- Email: [email protected]
Sentinel Node Localization and Biopsy in Breast Cancer
Clinical Oncology
Clinical Oncology (1999)11:111–117 # The Royal College of Radiologists
Annual Undergraduate Prize Royal College of Radiologists Annual Undergraduate Prize in Clinical Oncology: Prize-Winning Entries 1998. It has been a delight once again to work with the authors of the winning submissions for the Royal College of Radiologists Annual Undergraduate Prize in Clinical Oncology. The first and second prize winning entries follow and their excellence speaks for themselves. Congratulations to both authors. W. G. Jones Editor
Sentinel Node Localization and Biopsy in Breast Cancer* G. Howard-Alpe Royal Free Hospital School of Medicine, London, UK Abstract. Currently, it is routine practice to carry out axillary lymph node dissection at the time of surgical removal of a malignant primary breast tumour. As breast cancer is being diagnosed at an earlier stage in a growing percentage of cases, this procedure is proving unnecessary in a proportion of patients. Axillary lymph node dissection carries a risk of side effects and ideally we should identify the patients who actually require the procedure. The concept of the sentinel node is not new and was developed over twenty years ago. The technique of lymphoscintography was initially used to identify the sentinel node/s in patients with malignant melanoma. A proposed alternative management strategy to routine axillary lymph node dissection in patients with breast cancer is sentinel lymph node localization using lymphoscintography and biopsy of the identified sentinel node. The aim being to accurately predict the disease status of the axilla and consequently determine whether axillary lymph node dissection is indicated. The technique is currently undergoing a multi-centre trial in the UK. During my elective at St Luke’s Cancer Centre, Royal Surrey County Hospital, Guildford I was fortunate to observe this procedure being tested, with patients undergoing both lymphoscintography and sentinel node biopsy as well as axillary lymph node dissection. The results seen during my stay were extremely encouraging. Keywords: Axillary lymph node dissection; Biopsy; Breast cancer; Lymphoscintography; Sentinel node
Introduction Despite research into tumour-associated prognostic indicators in breast cancer, including oncogene expression, hormone receptor status, ploidy and Sphase, lymph node disease status remains the most accurate prognostic indicator of overall survival. The current practice of routine axillary lymph node dissection serves four purposes: it assures locoregional control, improves survival, and provides important information for staging and for prognosis. The status of the axilla affects treatment options including whether or not adjuvant chemotherapy is given to *Royal College of Radiologists Annual Undergraduate Prize: First Prize Winner Correspondence and offprint requests to: Dr G. Howard-Alpe, Flat 6, Westminster Bridge House, 7 Westminster Bridge Road, London SE1 7XP, UK.
premenopausal patients. Axillary lymph node dissection is therefore an important procedure in the treatment of breast cancer, and has been shown to be beneficial even with a T1A lesion [1]. The procedure of routine axillary lymph node dissection has side effects that can be disabling and painful, and impair a patient’s quality of life. The main side effect is lymphoedema of the arm. The rate of occurrence is dependent on whether radiotherapy is also administered to the axillary area and on surgical skill at the time of the operation. The incidence of lymphoedema varies directly with the amount of nodal tissue removed [2]; level one dissection or blind biopsy of a few nodes is associated with the lowest incidence of lymphoedema, but this carries a high incidence of false negative axillary staging. Additional side effects include vascular or brachial plexus injury but these are much rarer.
Owing to advances allowing the earlier detection of tumours through both screening programmes and public awareness campaigns, the question of whether axillary lymph node dissection is necessary as a routine procedure is currently being raised. It certainly appears that many women who have early tumours are unnecessarily undergoing a procedure that proves the lymph node status to be negative and causes a proportion of them to experience morbidity. This calls for research into less invasive ways of staging the axilla. A recent study by Singhal et al. in Ontario concluded that women older than 60 years had a low probability of nodal positivity (0%–8.7%), but that there is insufficient information in this group to give a 95% or better prediction of nodal status at the time of surgery. It called for studies into minimally invasive techniques such as sentinel node biopsy to minimize surgical morbidity in these women, who, as a result of early diagnosis, have an excellent long term outlook [3]. The concept of sequential lymphatic dissemination of a primary tumour was initially developed by Cabanas in 1977 [4]. He suggested that squamous cell carcinoma of the penis initially drained to a particular lymph node in the groin, which was always in the same location and termed the ‘sentinel node’. Tumour cells travelling in the lymphatics lodged in this first sentinel node and then, at a later stage, travelled on to other nodes in the direct drainage pathway. He hypothesized that if this sentinel node could be identified, removed surgically and examined, the histological status of the node acted as an indicator of the status of the entire lymphatic field. Therefore, if metastatic disease was found in the sentinel node there was a strong chance that other lymph nodes within that field would contain disease and regional lymph node dissection should be advised. When the sentinel node was disease free it was unlikely that other nodes would be involved and regional lymph node dissection would be inappropriate. For penile cancer this assumption is plausible because it is always in exactly the same part of the body. The issue becomes more complex when it is applied to other tumours whose location within the body is less precise. This theory was revived in 1992 by Morton et al. [5], who developed a technique of lymphatic mapping by lymphoscintography, allowing identification and histological examination of the sentinel node for melanoma patients. They proposed that the sentinel node could be any one node within a particular lymph basin dependent on the location of the primary tumour. Since then, much work has been carried out in perfecting the technique of lymphoscintography and sentinel node biopsy applied to melanoma. It was, however, only in 1994 that this technique was applied to breast cancer [6] and followed up by further research in 1997 [7]. I spent my elective at St Luke’s Cancer Centre, Royal Surrey County Hospital, where a consultant 112
G. Howard-Alpe
oncologist had teamed up with a consultant breast surgeon to investigate the value of sentinel node biopsy. As was demonstrated by Nieweg et al. [8], the centre decided to try to identify the sentinel node or nodes in patients newly diagnosed with breast cancer prior to surgery. After removal of the primary tumour the surgeon would then remove the sentinel node/s before proceeding to carry out an axillary dissection. The sentinel node/s were sent for separate histological examination. Later, the histology reports were compared to see whether the disease status of the sentinel node accurately predicted the disease status of the entire axillary lymph field. The use of Patent Blue V dye was also tested, as described by Nieweg, to assist in the localization of the sentinel node at operation [8].
Patients and Methods The patients all had fine-needle aspiration biopsyconfirmed malignant breast tumours. All were aware of their diagnosis and had been advised that surgical removal of the tumour was the best treatment option, for which they had given their consent. The sentinel node procedure was explained to them and consent obtained. During my stay at St Luke’s, six patients underwent the procedure. The following acquisition protocol was used for sentinel node imaging: 1. Patients were admitted to hospital prior to operation, allowing enough time for scanning either on the day of operation or the day previously. 2. The patient was brought to the Department of Nuclear Medicine. 3. Necessary equipment: Inco-pad (with a hole cut out); gloves; Mediswabs; 1 ml syringe with 23G needle; cobalt-57 flood source; cobalt-57 markers. 4. Radiopharmaceutical preparation; 99mTc NanoColloid made up in one of the following ways: (a) for patients going for surgery that day: 1000 MBq in 5 ml; (b) for patients going for surgery the following day: 2000 MBq in 5 ml. 5. Activity for injection: (a) for patients going for surgery that day: 15 MBq in 0.3 ml (this allows 5 MBq in needle dead space); (b) for patients going for surgery the following day: 60 MBq in 0.3 ml (this allows 20 MBq in needle dead space). 6. A gamma camera with a low energy, high resolution collimator was used. 7. The patient was supine on the bed, feet to gantry, arms to sides. 8. A senior doctor, either the consultant or his/her registrar then administered the injection directly into the tumour. 9. The camera was lowered immediately over the area to be imaged. There should be minimal delay
10.
11. 12. 13.
to the start of imaging after the injection. The cobalt-57 flood source was placed beneath the patient on the floor for all views except the relevant lateral. The imaging views included: (a) dynamic study: scanning for approximately 20 min performed immediately after the injection had been administered; (b) static studies: an anterior and relevant lateral view taken 30 min, 2 h and 4 h after injection. Scanning was for 5 min for each view. Effective dose equivalent to the patient: 0.1–0.75 mSv. There should be a total of seven images for each patient. If a sentinel node was visible on either the dynamic image or one of the static images, the senior doctor was informed so that he/she could indicate with a marker pen the site of the sentinel node/s as accurately as possible on the patient.
The images produced and the marks drawn on the patient should enable the surgeon to locate accurately the position of the sentinel node. The surgeon injects Patent Blue V dye into the primary tumour prior to its removal. The blue dye should drain first to the same node identified by the nuclear medicine imaging and allow the surgeon to follow the blue lymphatic trail to the node, therefore confirming the sentinel node. An intraoperative gamma probe is also used to measure the radioactivity in both the tumour and the blue-dyed sentinel node; this should be higher than in the surrounding tissue and other nodes. Once the breast tumour has been removed and the sentinel node/s have been identified and removed, the surgeon then proceeds to carry out a full axillary lymph node dissection. The sentinel node/s are sent for histological examination separately from the other axillary lymph nodes.
Results Figure 1 shows the operative findings in patient 2. Three lymph nodes filled with dye and a lymph drainage channel running from the primary tumour can be seen clearly. Figure 2 shows the nuclear medicine sentinel node scan of patient 4.
The Sentinel Nodes Localized The sentinel node scan on patient 1 visualized two nodes, the lower of which was termed the sentinel node. These two nodes were found at the time of operation to demonstrate increased radioactivity but they did not contain the blue dye. Both nodes were sent for histological examination. Three nodes were seen in the axilla of patient 2 and the first node seen was described as the sentinel node; this was located at the top of the level one axillary nodes. At operation,
Fig. 1. Operative findings in patient 2, demonstrating three dyefilled lymph nodes and a lymph drainage channel running from the primary tumour.
Fig. 2. Nuclear medicine sentinel node scan (patient 4). page 1
all three nodes contained blue dye but only two had increased radioactivity. All three nodes were sent for histological examination. Patient 3 had a total of five nodes visualized, which were located in both the axilla and the internal mammary chain. The sentinel node was taken to be the lowest node visualized, which was in the internal mammary chain, closely related to the primary tumour site. At operation, however, only one axillary node demonstrated increased radioactivity, but it contained no blue dye. This node was sent for histological examination. Only one node was visualized in patient 4; this was found in the internal mammary chain and was taken to be the sentinel node. No node was located at operation. The scans of patients 5 and 6 visualized no nodes. At operation, no nodes were localized in patient 5 but three nodes were found to have increased radioactivity in patient 6. Two of these nodes also contained blue dye and all three were sent for histological examination. Sentinel Node Localization and Biopsy in Breast Cancer
113
Table 1. Operative findings Patient no.
Background radioactivity level
No. of nodes with increased radioactivity level
Radioactivity level in the identified nodes
No. of nodes containing blue dye
1
~30 counts
2 Both axillary nodes
2 The same nodes that also contained radioactivity
2
~30 counts
2 Both axillary nodes
3
~60 counts
4
~20 counts
1 Axilla ?
Node A (lower node) ~60 counts Node B (higher node) ~120 counts Node A (lower node) ~100 counts Node B (higher node) ~100 counts ~300 counts
0
5 6
~50 counts ~40 counts
Area of increased counts found in internal mammary chain over 4th intercostal space; no node excised due to operative difficulties – Node A ~60 counts Node B ~110 counts Node C ~130 counts
0 3 All axillary nodes
3 The two nodes containing radioactivity plus one other node higher up in the axilla 0
0 2 Node A = blue Node B 0 blue Node c = not blue
Table 2. Histological results for patients with identifiable sentinel nodes Patient no.
Histology of primary lesion
Histology of sentinel node/s
Histology of other nodes removed in axillary clearance
1
Wide local excision: 9 mm diameter Grade I (5) invasive ductal carcinoma. 8 mm low and intermediate grade in situ carcinoma. Invasive and in situ disease excised by 2 mm. Invasive ductal carcinoma Grade II: 6 mm diameter, completely excised. Extensive intermediate grade ductal carcinoma in situ: 50 mm max. dimension extending close to deep margin. Mucinous carcinoma Grade III (7): 35 mm max. dimension extending within 2 mm of the deep margin. Areas of cribiform and micropapillary intermediate and high-grade in situ carcinoma present: max. dimension 1–2 mm. Pleomorphic invasive lobular carcinoma Grade II (7): 85 mm diameter, extending with in 5 mm of the deep margin. Associated with extensive in situ carcinoma.
Low blue sentinel node and high blue sentinel node both tumour free.
All other nodes also tumour free.
Four sentinel nodes all tumour free
All other nodes also tumour free.
Sentinel node tumour free
All other nodes also tumour free.
2
3
6
Sentinel node A (100 counts) tumour A further four nodes in the axilla free. Sentinal node B (110 counts) contained metastatic disease (total contained metastatic disease. of 6/27 nodes positive). Sentinel node C (130 counts) contained metastatic disease.
Table 3. Histological results for patients with no identifiable sentinel node/s Patient no.
Histology of primary lesion
Histology of axillary lymph nodes
4
Multifocal invasive carcinoma found predominantly in three main areas. Two areas of ductal carcinoma Grade II (7): max. dimension 31 mm, and one area of invasive lobular carcinoma Grade II (6): max. dimension 10 mm. Also areas of lobular in situ carcinoma: max. dimension 5 mm. Minimum observed clearance 15 mm. Invasive classical lobular carcinoma Grade I (5): 30 mm max. dimension. Extending to deep margin.
2/30 axillary nodes contained carcinoma (ductal type). No lymph node found in tissue from the internal mammary sample.
5
114
G. Howard-Alpe
All axillary lymph nodes tumour free.
Table 4. Implications of the results Patient no.
Histological confirmation of operative findings
1 2 3 4 5 6
True negative True negative True negative No sentinel node identified No sentinel node identified True positive
The operative findings are indicated in Table 1. The histological results for patients with identifiable sentinel node/s are shown in Table 2. The histological results for patients with no identifiable sentinel node/s are given in Table 3. The implications of these results are given in Table 4. It is difficult to draw conclusions concerning the specificity and sensitivity of these results owing to the extremely small number of patients included. However, it is interesting to note that there were no false negative results, and also that in two of the six patients who underwent the procedure, no sentinel node could be localized.
Discussion This examination of the procedure of sentinel node localization and biopsy raises interesting points about the technique and its benefits to patient care.
The Technique of Sentinel Node Localization A combination of three different ways of identifying the sentinel node was used. First, by nuclear medicine scanning after injection of a radiopharmaceutical preparation and marking of the site of the node on the patient. Secondly, by the use of the intraoperative gamma probe and, thirdly, by the injection of blue dye into the primary tumour in the operating theatre. This maximizes the chances of locating a sentinel node [9]. The investigating groups who have tested the procedure have used different techniques, which have lead to varying success rates in identifying the sentinel node. This serves to make the literature slightly confusing. After the six patients reported here, a change was made to the technique. Instead of an injection volume of 0.3 ml being used, this was increased to 1.0 ml. Subsequently, the rate of identification of sentinel nodes increased.
The De®nition of the Sentinel Node Different techniques in localization also bring into question the definition of the sentinel node. Cabanas [4] proposed that it was the first node to which lymph drained from the tumour site. However, it has also
been shown that there may be more than one sentinel node. Publications vary on their definition, so just how do we classify what is a sentinel node and what is not? The correct definition is the first lymph node to which lymph drains from the primary site. If there is a chain of connected lymph nodes draining lymph from the primary, the sentinel node is the first node in that chain. The other nodes in the chain are not sentinel nodes. Therefore, a sentinel node does not have to be the lymph node with the highest radioactivity count at operation or be the lymph node closest to the primary site.
The Injection Site Publications vary on whether injection of the radiopharmaceutical preparation should be directly into or around the site of the primary tumour. Concern has been expressed that an injection into the primary tumour causes increased pressure within the tumour and may result in micrometastases being released. This is a worrying concept but the possibility of this happening is minimized by the small volume of fluid injected. The same applies to the injection of the blue dye. To assess fully whether there is any increased risk of micrometastases, a trial would have to be set up comparing women who had sentinel node localization involving direct injection into the primary tumour with women who had not undergone this procedure. This would be difficult because breast cancer and the spread of secondary disease from the primary tumour is multifactorial. It would, therefore, be hard to find two comparable groups of women to enter such a trial.
Palpable Versus Impalpable Tumours Currently, this technique is limited to palpable tumours, which are obviously larger than impalpable tumours and therefore more advanced, with an increased likelihood of lymph node metastases. The procedure is not useful for women who have axillary lymph node metastases at presentation. They will require axillary lymph node dissection anyway and the benefit of the procedure is to identify the women for whom such dissection is an unnecessary procedure. The reason that the technique is limited to patients with palpable tumours is that the doctor administering the radiopharmaceutical injection has to be able to feel the tumour. This means that the patients who could most benefit from the procedure, the early screen-detected cases, are currently excluded. The way to rectify this would be to carry out a radiologically or ultrasound guided injection. This unfortunately only adds to the cost of the procedure in terms of finance, healthcare worker time and stress for the patient, but the procedure would avoid the complications associated with axillary clearance for many women. Sentinel Node Localization and Biopsy in Breast Cancer
115
The Signi®cance of no Visualized Nodes What conclusions can we draw about the patients in whom no sentinel node is found? 1. Human error? The injection was not administered correctly and missed the tumour site. 2. Tumour-filled lymph nodes? All of the lymph nodes already contain disease and therefore cannot drain lymph from the primary tumour site [10]. 3. Histological type of the tumour? Perhaps certain types of tumours are more likely to drain to different lymph fields. Some types may be more or less likely to metastasize via the lymphatics and this may have implications for sentinel node identification. These patients are just as interesting and should undergo full axillary lymph dissection to determine their lymph node status and to attempt to understand why no sentinel node could be localized.
Drainage to Lymphatic Systems Other than the Axillary Nodes In patient 4, the sentinel node scans identified drainage to lymph nodes in the internal mammary chain. This was an unexpected finding and resulted in an alteration to the subsequent management of this patient. The postoperative radiotherapy fields were extended to cover the lymph nodes in the internal mammary chain. This hopefully improved the outcome for this patient. It would be interesting to analyze the proportion of patients who, on investigation with a sentinel node scan, had lymph drainage to nodes other than in the ipsilateral axilla. Perhaps risk factors for lymph drainage other than to the axilla could be identified, such as tumour location or histological tumour type.
Disadvantages What are the disadvantages of this procedure? 1. It is another procedure for the patient to undergo, which all adds to the psychological stress of the treatment. Patients initially have a great deal of information to assimilate, with the concepts of surgery, postoperative radiotherapy and possibly even adjuvant chemotherapy. 2. There may be false negative results. The incidence of this has so far been shown to be very low, with many groups achieving a 0% false negative rate [2], however, any procedure, if performed frequently enough, will carry a false negative rate that could have devastating consequences for the patients involved. 3. Currently, the procedure is being carried out as one operation. The sentinel node and the rest of the axillary lymph nodes are all removed at the same time to assess whether the sentinel node could have predicted accurately the disease status of the 116
G. Howard-Alpe
axilla. This involves no change in the normal surgical management of the patient. In future, the procedure is being proposed as a two-step process. The patient would initially be taken to theatre for surgical removal of the primary tumour plus removal of the sentinel node. If the results of histological analysis of the sentinel node show that it contains disease, the patient would then have to return to theatre for axillary lymph node dissection. This means two operative procedures for the patient, increasing the risk of operative, postoperative and anaesthetic complications. The uncertainty also adds to the psychological stress for the patient. Studies have been carried out in which the sentinel node is examined histologically by frozen section while the patient is in theatre. Concern has been expressed that this is not an accurate enough histological examination, as sentinel node analysis seeks micrometastatic deposits within the lymph node, not necessarily a completely tumour filled node, and requires a more careful, detailed examination by the pathologist. 4. Concerns have also been raised that injection of blue dye into the tumour site can result in permanent discoloration of the surrounding skin [11]. This appears to be a relatively rare complication but should be noted and extreme care taken when administering the injection.
Advantages There are many advantages to the patient of sentinel node biopsy as an alternative to routine axillary dissection because it reduces the surgical morbidity to patients by identifying those for whom the procedure is unnecessary. This reduces the incidence of debilitating side effects such as lymphoedema of the arm in patients who do not require axillary dissection. So far, the results of the different groups testing the procedure have been encouraging. Many groups have shown rates of identifying the sentinel node reaching 100% [2]. The rate of false negative results appears to be extremely low and the complications of the procedure are small compared with the benefits gained. The research currently available supports the view that sentinel node lymphoscintography, intraoperative lymphatic mapping and sentinel lymphadenectomy accurately identify the sentinel lymph node/s most likely to contain metastatic disease. There appears to be a procedural learning curve, but the technique has been found to be feasible, economical and reproducible, and the demands placed upon nuclear medicine personnel, pathologists, surgeons and operating staff are not unreasonable. If the procedure of lymphoscintography and blue dye fails to identify a sentinel node, this has been found to be a risk factor for nodal replacement with tumour [10].
Acknowledgements. I would like to express my thanks to St Luke’s Cancer Centre, Royal Surrey County Hospital, Guildford and especially to Dr Bill White for the help I received during my stay there.
References 1. Moore MP, Kinne DW. Axillary lymphadenopathy: a diagnostic and therapeutic procedure. J Surg Oncol 1997;66:2–6. 2. Giuliano AE, Jones RC, Brennan M, et al. Sentinel lymphadenectomy in breast cancer. J Clin Oncol 1997;15:2345–50. 3. Singhal H, O’Malley FP, Tweedie E, et al. Axillary node dissection in patients with breast cancer diagnosed through Ontario Breast Screening Program: a need for minimally invasive techniques. Can J Surg 1997;40:377–82. 4. Cabanas RM. An approach for the treatment of penile carcinoma. Cancer 1977;39:456–66. 5. Morton DL, Wen D-R, Won JH, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992;127:392–9.
6. Giuliano AE, Kirgan DM, Guenther JM, et al. Lymphatic mapping and sentinel lymphadenectomy for breast cancer. Ann Surg 1994;220:391–401. 7. 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;349:1864–7. 8. Nieweg OL. Lymphatic mapping and sentinel node biopsy course 1997–1998. 9. Wong JH, Truelove K, Ko P, et al. Localisation and resection of an in transit sentinel node by use of lymphoscintography, intraoperative lymphatic mapping and a hand held gamma probe. Surgery 1996;120:114–6. 10. Guenther JM, Krishnamoorthy M, Tan LR. Sentinel lymphadenectomy for breast cancer in a community managed care setting. Cancer J Sci Am 1997;3:336–40. 11. Giuliano AE. Intradermal blue dye to identify sentinel lymph node in breast cancer [letter]. Lancet 1997;350:958.
Received for publication July 1998 Accepted following revision October 1998
Sentinel Node Localization and Biopsy in Breast Cancer
117
Clinical Oncology (1999)11:111–117 # The Royal College of Radiologists
Annual Undergraduate Prize Royal College of Radiologists Annual Undergraduate Prize in Clinical Oncology: Prize-Winning Entries 1998. It has been a delight once again to work with the authors of the winning submissions for the Royal College of Radiologists Annual Undergraduate Prize in Clinical Oncology. The first and second prize winning entries follow and their excellence speaks for themselves. Congratulations to both authors. W. G. Jones Editor
Sentinel Node Localization and Biopsy in Breast Cancer* G. Howard-Alpe Royal Free Hospital School of Medicine, London, UK Abstract. Currently, it is routine practice to carry out axillary lymph node dissection at the time of surgical removal of a malignant primary breast tumour. As breast cancer is being diagnosed at an earlier stage in a growing percentage of cases, this procedure is proving unnecessary in a proportion of patients. Axillary lymph node dissection carries a risk of side effects and ideally we should identify the patients who actually require the procedure. The concept of the sentinel node is not new and was developed over twenty years ago. The technique of lymphoscintography was initially used to identify the sentinel node/s in patients with malignant melanoma. A proposed alternative management strategy to routine axillary lymph node dissection in patients with breast cancer is sentinel lymph node localization using lymphoscintography and biopsy of the identified sentinel node. The aim being to accurately predict the disease status of the axilla and consequently determine whether axillary lymph node dissection is indicated. The technique is currently undergoing a multi-centre trial in the UK. During my elective at St Luke’s Cancer Centre, Royal Surrey County Hospital, Guildford I was fortunate to observe this procedure being tested, with patients undergoing both lymphoscintography and sentinel node biopsy as well as axillary lymph node dissection. The results seen during my stay were extremely encouraging. Keywords: Axillary lymph node dissection; Biopsy; Breast cancer; Lymphoscintography; Sentinel node
Introduction Despite research into tumour-associated prognostic indicators in breast cancer, including oncogene expression, hormone receptor status, ploidy and Sphase, lymph node disease status remains the most accurate prognostic indicator of overall survival. The current practice of routine axillary lymph node dissection serves four purposes: it assures locoregional control, improves survival, and provides important information for staging and for prognosis. The status of the axilla affects treatment options including whether or not adjuvant chemotherapy is given to *Royal College of Radiologists Annual Undergraduate Prize: First Prize Winner Correspondence and offprint requests to: Dr G. Howard-Alpe, Flat 6, Westminster Bridge House, 7 Westminster Bridge Road, London SE1 7XP, UK.
premenopausal patients. Axillary lymph node dissection is therefore an important procedure in the treatment of breast cancer, and has been shown to be beneficial even with a T1A lesion [1]. The procedure of routine axillary lymph node dissection has side effects that can be disabling and painful, and impair a patient’s quality of life. The main side effect is lymphoedema of the arm. The rate of occurrence is dependent on whether radiotherapy is also administered to the axillary area and on surgical skill at the time of the operation. The incidence of lymphoedema varies directly with the amount of nodal tissue removed [2]; level one dissection or blind biopsy of a few nodes is associated with the lowest incidence of lymphoedema, but this carries a high incidence of false negative axillary staging. Additional side effects include vascular or brachial plexus injury but these are much rarer.
Owing to advances allowing the earlier detection of tumours through both screening programmes and public awareness campaigns, the question of whether axillary lymph node dissection is necessary as a routine procedure is currently being raised. It certainly appears that many women who have early tumours are unnecessarily undergoing a procedure that proves the lymph node status to be negative and causes a proportion of them to experience morbidity. This calls for research into less invasive ways of staging the axilla. A recent study by Singhal et al. in Ontario concluded that women older than 60 years had a low probability of nodal positivity (0%–8.7%), but that there is insufficient information in this group to give a 95% or better prediction of nodal status at the time of surgery. It called for studies into minimally invasive techniques such as sentinel node biopsy to minimize surgical morbidity in these women, who, as a result of early diagnosis, have an excellent long term outlook [3]. The concept of sequential lymphatic dissemination of a primary tumour was initially developed by Cabanas in 1977 [4]. He suggested that squamous cell carcinoma of the penis initially drained to a particular lymph node in the groin, which was always in the same location and termed the ‘sentinel node’. Tumour cells travelling in the lymphatics lodged in this first sentinel node and then, at a later stage, travelled on to other nodes in the direct drainage pathway. He hypothesized that if this sentinel node could be identified, removed surgically and examined, the histological status of the node acted as an indicator of the status of the entire lymphatic field. Therefore, if metastatic disease was found in the sentinel node there was a strong chance that other lymph nodes within that field would contain disease and regional lymph node dissection should be advised. When the sentinel node was disease free it was unlikely that other nodes would be involved and regional lymph node dissection would be inappropriate. For penile cancer this assumption is plausible because it is always in exactly the same part of the body. The issue becomes more complex when it is applied to other tumours whose location within the body is less precise. This theory was revived in 1992 by Morton et al. [5], who developed a technique of lymphatic mapping by lymphoscintography, allowing identification and histological examination of the sentinel node for melanoma patients. They proposed that the sentinel node could be any one node within a particular lymph basin dependent on the location of the primary tumour. Since then, much work has been carried out in perfecting the technique of lymphoscintography and sentinel node biopsy applied to melanoma. It was, however, only in 1994 that this technique was applied to breast cancer [6] and followed up by further research in 1997 [7]. I spent my elective at St Luke’s Cancer Centre, Royal Surrey County Hospital, where a consultant 112
G. Howard-Alpe
oncologist had teamed up with a consultant breast surgeon to investigate the value of sentinel node biopsy. As was demonstrated by Nieweg et al. [8], the centre decided to try to identify the sentinel node or nodes in patients newly diagnosed with breast cancer prior to surgery. After removal of the primary tumour the surgeon would then remove the sentinel node/s before proceeding to carry out an axillary dissection. The sentinel node/s were sent for separate histological examination. Later, the histology reports were compared to see whether the disease status of the sentinel node accurately predicted the disease status of the entire axillary lymph field. The use of Patent Blue V dye was also tested, as described by Nieweg, to assist in the localization of the sentinel node at operation [8].
Patients and Methods The patients all had fine-needle aspiration biopsyconfirmed malignant breast tumours. All were aware of their diagnosis and had been advised that surgical removal of the tumour was the best treatment option, for which they had given their consent. The sentinel node procedure was explained to them and consent obtained. During my stay at St Luke’s, six patients underwent the procedure. The following acquisition protocol was used for sentinel node imaging: 1. Patients were admitted to hospital prior to operation, allowing enough time for scanning either on the day of operation or the day previously. 2. The patient was brought to the Department of Nuclear Medicine. 3. Necessary equipment: Inco-pad (with a hole cut out); gloves; Mediswabs; 1 ml syringe with 23G needle; cobalt-57 flood source; cobalt-57 markers. 4. Radiopharmaceutical preparation; 99mTc NanoColloid made up in one of the following ways: (a) for patients going for surgery that day: 1000 MBq in 5 ml; (b) for patients going for surgery the following day: 2000 MBq in 5 ml. 5. Activity for injection: (a) for patients going for surgery that day: 15 MBq in 0.3 ml (this allows 5 MBq in needle dead space); (b) for patients going for surgery the following day: 60 MBq in 0.3 ml (this allows 20 MBq in needle dead space). 6. A gamma camera with a low energy, high resolution collimator was used. 7. The patient was supine on the bed, feet to gantry, arms to sides. 8. A senior doctor, either the consultant or his/her registrar then administered the injection directly into the tumour. 9. The camera was lowered immediately over the area to be imaged. There should be minimal delay
10.
11. 12. 13.
to the start of imaging after the injection. The cobalt-57 flood source was placed beneath the patient on the floor for all views except the relevant lateral. The imaging views included: (a) dynamic study: scanning for approximately 20 min performed immediately after the injection had been administered; (b) static studies: an anterior and relevant lateral view taken 30 min, 2 h and 4 h after injection. Scanning was for 5 min for each view. Effective dose equivalent to the patient: 0.1–0.75 mSv. There should be a total of seven images for each patient. If a sentinel node was visible on either the dynamic image or one of the static images, the senior doctor was informed so that he/she could indicate with a marker pen the site of the sentinel node/s as accurately as possible on the patient.
The images produced and the marks drawn on the patient should enable the surgeon to locate accurately the position of the sentinel node. The surgeon injects Patent Blue V dye into the primary tumour prior to its removal. The blue dye should drain first to the same node identified by the nuclear medicine imaging and allow the surgeon to follow the blue lymphatic trail to the node, therefore confirming the sentinel node. An intraoperative gamma probe is also used to measure the radioactivity in both the tumour and the blue-dyed sentinel node; this should be higher than in the surrounding tissue and other nodes. Once the breast tumour has been removed and the sentinel node/s have been identified and removed, the surgeon then proceeds to carry out a full axillary lymph node dissection. The sentinel node/s are sent for histological examination separately from the other axillary lymph nodes.
Results Figure 1 shows the operative findings in patient 2. Three lymph nodes filled with dye and a lymph drainage channel running from the primary tumour can be seen clearly. Figure 2 shows the nuclear medicine sentinel node scan of patient 4.
The Sentinel Nodes Localized The sentinel node scan on patient 1 visualized two nodes, the lower of which was termed the sentinel node. These two nodes were found at the time of operation to demonstrate increased radioactivity but they did not contain the blue dye. Both nodes were sent for histological examination. Three nodes were seen in the axilla of patient 2 and the first node seen was described as the sentinel node; this was located at the top of the level one axillary nodes. At operation,
Fig. 1. Operative findings in patient 2, demonstrating three dyefilled lymph nodes and a lymph drainage channel running from the primary tumour.
Fig. 2. Nuclear medicine sentinel node scan (patient 4). page 1
all three nodes contained blue dye but only two had increased radioactivity. All three nodes were sent for histological examination. Patient 3 had a total of five nodes visualized, which were located in both the axilla and the internal mammary chain. The sentinel node was taken to be the lowest node visualized, which was in the internal mammary chain, closely related to the primary tumour site. At operation, however, only one axillary node demonstrated increased radioactivity, but it contained no blue dye. This node was sent for histological examination. Only one node was visualized in patient 4; this was found in the internal mammary chain and was taken to be the sentinel node. No node was located at operation. The scans of patients 5 and 6 visualized no nodes. At operation, no nodes were localized in patient 5 but three nodes were found to have increased radioactivity in patient 6. Two of these nodes also contained blue dye and all three were sent for histological examination. Sentinel Node Localization and Biopsy in Breast Cancer
113
Table 1. Operative findings Patient no.
Background radioactivity level
No. of nodes with increased radioactivity level
Radioactivity level in the identified nodes
No. of nodes containing blue dye
1
~30 counts
2 Both axillary nodes
2 The same nodes that also contained radioactivity
2
~30 counts
2 Both axillary nodes
3
~60 counts
4
~20 counts
1 Axilla ?
Node A (lower node) ~60 counts Node B (higher node) ~120 counts Node A (lower node) ~100 counts Node B (higher node) ~100 counts ~300 counts
0
5 6
~50 counts ~40 counts
Area of increased counts found in internal mammary chain over 4th intercostal space; no node excised due to operative difficulties – Node A ~60 counts Node B ~110 counts Node C ~130 counts
0 3 All axillary nodes
3 The two nodes containing radioactivity plus one other node higher up in the axilla 0
0 2 Node A = blue Node B 0 blue Node c = not blue
Table 2. Histological results for patients with identifiable sentinel nodes Patient no.
Histology of primary lesion
Histology of sentinel node/s
Histology of other nodes removed in axillary clearance
1
Wide local excision: 9 mm diameter Grade I (5) invasive ductal carcinoma. 8 mm low and intermediate grade in situ carcinoma. Invasive and in situ disease excised by 2 mm. Invasive ductal carcinoma Grade II: 6 mm diameter, completely excised. Extensive intermediate grade ductal carcinoma in situ: 50 mm max. dimension extending close to deep margin. Mucinous carcinoma Grade III (7): 35 mm max. dimension extending within 2 mm of the deep margin. Areas of cribiform and micropapillary intermediate and high-grade in situ carcinoma present: max. dimension 1–2 mm. Pleomorphic invasive lobular carcinoma Grade II (7): 85 mm diameter, extending with in 5 mm of the deep margin. Associated with extensive in situ carcinoma.
Low blue sentinel node and high blue sentinel node both tumour free.
All other nodes also tumour free.
Four sentinel nodes all tumour free
All other nodes also tumour free.
Sentinel node tumour free
All other nodes also tumour free.
2
3
6
Sentinel node A (100 counts) tumour A further four nodes in the axilla free. Sentinal node B (110 counts) contained metastatic disease (total contained metastatic disease. of 6/27 nodes positive). Sentinel node C (130 counts) contained metastatic disease.
Table 3. Histological results for patients with no identifiable sentinel node/s Patient no.
Histology of primary lesion
Histology of axillary lymph nodes
4
Multifocal invasive carcinoma found predominantly in three main areas. Two areas of ductal carcinoma Grade II (7): max. dimension 31 mm, and one area of invasive lobular carcinoma Grade II (6): max. dimension 10 mm. Also areas of lobular in situ carcinoma: max. dimension 5 mm. Minimum observed clearance 15 mm. Invasive classical lobular carcinoma Grade I (5): 30 mm max. dimension. Extending to deep margin.
2/30 axillary nodes contained carcinoma (ductal type). No lymph node found in tissue from the internal mammary sample.
5
114
G. Howard-Alpe
All axillary lymph nodes tumour free.
Table 4. Implications of the results Patient no.
Histological confirmation of operative findings
1 2 3 4 5 6
True negative True negative True negative No sentinel node identified No sentinel node identified True positive
The operative findings are indicated in Table 1. The histological results for patients with identifiable sentinel node/s are shown in Table 2. The histological results for patients with no identifiable sentinel node/s are given in Table 3. The implications of these results are given in Table 4. It is difficult to draw conclusions concerning the specificity and sensitivity of these results owing to the extremely small number of patients included. However, it is interesting to note that there were no false negative results, and also that in two of the six patients who underwent the procedure, no sentinel node could be localized.
Discussion This examination of the procedure of sentinel node localization and biopsy raises interesting points about the technique and its benefits to patient care.
The Technique of Sentinel Node Localization A combination of three different ways of identifying the sentinel node was used. First, by nuclear medicine scanning after injection of a radiopharmaceutical preparation and marking of the site of the node on the patient. Secondly, by the use of the intraoperative gamma probe and, thirdly, by the injection of blue dye into the primary tumour in the operating theatre. This maximizes the chances of locating a sentinel node [9]. The investigating groups who have tested the procedure have used different techniques, which have lead to varying success rates in identifying the sentinel node. This serves to make the literature slightly confusing. After the six patients reported here, a change was made to the technique. Instead of an injection volume of 0.3 ml being used, this was increased to 1.0 ml. Subsequently, the rate of identification of sentinel nodes increased.
The De®nition of the Sentinel Node Different techniques in localization also bring into question the definition of the sentinel node. Cabanas [4] proposed that it was the first node to which lymph drained from the tumour site. However, it has also
been shown that there may be more than one sentinel node. Publications vary on their definition, so just how do we classify what is a sentinel node and what is not? The correct definition is the first lymph node to which lymph drains from the primary site. If there is a chain of connected lymph nodes draining lymph from the primary, the sentinel node is the first node in that chain. The other nodes in the chain are not sentinel nodes. Therefore, a sentinel node does not have to be the lymph node with the highest radioactivity count at operation or be the lymph node closest to the primary site.
The Injection Site Publications vary on whether injection of the radiopharmaceutical preparation should be directly into or around the site of the primary tumour. Concern has been expressed that an injection into the primary tumour causes increased pressure within the tumour and may result in micrometastases being released. This is a worrying concept but the possibility of this happening is minimized by the small volume of fluid injected. The same applies to the injection of the blue dye. To assess fully whether there is any increased risk of micrometastases, a trial would have to be set up comparing women who had sentinel node localization involving direct injection into the primary tumour with women who had not undergone this procedure. This would be difficult because breast cancer and the spread of secondary disease from the primary tumour is multifactorial. It would, therefore, be hard to find two comparable groups of women to enter such a trial.
Palpable Versus Impalpable Tumours Currently, this technique is limited to palpable tumours, which are obviously larger than impalpable tumours and therefore more advanced, with an increased likelihood of lymph node metastases. The procedure is not useful for women who have axillary lymph node metastases at presentation. They will require axillary lymph node dissection anyway and the benefit of the procedure is to identify the women for whom such dissection is an unnecessary procedure. The reason that the technique is limited to patients with palpable tumours is that the doctor administering the radiopharmaceutical injection has to be able to feel the tumour. This means that the patients who could most benefit from the procedure, the early screen-detected cases, are currently excluded. The way to rectify this would be to carry out a radiologically or ultrasound guided injection. This unfortunately only adds to the cost of the procedure in terms of finance, healthcare worker time and stress for the patient, but the procedure would avoid the complications associated with axillary clearance for many women. Sentinel Node Localization and Biopsy in Breast Cancer
115
The Signi®cance of no Visualized Nodes What conclusions can we draw about the patients in whom no sentinel node is found? 1. Human error? The injection was not administered correctly and missed the tumour site. 2. Tumour-filled lymph nodes? All of the lymph nodes already contain disease and therefore cannot drain lymph from the primary tumour site [10]. 3. Histological type of the tumour? Perhaps certain types of tumours are more likely to drain to different lymph fields. Some types may be more or less likely to metastasize via the lymphatics and this may have implications for sentinel node identification. These patients are just as interesting and should undergo full axillary lymph dissection to determine their lymph node status and to attempt to understand why no sentinel node could be localized.
Drainage to Lymphatic Systems Other than the Axillary Nodes In patient 4, the sentinel node scans identified drainage to lymph nodes in the internal mammary chain. This was an unexpected finding and resulted in an alteration to the subsequent management of this patient. The postoperative radiotherapy fields were extended to cover the lymph nodes in the internal mammary chain. This hopefully improved the outcome for this patient. It would be interesting to analyze the proportion of patients who, on investigation with a sentinel node scan, had lymph drainage to nodes other than in the ipsilateral axilla. Perhaps risk factors for lymph drainage other than to the axilla could be identified, such as tumour location or histological tumour type.
Disadvantages What are the disadvantages of this procedure? 1. It is another procedure for the patient to undergo, which all adds to the psychological stress of the treatment. Patients initially have a great deal of information to assimilate, with the concepts of surgery, postoperative radiotherapy and possibly even adjuvant chemotherapy. 2. There may be false negative results. The incidence of this has so far been shown to be very low, with many groups achieving a 0% false negative rate [2], however, any procedure, if performed frequently enough, will carry a false negative rate that could have devastating consequences for the patients involved. 3. Currently, the procedure is being carried out as one operation. The sentinel node and the rest of the axillary lymph nodes are all removed at the same time to assess whether the sentinel node could have predicted accurately the disease status of the 116
G. Howard-Alpe
axilla. This involves no change in the normal surgical management of the patient. In future, the procedure is being proposed as a two-step process. The patient would initially be taken to theatre for surgical removal of the primary tumour plus removal of the sentinel node. If the results of histological analysis of the sentinel node show that it contains disease, the patient would then have to return to theatre for axillary lymph node dissection. This means two operative procedures for the patient, increasing the risk of operative, postoperative and anaesthetic complications. The uncertainty also adds to the psychological stress for the patient. Studies have been carried out in which the sentinel node is examined histologically by frozen section while the patient is in theatre. Concern has been expressed that this is not an accurate enough histological examination, as sentinel node analysis seeks micrometastatic deposits within the lymph node, not necessarily a completely tumour filled node, and requires a more careful, detailed examination by the pathologist. 4. Concerns have also been raised that injection of blue dye into the tumour site can result in permanent discoloration of the surrounding skin [11]. This appears to be a relatively rare complication but should be noted and extreme care taken when administering the injection.
Advantages There are many advantages to the patient of sentinel node biopsy as an alternative to routine axillary dissection because it reduces the surgical morbidity to patients by identifying those for whom the procedure is unnecessary. This reduces the incidence of debilitating side effects such as lymphoedema of the arm in patients who do not require axillary dissection. So far, the results of the different groups testing the procedure have been encouraging. Many groups have shown rates of identifying the sentinel node reaching 100% [2]. The rate of false negative results appears to be extremely low and the complications of the procedure are small compared with the benefits gained. The research currently available supports the view that sentinel node lymphoscintography, intraoperative lymphatic mapping and sentinel lymphadenectomy accurately identify the sentinel lymph node/s most likely to contain metastatic disease. There appears to be a procedural learning curve, but the technique has been found to be feasible, economical and reproducible, and the demands placed upon nuclear medicine personnel, pathologists, surgeons and operating staff are not unreasonable. If the procedure of lymphoscintography and blue dye fails to identify a sentinel node, this has been found to be a risk factor for nodal replacement with tumour [10].
Acknowledgements. I would like to express my thanks to St Luke’s Cancer Centre, Royal Surrey County Hospital, Guildford and especially to Dr Bill White for the help I received during my stay there.
References 1. Moore MP, Kinne DW. Axillary lymphadenopathy: a diagnostic and therapeutic procedure. J Surg Oncol 1997;66:2–6. 2. Giuliano AE, Jones RC, Brennan M, et al. Sentinel lymphadenectomy in breast cancer. J Clin Oncol 1997;15:2345–50. 3. Singhal H, O’Malley FP, Tweedie E, et al. Axillary node dissection in patients with breast cancer diagnosed through Ontario Breast Screening Program: a need for minimally invasive techniques. Can J Surg 1997;40:377–82. 4. Cabanas RM. An approach for the treatment of penile carcinoma. Cancer 1977;39:456–66. 5. Morton DL, Wen D-R, Won JH, et al. Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 1992;127:392–9.
6. Giuliano AE, Kirgan DM, Guenther JM, et al. Lymphatic mapping and sentinel lymphadenectomy for breast cancer. Ann Surg 1994;220:391–401. 7. 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;349:1864–7. 8. Nieweg OL. Lymphatic mapping and sentinel node biopsy course 1997–1998. 9. Wong JH, Truelove K, Ko P, et al. Localisation and resection of an in transit sentinel node by use of lymphoscintography, intraoperative lymphatic mapping and a hand held gamma probe. Surgery 1996;120:114–6. 10. Guenther JM, Krishnamoorthy M, Tan LR. Sentinel lymphadenectomy for breast cancer in a community managed care setting. Cancer J Sci Am 1997;3:336–40. 11. Giuliano AE. Intradermal blue dye to identify sentinel lymph node in breast cancer [letter]. Lancet 1997;350:958.
Received for publication July 1998 Accepted following revision October 1998
Sentinel Node Localization and Biopsy in Breast Cancer
117