Clinical efficacy and problems with CT lymphography in identifying the sentinel node in breast cancer

Sentinel node biopsy (SNB) has become a standard surgical procedure for patients with early-stage breast cancer [1, 2]. The sentinel nodes (SN) can be identified by a double-mapping procedure based on a gamma probe-guided method and a dye-guided method using a radioisotope (RI), or by a triple-mapping procedure that includes lymphoscintigraphy and is even more effective [1‐3]. However, the RI method can be performed only at institutions that are trained and licensed to use RI, and other institutions must rely on dye methods alone for SN identification [4‐8] On the other hand, for obtaining images of the lymph vessels and nodes, indirect lymphography seems to be a more convenient than direct intralymphatic administration of a contrast medium. Several studies of indirect lymphography were reported in the 1980s [9‐12], and Suga et al., reported successfully identifying SN by three-dimensional computed tomographic lymphography (CTLG) using a nonionic contrast medium [6.7]. We have also been identifying SN in breast cancer patients by CTLG and a dye-guided method since February 2003 as a clinical trial [5]. Here, we report our findings to date regarding the clinical efficacy and problems associated with SN identification by the combination of CTLG and a dye-guided method.

All studies in this paper were approved by the ethics committee of National Higashi-Tokushima Hospital. After presenting a detailed explanation of this clinical trial, written informed consent was obtained from all patients.

In breast cancer surgery, dissection of the axillary lymph nodes is considered effective for the objectives of performing staging and achieving local control [1, 13]. Therefore, in recent years SNB has become a standard procedure for patients with no metastasis of the axillary lymph nodes since it avoids unnecessary dissection [1, 2]. Combined use of an RI (gamma probe-guided method and lymphoscintigraphy) and a dye is currently considered to be more efficient for identification of the SN(s) than single use of either of these methods [1‐3]. In Japan, there are many institutions that do not have the necessary facilities for using RIs and thus must use the dye-guided method alone to identify SNs [4, 5]. On the other hand, Suga et al. proposed using a method called CTLG, which employs a nonionic contrast medium [6, 7], and we have applied that method for SN identification since February 2003 [5]. In this paper we have reported our findings regarding the usefulness and problems associated with combined application of CTLG and the dye-guided method for SN identification.

Our experimental subjects were 219 Japanese patients with primary breast cancer that was thought to be clinically free of axillary lymph node metastasis. CTLG and the dye-guided method were employed in an attempt to identify the SN(s) in each of the patients. The SN identification rates were 96% with CTLG alone, 92% with the dye-guided method alone and 99% when the findings with the two methods were combined. Those results are good when compared with the published data for combined use of an RI and the dye-guided method [2, 3]. In particular, our SN identification rate with the dye-guided method is better than the rates that have been reported to date [2, 3, 14]. We think that this is because the location of the SN had already been determined by the CTLG, making it easy to identify the lymph vessels and lymph nodes. The breakdown of the identification by CTLG showed clear enhancing of the lymph nodes in 88% of the patients, while in 5% of the patients identification of the SN was possible on the basis of the CT value in spite of the fact that the lymph nodes were not clearly enhanced. Because the observations were macroscopic, when the dye-guided method was used alone, SN identification was difficult unless the lymph node was stained to a sufficient degree. With CTLG, on the other hand, even if the enhancement is not very striking it can be surmised that the SN identification rate will be improved since the CT value can be taken into consideration.

In 14 patients, the SN was identified by CTLG, but not by the dye-guided method. The location of the SN was confirmed by ultrasonography just before the operation in all of these patients. In these 14 patients, not only the SN but also neighboring nodes were sampled at the same time. Therefore, it seemed that the SN biopsies had been performed accurately.

The sensitivity of CTLG could not be investigated in our patient population because backup axillary dissection was not performed for patients who were metastasis-negative in the SNB. In addition, none of the patients experienced postoperative recurrence in the axillary lymph nodes, but that does not serve as a basis for claiming that there were few false-negatives [1]. However, Tangoku et al., reported the sensitivity of CTLG to be 98%, and for that reason it can be thought that the sensitivity of CTLG is not inferior to that of the RI method [15].

With regard to correlations between the clinicopathological findings and the SN identification rate, the patient age, the location of the tumor, whether or not surgical biopsy was performed, the tumor size and the presence/absence of lymph node metastasis have been reported to influence identification of the SN [1, 16‐18]. In our patient series, the results showed no differences as a function of the age, menopausal status, tumor diameter or histopathological type. However, with the dye-guided method the SN identification rate was significantly lower in patients with a BMI of 25 or higher, whereas with CTLG and the combined method the SN identification rate was not influenced by the BMI. As the reason for this difference it is noted that with the dye-guided method it can be difficult to discern staining of lymph vessels and nodes with the naked eye if considerable subcutaneous fat is present, whereas with CTLG the subcutaneous fat plays no role since the observation is done by CT. In addition to the BMI, discrepancies between the SN identification rates with these three methods were observed as a function of the presence/absence of lymph node metastasis, the site of excisional biopsy (i.e., the lateral upper region and regions other than the lateral upper region), and the presence/absence of vascular invasion. Especially, the identification rates with CTLG and the combined method were significantly lower in node-positive patients compared to node-negative patients, and significantly lower with the combined method in vascular invasion-positive patients compared to negative patients. The size of injected particles, the injected volume and the injection site have been reported as examination factors that can influence SN identification [2]. The nonionic contrast media that are used in CTLG have a larger particle size than dyes [19]. For that reason it can be hypothesized that, in patients with lymph node metastasis or vascular invasion and in patients who had undergone lateral upper region excisional biopsy, who can be predicted to have lymph vessel occlusion, the movement of a nonionic contrast medium would be impeded compared with that of a dye, thus resulting in a lower SN identification rate. On the other hand, the SN identification rate with the dye-guided method was lower in the patients with excisional biopsy in any region than in patients without excisional biopsy. It can be hypothesized that this is probably because excisional biopsy leads to edema of the connective tissue, which makes it difficult to distinguish the dye.

We also investigated the number of SNs and the number of lymph vessels leading to them [20‐22]. These evaluations are difficult to achieve by the gamma probe-guided method and the dye-guided method, but CTLG permits detailed investigation. Our results showed that there was a single route leading to a single SN in 68% of the patients, multiple routes leading to a single SN in 4%, a single route leading to multiple SNs in 10% and multiple routes leading to multiple SNs in 8%. These results are in agreement with those reported by Tangoku et al., [23]. Thus, approximately 18% of the patients in our series had multiple SNs, and it can be hypothesized that it would be difficult to biopsy all of them if only the dye-guided method were employed. We think that combined use of CTLG with the dye-guided method would permit accurate biopsy, and for that reason we anticipate that combined performance of CTLG will prove useful.

CTLG is a diagnostic test that is performed prior to surgery for breast cancer. Studies are warranted to determine whether the CT findings or performance of fine-needle aspiration cytology of the SN will make it possible to achieve diagnosis of SN metastasis preoperatively and then decide whether or not SNB should be performed [24]. Such diagnosis leading to avoidance of unnecessary surgical procedures would represent a great clinical advantage by reducing the burden on the patient. Moreover, in the future it will be necessary to compare the usefulness of CTLG with the RI method.

Combined performance of CTLG with the dye-guided method permits better elucidation of the location of the SN(s) in breast cancer. This makes it easier to identify the SN(s) and results in a higher SN identification rate compared with application of the dye-guided method alone. Combination of CTLG and the dye-guided method was especially useful in obese patients. In addition, in patients with multiple SNs, it is advantageous to be able to perform SNB accurately. However, in patients with occlusion of the lymph vessels due to lymph node metastasis, having undergone lateral upper region lumpectomy or the presence of vascular invasion, there is a possibility that false-negative diagnostic results will be generated with CTLG.