- Email: [email protected]
Gamma probe-guided sentinel node biopsy—optimal timing for injection
European Journal of Surgk'al Oncolog.t, 1998; 24:515-519
Gamma probe-guided sentinel node biopsy--optimal timing for injection S. Schneebaum, J. Stadler, M. Cohen, D. Yaniv, J. Baron* and Y. Skornick Department of Surgery " A " and *Department of Nuclear Medicine, Ichilot, Hospital, Tel-At,iv Sourasky Medical Centel; Tel Avi~; Israel
Aims. We initiated a Phase I feasibility study using a gamma-detecting probe (GDP) and radiolabelled colloid to localize the sentinel lymph node (SLN) in breast cancer. The aim of the study was to establish the ideal timing for injection and examine any possible exclusion criteria for this method. Methods. Thirty breast cancer patients diagnosed by fine needle aspiration (FNA) were included in this study. All were injected with 60 MBq rhenium colloid labelled with °'~mTc(Tck-I 7). Scintigraphy was done 20 min, 2, 6 and 25 hours post-injection. Patients were then taken to surgery where they were injected with patent blue dye. During surgery, the SLN was located with a G D P (Neoprobe® Model 1000). In 28 patients, the SLN was identified by scintigraphy 2 hours after injection, identical to the images seen after 24 hours. Results. In all 28 patients, the SLN'was found by the GDP during surgery. In 26 patients the SLN was dyed blue. The two patients with no SLN localization had received prior radiation. Pathology disclosed SLNs with metastases in seven patients. Two patients had a negative SLN but had an axillary lymph node replaced by tumour. Conclusions. Two to 24 hours prior to surgery is suitable timing for injection. Previous radiotherapy predicts failure for this procedure. Further studies are needed to find the exact false-negative rate of this method for breast cancer.
Key words: sentinel node; axillary lymph node dissection; gamma-guided probe.
Introduction The incidence of breast cancer in Israel is on the rise: from 45/100,000 in 1960-1966 to 72/100,000 in 1967-1969) Lymph-node (LN) involvement is a major prognostic factor and the finding of an involved lymph node indicates the need for axillary dissection. Morton 2 reported a new procedure using vital dyes that enables intraoperative identification of the sentinel lymph node (SLN), the LN nearest the site of the primary melanoma on the direct drainage pathway, the most likely site of early metastasis. They identified the SLN(s) in 194 of 237 lymphatic basins and detected metastases in 40 specimens (21%). Metastases were present in 47 of 259 (18%) SLNs, while non-sentinel nodes were tile sole site of metastasis in only two of 3079 nodes from 194 lymphadenectomy specimens with an identifiable SLN; a false-negative rate of <1%. This technique requires meticulous dissection and there seems to be a long learning curve. In Morton's report, surgeon A's success in identifying the SLN (36 surgeries) was about 60%, compared with almost 100% for surgeon C (117 surgeries). A pilot trial examining this technique for breast cancer 3 Correspondence to: Dr Schlomo Schneebaum, Radioimmunoguided Surgery Unit, Department of Surgery "A". TeI-Aviv Sourasky Medical Center, 6 Weizmann Street, TeI-Aviv, Israel. Fax: 972-3-6974820. 0748-7983/I998•060515 + 05 $12.00•0
identified the SLN in 18 out of 22 patients. Seven SLNs were positive, and in three of them the SLN was the only LN with metastatic cancer. In a larger study of 174 breast cancer patients reported by Giuliano et al.,4metastases were found in 62 of the 114 SLNs identified. Pathology was consistent with that of total axillary dissection in 95.6% of the cases. Again, the success rate of SLN identification was dependent on the number of cases, detecting 58.6% of the nodes in the first 96 patients, 72.4% in the remaining patients and 78% in the last 50 patients. In order to increase the success rate, the intraoperative use of a gamma-detecting probe (GDP) was tried, s The main lymphatic drainage basin was located with lymphoscintigraphy using a nanocolloid labelled with technetium. Identification of the SLN was accomplished by two methods: (1) blue dye and (2) use of G D P after a second colloid injection. Nieweg et aL6 suggested the use of only one colloid injection 24 hours before surgery, for both lymphoscintigraphy and intraoperative LN detection. Blue dye was administered prior to commencement of surgery. In four patients, seven SLNs were identified by lymphoscintigraphy. Six of these nodes were found intraoperatively by the blue stain; all seven were identified with the GDP. There are conflicting data on whether the colloid remains after 24 hours (necessitating a second injection). There is also a major discrepancy between previous scintigraphy data 7 maintaining that the lymphatic pattern of the breast © 1998W.B.SaundersCompany Limited
S. Sclmeebaum et al.
516
Table !. Demographics, pathology and probe count ratio Patient
Age (years)
Operation
Sentinel nodes n
H&E pos.
Axillary nodes ' n
SIN background ratio
H&E pos.
I
59
L
I
I
9
0
I
74
M
I
0
12
0
71 62 73 70 60 31 49 66 64 75 70 33 36 40 65 57 40 75 45 68 50 74 66 70 25 44 74 69
L L M M M M L M M L M M L L L L L M L L L L L L L L M L
I 1 4 1 I -3 -I 2 I I 2 4 I I I 3 I I 2 2 2 4 1 I 2 1
0 0 3 1 0 -0 -0 0 0 I 0 0 0 0 0 0 0 1 0 0 0 3 0 0 0 I
7 17 15 10 15 10 3 8 5 14 8 8 9 22 14 15 12 13 9 11 10 13 13 10 8 14 22 14
0 0 I I 4 0 0 3 0 0 0 1 0 0 0 0 0 0 0 10 I 0 0 I 0 0 0 7
3 4 5 6 7 8* 9 10"*. I1 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Histology
I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. Lob. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. Paget's Dis Muc. Ca. Muc. Ca. Lob. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. Muc. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca.
9.96: 8.01: 7.04: 4.23: 5.75: 7.14: 8.38: 65.1:1 5.0: 5.7: 18.1: 30.9: 13.07: 27.8: 3.67: 5.08: 10.75: 51.7: 3.48: 7.6: 40.5: 8.7: 4.25: 23.0: 3.07: 20.8: 6.5:
Abbreviations: S/N background ratio: a ratio betweem probe counts of SN compared to axillary bed of resection probe counts; L = lumpectomy +axillary dissection; I.D. Ca. = invasive ductal carcinoma; M = modified radical mastectomy; Lob. Ca.=lobular ca; Muc. Ca =muci,ltSus carcinoma. Previous radiotherapy: *non-Hodgkin's disease; **neo-adjuvant.
is pathological only if there is metastatic disease in the axillary LNs, and the SLN data which suggest that the node may be identified in all patients, particularly those with metastasis. We designed this study to: (!) examine the feasibility of identifying the SLN in breast cancer patients, and (2) to establish the optimal time for injection prior to surgery.
Materials and methods
Thirty patients with palpable tumour diagnosed as having breast cancer by fine-needle aspiration (FNA) were included in a Phase IIII feasibility study. All patients signed an informed consent approved by our Institutional Review Board. All patients were injected with 60 MBq Tck-17 (C1S Bio International, G i f Sur Yvette Cedex, France) 99mTc (Re) rhenium colloid (rhenium sulfide 0.15 mg, gelatine 9.6 mg, ascorbic acid 7.0 mg, water for injection 1.0 mg). All patients underwent a series of scintigraphy, which was done at 20 min, 2, 6 and 24 hours post-injection. Should the findings no longer be traceable at 6 hours post-injection (in at least four patients), the protocol .would be adjusted accordingly and the injection given 4 hours before the operation. In the
event of loss of image 24 hours post-injection (in at least foul" patients) the protocol would be altered and the injection given on the morning of the operation. After the 24-hour scintigraphy, the patient went to surgery where an injection of 2 ml patient blue V (Guerbet, France) was given 10 min prior to the start of operation. A hand-held gamma-detecting probe, N e o p r o b e ® 1000 (Dublin, Ohio), was used before and during surgery to locate the SLN. In several patients counts were taken 20 rain, 2, and 24 hours after injection (Table I). The probe has a cadmium telluride crystal and can be used in a continuous survey mode, or for taking counts per second (cps). A specially designed collimator is attached to the probe for use with technetium (99mTc). A special small probe (experimental 510K) was also used in some of the patients. Counts were taken in triplicate and were noted on worksheets specially constructed for this study. In surgery, the probe was used to locate the SLN (Fig. 1). Counts were taken in vivo, then e x vivo to verify that the right node had been removed and then the bed of resection was surveyed to ascertain that there was no second SLN. The S L N was counted as well as the axillary bed of resection giving the SLN/axillary ratio. The SLN was sent to frozen section. SLN biopsy was followed by axillary dissection in all patients. The axillary dissection specimen was examined by the probe to confirm
G a m m a p r o b e , g u i d e d sentinel node biopsy
517
posttive for:-cancerafter~rogen S~e~!~0n.~d!~ne':W~ p0sttlve after paral~fin fixatiOfi,:Iti:~il.l:il :sL~ffwere:.prs[~i~ei~:sev;en l~atients. In One patient the SLN was the.ofily: L~n~olYed (Table 1). One pa~tient had aaegafive sLN.,on hematoxylin and eosin, but f o u r ofthe axiii~ FNs Oontaine_d ~ance~ This was the only patient w.[th SUspj~iO~: !4NS off physical examination:. Another patient had. a.,negative SLNibiit one of her axillary lymph nodes:had metastasiS.
Discussion
Fig. l. The gloved probe (P), the SLN indicated by surgical instrument (lower arrow), and surgeon's finger (F) compressing fatty tissue to show lymphatic vessel (upper arrow).
that no additional SLN was present and then the axillary LN bed of resection was surveyed with the probe.
Results
Thirty patients were enrolled, of mean age 51 years (range 33-75). Scintigraphy identified the SLN in 28 of the 30 patients• The SLN was visualized after the 2-hour scan and remained identifiable at the 24-hour scan (Fig. 2). In some patients the 20-min scintigraphy did not identify the SLN. Lateral and oblique anterior views helped in identifying the SLN and, in some, moving the breast medially was useful. Two patients in whom no SLN was identified had previously undergone radiation treatment (Table 1): one for a nonHodgkin's lymphoma diagnosed 19 years earlier (patient 8) and one as part ofa neoadjuvant treatment protocol (patient 10). The probe detected the SLN in all 28 patients in whom scintigraphy had identified the-SLN. However, only 26 were eoloured blue. The mean SLN/axillary ratio was 14.5. (range 3!07-65.1) with a median value of 8;01. The small probe'showed higher ratios than. the larger probe; 27.5 vs. 4.1. There was no problem in identifying the SN despite the lateral location of some of the tumours. All SLNs were sent for frozen section. Ten SLNs were
Axillary LN involvement is a .crifii~al .factOr ii~. planning treatment for breast cancer patientS. As detection 0faxiqlary involvemen't b y physical examination has a tiigh fdlse~ positiwe rate :of 25-30% and a .high false-negativerate Of 28-30%,8 the SLN concept for breast canceris a p p e ~ n g , In: order to test this hypothesis, ~ e te¢lanique shoald be safe, simple and as surgeon-independent as possible. The accuracy Of the method reported by Giuliano et al: 4 i n a •:groUp of 174 patients (95%)showedan ,overall, detection rate of 65%, which :increased to 78% :in the last 50 patients. With further experience (400 9atients),. this detecti0n rate increased to over 90%, 9 In 0rdet to decrease ~ i s prolonged learning cur~e, the use of a radiocolloid was suggested. Van der Veen et- al) 9 were the first to report their experience.ofa Sulphur colloid injection for melanoma. They used two injections, one for lymphoscintigraphy and a later one for intraoperative detection with a GDP, and successfully detected the SLN in all their 11 patients• Nieweg et al. ~ reported the use of one injection and in four patients seven SLNs were identified by scintigraphy; six of these were' found intraoperatively by blue stain; all seven were identified with the GDP. In a recent presentation at ECCO (European Cancer Conference), Kroon et al. ~° reported their experience in 55 patients• Detection of SLNs with blue dye was successful in'76%, and 100% with the GDP. Albertini et al. reported their experience with both methods resulting in successful identification of 57 out of 62 patients (95%)." In a larger sample of 178 patients, the same group reported a similar success rate in 164 patients (92%). t2 Veronesi et al. reported a success rate o f 98% in 160 patientsusing both methodsJ ~ Using a GDP for breast. SLN identification has. several limitations as the injection site can be :very near to the axillary LNs as in laterally located lesions, i.e. collimation becomes of major importance. Another concern, is the possibility of a radioactive injection causing contamination in the operation room. The fact that the blue ;dye is absorbed quic~y after injection signifies that the size of the colloid particle has to be bigger than the blue dye particle. The particle, size.is known to be.of major importance~ A particle size~of 3-~30nm is considered favourable for lymphoscintigraptty. 99roTe, antimony trisulphide.colloid (~Tc-ATC)-was pre~iously ~the most frequently used radiopharmaceutical for lymphoscintigraphy in..the diagnosis: of the :oedematous extremity.~° Ho,ivever,. this drug .is: not :approved' for .usage in the USA~ ~ u n g et aL ,used teelane.ti~.tm ~e~siflphug: colloid filtered to particle ~size ~<50am, e~actly 39~0~.-3,:0, This showed good results.,in ~.scintigrap~y,, as~ thw ~arge
518
S. S c h n e e b a u m
et al.
Fig. 2. Scintigraphy 2 and 24 hours after injection, showing same SLNs.
particles trapped in the injection site were filtered. Radioactive sulphur colloid produced by the hydrogen sulphide method (i.e. 99mTc sulphur minicolioid) has small particles (
involved. Eighteen of the total 285 nodes examined in these 14 cases were SLNs. Tumours were found in 13 out of 18 SLNs (72.2%) and in only one of .267 non-SLNs (0.37"/o) (P<0.0001). A similar finding was reported by Reintgen et al.~7 in melanoma patients, with a probability of the SLN being a random event of 3.7 x 10 -8. Sensitivities of LN mapping and random biopsy were proven by Guiliano et al. 4 to be 61.9 and 17.5%, respectively (P<0.0001). We identified the SLN with the G D P in 100% of the patients with no prior radiation treatment and in 95% with the blue dye. However, our surgical technique differs as we first locate the SLN with the GDP, then verify whether it is dyed blue and then dissect to see whether there is a more proximal node. Our experience has shown that it is always the same,-node and a more proximal node was never located. The absence of lymphatic pathways after radiation was surprising and the fact that one patient had axillary LN metastases may indicate their presence prior to radiation (neoadjuvant protocol). We confined our LN examination to frozen section and H&E examination, which is the current practice in Israel. Our limited experience suggests that frozen section may not be enough. Whether more extensive testing such as immunohistochemistry or molecular genetic techniques should be used requires further study as these tests are not routinely done at our institution. Assuming the SLN concept has almost no false-negatives and that we have a perfect technique for its detection, the question of whether removal of negative axillary LNs is of value to the patients still
G a m m a probe-guided sentfllel node biopsy
remains unanswered, in view of the possibility of micrometastases not detected by our current studies. Several studies have shown that the findings of occult tumour are biologically significant. The International Ludwig Breast Cancer Study Group t~ reported that serial sectioning of ipsilateral axillary LNs declared disease-free after routine histological examination revealed micrometastases in 83 out of 921 (9%) patients with breast cancer. These patients had a poorer disease-free and overall survival after a median of 5 years follow-up than patients whose LNs remained negative after serial sectioning. Polymerase chain reaction (PCR) methods have been reported to increase the sensitivity of occult tumour cell detection in LNs. Noguchi et al. ~9 examined 50 axillary LNs from patients with primary breast carcinomas. Nine of these nodes were found to contain metastasis by immunohistochemical teclmiques. In all nine cases, MUC1 m R N A could be amplified by reversetranscriptase PCR. However, in the 41 LNs without detectable metastasis, six contained detectable M U C I mRNA. Schoenfeld et al. 2° reported a similar experience in 57 LNs in which the use of immunohistochemistry with anticytokeratin antibodies identified 18 LNs with cancer, all verified by PCR using K I 9 primers. Iil addition, four of the 39 (10%,) histologically negative nodes gave a K I 9 product detectable with ethidium staining. After 40 cycles of PCR on Southern blotting and hybridization with ~'-Plabelled K I 9 cDNA, a further 10 nodes (28'7.) gave positive signals. A possible benefit of identifying the SLN may be in directing studies to these specific LNs in order to assess their clinical significance. It is interesting that in level II alone Giuliano et al. 8 found a SLN in 23.3% of the dissections. Acknowledging that 38% of SLNs are the only LNs with metastasis suggests that metastatic spread to the axilla is determined by the specific lymphatic drainage of the primary tumour which in turn depends on each patient's lymphatic anatomy. Our experience is similar, but we cannot substantiate it with our smaller number of patients and continue to adopt the currently suggested approach that an axilliary dissection should include levels I and 1I, as the risk of a false-negative with such magnitude of dissection is only I--9%o.8'21 We conclude that 2-24 hours post-injection is suitable timing for injection. The 24-hour period also favourably decreases operation room contamination. Patients with previous radiation therapy are not suitable for SLN mapping. Based on our data and review of the current literature, we feel that the SLN in breast cancer is an apppealing concept and continue with our investigations. References 1. Ministry of Health, Department of Epidemiology, State of Israel. Momh O, Epidemiol Record, March 1994.
519
2. Morton DL, Wen D-R, Wong JH, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992; 127: 392-9. 3. Krag D, Weaver D,'Alex J, Fairbank J. Surgical resection and radiolocalization of sentinel lymph node in breast cancer using a gamma probe. SSO Abstract Book, 1994: 32. 4. Guiliano AE, Kirgan DM, Guenther JM, Morton DL. Lymphatic mapping and sentinel lymphadenectomy for breast cancer. Ann Surg 1994; 220: 391-401. 5. Van der Veen H, Hoekstra OS, Paul MA, et al. Gamma probeguided sentinel node biopsy to select patients with melanoma for lymphadenectomy. Br J Surg 1994; 81: 1769-70. 6. Nieweg OE, Baidjnath Panday RKL, Muller S, et al. Lymphoscintigraphy, lymphatic mapping and sentinel node biopsy in melanoma patients using a single dose of ~'q'ccolloid. European Society of Surgical Oncology, Heidelberg, June 1994. 7. Rosenblum 1, Wischnitzer Th, Stadler JA, Bar-On J. Lymphography of the breast as an accessory in the early diagnosis and detection of carcinoma of the breast: lymphography as a prognostic tool. Worm J Surg 1982; 6: 126-9. 8. Recht A, Houlihan MJ. Axillary lymph nodes and breast cancer. Cancer 1995; 76: 1491-512. 9. Guilliano AE. Lymphatic mapping and sentinel node biopsy in breast cancer. J A M A 1997; 277:791 (letter). 10. Kroon BBR, Kapteijn BAE, Baidjnath P, et al. Results of sentinel node biopsy in cutaneous melanoma. European Cancer Conference 8, Paris. Eur J Cancer 1995; 31A(Suppl 5): 133-4, abstract 637. II. Albertini J J, Lyman OH, Cox C, et al. Lymphatic mal~ping and sentinel node biopsy in the patient with breast cancer. J A M A 1998; 276: 1818-22. 12. Joseph E, Cox C, Clark R, et al. Radioguided surgery for the ultrastaging of breast cancer patients. American Society o f Clinical Oncolog)'. 1997. 13. Veronesi U, Paganelli G, Galimberti V, et al. Sentinel node biopsy to avoid axillary dissection in breast cancer with clinically negative lymph nodes. Lancet 1997-"348: 1864--7. 14. Hung JC, Wiseman GA, Wahner HW, et aL Filtered technetium99-m-sulfur colloid evaluated for lymphoscintjgraphy. J Nucl Med 1995; 36: 1895-901. 15. Dunson GL, Thrall JH, Stevenson JS, Pinsky SM. 99~I'c minicolloid for radionuclide lymphography. Radiology 1973; 109: 387-92. 16..Richards TB, McBiles M, Collins PS. An easy method for diagnosis of lymphedema. Amt Vasc Surg 1990; 4: 255-9. 17. Reinigen D, Cruse CW, Wells K, et al. The orderly progression of melanoma nodal metastases. Ann Surg 1994; 220: 759-67. 18. International (Ludwig) Breast Cancer Study Group. Lancet 1990; 335: 1565-8. 19. Noguchi S, Aihara T, Nakamoni S, et al. The detection of breast carcinoma micrometastases in axillary lymph nodes by means of reverse transcriptase-polymerase chain reaction. Cancer 1994; 74: 1595-600. 20. Schoenfeld A, Lugmani Y, Smith D, et aL Detection of breast cancer micrometastases in axillary lymph nodes by using polymerase chain reaction. Cancer Res 1994; 54: 2966-90. 21. Fisher B, Wolmark N, Bauer M, et al. The accuracy of clinical node staging and of limited axillary dissection as a determinant of histological nodal status in carcinoma of the breast. Sarg Gynecol Obstet 1981; 152: 765-72.
Accepted for publication 23 June 1998
Gamma probe-guided sentinel node biopsy--optimal timing for injection S. Schneebaum, J. Stadler, M. Cohen, D. Yaniv, J. Baron* and Y. Skornick Department of Surgery " A " and *Department of Nuclear Medicine, Ichilot, Hospital, Tel-At,iv Sourasky Medical Centel; Tel Avi~; Israel
Aims. We initiated a Phase I feasibility study using a gamma-detecting probe (GDP) and radiolabelled colloid to localize the sentinel lymph node (SLN) in breast cancer. The aim of the study was to establish the ideal timing for injection and examine any possible exclusion criteria for this method. Methods. Thirty breast cancer patients diagnosed by fine needle aspiration (FNA) were included in this study. All were injected with 60 MBq rhenium colloid labelled with °'~mTc(Tck-I 7). Scintigraphy was done 20 min, 2, 6 and 25 hours post-injection. Patients were then taken to surgery where they were injected with patent blue dye. During surgery, the SLN was located with a G D P (Neoprobe® Model 1000). In 28 patients, the SLN was identified by scintigraphy 2 hours after injection, identical to the images seen after 24 hours. Results. In all 28 patients, the SLN'was found by the GDP during surgery. In 26 patients the SLN was dyed blue. The two patients with no SLN localization had received prior radiation. Pathology disclosed SLNs with metastases in seven patients. Two patients had a negative SLN but had an axillary lymph node replaced by tumour. Conclusions. Two to 24 hours prior to surgery is suitable timing for injection. Previous radiotherapy predicts failure for this procedure. Further studies are needed to find the exact false-negative rate of this method for breast cancer.
Key words: sentinel node; axillary lymph node dissection; gamma-guided probe.
Introduction The incidence of breast cancer in Israel is on the rise: from 45/100,000 in 1960-1966 to 72/100,000 in 1967-1969) Lymph-node (LN) involvement is a major prognostic factor and the finding of an involved lymph node indicates the need for axillary dissection. Morton 2 reported a new procedure using vital dyes that enables intraoperative identification of the sentinel lymph node (SLN), the LN nearest the site of the primary melanoma on the direct drainage pathway, the most likely site of early metastasis. They identified the SLN(s) in 194 of 237 lymphatic basins and detected metastases in 40 specimens (21%). Metastases were present in 47 of 259 (18%) SLNs, while non-sentinel nodes were tile sole site of metastasis in only two of 3079 nodes from 194 lymphadenectomy specimens with an identifiable SLN; a false-negative rate of <1%. This technique requires meticulous dissection and there seems to be a long learning curve. In Morton's report, surgeon A's success in identifying the SLN (36 surgeries) was about 60%, compared with almost 100% for surgeon C (117 surgeries). A pilot trial examining this technique for breast cancer 3 Correspondence to: Dr Schlomo Schneebaum, Radioimmunoguided Surgery Unit, Department of Surgery "A". TeI-Aviv Sourasky Medical Center, 6 Weizmann Street, TeI-Aviv, Israel. Fax: 972-3-6974820. 0748-7983/I998•060515 + 05 $12.00•0
identified the SLN in 18 out of 22 patients. Seven SLNs were positive, and in three of them the SLN was the only LN with metastatic cancer. In a larger study of 174 breast cancer patients reported by Giuliano et al.,4metastases were found in 62 of the 114 SLNs identified. Pathology was consistent with that of total axillary dissection in 95.6% of the cases. Again, the success rate of SLN identification was dependent on the number of cases, detecting 58.6% of the nodes in the first 96 patients, 72.4% in the remaining patients and 78% in the last 50 patients. In order to increase the success rate, the intraoperative use of a gamma-detecting probe (GDP) was tried, s The main lymphatic drainage basin was located with lymphoscintigraphy using a nanocolloid labelled with technetium. Identification of the SLN was accomplished by two methods: (1) blue dye and (2) use of G D P after a second colloid injection. Nieweg et aL6 suggested the use of only one colloid injection 24 hours before surgery, for both lymphoscintigraphy and intraoperative LN detection. Blue dye was administered prior to commencement of surgery. In four patients, seven SLNs were identified by lymphoscintigraphy. Six of these nodes were found intraoperatively by the blue stain; all seven were identified with the GDP. There are conflicting data on whether the colloid remains after 24 hours (necessitating a second injection). There is also a major discrepancy between previous scintigraphy data 7 maintaining that the lymphatic pattern of the breast © 1998W.B.SaundersCompany Limited
S. Sclmeebaum et al.
516
Table !. Demographics, pathology and probe count ratio Patient
Age (years)
Operation
Sentinel nodes n
H&E pos.
Axillary nodes ' n
SIN background ratio
H&E pos.
I
59
L
I
I
9
0
I
74
M
I
0
12
0
71 62 73 70 60 31 49 66 64 75 70 33 36 40 65 57 40 75 45 68 50 74 66 70 25 44 74 69
L L M M M M L M M L M M L L L L L M L L L L L L L L M L
I 1 4 1 I -3 -I 2 I I 2 4 I I I 3 I I 2 2 2 4 1 I 2 1
0 0 3 1 0 -0 -0 0 0 I 0 0 0 0 0 0 0 1 0 0 0 3 0 0 0 I
7 17 15 10 15 10 3 8 5 14 8 8 9 22 14 15 12 13 9 11 10 13 13 10 8 14 22 14
0 0 I I 4 0 0 3 0 0 0 1 0 0 0 0 0 0 0 10 I 0 0 I 0 0 0 7
3 4 5 6 7 8* 9 10"*. I1 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Histology
I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. Lob. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. Paget's Dis Muc. Ca. Muc. Ca. Lob. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. Muc. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca. I.D. Ca.
9.96: 8.01: 7.04: 4.23: 5.75: 7.14: 8.38: 65.1:1 5.0: 5.7: 18.1: 30.9: 13.07: 27.8: 3.67: 5.08: 10.75: 51.7: 3.48: 7.6: 40.5: 8.7: 4.25: 23.0: 3.07: 20.8: 6.5:
Abbreviations: S/N background ratio: a ratio betweem probe counts of SN compared to axillary bed of resection probe counts; L = lumpectomy +axillary dissection; I.D. Ca. = invasive ductal carcinoma; M = modified radical mastectomy; Lob. Ca.=lobular ca; Muc. Ca =muci,ltSus carcinoma. Previous radiotherapy: *non-Hodgkin's disease; **neo-adjuvant.
is pathological only if there is metastatic disease in the axillary LNs, and the SLN data which suggest that the node may be identified in all patients, particularly those with metastasis. We designed this study to: (!) examine the feasibility of identifying the SLN in breast cancer patients, and (2) to establish the optimal time for injection prior to surgery.
Materials and methods
Thirty patients with palpable tumour diagnosed as having breast cancer by fine-needle aspiration (FNA) were included in a Phase IIII feasibility study. All patients signed an informed consent approved by our Institutional Review Board. All patients were injected with 60 MBq Tck-17 (C1S Bio International, G i f Sur Yvette Cedex, France) 99mTc (Re) rhenium colloid (rhenium sulfide 0.15 mg, gelatine 9.6 mg, ascorbic acid 7.0 mg, water for injection 1.0 mg). All patients underwent a series of scintigraphy, which was done at 20 min, 2, 6 and 24 hours post-injection. Should the findings no longer be traceable at 6 hours post-injection (in at least four patients), the protocol .would be adjusted accordingly and the injection given 4 hours before the operation. In the
event of loss of image 24 hours post-injection (in at least foul" patients) the protocol would be altered and the injection given on the morning of the operation. After the 24-hour scintigraphy, the patient went to surgery where an injection of 2 ml patient blue V (Guerbet, France) was given 10 min prior to the start of operation. A hand-held gamma-detecting probe, N e o p r o b e ® 1000 (Dublin, Ohio), was used before and during surgery to locate the SLN. In several patients counts were taken 20 rain, 2, and 24 hours after injection (Table I). The probe has a cadmium telluride crystal and can be used in a continuous survey mode, or for taking counts per second (cps). A specially designed collimator is attached to the probe for use with technetium (99mTc). A special small probe (experimental 510K) was also used in some of the patients. Counts were taken in triplicate and were noted on worksheets specially constructed for this study. In surgery, the probe was used to locate the SLN (Fig. 1). Counts were taken in vivo, then e x vivo to verify that the right node had been removed and then the bed of resection was surveyed to ascertain that there was no second SLN. The S L N was counted as well as the axillary bed of resection giving the SLN/axillary ratio. The SLN was sent to frozen section. SLN biopsy was followed by axillary dissection in all patients. The axillary dissection specimen was examined by the probe to confirm
G a m m a p r o b e , g u i d e d sentinel node biopsy
517
posttive for:-cancerafter~rogen S~e~!~0n.~d!~ne':W~ p0sttlve after paral~fin fixatiOfi,:Iti:~il.l:il :sL~ffwere:.prs[~i~ei~:sev;en l~atients. In One patient the SLN was the.ofily: L~n~olYed (Table 1). One pa~tient had aaegafive sLN.,on hematoxylin and eosin, but f o u r ofthe axiii~ FNs Oontaine_d ~ance~ This was the only patient w.[th SUspj~iO~: !4NS off physical examination:. Another patient had. a.,negative SLNibiit one of her axillary lymph nodes:had metastasiS.
Discussion
Fig. l. The gloved probe (P), the SLN indicated by surgical instrument (lower arrow), and surgeon's finger (F) compressing fatty tissue to show lymphatic vessel (upper arrow).
that no additional SLN was present and then the axillary LN bed of resection was surveyed with the probe.
Results
Thirty patients were enrolled, of mean age 51 years (range 33-75). Scintigraphy identified the SLN in 28 of the 30 patients• The SLN was visualized after the 2-hour scan and remained identifiable at the 24-hour scan (Fig. 2). In some patients the 20-min scintigraphy did not identify the SLN. Lateral and oblique anterior views helped in identifying the SLN and, in some, moving the breast medially was useful. Two patients in whom no SLN was identified had previously undergone radiation treatment (Table 1): one for a nonHodgkin's lymphoma diagnosed 19 years earlier (patient 8) and one as part ofa neoadjuvant treatment protocol (patient 10). The probe detected the SLN in all 28 patients in whom scintigraphy had identified the-SLN. However, only 26 were eoloured blue. The mean SLN/axillary ratio was 14.5. (range 3!07-65.1) with a median value of 8;01. The small probe'showed higher ratios than. the larger probe; 27.5 vs. 4.1. There was no problem in identifying the SN despite the lateral location of some of the tumours. All SLNs were sent for frozen section. Ten SLNs were
Axillary LN involvement is a .crifii~al .factOr ii~. planning treatment for breast cancer patientS. As detection 0faxiqlary involvemen't b y physical examination has a tiigh fdlse~ positiwe rate :of 25-30% and a .high false-negativerate Of 28-30%,8 the SLN concept for breast canceris a p p e ~ n g , In: order to test this hypothesis, ~ e te¢lanique shoald be safe, simple and as surgeon-independent as possible. The accuracy Of the method reported by Giuliano et al: 4 i n a •:groUp of 174 patients (95%)showedan ,overall, detection rate of 65%, which :increased to 78% :in the last 50 patients. With further experience (400 9atients),. this detecti0n rate increased to over 90%, 9 In 0rdet to decrease ~ i s prolonged learning cur~e, the use of a radiocolloid was suggested. Van der Veen et- al) 9 were the first to report their experience.ofa Sulphur colloid injection for melanoma. They used two injections, one for lymphoscintigraphy and a later one for intraoperative detection with a GDP, and successfully detected the SLN in all their 11 patients• Nieweg et al. ~ reported the use of one injection and in four patients seven SLNs were identified by scintigraphy; six of these were' found intraoperatively by blue stain; all seven were identified with the GDP. In a recent presentation at ECCO (European Cancer Conference), Kroon et al. ~° reported their experience in 55 patients• Detection of SLNs with blue dye was successful in'76%, and 100% with the GDP. Albertini et al. reported their experience with both methods resulting in successful identification of 57 out of 62 patients (95%)." In a larger sample of 178 patients, the same group reported a similar success rate in 164 patients (92%). t2 Veronesi et al. reported a success rate o f 98% in 160 patientsusing both methodsJ ~ Using a GDP for breast. SLN identification has. several limitations as the injection site can be :very near to the axillary LNs as in laterally located lesions, i.e. collimation becomes of major importance. Another concern, is the possibility of a radioactive injection causing contamination in the operation room. The fact that the blue ;dye is absorbed quic~y after injection signifies that the size of the colloid particle has to be bigger than the blue dye particle. The particle, size.is known to be.of major importance~ A particle size~of 3-~30nm is considered favourable for lymphoscintigraptty. 99roTe, antimony trisulphide.colloid (~Tc-ATC)-was pre~iously ~the most frequently used radiopharmaceutical for lymphoscintigraphy in..the diagnosis: of the :oedematous extremity.~° Ho,ivever,. this drug .is: not :approved' for .usage in the USA~ ~ u n g et aL ,used teelane.ti~.tm ~e~siflphug: colloid filtered to particle ~size ~<50am, e~actly 39~0~.-3,:0, This showed good results.,in ~.scintigrap~y,, as~ thw ~arge
518
S. S c h n e e b a u m
et al.
Fig. 2. Scintigraphy 2 and 24 hours after injection, showing same SLNs.
particles trapped in the injection site were filtered. Radioactive sulphur colloid produced by the hydrogen sulphide method (i.e. 99mTc sulphur minicolioid) has small particles (
involved. Eighteen of the total 285 nodes examined in these 14 cases were SLNs. Tumours were found in 13 out of 18 SLNs (72.2%) and in only one of .267 non-SLNs (0.37"/o) (P<0.0001). A similar finding was reported by Reintgen et al.~7 in melanoma patients, with a probability of the SLN being a random event of 3.7 x 10 -8. Sensitivities of LN mapping and random biopsy were proven by Guiliano et al. 4 to be 61.9 and 17.5%, respectively (P<0.0001). We identified the SLN with the G D P in 100% of the patients with no prior radiation treatment and in 95% with the blue dye. However, our surgical technique differs as we first locate the SLN with the GDP, then verify whether it is dyed blue and then dissect to see whether there is a more proximal node. Our experience has shown that it is always the same,-node and a more proximal node was never located. The absence of lymphatic pathways after radiation was surprising and the fact that one patient had axillary LN metastases may indicate their presence prior to radiation (neoadjuvant protocol). We confined our LN examination to frozen section and H&E examination, which is the current practice in Israel. Our limited experience suggests that frozen section may not be enough. Whether more extensive testing such as immunohistochemistry or molecular genetic techniques should be used requires further study as these tests are not routinely done at our institution. Assuming the SLN concept has almost no false-negatives and that we have a perfect technique for its detection, the question of whether removal of negative axillary LNs is of value to the patients still
G a m m a probe-guided sentfllel node biopsy
remains unanswered, in view of the possibility of micrometastases not detected by our current studies. Several studies have shown that the findings of occult tumour are biologically significant. The International Ludwig Breast Cancer Study Group t~ reported that serial sectioning of ipsilateral axillary LNs declared disease-free after routine histological examination revealed micrometastases in 83 out of 921 (9%) patients with breast cancer. These patients had a poorer disease-free and overall survival after a median of 5 years follow-up than patients whose LNs remained negative after serial sectioning. Polymerase chain reaction (PCR) methods have been reported to increase the sensitivity of occult tumour cell detection in LNs. Noguchi et al. ~9 examined 50 axillary LNs from patients with primary breast carcinomas. Nine of these nodes were found to contain metastasis by immunohistochemical teclmiques. In all nine cases, MUC1 m R N A could be amplified by reversetranscriptase PCR. However, in the 41 LNs without detectable metastasis, six contained detectable M U C I mRNA. Schoenfeld et al. 2° reported a similar experience in 57 LNs in which the use of immunohistochemistry with anticytokeratin antibodies identified 18 LNs with cancer, all verified by PCR using K I 9 primers. Iil addition, four of the 39 (10%,) histologically negative nodes gave a K I 9 product detectable with ethidium staining. After 40 cycles of PCR on Southern blotting and hybridization with ~'-Plabelled K I 9 cDNA, a further 10 nodes (28'7.) gave positive signals. A possible benefit of identifying the SLN may be in directing studies to these specific LNs in order to assess their clinical significance. It is interesting that in level II alone Giuliano et al. 8 found a SLN in 23.3% of the dissections. Acknowledging that 38% of SLNs are the only LNs with metastasis suggests that metastatic spread to the axilla is determined by the specific lymphatic drainage of the primary tumour which in turn depends on each patient's lymphatic anatomy. Our experience is similar, but we cannot substantiate it with our smaller number of patients and continue to adopt the currently suggested approach that an axilliary dissection should include levels I and 1I, as the risk of a false-negative with such magnitude of dissection is only I--9%o.8'21 We conclude that 2-24 hours post-injection is suitable timing for injection. The 24-hour period also favourably decreases operation room contamination. Patients with previous radiation therapy are not suitable for SLN mapping. Based on our data and review of the current literature, we feel that the SLN in breast cancer is an apppealing concept and continue with our investigations. References 1. Ministry of Health, Department of Epidemiology, State of Israel. Momh O, Epidemiol Record, March 1994.
519
2. Morton DL, Wen D-R, Wong JH, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992; 127: 392-9. 3. Krag D, Weaver D,'Alex J, Fairbank J. Surgical resection and radiolocalization of sentinel lymph node in breast cancer using a gamma probe. SSO Abstract Book, 1994: 32. 4. Guiliano AE, Kirgan DM, Guenther JM, Morton DL. Lymphatic mapping and sentinel lymphadenectomy for breast cancer. Ann Surg 1994; 220: 391-401. 5. Van der Veen H, Hoekstra OS, Paul MA, et al. Gamma probeguided sentinel node biopsy to select patients with melanoma for lymphadenectomy. Br J Surg 1994; 81: 1769-70. 6. Nieweg OE, Baidjnath Panday RKL, Muller S, et al. Lymphoscintigraphy, lymphatic mapping and sentinel node biopsy in melanoma patients using a single dose of ~'q'ccolloid. European Society of Surgical Oncology, Heidelberg, June 1994. 7. Rosenblum 1, Wischnitzer Th, Stadler JA, Bar-On J. Lymphography of the breast as an accessory in the early diagnosis and detection of carcinoma of the breast: lymphography as a prognostic tool. Worm J Surg 1982; 6: 126-9. 8. Recht A, Houlihan MJ. Axillary lymph nodes and breast cancer. Cancer 1995; 76: 1491-512. 9. Guilliano AE. Lymphatic mapping and sentinel node biopsy in breast cancer. J A M A 1997; 277:791 (letter). 10. Kroon BBR, Kapteijn BAE, Baidjnath P, et al. Results of sentinel node biopsy in cutaneous melanoma. European Cancer Conference 8, Paris. Eur J Cancer 1995; 31A(Suppl 5): 133-4, abstract 637. II. Albertini J J, Lyman OH, Cox C, et al. Lymphatic mal~ping and sentinel node biopsy in the patient with breast cancer. J A M A 1998; 276: 1818-22. 12. Joseph E, Cox C, Clark R, et al. Radioguided surgery for the ultrastaging of breast cancer patients. American Society o f Clinical Oncolog)'. 1997. 13. Veronesi U, Paganelli G, Galimberti V, et al. Sentinel node biopsy to avoid axillary dissection in breast cancer with clinically negative lymph nodes. Lancet 1997-"348: 1864--7. 14. Hung JC, Wiseman GA, Wahner HW, et aL Filtered technetium99-m-sulfur colloid evaluated for lymphoscintjgraphy. J Nucl Med 1995; 36: 1895-901. 15. Dunson GL, Thrall JH, Stevenson JS, Pinsky SM. 99~I'c minicolloid for radionuclide lymphography. Radiology 1973; 109: 387-92. 16..Richards TB, McBiles M, Collins PS. An easy method for diagnosis of lymphedema. Amt Vasc Surg 1990; 4: 255-9. 17. Reinigen D, Cruse CW, Wells K, et al. The orderly progression of melanoma nodal metastases. Ann Surg 1994; 220: 759-67. 18. International (Ludwig) Breast Cancer Study Group. Lancet 1990; 335: 1565-8. 19. Noguchi S, Aihara T, Nakamoni S, et al. The detection of breast carcinoma micrometastases in axillary lymph nodes by means of reverse transcriptase-polymerase chain reaction. Cancer 1994; 74: 1595-600. 20. Schoenfeld A, Lugmani Y, Smith D, et aL Detection of breast cancer micrometastases in axillary lymph nodes by using polymerase chain reaction. Cancer Res 1994; 54: 2966-90. 21. Fisher B, Wolmark N, Bauer M, et al. The accuracy of clinical node staging and of limited axillary dissection as a determinant of histological nodal status in carcinoma of the breast. Sarg Gynecol Obstet 1981; 152: 765-72.
Accepted for publication 23 June 1998