The sentinel node concept in prostate cancer: Present reality and future prospects

A sentinel node (SN) is de (cid:222) ned as the (cid:222) rst site where cancer cells are carried by lymph (cid:223) ow from a tumor. If this de (cid:222) nition (SN concept) correctly re (cid:223) ects the clinical reality, intraoperative SN biopsy would facilitate precise nodal staging. In malignant melanoma, a prolonged survival has been evidenced by a large-scale randomized controlled study. On the contrary, research on SN concept in deeply located cancers including prostate cancer, is still investigative, and no concrete data from clinical trials are yet available. Since 1993, several investigators have demonstrated that the SN concept could be applied in prostate cancer patients as well with high accuracy. Although promising and technically feasible in pre-clinical settings, many hurdles remain to be cleared before clinical application can be recommended. This review addresses the current status and related issues of the SN concept in prostate cancer, and discusses the future directions. ABSTRACT primary lesion (or regardless of its state), with the presence/absence of LN metastases diagnosed based on the pathological results of the


INTRODUCTION
When resecting malignant foci surgeons must always take into account the presence of invisible micrometastases so as to achieve the best possible prophylactic and systematic lymph node (LN) dissection. However, the more extensive the LN dissection the greater becomes the degree of surgical invasiveness, resulting in increasing organ functional impairment postoperatively. It would be ideal to be able to differentiate beforehand cases requiring and not requiring LN dissection, although the precision of currently available diagnostic imaging modalities is not yet sufÞ cient for this purpose. [1] As one approach to help overcome this problem the clinical application of the sentinel node (SN) concept has been attracting increasing attention. This concept was Þ rst proposed in the early 1990s, [2] and as is well known, has been widely applied internationally especially for breast cancer [3] and malignant melanoma. [4] On the other hand, in the case of malignancies situated deep within the body, like prostate cancer, the applicability of the SN concept is still in the investigative stage with no Þ rm conclusion yet reached as to its validity. [5,6] SN is deÞ ned as the Þ rst LN to receive lymph ß ow from a tumor. The SN concept is based on the assumption that metastases would Þ rst develop in this SN. In other words, if no metastasis is present in this SN, metastases can be judged to be absent in other LN, and gratuitous LN dissection can be omitted thereby reducing surgical invasiveness. Conversely, when metastasis is present in the SN, the LN dissection can be extended facilitating a switch to a more radical surgical method. In this way, SN navigation surgery (SNNS) can be used to determine the indications for radical surgery according to the presence/absence of SN metastases and decide the extent of the LN dissection required, but in prostate cancer this is still limited to the basic science or clinical research stage and has in general not yet been applied clinically. This review Þ rst discusses the various issues related to LN dissection in prostate cancer, then, notes the basic research results validating the SN concept, and Þ nally outlines the results of recent clinical research and future prospects.

VARIOUS ISSUES RELATED TO LN DISSECTION IN PROSTATE CANCER
In contrast to numerous other malignancies, in prostate cancer various issues remain to be resolved regarding LN dissection.

The extent of dissection
Usually, in any kind of malignancy the extent of LN dissection is determined by the state of the primary lesion (or regardless of its state), with the presence/absence of LN metastases diagnosed based on the pathological results of the Departments of Urology, 1 Nuclear Medicine, 2 Clinical Pathology, Tonami General Hospital, Tonami, Japan dissected LN. If this deÞ nition is not clariÞ ed comparisons of the results of studies on LN metastases conducted by different surgeons at different institutions will be problematic. When commencing research on LN in prostate cancer this is the Þ rst issue that has to be confronted. Even if a study speciÞ es 'limited dissection', 'standard dissection', and 'extended dissection', in practice in many cases it will be impossible to know the actual extent of the dissection. Moreover, in some papers the extent of the dissection is not speciÞ ed at all.

The staging significance of lymph node dissection
Since metastatic LNs of prostate cancer are not markedly enlarged and are usually microscopic in nature, the usefulness of diagnostic imaging in detecting them is low.
Recently, modalities such as Positron Emission Tomography-Computerized Tomography (PET-CT) are being used, but their diagnostic rate of clinically localized prostate cancer is unfavorable. [7] Accordingly, although extending the range of the dissection would theoretically facilitate a more accurate diagnosis of LN metastases, the burden of complications such as edema of the lower extremities and lymphorrhea might be increased. [8,9] On the other hand, if according to nomograms such as Partin Tables [10] and Kattan nomogram [11] the probability of LN metastases is below a given level, omission of LN dissection at the time of radical prostatectomy (RPx) may also be a reasonable option at present . [12] However, in studies in which extended LN dissection was performed for clinically localized cancer, metastases limited to beyond the range of standard dissection were found at high rates, namely 35% [13] and 19% [14] of metastasis-positive cases. Moreover, discrepancies in the histological Þ ndings are occasionally noted between biopsy specimens and totally resected specimens, with underestimation of the pathological stage becoming a problem when nomograms are used. [15] Therapeutic significance of lymph node dissection Bhatta-Dar et al., performed RPx in 336 patients with lowrisk prostate cancer (with dissection 140 cases/without dissection 196 cases) and retrospectively noted no difference in the six-year biochemical recurrence-free survival rate between the two groups. [16] In the study of Salomon et al., in which 43 cases receiving perineal RPx without dissection and 25 cases undergoing retropubic RPx with dissection were compared, no difference in the Þ ve-year recurrence rate was found either, although the numbers were small. [17] On the other hand, Bader et al., who performed extended dissection in 367 patients with prostate cancer, reported that in the 40% of cases with only a single pelvic LN metastasis no biochemical recurrence was detected during a 45-month follow-up. [18] Similarly, Han et al. and Daneshmand et al., reported that in patients with LN metastases biochemical nonrecurrence rates of 15% and 21% at ≥10 years respectively were observed. [19,20] In this way, it is clear that no consensus has yet been reached with respect to the therapeutic signiÞ cance of pelvic LN dissection in prostate cancer.

Significance of SN in prostate cancer
In prostate cancer, not only the signiÞ cance of the dissection itself but also the optimal method to achieve it have yet to be resolved, especially since it is difÞ cult to accurately grasp the presence/absence of metastases. For example, even if the nomograms are used, the original data reß ect the pathological results of only some of the intrapelvic LN, precluding the making of accurate predictions. Also, it is selfevident that the more limited the extent of the dissection is the less accurate will the diagnosis of metastases be.
In this context, it is very important to validate the SN concept as veriÞ ed by extended dissection. If like in breast cancer and malignant melanoma the SN concept is validated, accurate information about the LN will be obtainable by biopsy of the SN alone. Moreover, based on the information on the state of LN metastases provided by the SN biopsy, SNNS can be performed to determine the appropriateness of radical treatment. In cases negative for LN metastases in which a long-term course can be anticipated, prostatectomy or radiotherapy can be applied, whereas in cases found to be metastasis-positive conservative treatment such as hormone therapy can be resorted to so as to spare the patient gratuitous invasiveness.

Validation of the SN concept in prostate cancer
Since the 1999 report of Wawroschek et al., several groups have published papers regarding validation of the SN concept. [5,[21][22][23][24][25] The largest scale study was that of Wawroschek et al., in which the results of 1055 cases were analyzed. [6] Of these, LN metastases were found in 207 cases (19.6%), with false negative results (non-SN metastases found in the absence of SN metastases) noted in only 2/207 cases (1%). In 131/207 cases (63.3%), metastases were found beyond the range of routine dissection (external iliac LN + obturator LN). Other groups including our own have obtained similar results in smaller scale investigations [ Table 1]. Techniques to validate the SN concept are outlined below.

Tracers and their method of administration
Tracers used in this context include Technetium (Tc)-labeled colloid, which is a radioisotope, and dyes. Dyes are seldom used alone and in most cases are used in combination with radioisotopes. Since dyes are taken up by lymph ß ow almost immediately after injection into the tissues, intraoperative observation under direct vision is required. When LNs are destroyed by surgical manipulations, the dye will ß ow out of the LN making accurate identiÞ cation of the SN impossible. In surface cancers such as breast cancer and malignant melanoma this dye method is extremely useful, whereas in prostate cancer identiÞ cation of the SN is extremely difÞ cult using this dye. This is because when the peritoneum is not detached from the pelvic ß oor LN stained with the dye cannot be veriÞ ed visually. To achieve this, the peritoneum should be detached from the pelvic ß oor beforehand, although it is difÞ cult to completely avoid damaging the LNs during this procedure. For this reason, use of radioisotopes is indicated for the identiÞ cation of SN in prostate cancer.
Colloid particle diameter is an important factor determining the speed at which the tracer passes through the LN and the degree of entrapment of the colloid within the LN. [26] According to the type of colloid used as well as the timing of SN detection after tracer administration, the number of SN that can be pointed out will vary, and thus the number and distribution of SN will differ according to the type of colloid used and the individual investigator. Tracers are usually injected transrectally to the prostate, with one injection each to the left and right lobes in most cases. The administered dose also differs according to the individual investigator with the total dose usually in the range of about 0.2 cc to 3 cc. Our group injects tracer ( 99m Tc-phytate: 40 MBq/0.1 ml/site) under transrectal ultrasonography 5-6 h prior to surgery into the right and left peripheral zones of the prostate.

Identification of SN
Here, we introduce our SN identiÞ cation method using 99m Tc-phytate. [27] One hundred and eighty minutes after tracer injection lymphoscintigraphy is performed, and LN in which radioisotopes have been taken up, namely SN, are depicted on the Þ lm. The number of SN will vary according to the time elapsing from the time of tracer injection to imaging, with the optimal time until imaging differing according to the type of colloid used.
Lymphoscintigrams can be displayed as planar images or single photon emission computed tomographs (SPECT), but SPECT imaging is preferable to discern the threedimensional location of SN. Moreover if CT-SPECT fusion images are prepared the anatomical position of SN will be even better appreciated [ Figure 1]. Three sites including the umbilicus are marked on CT and SPECT, and a fusion image is obtained by overlapping the marked sites of the two images on a computer. [28,29] At the time of intraoperative probing the information provided by this fusion image facilitates more accurate intraoperative detection of SN, and reduces the operating time for fusion-image supported SN resection. [30] Since validation of the SN concept requires knowledge of the state of not only SN but also non-SN metastases, backup extended dissection is usually performed after SN biopsy. Jeschke et al., perform SN biopsy and extended   dissection laparoscopically, and to avoid the shine-through phenomenon, use a gamma probe that can detect gamma rays from 90 degrees laterally. [31] Although a strict learning curve exists, laparoscopic extended pelvic lymphadenectomy is worth performing for validation of the SN concept because this procedure yields an equivalent number of resected LNs with less morbidity. [32] Since in most other reports, validation has been performed under laparotomy, we introduce the method adopted by us here.
In short, through a lower abdominal midline incision the LN groups below the aortic bifurcation are located, and the SN is identiÞ ed with a gamma probe. The gamma probe used intraoperatively is usually a handy type weighing about 200-300 g, which measures gamma rays released from the radionuclide at the tip detector portion [ Figure 2]. LN identiÞ ed by the gamma probe as SN are sampled, after which backup extended dissection is conducted, with the right and left common iliac vessels, external iliac vessels, internal iliac vessels, obturator fossa and presacral LN groups all resected separately. On a back table the LNs are mapped separately, and the SN more accurately identiÞ ed [ Figure  3]. The prostate injected with tracer emits high levels of radioactivity, which interfere with the detection of weak gamma rays from the SN, with this referred to as the shinethrough phenomenon. To prevent the occurrence of this phenomenon, the prostate is covered with a lead shield.

Pathologic diagnosis
The pathologic method used to diagnose resected LN differs with the individual investigator. In Wawroschek et al.'s method, the resected SN is cut into 2-mm slices, with Þ ve serial sections prepared from each slice. One of these slices is subjected to immunohistochemical staining using anti-cytokeratin antibody, and the remaining slices to routine hematoxylin and eosin staining. [33] To achieve a more detailed diagnosis of metastases we prepare serial sections from slices cut at 250 micrometer thickness, and like Wawroschek et al., perform immunohistochemical and H and E staining. Although the deÞ nition of micrometastases and their biological signiÞ cance remain controversial, it is recognized that the diagnostic efÞ cacy rate can be enhanced by 10-30% by adding immunohistochemical to H and E staining. [34]

RESULTS
As noted above, the results of the largest scale study conducted by Wawroschek   was added. On the other hand, in the group in which NHT was given (n = 27), four cases were diagnosed as metastasis positive by routine pathologic examination (14.8%), with this number rising by Þ ve to a total of nine (33.3%) when micrometastasis diagnosis was added [ Table 2]. In other words, these results imply that had NHT been given to the former group, micrometastases would have been overlooked in Þ ve cases. Thus, despite the fact that the deÞ nition of micrometastasis is unclear and their inß uence on prognosis is unknown, NHT was shown to decrease the diagnostic efÞ cacy rate of LN metastases.

FUTURE PROSPECTS
Based on the report of Wawroschek et al., and our own results, if the SN is metastasis-free the patient can be deemed to be more likely negative for LN metastases. In this way, the SN concept can be considered to be valid in prostate cancer as well. In contrast, when metastases are present in the SN, it is thought that not all LN metastases can be resected with SN biopsy alone since in numerous cases affected LN will be located beyond the range of standard dissection, and moreover, metastases may be present in non-SN as well. However, all hitherto conducted research on SN in prostate cancer has been investigative in nature and undertaken at single institutions. In future, to promote research in this Þ eld and to aim for widespread clinical application various problems will Þ rst have to be overcome. First, multiinstitutional studies will be needed, and at a minimum there will have to be standardization of various inß uencing factors such as the type of tracer used, length of the period from tracer administration to SN biopsy, and extent of the dissection for the sake of veriÞ cation. In future, in addition to technical improvements and accumulation of experience, the planning of well-designed collaborative studies and randomized comparative trials will be needed to achieve further advances in research on SN in prostate cancer.
The primary goal of SN research is to be able to make available to patients the least invasive surgical intervention possible. In future, it would be desirable to realize SNNS with a less invasive method. A representative low-invasive method currently applicable is laparoscopic surgery. Our group has adopted endoscopy-assisted minilaparotomy, in which a small 5cm lower abdominal midline incision is made, through which all surgical tools including the endoscope are inserted and SN biopsy and RPx performed. [35] To perform SN biopsy through this small hole, it is useful to obtain detailed information beforehand about the anatomical position of the SN from the above-mentioned SPECT-CT fusion image. For SNNS to be widely applied clinically many research issues remain to be addressed, but various technical problems have already been overcome and have facilitated the adoption of less invasive surgical methods. Further advances in research are awaited in the near future.