN biopsy surgery within the time limit imposed by the radioactive decay of the 99mTc label attached to the colloid particles. In this situation, the surgery is rescheduled and LS repeated to relabel the SLNs so that they may be accurately biopsied. This has happened on 21 occasions at the Sydney Melanoma Unit and we have performed a retrospective analysis of the reproducibility of the LS results. In 19 patients, the same SLNs were shown in the same locations on the two studies. Two patients had discrepant results. One showed two extra interval nodes on the back as well as concordant flow to SLNs in each axilla. The other with a leg melanoma showed the same groin SLNs but failed to relabel the two popliteal SLNs on the second study. SLN locations were identical during 95%, and SLNs were identical 94% of the time. These results indicate that in routine clinical practice LS is a highly reproducible procedure to locate and radiolabel the SLNs prior to biopsy in patients with melanoma.
Key Words: Lymphoscintigraphy • Reproducibility • Cutaneous melanoma
|
INTRODUCTION
|
|---|
Over the past 15 years we have performed lymphoscintigraphy (LS) in over 5000 patients with cutaneous melanoma to identify and mark the surface location of sentinel lymph nodes (SLNs) prior to their surgical removal during selective SLN biopsy. Like others, we have found that SLNs removed in this way accurately stage the lymph node fields in about 98% of the patients.1,2 The reproducibility of LS for lymphatic mapping in melanoma is assumed to be an important factor in achieving such accuracy but conflicting results have been published.3–5 In one report reproducibility was only 50%,5 but these authors did achieve a better result with 84% reproducibility when more patients were studied.6 All these studies were performed prospectively to test reproducibility under ideal conditions. Repeat LS will sometimes need to be performed in routine clinical practice, however, with different nuclear medicine physicians and technologists involved. This situation may introduce variables that could adversely affect the reproducibility of LS. To further examine the reproducibility of the LS technique in routine clinical practice we have examined our database retrospectively to identify patients in whom LS was repeated.
These repeat LS procedures have been necessary because very occasionally the SLN biopsy procedure could not be completed within the appropriate time window of around 24 h provided by the radioactive decay and 6-h half-life of 99mTechnetium (Tc).7 Reasons for this have included cancelled operating sessions, medical events occurring after the LS but before the scheduled time for surgery and, in one patient, an allergic reaction to blue dye injection pre-operatively. In all these circumstances a repeat LS has been performed to relabel the SLNs prior to the rescheduled surgery. When this has happened we have reimaged each patient to document once more the path of the lymphatic collecting vessels and the exact location of all SLNs.
|
PATIENTS AND METHODS
|
|---|
Patients
Our database containing details of 5254 LS procedures performed in patients with melanoma over the past 15 years was searched to identify those who had undergone two LS studies. The database records the exact location of the primary melanoma site in each patient using "x" and "y" coordinates on one of six standard maps that cover completely the skin of the human body. Patients were selected when radiocolloid injection was given at the same skin site, with no surgical intervention between the two studies. Patients who had repeat LS performed to map drainage from a local or intransit recurrence at the same skin site were excluded. A total of 21 patients were found to satisfy the search criteria. The delay from excision biopsy of the primary cutaneous melanoma to the date of the first LS ranged from 6 to 51 days, with a mean of 25 days (SD = 11 days). The gap between the two LS studies ranged from 3 to 34 days, with a mean of 16 days (SD = 8 days). The two LS studies in each patient were compared to document the exact location and number of all SLNs. The relative "brightness" of the SLNs and the number of lymphatic collecting vessels draining the injection site on each occasion were also noted.
Imaging
The radiocolloid used at the Sydney Melanoma Unit is 99mTc antimony sulfide colloid and the activity injected intradermally (10–40 MBq in 0.05 ml per injection) is varied depending on the expected delay between tracer injection and surgery. The number of intradermal injections also varies from patient to patient but we aim to give the smallest number of injections consistent with obtaining an accurate map of lymph drainage from the primary melanoma site. Most commonly two or four injections are given, one or two on each side of the excision biopsy site in the middle of the scar. With a long excision biopsy, scar injections are clustered around the middle of the biopsy scar and we never inject at the ends of the biopsy scar since our intention is to inject the skin that was immediately adjacent to the melanoma when it was in situ. The physician performing the second LS was never aware of the precise site of intradermal injection of tracer during the first LS but followed our normal protocol to inject as close to the center of the scar as possible. The physician was aware through the report of the first scan of the exact number of intradermal injections given in each patient. The same number of injections was given on the second study in 13 patients while in 6 patients the exact number of injections was not recorded on the second study. In two patients, there were fewer injections on the second study. One had two injections given proximal to the lesion site on the leg when a third injection had been given distal to the lesion on the first study and the other deliberately had been given only two injections of a solution that contained only 50% of the radio-colloid particles present in the four injections given for the first study. In all studies, an ultra-high resolution microcast collimator is used with a large field of view gamma camera. Early dynamic imaging at one frame per minute for 10 min in a 256 by 256 matrix is performed to identify the lymphatic collecting vessels as they pass to the draining SLNs. Delayed imaging is performed 1–2 h later and the surface location of all SLNs is marked on the overlying skin with a temporary cross of Castellani’s paint and a permanent point tattoo of carbon black. Again a 256 by 256 matrix is used and images are acquired for 5–10 min over the site of all possible draining SLNs. Delayed images are acquired both with and without a transmission source behind the patient to highlight the body outline. The LS is routinely performed either on the morning of the surgery or during the previous afternoon. In both these situations there is no need for any further injection of radio-colloid, as adequate tracer remains in the node at the time of surgery to allow intra-operative location of SLNs using a hand-held gamma detecting probe.7
|
RESULTS
|
|---|
In the 21 patients, a total of 63 SLNs were identi-fied in 37 locations(34 standard node fields and 3 interval node sites).
In 19 of the 21 patients (90%) the results of the two LS studies were identical, including both the total number of SLNs in each patient and their location (Fig. 1
). In two patients the results were not identical, patient numbers 6 and 15 in Table 1. Patient 6 (Fig. 2) had a melanoma excision biopsy site on the upper back in the midline and on the first LS showed two lymphatic collecting vessels passing to the right axilla, where they converged on a single SLN, and two lymphatic collecting vessels passing to the left axilla to converge on a single SLN. Faint drainage also occurred to an interval SLN higher on the back. On the second study performed 21 days later, the same lymphatic collecting vessels were seen converging on the same SLNs in each axilla though all of the channels were better visualized on the scan. Two bright channels were also seen passing up to three bright interval SLNs on the back. Two extra SLNs were therefore seen on the repeat LS study in this patient, both interval nodes on the back. In this patient two intradermal injections had been given on the first study and the number of injections given on the second study was not recorded. The other patient whose scan results were discordant, patient 15 (Fig. 3), had a melanoma excision biopsy site on the left lateral shin above the ankle. The first LS showed two SLNs in the left groin and separate drainage to two SLNs in the left popliteal fossa. The second LS performed 14 days later showed the two SLNs in the left groin but no drainage was identified to the pop-liteal fossa SLNs. This patient had received three intradermal injections for both studies, two proximal to and one distal to the melanoma site on the left shin.
View larger version (84K):
[in this window]
[in a new window]
|
FIG. 1. Delayed lymphoscintigraphy images anteriorly over the pelvis and posteriorly over the popliteal fossa of Patient 17 performed 2 hours after the intradermal injection of 99mTc antimony sulfide colloid. The top panels show the first study performed after intradermal injections at two points, one on each side of the excision biopsy site on the sole of the right foot. The bottom panels show the second study performed 14 days later. The two studies are identical, showing no drainage to the popliteal fossa and two sentinel nodes side by side in the right groin (vertical arrow) with an identical distribution of tracer in second tier nodes higher in the right groin and pelvis. The relative brightness of all of the nodes is also identical in the two studies.
|
|
View larger version (124K):
[in this window]
[in a new window]
|
FIG. 2. Lymphoscintigraphy studies of patient 6 with a melanoma excision biopsy site on the upper back just to the right of midline after the intradermal injection of 99mTc antimony sulfide colloid. The left panels show the summed 10 minute dynamic image posteriorly over the shoulders and the middle and right panels show the delayed scans in the anterior and posterior projections over the chest and shoulders. The first study (top row) was performed after 2 injections and shows a single sentinel node in each axilla (curved arrows) and a single faint interval SLN (vertical arrow) on the back above the injection site. The repeat study done 21 days later (bottom row) again shows a single sentinel node in each axilla but now shows 3 bright interval SLNs on the back (vertical arrow). All the lymphatic collecting vessels and the sentinel nodes are brighter on the second study.
|
|
View larger version (112K):
[in this window]
[in a new window]
|
FIG. 3. Delayed lymphoscintigraphy images anteriorly over the pelvis and posteriorly over the popliteal fossa of patient 15 performed 2 hours after the intradermal injection of 99mTc antimony sulfide colloid. Three intradermal injections were given for each study, two proximal and one distal to the lesion site on the lateral shin above the ankle. The first study (top row) shows two sentinel nodes (horizontal arrow) in the left popliteal fossa and two sentinel nodes (diagonal arrow) in the left groin with some second tier activity in nodes higher in the groin and pelvis. The second study (bottom row) shows no sentinel nodes in the popliteal fossa and two brighter sentinel nodes in the left groin (diagonal arrow). The relative brightness of second tier nodes higher in the groin and pelvis is also different on the second study.
|
|
The number of intradermal injections given for the repeat study was known to be different in two patients, patients 4 and 16. The results were identical in patient 4 and even in patient 16 when only 25% of the total number of radiocolloid particles was injected at two points instead of four, the same sentinel nodes were radiolabeled though the popliteal nodes were fainter on the second study.
Thirty-five of the 37 SLN locations (95%) and 59 of the 63 SLNs (94%) were identical in the two LS studies.
|
DISCUSSION
|
|---|
Since SLN biopsy using pre-operative LS has been shown to accurately stage the regional lymph nodes, there has been an assumption that LS itself is a reliable and reproducible method of mapping lymphatic drainage from a melanoma site to the draining SLNs. Several prospective studies have supported this contention,3,4,6 however, such reproducibility has not been documented in the day-to-day routine of clinical practice. Usually our patients have their SLN biopsy within 24 h of their LS so that the SLNs can be located during surgery using a gamma detecting probe as well as blue dye without a further injection of radionuclide. Occasionally, however, as mentioned earlier, surgery is not completed within the time that the SLNs remain radiolabeled and these patients have had their surgery rescheduled. In this circumstance, we have always repeated the LS to relabel the SLNs for surgical removal. In some of our patients different physicians and technologists were involved in the repeat studies which had the potential to alter the results of LS. Our retrospective analysis of patients who have had LS repeated prior to SLN biopsy has documented a high degree of reproducibility. This confirms the robust and reliable nature of LS for locating the SLNs prior to SLN biopsy. Our results are similar to those reported in a prospective study of LS reproducibility by Kapteijn et al.,3 who identified the same SLNs in 88% of their patients. Further analysis of their data showed that they identified 44 of 47 SLNs in their 25 patients on both LS studies, an accuracy of 94%.
The only differences that we observed were two additional interval SLNs on the back in one patient and failure to radiolabel two popliteal SLNs in another. Some regard nodes in both these locations as being "interval nodes" outside standard node fields8 but it has always been our practice to include the popliteal fossa and the epitrochlear region in the standard recognized node field category. We regard an interval node as a node lying along the path of a lymphatic collecting vessel between a primary tumor site and a standard node field.9 It remains important for the patient, however, that all SLNs identified are marked on LS, including interval nodes, since we have found metastases in interval nodes with about the same frequency as we have found metastases in other SLNs.9
In the overall experience of the Sydney Melanoma Unit, 17% of patients with intermediate thickness melanomas have been found to have metastases in their SLNs.10 However, the chance that any individual SLN will contain metastasis is less than this since many patients have more than one SLN and there is usually only one positive SLN in a patient.
If there is a 10% chance of any individual SLN harboring metastasis, the implication is that LS will fail to identify a true SLN containing metastasis in less than 1% of the patients with melanoma. This error rate is greater than that reported using LS in some other tumors such as breast cancer11 and penile cancer12 where 100% reproducibility was claimed; however, this is an adequate level of accuracy to recommend LS prior to SLN biopsy to enhance the reliability of the method and is certainly much more accurate than any of the alternative approaches that are available at the present time. Pre-operative blue dye injection is also used routinely as a part of the SLN biopsy procedure in most centers and this should make up for the very occasional deficiency of LS and result in accurate SLN identification in almost all patients.
Although the site and number of SLNs seen on LS did not vary between the two LS studies in most patients, we did in some patients notice that some of the SLNs were "brighter" or "fainter" on one study versus the other. In patient 16, we presume this was due to fewer injections and fewer radiocolloid particles being injected and in six other patients the precise number of injections given on the second study was not recorded. It is possible in some others that a longer period of healing following the original excision biopsy of the melanoma led to better access to the initial lymphatic capillaries for the radiocolloid particles on the second LS study and caused better visualization of the lymphatic collecting vessels and SLNs. Part of the normal function of the lymphatic system is the removal of cellular debris and products of metabolism and this could partially overload the lymphatic system regionally when it responds to an excision biopsy and the inflammation that often follows. A possible further cause of slight differences between the two studies is that the physician performing the second LS was not aware of the precise point of intradermal injection on the first study in relation to the excision biopsy scar. Injections given at slightly different depths could also cause slight differences in the precise number of radiocolloid particles gaining access to the initial lymphatic capillaries.
Other variables that may have contributed to the slight differences observed in sequential studies include the ambient temperature of the day and the level of physical activity undertaken by the patient that day. There may have been slight differences in the intensity of local massage performed by the injector on the sequential studies. All of these factors are known to affect lymphatic drainage of the skin.13
These findings underline our contention that the degree of radiolabeling of a lymph node should not be used as a means of identifying that node as a SLN. Rather, the documentation of direct lymphatic drainage to a lymph node should be the essential requirement for defining that node as an SLN.14 It has been previously documented that the "hottest" or "brightest" SLN in a node field is not always the SLN that harbors the metastatic disease15 and that any SLN may do so.
|
CONCLUSIONS
|
|---|
The reproducibility of LS in accurately identifying the location of SLNs in patients with cutaneous melanoma after excision biopsy is remarkably good and this accuracy is maintained even in the routine clinical practice environment. We did observe minor differences in the sequential LS studies in terms of the number and visibility of the draining lymphatic collecting vessels on dynamic imaging and the "brightness" of the radiolabeling of the SLNs, on delayed imaging. This may have been due to a variety of factors as discussed earlier but did not affect the accuracy of the technique in locating the SLNs.
We estimate that variations in the reproducibility of LS will cause the non-radiolabeling of an SLN that harbors metastatic disease in less than 1% of the patients with melanoma. This error rate is sufficiently small to allow confident application of the technique of LS to provide lymphatic mapping for SLN biopsy in patients with cutaneous melanoma since one would expect these very occasional discrepancies would be corrected by the concomitant use of blue dye at the time of surgery.
|
ACKNOWLEDGMENTS
|
|---|
We would like to thank the surgeons of the Sydney Melanoma Unit for referring their patients for lymphatic mapping with LS prior to SLN biopsy and our nuclear medicine technologists who are responsible for the high quality of our scans. These currently include: Kim Ioannou, Angelique Nguyen, Rhodora Fitzgerald, Christos Koliris and in the past Ian Dyer, Tracey Smith, and Sally Raymond. We also thank Katherine Uren of Kuillustration.com for the medical illustrations in each of the figures.
Received for publication April 4, 2006.
Accepted for publication August 8, 2006.
|
REFERENCES
|
|---|
- Thompson JF, McCarthy WH, Bosch CM, et al. Sentinel lymph node status as an indicator of the presence of metastatic melanoma in regional lymph nodes. Melanoma Res 1995; 5:255–260.[Medline]
- Reintgen D, Cruse CW, Wells K, et al. The orderly progression of melanoma nodal metastases. Ann Surg 1994; 220:759–767.[Medline]
- Kapteijn BA, Nieweg OE, Valdes Olmos RA, et al. Reproducibility of lymphoscintigraphy for lymphatic mapping in cutaneous melanoma. J Nucl Med 1996; 37:972–975.[Abstract/Free Full Text]
- Tonakie A, Sondak V, Yahanda A, Wahl RL. Reproducibility of lymphoscintigraphic drainage patterns in sequential 99mTc human serum albumin and 99mTc sulfur colloid studies: implications for sentinel node identification in melanoma. Surgery 1999; 126:955–962.[CrossRef][Medline]
- Mudun A, Murray D, Herda S, et al. Reproducibility of lymphoscintigraphy and intraoperative surgical probe use to identify the sentinel node in patients with melanoma [Abstract]. J Nucl Med 1995; 36(suppl):263P.
- Mudun A, Murray DR, Herda S, et al. Early stage melanoma: lymphoscintigraphy, reproducibility of sentinel node detection, and effectiveness of the intraoperative gamma probe. Radiology 1996; 199:171–175.[Abstract/Free Full Text]
- Thompson JF, Niewind P, Uren RF, Bosch CM, Howman-Giles R, Vrouenraets BC. Single-dose isotope injection for both pre-operative lymphoscintigraphy and intraoperative sentinel lymph node identification in melanoma patients. Melanoma Res 1997; 7:500–506.[CrossRef][Medline]
- McMasters KM, Chao C, Wong SL, et al. Interval sentinel lymph nodes in melanoma. Arch Surg 2002; 137:543–549.[Abstract/Free Full Text]
- Uren RF, Howman-Giles R, Thompson JF, et al. Interval nodes: the forgotten sentinel nodes in patients with melanoma. Arch Surg 2000; 135:1168–1172.[Abstract/Free Full Text]
- Yee VS, Thompson JF, McKinnon JG, et al. Outcome in 846 cutaneous melanoma patients from a single center after a negative sentinel node biopsy. Ann Surg Oncol 2005; 12:429–439.[Abstract/Free Full Text]
- Thompson JF, Uren RF. What is a ‘sentinel’ lymph node? Eur J Surg Oncol 2000; 26:103–104.[CrossRef][Medline]
- Tanis PJ, Valdes Olmos RA, Muller SH, Nieweg OE. Lymphatic mapping in patients with breast carcinoma: reproducibility of lymphoscintigraphic results. Radiology 2003; 228:546–551.[Abstract/Free Full Text]
- Kroon BK, Valdes Olmos RA, van Tinteren H, Nieweg OE, Horenblas S. Reproducibility of lymphoscintigraphy for lymphatic mapping in patients with penile carcinoma. J Urol 2005; 174:2214–2217.[CrossRef][Medline]
- Uren RF, Thompson JF, Howman-Giles RB. (1999) Lymphatic drainage of the skin and breast. Locating the sentinel nodes. Amsterdam: Harwood Academic Publishers.
- Uren RF, Howman-Giles RB, Shaw HM, Thompson JF, McCarthy WH. Lymphoscintigraphy in high-risk melanoma of the trunk: predicting draining node groups, defining lymphatic channels and locating the sentinel node. J Nucl Med 1993; 34:1435–1440.[Abstract/Free Full Text]
- Jacobs IA, Chang CK, DasGupta TK, Salti G. High isotope counts and sentinel node positivity in patients with melanoma. Arch Surg 2003; 138:63–66.[Abstract/Free Full Text]
TOP
ABSTRACT
INTRODUCTION
