Int J Clin Oncol (2006) 11:214–220 DOI 10.1007/s10147-005-0554-1
© The Japan Society of Clinical Oncology 2006
ORIGINAL ARTICLE Tsuyoshi Ishihara · Atsushi Kaguchi · Shigeto Matsushita Shinya Shiraishi · Seiji Tomiguchi · Yasuyuki Yamashita Toshiro Kageshita · Tomomichi Ono
Management of sentinel lymph nodes in malignant skin tumors using dynamic lymphoscintigraphy and the single-photon-emission computed tomography/computed tomography combined system
Received: June 30, 2005 / Accepted: December 12, 2005
Abstract Background. The differentiation of true sentinel lymph nodes from nonsentinel lymph nodes is difficult in cases of multiple radiolabeled or dyed lymph nodes. Methods. We examined the locations of sentinel lymph nodes in melanoma and other malignant skin tumors by using dynamic lymphoscintigraphy and the single-photonemission computed tomography/computed tomography (SPECT/CT) combined system. Results. Sentinel lymph nodes were detected in 45 of the 53 patients examined using only the ordinary blue dye method (85%), and were detected in all 35 patients examined using the SPECT/CT method (100%). Twenty of the 35 patients mentioned above had one sentinel lymph node. Multiple sentinel lymph nodes were demonstrated in the head and neck areas using the SPECT/CT method. Significant differences (P = 0.0015) in the numbers of sentinel lymph nodes were found between the blue dye method only and the SPECT/CT method in the neck area. Popliteal sentinel lymph nodes were recognized in three patients, and cubital sentinel lymph nodes were recognized in two patients. Two patients had plural regional lymph nodes: one had popliteal and groin sentinel lymph nodes, while the other had cubital and axillary sentinel lymph nodes. The probe counts of the popliteus and cubitus were significantly lower (P = 0.0241)
T. Ishihara (*) · A. Kaguchi · S. Matsushita Department of Plastic and Reconstructive Surgery, Graduate School of Medical Sciences, Kumamoto University, 1-1-1 Honjo, Kumamoto 860-0811, Japan Tel. +81-96-373-5233; Fax +81-96-373-5235 e-mail:
[email protected] S. Shiraishi · S. Tomiguchi · Y. Yamashita Department of Diagnostic Radiology, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, Japan T. Kageshita Department of Dermatology, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, Japan T. Ono Graduate School of Medical Sciences, Kumamoto University, Kumamoto, Japan
than the counts in the groin, axilla, and neck areas. Micrometastatic sentinel lymph nodes were recognized in four patients, and two patients had metastases in both sentinel and nonsentinel lymph nodes. Conclusions. Dynamic lymphoscintigraphy was useful when we were concerned about cubital and popliteal lymph nodes. The SPECT/CT combined system was useful in recognizing the anatomical location of sentinel lymph nodes before biopsy. The detection rate of sentinel lymph nodes using the SPECT/CT method was always better than that with the blue dye method (P = 0.0197). Key words Sentinel lymph node · Dynamic lymphoscintigraphy · Single-photon-emission computed tomography · Combined system · Melanoma · Skin cancer
Introduction The detection rate of sentinel lymph nodes (SLNs) in cancer is improved by using the blue dye method with the radioisotope method.1 However, identifying true SLNs from nonSLNs is difficult in cases of multiple radiolabeled or dyed lymph nodes. In particular, the SLN count by the ordinary gamma probe method varies by institution, so that method is not reliable for detecting SLNs. The regional lymph nodes of breast cancer are in the axillary region only. However, some melanoma and other skin cancers have plural regional lymph nodes. Skin cancer, such as that found primarily on the trunk, may have both axillary and groin lymph channels. Melanoma in the lumbar area that had both axillary and groin SLNs has also been seen. Melanoma and other skin cancers can occur in any area of the skin, and true SLNs may be overlooked if the ordinary gamma probe method is used to study skin tumors. Few articles have reported the location of SLNs in relation to skin tumor location, and a combined image of lymphoscintigraphy and computed tomography was developed to detect SLNs anatomically. For the first time,
215 Table 1. Patient information Patient no.
Age (years)
Sex
Diagnosis
Tumor location
SLN location
No. of SLNs
LN dissection
Metastasis SLN
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35
65 75 75 57 87 78 74 35 88 80 54 75 70 82 51 72 37 82 81 73 68 65 79 61 80 69 88 75 43 65 82 60 63 78 71
M M M M F F F F F F F F M F F M M F F M F M M F M M F F M F F M M F M
ALM ALM LMM ALM ALM ALM SSM SSM Porocarcinoma SCC SSM Bowen ca. ALM ALM ALM ALM SCC ALM ALM SSM ALM LMM ALM NM SCC SSM ALM Mucous M SSM Bowen ca. LMM SCC SCC SCC SSM
Planta Heel Cheek 5th toe Planta Heel Posterior lower leg Posterior lower leg Palm Temple Popliteus Posterior thigh Lateral heel Index finger Planta Planta Medial malleolus Big toe nail Thumb nail Planta Thumb nail Nose top Thumb nail Post auricule Auricule Post lateral neck Lateral planta Genitalia Lower leg 3rd toe Cheek Under lip Posterior neck Ring finger Back
Groin Groin, popliteus Submental, upper neck Groin Groin Groin Groin Groin Cubitus Preauricla Groin Groin Popliteus Axilla Groin Popliteus Groin Groin Axilla Groin Axilla Submandible Axilla Neck Neck Neck Groin Bilateral groin Groin Groin Submandible Submental Neck Axilla, cubitus Axilla
1 2 4 1 1 1 1 1 1 1 1 1 1 2 1 2 1 1 1 2 1 2 2 2 2 2 2 2 1 1 2 2 1 2 1
− − − + − − + + − − − − − − − − − − − − + − − + − − − + − − − − − − +
− − − + − − − + − − − − − − − − − − − − − − − − − + − − − − − − − − +
Other LN (no.)
+(1/11) −(0/13) +(2/12) −(0/2)
−(0/13) −(0/26)
−(0/17)
−(0/3) −(0/20)
M, male; F, female; ALM, acral lentiginous melanoma; LMM, lentigo maligna melanoma; SSM, superficial spreading melanoma; SCC, squamous cell carcinoma; SLN, sentinel lymph node; LN, lymph node
we examined SLNs in melanoma and other malignant skin tumors using dynamic lymphoscintigraphy and the singlephoton-emission computed tomography/computed tomography (SPECT/CT) combined system, and we report on the significance of this method on SLN biopsies.
Patients and methods Patients During the period from June 2003 to March 2005, 35 patients were examined for SLNs in malignant skin tumors using the SPECT/CT method at the Department of Dermatology, Graduate School of Medical Sciences, Kumamoto University. There were 15 men and 20 women with an age range of 35–88 years (mean 70 years). The diagnoses were as follows: melanoma, 26 cases; squamous cell carcinoma, 6 cases; Bowen carcinoma, 2 cases; porocarcinoma, 1 case. The tumor locations are shown in Table 1. Lymph node dissection was performed in patients in whom the mela-
noma clinically exhibited a nodule or an ulcer, and also in patients in whom micrometastases in the SLN was recognized on histopathology. Written informed consents were obtained from each patient before SLN biopsy. In one case in which micrometastases in the SLN was recognized, lymph node dissection was not performed because the patient did not consent. During the period from 1998 to 2002, 53 patients were examined for SLNs in malignant skin tumors using the blue dye method only in our department. Statistical analyses were performed on the blue dye method only and the SPECT/CT method. The statistical analyses were performed using the Mann–Whitney U-test and Fisher’s exact probability test software (Statview-J, Version 5.0, SAS Institute, Cary, NC, USA). System design We used a combined SPECT/CT system that incorporated a commercially available gantry-free SPECT with dual-head
216 Fig. 1. The SPECT/CT combined system. Single-photonemission computed tomography (left) and lymphoscintigraphy (middle) is fused, and the radiolabeled lymph node can be detected by computed tomography (right) in axial, coronal, and sagittal sections
Axial
Coronal
Sagittal
detectors (Skylight, ADAC Laboratories, Milpitas, CA, USA) and an eight multidetector-row CT scanner (LightSpeed Ultra Instrument, GE, Milwaukee, WI, USA). The two instruments were juxtaposed so that the CT table bearing the patient could be moved directly into the SPECT scanner before CT scanning. As a result, each patient was positioned identically for SPECT and CT imaging (Fig. 1).
Lymphoscintigraphy All patients underwent cutaneous lymphoscintigraphy on the day before surgery using 99mTc-phytate (DAI-ICHI Radioisotope Laboratories, Tokyo, Japan). The radioactive tracer was injected intradermally in four equal parts, each in a volume of approximately 0.2 m (2–5 mCi), around the primary skin tumor or biopsy site. Dynamic lymphoscintigraphy was performed with gantry-free SPECT with dual-head detectors imaging one frame every 60 s for 30 min to identify focal areas of accumulation, followed by multiple 5-min static images. In all patients, 3-h postinjection delayed images were also obtained.
Single-photon-emission computed tomography/computed tomography imaging Single-photon-emission computed tomography imaging started 3 h after the intradermal administration of 99mTcphytate. Single-photon-emission computed tomography data acquisition was performed with a vertex generalpurpose parallel-hole (VXGP) collimator. A 360° SPECT scan encompassing the thorax and upper abdomen was acquired. Computed tomography images without contrast administration were obtained in this study. Reconstructed CT images were processed into Digital Imaging and Communications in Medicine (DICOM) data and then transferred to Pegasys (ADAC Laboratories, Milpitas, CA, USA), which is a workstation for SPECT processing. One lumen of a three-way stopcock (inner diameter 4 mm, length 10 mm) containing an aqueous solution of 99m Tc-phytate and a contrast medium was used as an external fiducial marker. To obtain a precise record of both images, external fiducial markers were fixed to the common platform for SPECT and CT imaging. The two scans were performed sequentially. Fusion of the SPECT images
217
with the CT images was manually performed by aligning the external fiducial markers of the two images on a workstation. Surgery Measurements of radioactivity in the radiolabeled lymph nodes were made intraoperatively with a hand-held gamma probe (Crystal CXS-SG03; Anzai Medical, Tokyo, Japan). Sentinel lymph nodes showed increased focal radiotracer uptake, and counts were accumulated during 10-s intervals and recorded. Small incisions were made over areas of increased activity. Ex vivo counts of the SLNs were obtained and compared with the nodal bed counts after removal. Intraoperative SLN mapping was also performed with 1%–2% sulfan blue dye (patent blue violet; Chroma, Bonn, Germany). The blue dye, 0.5–1.0 m, was injected intradermally at the site of the tumor margin with a 5–10-mm interval. All harvested SLNs were examined histopathologically using routine hematoxylin–eosin staining, and we also performed immunochemical staining for gp100 and MART-1 in and around the SLNs of patients with melanoma. If the SLNs contained tumor cells, a complete lymphadenectomy was planned for a later date.
locations of the SLNs are shown in Fig. 2; more than one SLN was revealed in the head and neck area. Significant differences (P = 0.0015) in the numbers of SLNs were found between the blue dye method only and the SPECT/CT method in the neck area (Table 3). The probe counts of excised SLNs were generally higher than those of ex vivo SLNs but not significantly so (Fig. 3). In five cases, the probe count of SLNs was lower than that of ex vivo SLNs. Lymph node regions in these five cases were as follows: neck, 2 cases; groin, 2 cases; axilla, 1 case. Popliteal SLNs were recognized in 3 patients: two had popliteal SLNs only, and the third had both popliteal and groin SLNs. The probe counts of the popliteus and cubitus were significantly lower (P = 0.0241) than in the groin, axilla, and neck areas (Fig. 4). A cubital SLN was recognized in two patients: one of these had both a superficial cubital node and an axillary node. The details of the patient who had a profunda cubital SLN were as follows. The patient had a tumor originating on the palm, and three lymph nodes were detected by static lymphoscintigraphy. The lymph nodes were recognized in both the cubital and axillary regions: the two cubital SLNs were (1) a superficial cubital node which was located in the medial epicondyle of the humerus, and (2) a profunda cubital node which was located in the forearm (Fig. 5a). The profunda cubital node was detected first, the axillary node was detected second, and the superficial cubital node was detected last by dynamic scintigraphy (Fig. 5b). The lymph channel of the profunda cubital node and axillary node was
Results Using the SPECT/CT method, sentinel lymph nodes were detected in all patients: one SLN in 20 patients, two SLNs in 14 patients, and more than three SLNs in 1 patient. The detection rate of SLNs is shown in Table 2. The detection rate using the blue dye method only was 85%, and the rate using the SPECT/CT method was 100% (P = 0.0197). The
Table 2. The detection rate of sentinel lymph nodes. There was a significant difference (P = 0.0197) in the detection rate between the blue dye method only and the SPECT/CT method Method
Number of cases examined
Number of cases detected
Rate (%)
Blue dye only SPECT/CT
53 35
45 35
85 100
SPECT/CT, single-photon-emission computed tomography/computed tomography
Fig. 2. Classification of sentinel lymph node number by region. Gray columns, 1 LN (Lymph node); black columns, 2 LNs; white columns, more than 3 LNs
Table 3. The number of sentinel lymph nodes SLN number
Number of cases by LN location Groin
1 2 More than 3
Axilla
Neck
Blue dye only
SPECT/CT
Blue dye only
SPECT/CT
Blue dye only
SPECT/CTa
11 4 4
14 3 0
8 2 0
4 2 0
11 1 0
2 6 1
a Significant differences (P = 0.0015) in the number of SLNs were found between the blue dye method only and the SPECT/CT method in the neck area
Fig. 3. Gamma probe counts of ex vivo, postbiopsy, and excised radiolabeled lymph nodes. The probe counts of excised SLNs were higher than those of ex vivo SLNs but not significantly so (P = 0.4849). Diamonds, ex vivo count; squares, bed count after removal; triangles, excised LN count
Fig. 4. Difference in ex vivo counts by region. The probe counts of the popliteus and cubitus were significantly lower (P = 0.0241) than the counts in the groin, axilla, and neck areas. Medium gray, groin (n = 13); dark gray, popliteus (n = 3); white, axilla (n = 8); pale gray, cubitus (n = 3); black, neck (n = 13)
marker
axillary node superficial cubital node Profunda cubital node
profunda cubital node
a
c Fig. 5a–d. a Static lymphoscintigraphy of the upper extremity. The superficial cubital node, profunda cubital node, and axillary node were detected. b Dynamic lymphoscintigraphy of the upper extremity. Only
b
d the profunda cubital node was detected. c The SPECT/CT threedimensional image. Arrows indicate radiolabeled lymph nodes. d The profunda cubital node (arrow) during surgery
219
found to be continuous on dynamic lymphoscintigraphy. The probe counts of excised lymph nodes were as follows: profunda cubital nodes, 250; axillary nodes, 200; superficial cubital nodes, 80. As mentioned above, the profunda cubital node was a true SLN. All of the three radiolabeled lymph nodes were examined after surgery and had no micrometastases. Micrometastatic SLNs were recognized in 4 of the 35 patients. Seven of the 35 patients underwent lymph node dissections: of these 7 patients, two had metastases in both SLN and nonSLNs, four had no metastases in either SLN or nonSLNs, and one had metastasis in the SLN only.
Discussion Morton et al.2 originally described injecting blue dye into the dermis around a primary tumor. The blue dye method is easy and does not require the use of special tools. However, it is necessary to make a large incision in the skin when SLNs cannot be recognized in the predicted area. This technique requires experience to achieve a high success rate.3 The radioisotope method using a hand-held gamma probe is useful to compensate for the weakness of the blue dye method. Krag et al.4 and other researchers5–7 have described injecting technetium Tc 99m sulfur colloid around the primary tumor site and using a hand-held gamma probe to localize SLNs. Sentinel lymph node procedures have been reported to be even more successful when blue dye is used in combination with a gamma probe, rather than alone.8–11 We have used the radioisotope method of SPECT/CT plus the blue dye method after a study using blue dye only, and the success rate improved to 100%. Melanoma and other skin cancers may occur in any area of the skin. Some melanomas and other skin cancers have multiple regional lymph nodes, but little research has been carried out in the popliteal and cubital areas.12 The SLN count by the ordinary gamma probe method varies by institution, so that method is not reliable for detecting SLNs. True SLNs may be overlooked if the ordinary gamma probe method is used in cases that have SLNs in plural regional lymph nodes, or in lymph channels in the cubitus and popliteus. In this study, popliteal and cubital SLNs were specifically investigated using the combined system of lymphoscintigraphy and SPECT/CT. Previous studies using lymphoscintigraphy11,13,14 have shown that cutaneous lymphatic drainage pathways are variable among patients, especially in the head and neck area. Several studies have reported success rates of 90% or higher for SLN biopsies in patients with melanomas of the head and neck.11,13 Dynamic lymphoscintigraphy with a radioactive colloid allows the visualization of lymphatic channels from the injection site to the lymph nodes.15,16 This allows the identification of true SLNs and distinguishes multiple SLNs from nonSLNs. Dynamic lymphoscintigraphy is especially useful in the detection of popliteal and cubital SLNs. Static lymphoscintigraphy can visualize lymph channels and lymph nodes. However, it is difficult to
decide on true SLNs in cases of plural regional radiolabeled lymph nodes. Dynamic lymphoscintigraphy can visualize the flow of lymph with the passage of time and confirm a true SLN. In this series, popliteal SLNs were recognized in three patients and cubital SLNs were recognized in two patients using dynamic lymphoscintigraphy. In the two patients in whom popliteal SLNs were recognized, radiolabeled lymph nodes were found in the popliteus only. The SPECT/CT method is useful to compensate for the weaknesses of the gamma probe method and lymphoscintigraphy. Surgeons will be able to recognize SLNs easily in surgery because their anatomical location is identified before biopsy by using the SPECT/CT method. We analyzed three-dimensional SPECT/CT images and visualized the anatomical location of SLNs (Fig. 5c). This method allowed us to pinpoint the biopsy sites of SLNs. The SLN that was a profunda cubital node in the forearm region was clearly observed using this method (Fig. 5d). Regional lymph nodes of the upper extremities are axillary and superficial cubital nodes according to the classification of the American Joint Committee on Cancer (AJCC).17 A profunda cubital node is not classified as being in the upper extremities, but there is the possibility of a deep palmar skin tumor having a profunda cubital node as a SLN. The lymphatic drainage patterns of the head and neck are multiple and varied, and an average SLN number of 1.25–2.5 has been reported by investigators.13 Our average number of 2.0 SLNs per patient was almost the same. Significant differences in the number of SLNs were found between the blue dye method only and the SPECT/CT method in the neck area. SLNs of the neck area may be overlooked when the blue dye method is used exclusively. The SPECT/CT method was of limited use if the radiolabeled lymph node was near a primary skin tumor, such as in the head and neck area. The reason for this is that the radiolabeled lymph node is masked by highly radiolabeled primary areas. We studied a case of squamous cell carcinoma on the ear. The radiolabeled lymph nodes in this case were recognized in the upper neck area around the external jugular vein using the SPECT/CT method. However, this lymph node was not a true SLN. The true SLN was recognized in the postauricular area near the primary tumor. We decided the the postauricular node was the true SLN because it was stained by blue dye on surgery, and the efferent lymphatic of the stained lymph node was connected to the next lymph node that was radiolabeled using the SPECT/ CT method. Surgeons generally do not trust SPECT/CT images and everyone must pay attention during the SLN biopsy, allowing for the anatomical lymphatic flow from the primary area. The SPECT method has recently become widely accepted in Japan,18 and the SPECT image and CT image can be fused on a workstation using appropriate software. When compared with other forms of diagnosis such as MRIs and CTs, we believe that this new method will not pose an excessive financial burden. The ordinary gamma probe method must be utilized in the operating theater, while the SPECT/CT method is handled in the radiological department. There is no problem ethically or legally with
220
the SPECT/CT method if surgeons do not use a hand-held gamma probe in the operating theater. In the near future, every surgeon would ideally perform the biopsy of SLNs while looking at SPECT/CT images in the operating theater. Acknowledgment The authors thank Junko Kawasaki for her statistical analyses.
References 1. Brady MS, Coit DG (1997) Sentinel lymph node evaluation in melanoma. Arch Dermatol 133:1014–1020 2. Morton DL, Wen DR, Wrong JH, et al. (1992) Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 127:392–399 3. Ishihara T, Nakamura A, Kageshita T, et al. (2001) Sentinel lymph node study in 30 cases of melanoma. J Dermatol 28:347– 352 4. Krag DN, Meijer SJ, Weaver DL, et al. (1995) Minimal-access surgery for staging of malignant melanoma. Arch Surg 130:654– 658 5. Alex JC, Weaver DL, Fairbank JT, et al. (1993) Gamma-probeguided lymph node localization in malignant melanoma. Surg Oncol 2:303–306 6. Alex JC, Krag DN (1993) Gamma-probe-guided localization of lymph nodes. Surg Oncol 2:137–143 7. van der Veen H, Hoekstra OS, Paul MA, et al. (1994) Gammaprobe-guided sentinel node biopsy to select patients with melanoma for lymphadenectomy. Br J Surg 81:1769–1770
8. Bostick P, Essner R, Sarantou T, et al. (1997) Intraoperative lymphatic mapping for early-stage melanoma of the head and neck. Am J Surg 174:536–539 9. Wells KE, Rapaport DP, Cruse CW, et al. (1997) Sentinel lymph node biopsy in melanoma of the head and neck. Plast Reconstr Surg 100:591–599 10. Alex JC, Krag DN, Harlow SP, et al. (1998) Localization of regional lymph nodes in melanomas of the head and neck. Arch Otolaryngol Head Neck Surg 24:135–140 11. Carlson GW, Murray DR, Greenlee R, et al. (2000) Management of malignant melanoma of the head and neck using dynamic lymphoscintigraphy and gamma-probe-guided sentinel lymph node biopsy. Arch Otolaryngol Head Neck Surg 126:433– 437 12. Ishihara T, Kageshita T, Matsushita S, et al. (2003) Investigation of sentinel lymph nodes of the axillary and cubital regions in upperextremity malignant skin tumors: a series of 15 patients. Int J Clin Oncol 8:297–300 13. Wilt HW, Thompson JF, Uren RF, et al. (2004) Correlation between preoperative lymphoscintigraphy and metastatic nodal disease sites in 362 patients with cutaneous melanoma of the head and neck. Ann Surg 239:544–552 14. Albertini JJ, Cruse CW, Rapaport D, et al. (1996) Intraoperative radiolymphoscintigraphy improves sentinel lymph node identification for patients with melanoma. Ann Surg 223:217–224 15. Uren RF, Howman-Giles RB, Shaw HM, et al. (1993) Lymphoscintigraphy in high-risk melanoma of the trunk: predicting draining node groups, defining lymphatic channels, and locating the sentinel node. J Nucl Med 34:1435–1440 16. Pijpers R, Collet GJ, Meijer S, et al. (1995) The impact of dynamic lymphoscintigraphy and gamma-probe guidance on sentinel node biopsy in melanoma. Eur J Nucl Med 22:1238–1241 17. Greene FL, Morrow M, Balch C, et al. (eds) (2002) AJCC Cancer staging handbook, 6th edn. Springer, New York, p 229–254 18. Kubo A, Kinoshita F (2004) Nuclear medicine notebook, 4th edn. Kanehara, Tokyo, p 30–43