Application of sentinel node biopsy to gastric cancer surgery

Application of sentinel node biopsy to gastric cancer surgery Masahiro Hiratsuka, MD, Isao Miyashiro, MD, Osamu Ishikawa, MD, Hiroshi Furukawa, MD, Kazuyoshi Motomura, MD, Hiroaki Ohigashi, MD, Masao Kameyama, MD, Yo Sasaki, MD, Toshiyuki Kabuto, MD, Shingo Ishiguro, MD, Shingi Imaoka, MD, and Hiroki Koyama, MD, Osaka, Japan

Background. Sentinel node (SN) biopsy has been tried in the management of a variety of cancers with the hope that it would eliminate many unnecessary lymph node dissections, resulting in less morbidity. This important technique, however, has not been tried in gastric cancer surgery. The feasibility of SN biopsy and its accuracy in predicting the lymph node status in patients with gastric cancer were examined in the current study. Patients and methods. SN biopsy was performed in patients with T1 (n = 44) or T2 (n = 30) gastric cancers (ie, immediately after laparotomy, indocyanine green was injected around the primary tumor, and the green-stained nodes [SNs: 2.6 ± 1.7 nodes per patient] were removed). Then, gastrectomy with extended lymphadenectomy was performed. The unstained nodes (non-SNs: 39 ± 18 nodes per patient) were obtained from the resected specimens. Both SNs and non-SNs were subjected to histologic examination with hematoxylin-eosin. Results. SNs could be identified in 73 of 74 patients (success rate, 99%). Of these 73 patients, 10 had lymph node metastases in SNs or non-SNs, or both; 6 in both SNs and non-SNs; 3 in SNs alone; and 1 in non-SNs alone. The sensitivity of the SN status in the diagnosis of the lymph node status of the patient was 90% (9/10) and specificity was 100% (63/63). Sensitivity was 100% in the T1 group (n = 44) and 88% in the T2 group (n = 29). Conclusions. SN biopsy using indocyanine green can be performed with a high success rate, and the SN status can predict the lymph node status with a high degree of accuracy, especially in patients with T1 gastric cancer. (Surgery 2001;129:335-40.) From the Department of Surgery and Pathology, Osaka Medical Center for Cancer and Cardiovascular Diseases, Osaka, Japan

SURGICAL RESECTION remains the only potentially curative treatment for patients with gastric cancers. As a result of the widespread use of gastric endoscopy during the last 2 decades, there has been an increasing chance of resecting early-stage gastric cancers.1 Even though a primary tumor is small, a major operation, including proximal, distal, or total gastrectomy, is usually conducted as a “standard procedure” rather than wedge or segmental resection.2,3 The major operation seems to have a theoretical advantage in that it can concomitantly eradicate cancer cells, if present, in the perigastric lymph nodes along with the resection Accepted for publication September 10, 2000. Reprint requests: Masahiro Hiratsuka, MD, Department of Surgery, Osaka Medical Center for Cancer and Cardiovascular Diseases; 1-3-3 Nakamichi, Higashinari-ku, Osaka 537-8511, Japan. Copyright © 2001 by Mosby, Inc. 0039-6060/2001/$35.00 + 0 11/56/111699 doi:10.1067/msy.2001.111699

of the lesser and greater curvatures. However, the incidence of nodal involvement is reported to be as low as 2% to 18% when the depth of cancer invasion is limited in the mucosal or submucosal layer (T1) and about 50% when cancers invade the muscular or subserosal layer (T2).4 Thus, it is considered that we are performing a larger lymphadenectomy than necessary in a considerably high proportion of patients with T1 and T2 gastric cancers. If we had some reliable indicator that could predict the absence of lymph node metastases with a high accuracy, we would be able to not only preserve a larger volume of the stomach but also to eliminate the lymphadenectomy, resulting in less morbidity. In 1992, Morton et al5 reported that the first lymph node to receive drainage from the primary tumor, the sentinel node (SN), was successfully detected by dye injection in cutaneous melanoma. Since then, this technique has been confirmed in melanoma6 and applied to breast,7 thyroid,8 colorectal,9,10 and other cancers.11 A SURGERY 335

336 Hiratsuka et al

Surgery March 2001

Fig 1. SN detection technique. Indocyanine green was injected around the primary gastric cancer (A). The dye mostly moved toward the lesser curvature through some fine vessels, and, finally, stained 1 SN (B).

variety of tracers (radiolabeled particles and vital dyes) are commercially available12 and, as a result, there has been an increasing number of reports13 to support the idea that SN biopsy could provide us with a valuable tool to more efficiently determine the nodal status, appropriate extent of lymphadenectomy, and indication of adjuvant treatment. Turner14 indicated that the probability of breast cancer cells in non-SNs was less than 0.1% if the sentinel node is tumor-free by both hematoxylin and eosin and immunohistochemical staining of multiple sections. However, no report has been available on the SN biopsy for gastric cancers. The current study examined the feasibility of SN biopsy and its accuracy in predicting the status of non-SN lymph nodes in gastric cancer operation for T1 or T2 disease.

PATIENTS AND METHODS Operation. From October 1998 to September 1999, there were 74 consecutive patients with T1 or T2 gastric cancers who were enrolled in this study and treated at the Osaka Medical Center for Cancer and Cardiovascular Diseases. Informed consent was obtained from every patient preoperatively. At the time of the laparotomy, the stage of the disease (T1 or T2) was macroscopically confirmed by the absence of cancer invasion to the serosal layer of the stomach.15 There were 44 T1 cancers (T1m = 30, T1sm = 14) and 30 T2 cancers. A fine needle (26-gauge) was inserted into the subserosal layer at 4 different sites around the primary tumor, and dye, indocyanine green (ICG) (Diagnogreen; Dai-Ichi Pharm Co Ltd, Tokyo, Japan) was gently injected. The total amount of injected dye

Hiratsuka et al 337

Surgery Volume 129, Number 3

Fig 2. Micrometastases in SN. Metastases are observed in SN (arrow) (section, × 12.5).

Table. Sentinel and non-sentinel lymph node status in patients with T1 and T2 gastric cancer Metastases in SNs/non-SNs Depth of cancer invasion T1 T2 Total

Patients (No.) 44 30 74

Patients with SNs (No.) 44 (100%) 29* (97%) 73 (99%)

Neg/Neg

Pos/Neg

Pos/Pos

Neg/Pos

42 (95%) 21 (70%) 63 (85%)

2 (5%) 1 (3%) 3 (4%)

0 (0%) 6 (20%) 6 (8%)

0 (0%) 1 (3%) 1 (1%)

*SN could not be identified in 1 patient with T2 cancer. This patient had metastases in the unstained nodes (non-SNs).

was 5 mL (25 mg) for each patient (Fig 1). All lymph nodes (sentinel nodes [SN]) that were stained green within 5 minutes after the dye injection were removed before the distal gastrectomy (n = 55), proximal gastrectomy (n = 3), or total gastrectomy (n = 16). In parallel with the gastrectomy, all patients received an extended lymph node dissection (D2 procedure according to the Japanese classification of gastric carcinoma16) where not only perigastric nodes but also other peripheral nodes along the hepato-duodenal ligament, the common hepatic artery, left gastric artery, and, in some cases, the splenic artery were removed. Macroscopic and microscopic analysis. The resected stomach was opened by a longitudinal incision along the greater curvature for macroscopic inspection, fixed in 20% formalin, and extended on a flat board. The stained lymph nodes (SN) obtained during the operation and the unstained lymph nodes (non-SN) obtained from the resected stomach were also fixed in 20% formalin solution. The SNs were sliced into 2-mm sections. Non-SNs and the resected stomach were sliced into 5-mm sec-

tions. They were embedded in paraffin, and 5-µm sections were cut to be stained with hematoxylin and eosin for histologic examination. Location of lymph nodes. SNs and non-SNs were classified into 2 groups (ie, the perigastric lymph node group and the peripheral lymph node group). The perigastric group included the nodes along the greater and lesser curvature (No. 1-6 according to the Japanese classification of gastric carcinoma16), and the peripheral group included the nodes located at the periphery of the stomach (No. 7-20 according to the Japanese classification of gastric carcinoma16). Statistical methods. The 95% CIs, diagnostic accuracy, sensitivity, specificity, and false-negative rate were calculated by using the definitions given by Veronesi.17 RESULTS After ICG was injected around the gastric tumor, no patients developed allergic reaction or hypotension. The injected dye was observed entering 2 or 3 fine vessels and gradually moving toward the

338 Hiratsuka et al

Surgery March 2001

Fig 3. Location of SNs in patients with lymph node metastases in SNs or non-SNs, or both. No. 1 to 10: Patients with a successful identification of SNs and with lymph node metastases in SNs or non-SNs, or both. No. 11: SNs could not be identified in this patient who had metastases in the unstained nodes (non-SNs). SN, Sentinel node; non-SN, non-sentinel node; open circle, SN without metastases; closed circle, SN with metastases; x, non-SNs with metastases; ND, not detected. Fractions represent the number of metastatic lymph nodes/total number of lymph nodes examined.

greater or lesser curvatures, or both. Of 74 patients, the stained nodes were detectable in 73 (2.6 ± 1.7 nodes per patient; range, 1-9) and undetectable in only 1 patient with stage T2 disease (Table). The stained sentinel nodes mostly belonged to the perigastric lymph node group (2.3 ± 1.7 nodes; range, 0-8)(Fig 2) and rarely to the peripheral lymph node group (0.3 ± 0.7 nodes; range, 0-4). After gastrectomy, the resected stom-

ach was opened and inspected from both the serosal and mucosal aspects to confirm that the dye was injected into the subserosal and submucosal layers close to the gastric tumor in all 74 patients. From the resected specimens, 39 ± 18 (range, 13-114) unstained lymph nodes (non-SNs; 26 ± 12 perigastric and 13 ± 10 peripheral nodes) were collected. Of 74 patients with T1 or T2 cancers, SNs could be identified in 73 patients (success rate, 99%

Surgery Volume 129, Number 3

[95% CI: 96%-101%]). Comparison between the SN and non-SN status is shown in the Table. Of 73 patients with a successful identification of SNs, 10 were found to have lymph node metastases in SNs or non-SNs, or both; 6 in both SNs and non-SNs; 3 in SNs alone; and 1 in non-SNs alone. The sensitivity of the SN status in diagnosis of the lymph node status (SN plus non-SN) of the patient was 90% (9/10)(95% CI:71%-100%). Specificity and diagnostic accuracy were calculated to be 100% (63/63) and 99% (72/73)(95% CI:96%-100%), respectively. The false-negative rate was 10% (1/10)(95% CI:0%-28%). In the patients with T1 cancers, the SN status could diagnose the lymph node status of the patient with 100% accuracy (sensitivity, 100% [2/2] and specificity, 100% [42/42]), and there was no false-negative case (0/2). In the patients with T2 cancers, sensitivity, specificity, and diagnostic accuracy were 88% (7/8)(95% CI:65%100%), 100% (21/21), and 97% (28/29)(95% CI:90%-100%), respectively. There was 1 false-negative case (false negativity, 13% [1/8] [95% CI:0%36%]). Fig 3 illustrates the location of SNs and metastatic non-SNs in 10 patients with lymph node metastases in SNs or non-SNs, or both. Of the 35 SNs obtained from 10 patients (No. 1-10 in Fig 3), 34 (97%) were located in the perigastric area and 1 (3%) was in the peripheral area (No. 10 in Fig 3). The SNs were likely to be located along the curvature close to the primary tumor. The number of positive SNs per patient ranged from 1 to 3. DISCUSSION The SN is defined as the first node in the regional lymphatic basin that drains the primary tumor, and either radionucleotide tracer or vital dye has been used for detecting it. In gastric cancers, it was shown that the perigastric nodes close to the primary lesion were the most common sites of lymphatic metastasis.4,18 Thus, we postulated that it would be difficult to clearly differentiate the SNs by the radionucleotide method because of the interference by the gamma rays emitted from the primary tumor located very close to the SNs. Therefore, we used the dye method instead of the radionucleotide method. Although ICG was initially developed as the test drug for the liver and circulatory function,19,20 it has been used for color lymphography and, more recently, for SN detection.21,22 When this dye was injected around the tumor, it was soon bound to albumin and carried more specifically through the lymphatic vessels than indigo carmine or Evans blue.23 Although

Hiratsuka et al 339

there has been no previous report comparing ICG with other dyes (isosulfan blue,5 carbon dye,24 and others) for the SN staining, our intraoperative observation that ICG entered some fine vessels and moved more slowly than the blood flow suggests that the injected ICG was probably carried to the SNs by means of the lymphatic stream. In the current study, the success rate of SN detection was as high as 99%, and there was only 1 false-negative (1 patient had metastases in non-SNs but not in SNs). These results seem to be comparable with those reported in other cancers, where the SN detection rates ranged from 66% to 98% and the false-negative rates ranged from 0% to 12%.7,13,25 Our study, however, might be vulnerable to the criticism that only relatively early stage (T1 and T2) gastric cancers were included whereas other reports included cancers of various stages. It is well established that the lymphatic flow would be easily altered when some of the main lymphatic routes were involved and occluded.26 The fact that SNs failed to be detected in 1 patient (No. 11 in Fig 3) with 5 involved non-SNs might be explained by the lymphatic vessel occlusion caused by the massive lymph node metastases. Thus, it is possible that the success rate of SN detection might decrease and a false-negative rate might increase in more advanced gastric cancers (T3 and T4) where patients are more likely to have lymph node metastases. Furthermore, non-SNs were examined less carefully than SNs (2-mm sections) in our study because almost all studies on SN in breast cancers had been done in this way. A more careful examination of the non-SNs might result in a higher false-negative rate. Another limitation of the current study is a relatively low statistical power attributable to a small number of patients with lymph node metastases, resulting in a wide 95% CI of the false-negative rates. These important issues need to be investigated in future studies, including a larger number of patients with various stages and using a more careful method (serial sectioning, immunohistochemistry, genetic diagnosis, etc) for the detection of lymph node metastases in SNs and non-SNs. CONCLUSION Our preliminary study has shown that SN biopsy using ICG is feasible in patients with T1 and T2 gastric cancer, and it can predict the lymph node status of the patient with a high accuracy. Because we experienced one false-negative case with T2 gastric cancer, it is prudent to state that SN biopsy is applicable to T1 gastric cancer patients with

340 Hiratsuka et al

safety, but not in T2 gastric cancer patients. Recent advances in endoscopic techniques, including endoscopic ultrasonography, have made the detection of early gastric cancers easier27 and have also enabled the precise estimation of lymph node status, but it still remains difficult to predict the lymph node status with a high accuracy only by such techniques. Wedge resection or segmental gastrectomy is possible by laparoscopic operation, and curative resection of the superficial cancer within the mucosal layer is also possible by endoscopic surgery. Because our method of SN detection using ICG is possible under laparoscopic operation, this technique has a potential to be combined with such less invasive operations. This combination is expected to decrease the incidence of complications, shorten the period of hospital stay, and improve the postoperative quality of life without compensating the curability. Further study is needed to develop a method that can diagnose the SN status with accuracy during the operation. For this purpose, studies on the intraoperative detection of metastases in SN by frozen section or imprint cytology, or both, are ongoing in our institute.

REFERENCES 1. Perri F, Iuliano R, Valente G, Angelillo IF, Arrigoni A, Campra D, et al. Minute and small early gastric cancers in a Western population: a clinicopathologic study. Gastrointest Endosc 1995;41:475-80. 2. Siewert JR, Kestlmeier R, Busch R, Böttcher K, Roder JD, Müller J, et al. Benefits of D2 lymph node dissection for patients with gastric cancer and pN0 and pN1 lymph node metastases. Br J Surg 1996;83:1144-7. 3. Furukawa H, Hiratsuka M, Imaoka S, Ishikawa O, Kabuto T, Sasaki Y, et al. Phase II study of limited surgery for early gastric cancer: segmental gastric resection. Ann Surg Oncol 1999;6:166-70. 4. Sasako M, McCulloch P, Kinoshita T, Maruyama K. New method to evaluate the therapeutic value of lymph node dissection for gastric cancer. Br J Surg 1995;82:346-51. 5. Morton DL, Wen DR, Wong JH, Economou JS, Cagle LA, Storm FK, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992;127:392-9. 6. Lingam MK, Mackie RM, McKay AJ. Intraoperative identification of sentinel lymph node in patients with malignant melanoma. Br J Cancer 1997;75:1505-8. 7. Giuliano AE, Kirgan DM, Guenther JM, Morton DL. Lymphatic mapping and sentinel lymphadenectomy for breast cancer. Ann Surg 1994;220:391-401. 8. Bilchik AJ, Giuliano A, Essner R, Bostick P, Kelemen P, Foshag LJ, et al. Universal application of intraoperative lymphatic mapping and sentinel lymphadenectomy in solid neoplasms. Cancer J Sci Am 1998;4:351-8. 9. Joosten JJA, Strobbe LJA, Wauters CAP, Pruszczynski M, Wobbes TH, Ruers TJM. Intraoperative lymphatic mapping

Surgery March 2001

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

and the sentinel node concept in colorectal carcinoma. Br J Surg 1999;86:482-6. Chin PL, Medeiros J, Schwarz RE. Use of the sentinel lymph node to determine metastases of gastrointestinal malignancies: a word of caution. J Surg Oncol 1999;71:239-42. Hill ADK, Brady MS, Coit DG. Intraoperative lymphatic mapping and sentinel lymph node biopsy for Merkel cell carcinoma. Br J Surg 1999;86:518-21. Liberman L, Cody HS III, Hill ADK, Rosen PP, Yeh SDJ, Akhust T, et al. Sentinel lymph node biopsy after percutaneous diagnosis of nonpalpable breast cancer. Radiology 1999;211:835-44. Veronesi U, Paganelli G, Galimberti V, Viale G, Zurrida S, Bedoni M, et al. Sentinel-node biopsy to avoid axillary dissection in breast cancer with clinically negative lymphnodes. Lancet 1997;349:1864-7. Turner RR, Ollila DW, Krasne DL, Giuliano AE. Histopathologic validation of the sentinel lymph node hypothesis for breast carcinoma. Ann Surg 1997;226:271-8. Sobin LH, Wittekind CH. TNM classification of malignant tumours. 5th ed. New York: Wiley-Liss, International Union Against Cancer (UICC); 1997. p. 123-30. Japanese Gastric Cancer Association. Japanese classification of gastric carcinoma. 2nd English edition. Gastric Cancer 1998;1:10-24. Veronesi U, Paganelli G, Viale G, Galimberti V, Luini A, Zurrida S, et al. Sentinel lymph node biopsy and axillary dissection in breast cancer: results in a large series. J Natl Cancer Inst 1999;91:368-73. Maruyama K, Gunven P, Okabayashi K, Sasako M, Kinoshita T. Lymph node metastases of gastric cancer—general pattern in 1931 patients. Ann Surg 1989;210:596-602. Caesar J, Shaldon S, Chiandussi L, Guevara L, Sherlock S. The use of indocyanine green in the measurement of hepatic blood flow and as a test of hepatic function. Clin Sci 1961;21:43-57. Cherrick GR, Stein SW, Leevy CM, Davidson CS. Indocyanine green: observations on its physical properties, plasma decay and hepatic extraction. J Clin Invest 1960;39:592600. Motomura K, Inaji H, Komoike Y, Kasugai T, Nagumo S, Noguchi S, et al. Sentinel node biopsy in breast cancer patients with clinically negative lymph-nodes. Breast Cancer 1999;6:259-62. Motomura K, Inaji H, Komoike Y, Kasugai T, Noguchi S, Koyama H. Sentinel node biopsy guided by indocyanine green dye in breast cancer patients. Jpn J Clin Oncol 1999;29:604-7. Takayama S, Furuhama K, Ohura K, Onodera T, Akimoto T. Experimental studies on the usefullness of indocyanine green (ICG) as a lymphatic vital dye [Japanese]. Oyo Yakuri/Pharmacometrics 1980;19:603-14. Lucci A, Turner RR, Morton DL. Carbon dye as an adjunct to isosulfan blue dye for sentinel lymph node dissection. Surgery 1999;126:48-53. Giuliano AE, Jones RC, Brennan M, Statman R. Sentinel lymphadenectomy in breast cancer. J Clin Oncol 1997;15: 2345-50. Takahashi S, Takahashi T, Sawai K, Hagiwara A, Tokuda H, Kato G, et al. Studies on para-aortic metastatic lymph nodes in gastric cancer after endoscopic injection of activated carbon particles [Japanese]. Nippon Geka Gakkai Zasshi 1987;88:35-40. Pollack BJ, Chak A, Sivak MV Jr. Endoscopic ultrasonography. Semin Oncol 1996;23:336-46.