The incidence of malignant melanoma is steadily increasing in the UnitedStates. Although treatment options have been limited and are primarilysurgical, there have been some promising responses to immunotherapies, such asinterferon alfa-2b. As treatment options increase, so does the need foraccurate nodal staging of disease. The recent use of minimally invasivesurgical biopsy of the first node draining the tumor site has reduced the needfor extensive radical nodal dissection, with an associated morbidity. Althoughblue dye studies are an effective technique for identifying this"sentinel," or first draining node, lymphoscintigraphy has severaladvantages. In this article we will explore such advantages, as well as thetechnical details needed for the performance of lymphoscintigraphy. Inaddition, we will discuss the use of the intraoperative gamma probe.

Malignant melanoma

Pathology, grading, and staging-The four major types of malignantmelanoma, in order of prevalence, are superficial spreading, nodular, lentigomaligna, and acral lentiginous. Melanomas are graded by tumor thickness(Breslow grades) and/or the level of dermal invasion (Clark levels). Tumorspread is lymphatic and hematogeneous, with lymphatic spread predominating inearly metastatic disease. Additionally, the likelihood of nodal metastasis isproportional to tumor thickness. Balch and coworkers reported a 0% incidence ofregional metastases in patients with lesions less than 0.75 mm thick, a 25%incidence in patients with lesions between 0.76 to 1.50 mm thick, 51% forlesions 1.5 to 3.99 mm thick, and 62% for those with lesions greater than 4.0mm thick, based on a 3-year actuarial rate of recurrence.1 Staging of disease,as defined by the Joint Commission on Cancer, is summarized in table 1.

Treatment

In the absence of metastatic spread, treatment of malignant melanomaconsists of wide local resection of the primary lesion. Resection margins aredetermined by lesion thickness; 1 cm margin for lesions of less than 1 mm indepth and 2 to 4 cm for lesions that are 1 to 4 mm thick. Resection does notimprove survival for patients with lesions that are greater than 4 mm thick, asthese lesions have a high incidence of metastasis.

In cases of clinically palpable nodal disease, radical nodal dissection isindicated. For those cases in which the lesion is 1 to 4 mm thick and there isno evident palpable adenopathy, the use of radical node dissection to eliminatemicrometastatic disease has been controversial, and this procedure, therefore,has been referred to as "elective lymph node dissection"or ELND.2,3Recently, however, there is some evidence that ELND can prolong survival incertain groups of patients.4

The status of nodal involvement is an important prognostic indicator. Forexample, Balch and coworkers report a 48% 10-year survival rate in patientswith clinical stage I and II disease with occult nodal metastases versus a 65%10-year survival for those patients without occult nodal disease.5 Staging ofnodal involvement has become even more important with the discovery that somepatients with nodal disease have benefited from interferon therapy.6

Blue dye and the sentinel node

Lymphatic drainage, using a variety of blue dyes such as isofulfan blue andmethylene blue, has been studied for decades. The concept of identification andbiopsy of the sentinel node as a means for staging an entire nodal bed wasdeveloped in 1992 by Morton et al7 using blue dye. The sentinel node pathologyaccurately represents the pathology of the nodal bed.8 To locate the sentinelnode, the dye is injected intradermally around the tumor or biopsy site 15minutes prior to surgical exploration of the nodal basin. Timing is important,as nodal retention of the dye is transient.

Limitations of blue dye studies

The use of blue dye to locate the sentinel node is limited by the cliniciansability to predict the nodal basin that will drain lymph and cells from theprimary tumor site. Unfortunately, pathways of cutaneous lymphatic drainage arevariable. In their study of a large series of patients with melanoma, Slingluffand coworkers reported that nodal metastases were identified clinically incontralateral nodes for primary lesions outside of watershed regions in 10% ofcases.9 These watershed regions, from which lymphatic drainage is quitevariable, are found in areas such as the waist and the vertical midline of thetrunk and head (figure 1).

Studies by O'Brien et al and Wells et al have shown discordance of 34%and 84%, respectively, between the clinically suspected basin and that whichwas mapped by lymphoscintigraphy in lesions of the head and neck.10,11 A markedvariability in drainage patterns in truncal melanoma has been reported by Urenand coworkers:12 59% of their patients had drainage to two or more nodalbasins.

Sentinel node lymphoscintigraphy

In 1992, Alex and Krag adapted lymphoscintigraphy to locate the sentinelnode.13 This technique has been quite successful, as it permits preoperativeidentification of the nodal basin(s) containing sentinel node(s) (it is notuncommon to see simultaneous, or near simultaneous, drainage to more than onenodal basin). In addition, these researchers demonstrated the use of alightweight intraoperative gamma probe to locate the radioactive node withinthe wound. Using this method, timing of surgical exploration is less crucial,as most radiopharmaceuticals used for lymphoscintigraphy have a relatively longresidence time in nodal tissue (at least several hours).

Blue dye and lymphoscintigraphic techniques are not mutually exclusive. Infact, concomitant use of both these agents has been found to yield a greatersentinel node detection rate.14

Radiopharmaceuticals

The endothelial walls of the lymph capillaries are permeable to much largerparticles than blood capillaries, and radiopharmaceuticals that are smallenough to enter the lymphatics but too large to diffuse into the capillariesare preferred for lymphoscintigraphy.15 The most commonly usedradiopharmaceuticals are 99mTc-antimony sulphide colloid, 99mT-human serumalbumin, 99mTc-sulfur colloid, and 99mTc ultrafiltered sulfur colloid. Sulfurcolloid, ultrafiltered sulfur colloid, and human serum albumin (HSA)preparations are available for use in the United States. 99mTc-HSA is anon-particulate agent which demonstrates good lymph channel visualization andrapid clearance from the injection site. It appears to have a shorter residencetime in lymph nodes than ultrafiltered sulfur colloid, which may restricttiming of surgery post injection.16

Sulfur colloid is particulate and has a diameter between 100 and 1000 nm.17As particulate matter less than 100 nm is most readily mobilized from theinjection site,16 ultrafiltered sulfur colloid, prepared by slowly pushingsulfur colloid through 0.1 or 0.22 um (100 or 220 nm) filters, is the mostcommonly used radiopharmaceutical for lymphoscintigraphy. As with HSA, themigration of the ultrafiltered colloid is more rapid than the migration ofunfiltered sulfur colloid and, therefore, these smaller agents permit bettervisualization of the lymph channels. Visualization of the lymphatic channelsleading to the sentinel node improves the diagnostic certainty of thestudy.17,18 The following studies were performed with 99m-Tc-ultrafilteredsulfur colloid prepared with a 0.22 um filter.

Patient selection

It is recommended that sentinel node lymphoscintigraphy be used only onpatients with either intact primary lesions or those on whom only excisionalbiopsies have been performed, as it is generally felt that the extent ofsurgery required for wide local excision may significantly alter lymphaticdrainage in the tissue surrounding the scar.

Performance of the study

Two to four individual doses of 3.7 to 7.4 MBq (100 to 200 microCuries)ultrafiltered 99mTc-sulfur colloid in a volume of 0.1 to 0.2 cc each are drawninto 1 cc syringes. Small gauge needles (preferably size 25 or smaller) areattached to each syringe.

The patient is placed in a supine or prone position on the imaging table.The lesion or biopsy site is prepped with betadine and an alcohol pad. A drapewith a central hole is placed to expose an area of skin a few centimeters indiameter surrounding the lesion or local biopsy site. Draping the site isimportant to prevent contamination, as leakage from the site of injection iscommon. Occasionally the backpressure following injection is great enough tocause vertical spray; therefore, the person injecting the material shouldconsider covering the injection site with gauze prior to removing the needle.

To prepare the injectate, a small amount of air should be drawn into eachsyringe to "follow" the fluid administration. Small gauge needles maybe bent (using partially removed caps) at their middle at an approximately 30to 45 degree angle, bevel up. Injections are intradermal, and as such should besuperficial, as most of the epidermis is only 0.1 to 0.2 mm thick (though solescan be up to 0.7 mm thick). Two to four intradermal injections are placedaround the lesion, each approximately 1 to 2 cm from the lesion. In somepostoperative scars a more distant injection may be needed to avoid fibrotictissue. A successful intradermal injection is characterized by a painful, tightblanched wheal with accentuation of the hair follicles. A deeper injection intothe subcutaneous tissues will result in undesirable large, flat tissue swellingand poor lymphatic migration.

Imaging

Imaging should commence immediately after the injection because migrationwith ultrafiltered sulfur colloid (as well as 99mTc-HSA) is often rapid. Theinjection site should be imaged first to ascertain successful focal injectionsand to assess local contamination. Some of the skin contamination can beremoved with alcohol wipes; contaminated clothing and sheets should be removedfrom the field of view. Prior to imaging, the injection site can be shielded toimprove visualization of lymph channels, which can be faint. Rapid imaging isrequired to follow channels to nodal beds; at our institution, we obtainsequential 30 to 60 second static images. Nodes should be imaged and recordedin order of appearance. Channels should be followed quickly, as they are seenonly transiently. The sentinel node is generally visualized within the firstseveral minutes of imaging, and it is not uncommon to visualize the node within1 to 2 minutes post injection.

Nodes should be marked on the skin with indelible markers while the patientis posed, if possible, in the surgical position. For example, axillary nodesare marked with the arm extended perpendicularly from the patients side or, ifthe surgeon desires, above the head. When possible, lateral, oblique, ortomographic views should be used to estimate nodal depth. These views also willhelp differentiate nodal groups that may be superimposed on anterior orposterior views, such as subscapular and anterior axillary nodes andsupraclavicular and posterior cervical nodes. Two such cases are illustrated infigures 2 and 3.

Intraoperative gamma probes

Intraoperative probes are used to locate radioactive lymph nodes forremoval. These probes are lightweight hand held instruments containing ashielded detector attached to a charge sensitive pre-amplifier. The probe isconnected to a rechargeable control unit with visual and audio signal output.

Although some probes are designed with Nal crystals, the majority usesemiconductors (figure 4). Semiconductor detectors can be thought of as thesolid equivalent of a gas detector, with all the advantages of a solidmaterial. In gas detectors, ion pairs (a positively charged gas ion and anegatively charged electron) are created by ionizing radiation. Insemiconductors, electrons also are dislodged by ionizing radiation, but insteadof positively charged gas molecules, positive "holes" are createdwithin the crystalline structure (figure 5).

The primary advantage of these materials over their gaseous counterparts istheir relative density. Therefore, a gamma photon is more likely to interactwith a molecule in the semiconductor than in an equivalent volume of gas. Inaddition, a greater yield of charges will result from an interaction in asemiconductor because the electrons in the semiconductor are less tightly boundthan those in gas; approximately 3 keV of energy is needed to dislodged anelectron in a semiconductor material, compared to approximately 35 keV in agas.18

Semiconductor materials are less conductive than the metals which are usedfor carrying electric current (such as copper), hence the name"SEMIconductor". A few of the semiconductor compounds that are areavailable for use include cadmium telluride (CdTe), cadmium zinc telluride(CdZnTe), and zinc telluride (ZnTe).

The semiconductor (or crystal) detector is shielded and collimated with asleeve of tungsten so that only emissions directly in front of the probe arecounted. The tungsten shielding blocks 99 to 99.9% of 99mTc photons (figure 4).

The energy window for the semiconductor probe is set from above 80 keV (theenergy of tungsten fluorescence) up to approximately 200 keV (or higher). Theoverall sensitivity of the probe is approximately 60 to 70% for 99mTc.

Control units for the probes are often designed with audible output, inaddition to numerical count displays. Some manufacturers have modified theauditory output so that the user can differentiate count rates up to a fewthousand cts/sec. These modifications generally involve conversion of countrates into oscillating frequencies, or warbles, in which the pitch frequency isproportional to the number of counts.

When using the probe, remember to exclude photons arising from the injectionsites from the field of view when searching for or examining a suspectedradioactive node. Once located (figures 6 and 7), the sentinel node is excised,and its counts are measured while the probe is directed away from the patient.It may be helpful to compare counts in the node with those from a small sampleof non-nodal tissue excised from the site. The node should be at least two tothree times as radioactive as the background tissue, and is generally 10 to 100times as radioactive. Following excision of the node, it is wise to check thewound for residual radioactivity.

Radiation safety

Radiation exposure from nodal tissue is minimal, and even that from theprimary injection site is not excessive. In a preliminary study, Edmiston andcoworkers reported an average dose of 0.4 ± 0.9 mR/hr at the surface ofthe sentinel node and 17.7 ± 3.6 mR/hr at the surface of the injectionsite.19 Miner et al recorded the level of exposure to a surgeon's hands as9.4 ± 3.6 mRem/case.20

A controversy exists concerning the need for a delay prior to processingradioactive tissue in pathology. In addition to personnel exposure, somelaboratories prefer not to contaminate their cryotomes with radioactive tissue.Both Edmiston's study and Miner's recommend waiting 48 to 72 hours toprocess tissue containing the primary injection site.19,20

Summary

Lymphoscintigraphy is a valuable tool for accurate and rapid localization ofsentinel nodes, used for staging cutaneous malignant melanoma. In contrast toblue dye studies, lymphoscintigraphy is able to detect sentinel nodes inunsuspected locations and in multiple nodal beds. Successful use of thistechnique requires correct injection of the agent, rapid early imaging, lateraland oblique imaging for accurate localization of nodes, careful skin markingwith the patient in the surgical position, and correct use of theintraoperative probe. AR

Dr. Powsner is Assistant Professor of Radiology at Boston University Schoolof Medicine, Boston Medical Center, Boston, MA.

References

1. Balch MB, Murad TM, Soon SJ, et al: Tumor thickness as a guide tosurgical management of clinical stage 1 melanoma patients. Cancer 43:883-888,1979.

2. McCarthy WH, Shaw HM, Cascinelli N, et al: Elective lymph node dissectionfor melanoma: Two perspectives. World J Surg 16(2):203-213, 1992.

3. Cole DJ, Baron PL: Surgical management of patients with intermediatethickness melanoma: Current role of elective lymph node dissection. Semin Oncol23(6):719-724, 1996.

4. Balch CM, Soung SJ, Bartolucci AA, et al: Efficacy of an electiveregional lymph node dissection of 1 to 4 mm thick melanomas for patients 60years of age and younger. Ann Surg 224(3):255-266, 1996.

5. Balch CM, Soon SJ, Shaw HM, et al: An analysis of prognostic factors in8500 patients with cutaneous melanoma. In: Balch CM, Houghton AN, Milton GW, etal (eds): Cutaneous Melanoma, ed 2. Philadelphia, J.B. Lippincott Company,1992.

6. Kassas HE, Kirkwood JM: Adjuvant application of interferons. Semin Oncol23(6):737-743, 1996.

7. Morton DL, Wen DR, Wong JH, et al: Technical details of intraoperativelymphatic mapping for early stage melanoma. Arch Surg 127:392-399, 1992.

8. Reintgen D, Cruse CW, Wells K, et al: The orderly progression of melanomanodal metastases. Ann Surg 220(6):759-767, 1994.

9. Slingluff CL, Stidham KR, Ricci WM, et al: Surgical management ofregional lymph nodes in patients with melanoma. Ann Surg 219(2):120-130, 1994.

10. O'Brien CJ, Uren RF, Thompson JF, et al: Prediction of potentialmetastatic sites in cutaneous head and neck melanoma using lymphoscintigraphy.Am J Surg 170:461-466, 1995.

11. Wells KE, Cruse CW, Daniels S, et al: The use of lymphoscintigraphy inmelanoma of the head and neck. Plast Reconstr Surg 93(4):757-761, 1994.

12. Uren RF, Howman-Giles RB, Shaw HM, et al: Lymphoscintigraphy inhigh-risk melanoma of the trunk: Predicting draining node groups, defininglymphatic channels and locating the sentinel node. J Nucl Med 34(9):1435-1440,1993.

13. Alex JC, Krag DN: Gamma probe guided localization of lymph nodes. SurgOncol 2(3):137-144, 1993.

14. Albertini JJ, Cruse WC, Rapaport D, et al: Intraoperativeradiolymphoscintigraphy improves sentinel lymph node identification forpatients with melanoma. Ann Surg 223(2):217-224, 1996.

15. Bergqvist LP, Strand SE, Persson BR: Particle sizing and biokinetics ofinterstitial lymphoscintigraphic agents. Semin Nucl Med 13(1):9-19, 1983.

16. Glass EC, Essner R, Morton DL: Kinetics of three lymphoscintigraphicagents in patients with cutaneous melanoma. J Nucl Med 39(7):1185-1190, 1998.

17. Hung JC, Wiseman GA, Wahner HW, et al: Filtered technetium-99m-sulfurcolloid evaluated for lymphoscintigraphy. J Nucl Med 36(10):1895-1901, 1995.

18. Goldfarb LR, Alazraki NP, Eshima D, et al: Lymphoscintigraphicidentification of sentinel lymph nodes: Clinical evaluation of 0.22-umfiltration of Tc-99m sulfur colloid. Radiology 208(2):505-509, 1998.

19. Powsner RA, Powsner ER: Non-scintillation detectors. In: Essentials ofNuclear Medicine Physics, pp 59-60. Malden, Massachusetts, Blackwell Science,1998.

20. Edmiston K, Deye J, Talkington G, Seneca R: Management of sentinel lymphnode biopsy specimens: The radiation safety issues. Abstract 46, Surgicaloncology meetings, San Diego, March, 1998.

21. Miner T, Shriver C, Flicek P, et al: Additional guidelines for the safemanagement of radiation associated with sentinel lymph node biopsy. Abstract27, Surgical oncology meetings, San Diego, March, 1998.