Introduction

Sentinel node (SN) biopsy is a minimally invasive surgical approach to evaluate lymph node (LN) staging in clinically node-negative patients. The rationale of this technique is that, since SNs are the LNs with direct lymphatic drainage from the primary tumor and thus represent the first LNs receiving lymph-borne metastatic cells, their tumor status may predict whether the overall LNs in that lymphatic basin contain cancer or not. If the SN turns out tumor free on histopathological examination, all other LNs within the draining lymphatic basin are also assumed free of tumor, avoiding an unnecessary full lymphadenectomy and associated morbidity. Up to now, the most widely used radiopharmaceuticals for SN mapping in breast cancer, melanoma, and head and neck cancer have been 99mTc-labeled sulfur colloid in the USA and 99mTc-nanocolloidal albumin in Europe [1,2,3,4]. Compared to blue dye or fluorescent agents, the main strength of these radiopharmaceuticals is the possibility to perform preoperative lymphatic mapping, allowing a personalized surgical planning, and reducing surgical failures. The associated radiation dose of this technique is negligible at the patient level. The main drawbacks are the passive transport through the lymphatics, a relative slow clearance from the injection site resulting in a delayed migration to LNs, the pass-through to second-echelon LNs, and a possible masking of SNs that are close to the primary tumor (injection site). These limitations explain why there is a trend towards a novel radiopharmaceutical, 99mTc-tilmanocept, which contains relatively small macromolecules (molecular weight of 19,000 Da, diameter of 7 nm) that enable a rapid clearance from the injection site, resulting in a quick uptake in SNs and a more accurate detection of SNs close to the primary tumor. 99mTc-tilmanocept selectively binds to mannose receptors (CD206) expressed on the surface of LN macrophages and dendritic cells [5], thus reducing its migration to higher echelon nodes. 99mTc-tilmanocept was first used in clinical practice in breast cancer patients [6].

In the current systematic review and meta-analysis, we aimed to provide an update from a European expert panel on the performance of 99mTc-tilmanocept for SN mapping in breast cancer, melanoma, and head and neck cancer.

Materials and methods

Search strategy

A systematic literature search of the PubMed/MEDLINE and Embase databases was performed using the Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) guidelines [7]. The search string was built using synonyms of “Sentinel Lymph Node” and “99mTc-diethylene-triamine-pentaacetic acid (DTPA)-mannosyl-dextran,” including corresponding Mesh terms (Online Resource 1). The search was last updated on August 18, 2022. Search results were not limited by publication date.

Study selection

All retrieved articles were initially independently screened for eligibility by title and abstract by two authors (G.R. and A.C.). Full texts of potentially eligible studies meeting all inclusion criteria were retrieved for further evaluation. Original articles investigating the Food and Drug Administration (FDA)– and European Medicine Agency (EMA)–approved use of 99mTc-tilmanocept for SN identification in oncological patients in any tumor type were eligible for inclusion. Review articles, letters to the editor, editorials, and case reports were excluded. Articles not written in English were also excluded. A cross-reference check was performed. Any disagreements were resolved by discussion.

Quality assessment

The methodological quality of the included studies was evaluated using the Quality Assessment of Diagnostic Accuracy Studies tool version 2 (QUADAS-2) [8] to grade the risk of bias and methodological applicability on four key domains: patient selection, index test, reference standard, and flow and timing. Two authors (G.R. and E.J.d.K.) independently assessed the quality of the studies. Any disagreements were solved by consensus after re-evaluation and discussion of the respective references. The QUADAS-2 scores for all included studies were tabulated and a summary report was constructed.

Data extraction

The 99mTc-tilmanocept performance for SN mapping in breast cancer, melanoma, and head and neck cancer was evaluated by conducting separate meta-analyses per tumor type. For each study, appropriate data were extracted for both 99mTc-tilmanocept and the reference standard technique, including the number of patients, preoperative detection rate (DR), intraoperative DR, and the proportion of patients with LN metastasis at histopathology after dissection and/or follow-up (pN +) that were correctly staged by SN biopsy with 99mTc-tilmanocept (pN + sensitivity or SN + /pN + ratio). The preoperative and intraoperative DRs were defined as the percentage of patients in whom at least one SN was detected at lymphoscintigraphy or during surgery, respectively. In case relevant data were missing from studies to obtain these parameters, the corresponding authors were contacted through email.

According to data availability (Tables 1, 2, and 3), the estimated performance outcomes varied among the three main tumor types. For breast cancer, the 99mTc-tilmanocept pooled per-patient preoperative DR was estimated, as well as the per-patient and per-lesion (in case of bilateral disease) intraoperative DRs. Moreover, the per-patient intraoperative 99mTc-sulfur colloid DR was also estimated by pooling studies that used 99mTc-sulfur colloid as reference standard. For melanoma and head and neck cancer patients, the pooled preoperative and intraoperative 99mTc-tilmanocept DRs were estimated, while insufficient data were available regarding reference standard techniques. Finally, the pooled proportion of pN + patients that were correctly staged by 99mTc-tilmanocept SN biopsy (SN + /pN + ratio) was estimated for melanoma, whereas no sufficient data were available for breast cancer and head and neck cancer.

Table 1 Characteristics of the included studies on breast cancer
Full size table
Table 2 Characteristics of the included studies on melanoma
Full size table
Table 3 Characteristics of the included studies on head and neck cancer
Full size table

Statistical analysis

Meta-analysis was performed using Stata/MP, version 14.2 (StataCorp LP, College Station, TX) [9]. The metaprop command in Stata/MP and random-effects modeling were used to estimate the pooled pre-/intraoperative DRs and/or the pooled pN + sensitivity (i.e., SN + /pN + patients ratio) with corresponding 95% confidence intervals (CIs). Pooled results are presented in forest plots. Heterogeneity between studies was assessed by visual inspection of the forest plots and estimated using the inconsistency index I2.

Results

Literature search

The systematic literature search yielded a total of 174 articles. After removal of duplicates, screening of titles and abstracts resulted in 24 potentially eligible studies. After full-text retrieval and evaluation, no further studies were excluded. No additional studies were identified at cross-reference checking. As such, 24 articles were included (Fig. 1) [6, 10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32].

Fig. 1
figure 1

Flow chart showing the search strategy

Full size image

Quality assessment

The results of the QUADAS-2 assessment are shown in Supplemental Tables 1-3 (Online Resources 2–4). Studies mainly showed a low risk of bias and few concerns regarding applicability in the first two key domains (i.e., patient selection and index test). In contrast, some studies did not include 99mTc-sulfur colloid or 99mTc-nanocolloidal albumin and/or blue dye as reference standard in their design and were therefore judged as having a high risk of bias and/or applicability concerns in the last two key domains (i.e., reference standard, flow and timing) (see Supplemental Tables 1-3). Taking this risk of bias into account, these studies were included in the meta-analysis exclusively for the estimation of pooled endpoints regarding 99mTc-tilmanocept alone (pre-/intraoperative detection rates, pN + sensitivity), while their data could not be used for a pooled comparative analysis due to the absence of a reference standard. Considering all the above, the included studies were deemed of sufficient methodologic quality to be included in the systematic review and meta-analysis.

Study characteristics

The main characteristics of the included studies are summarized in Tables 1, 2, and 3. Fourteen studies had a prospective study design [6, 10,11,12,13,14,15,16, 18, 19, 21, 25, 31, 32], nine were retrospective [17, 20, 22,23,24, 26,27,28, 30], and one was both prospective and retrospective [29]. Most studies used both 99mTc-tilmanocept and blue dye for the SN biopsy procedure. Of the 12 studies that also compared 99mTc-tilmanocept performance with a reference standard mapping technique, ten used 9mTc-sulfur colloid [6, 10,11,12, 17, 21, 23, 24, 30, 32] while only two used 99mTc-nanocolloidal albumin [25, 29].

Twenty-one of 24 studies provided data on the pre-/intraoperative DR and/or the pN + sensitivity (SN + /pN + patients ratio) and were included in the meta-analysis [6, 10,11,12,13,14,15,16,17,18, 20,21,22,23,24, 26,27,28,29, 31, 32].

Breast cancer

Fourteen studies investigated the use of 99mTc-tilmanocept to identify SNs in breast cancer (Table 1) [6, 11,12,13, 15, 17, 19,20,21,22, 24, 26, 28, 32]. In five studies (n = 111 patients), the preoperative DR at a patient level varied between 76 and 100% [13, 20, 21, 24, 28]. In 11 studies (n = 1000 patients), the intraoperative DR [6, 11,12,13, 15, 17, 20,21,22, 28, 32] was consistently ≥ 99% with a single exception of a 94% DR reported by Leong et al. [13].

At the lesion level, the preoperative DR (two studies) [26, 28] and the intraoperative DR (five studies) [6, 15, 24, 26, 28] were higher than 93% and 97%, respectively. Eight studies (446 patients) compared 99mTc-tilmanocept SN biopsy performance with 99mTc-sulfur colloid [6, 11, 12, 17, 19, 21, 24, 32]. Both radiopharmaceuticals showed comparable intraoperative DR, close to 100% in most cases.

In more detail, Rietbergen et al. reported a preoperative and intraoperative SN DR of 100% in 13 patients (n = 14 tumors) [28]. At delayed images (2 h), the number of SNs was concordant with early images (15 min), but the injection site/SN ratio was 72% lower. Uptake in non-SNs was found in 7 of 15 (47%) basins, with an average SN/non-SN ratio of 6.6 (range: 2.3–15.6). A multicenter study conducted by Vidal-Sicart et al. in 344 patients (n = 352 lesions) reported per-lesion preoperative and intraoperative SN DRs for 99mTc-tilmanocept of 92.9% and 97.2%, respectively. Moreover, 80 of 81 metastatic SNs were detected, with no axillary recurrences during follow-up (mean 19 months, range 16–25) [26]. In previous studies by Unkart and Wallace et al., the preoperative and intraoperative DRs were 76.5% and 100%, respectively [20], with a comparable mean number of removed SNs between the 1-day and the 2-day injection protocols (3.0 ± 1.9 vs 2.7 ± 1.4, respectively) [22]. Similar results were also reported by Leong et al., who also described a relative difference between the 1-day and the 2-day protocol (92.6% vs 83.3%) possibly attributed to the small sample size of the 2-day group [13].

Regarding the comparison between 99mTc-tilmanocept and 99mTc-sulfur colloid, Ferez-Pinzon et al. reported no significant differences in the number of harvested SNs (2.0 vs 1.8, respectively) and in the time to transcutaneous localization (3.3 ± 2.0 vs 3.9 ± 2.3 min) in 86 patients [32]. Similarly, Murphy et al. found no statistical difference in harvested SNs, number of identified pN + patients, or pain scores after intradermal injection [24] in a cohort of 76 patients. Previously, Unkart et al. reported the first randomized, double-blinded trial, showing a comparable preoperative DR (23/25 vs 26/27 patients) and number of harvested SNs (54 vs 61) [21], but reduced injection pain for 99mTc-tilmanocept [19]. On the contrary, in the cohort of 199 patients by Baker et al., 99mTc-tilmanocept led to fewer removed SNs per patient (mean 1.85 ± 0.78 vs 3.24 ± 1.62, p < 0.001), while also showing a higher sensitivity for nodal metastasis [17] and an equal intraoperative SN DR (100%). A previous study by Wallace et al., using a 2-day SN mapping protocol, found 99mTc-tilmanocept to have a significantly faster injection site clearance rate than 99mTc-sulfur colloid (mean clearance half-time: 2.18 ± 1.09 h vs 57.4 ± 92.8 h, p < 0.001), but an equivalent mean SN uptake and intraoperative DR (4/4 vs 3/5 patients, respectively) [12]. Similar results have been reported in two previous studies by Wallace et al. in 12 and 10 patients, respectively, using a 1-day protocol [6, 11]. Finally, Wallace et al. also compared 99mTc-tilmanocept to vital blue dye, showing the higher intraoperative DR (98.6% vs 88.5%, p < 0.0001) and greater metastatic nodal sensitivity (31/33 vs 25/33 LNs, p = 0.03) of the former [15].

Melanoma

Seven studies investigated the use of 99mTc-tilmanocept in melanoma patients (Table 2) [10, 13, 14, 23, 28, 30, 31]. The preoperative and intraoperative DRs ranged from 97 to 100%. Three studies compared 99mTc-tilmanocept with 99mTc-sulfur colloid, with preoperative and intraoperative DRs ranging between 96 and 100% for both radiotracers [10, 23, 30].

More in depth, both Rietbergen et al. and Mwagiru et al. reported a 99mTc-tilmanocept preoperative and intraoperative SN DR of 100% in 26 and 15 patients, respectively [28, 31]. Rietbergen et al. also found non-SN uptake in 11 of 20 (55%) basins, with a mean SN/non-SN ratio of 6.7 (range 3.0–15.4). Moreover, at delayed images (2 h) the injection site/SN ratio per-LN basin was 61% lower compared to early images (15 min). Comparable results were also reported by Leong et al., with a 98% preoperative and intraoperative DR, a pN + sensitivity of 91.7%, no relevant adverse events, and no significant difference between the same-day and next-day surgery protocols [13].

Regarding the comparison between 99mTc-tilmanocept and 99mTc-sulfur colloid, Eckhoff et al. (n = 391 patients) found significantly shorter lymphoscintigraphy mapping times for 99mTc-tilmanocept (51.8 vs 195.1 min, p < 0.01), but no significant differences in terms of SNs identified at lymphoscintigraphy (3.56 vs 3.28), removed SNs (3.05 vs 3.1), percentage of metastatic patients (11.3% vs 17.4%), and false negative rates (25% vs 22%) at a median follow-up of 6 months and 25 months for 99mTc-tilmanocept and 99mTc-sulfur colloid, respectively [30]. In a previous study by Silvestri et al. (n = 370 patients), 99mTc-tilmanocept and 99mTc-sulfur colloid showed preoperative DRs greater than 96%, intraoperative DRs of 100%, and no significant differences in sensitivity for nodal metastasis (23/25 vs 28/31 patients) and false negative cases (1.0% vs 1.6%) at a median follow-up of 5.4 months and 25.2 months, respectively [23]. However, 99mTc-tilmanocept led to lower radiation doses (503 vs 580 µCi, p < 0.0001), shorter mapping times (31 vs 34 min, p = 0.008), and fewer excised SNs (2.5 vs 2.9, p = 0.04) compared to 99mTc-sulfur colloids. Similarly, in a phase I clinical trial by Wallace et al. (n = 24 patients), 99mTc-tilmanocept exhibited a significantly faster injection site clearance rate at all dose levels (1.0, 5.0, or 10.0 nmol) (mean clearance half-time: 2.17 ± 0.96 h vs 14.7 ± 6.3 h, p < 0.001) and a comparable mean SN uptake (0.73% ± 0.94% vs 0.85% ± 1.19%), but a lower number of SNs visualized per basin (1.6 vs 1.9) than 99mTc-sulfur colloids [10]. Both tracers showed an intraoperative DR of 100%, with lower radiation absorbed doses for 99mTc-tilmanocept and no relevant adverse events. Finally, a study by Sondak et al. compared 99mTc-tilmanocept to blue dye showing higher intraoperative DR (97% vs 90%, p = 0.002), per-LN concordance (98.7% vs 63.7%, p < 0.001), per-patient concordance (97.8% vs 50.7%, p < 0.001), and number of identified metastatic SNs (45/45 vs 36/45, p = 0.004) [14].

Head and neck cancer

Six studies investigated the use of 99mTc-tilmanocept in patients with head and neck cancer (Table 3) [16, 18, 25, 27, 29, 31]. Besides the 97% preoperative DR reported by Mahieu et al. [29], four studies reported a DR of 100% [16, 25, 27, 29, 31]. Similarly, the intraoperative DR was 100% in five studies [16, 25, 27, 29, 31], with a slightly lower value of 98% reported by Agrawal et al. [18]. Only two studies compared 99mTc-tilmanocept to 99mTc-nanocolloid, showing a similar performance for both tracers [25, 29].

In more detail, a prospective cross-sectional study by Mwagiru et al. in oral squamous cell carcinoma (OSCC) patients (n = 9) reported a preoperative and intraoperative DR of 100% [31]. Equal DRs were also described by Doll et al. in floor of mouth (n = 6), tongue (n = 5), and upper alveolar crest/hard palate (n = 2) tumors [27], with no recurrences in SN-negative patients during follow-up (mean 20.3 months, range 10–28). Similar results were obtained by Agrawal et al. in intraoral (n = 78) and cutaneous (n = 5) squamous cell carcinoma undergoing elective neck dissection [18]. 99mTc-tilmanocept showed an intraoperative DR of 97.6%, a false negative rate of 2.56%, a negative predictive value of 97.8%, and an overall accuracy of 98.8%, with no significant differences related to the injection timing (1-day vs 2-day protocol). Notably, in the subcohort of patients with floor of mouth tumors (n = 20), 99mTc-tilmanocept DR and staging accuracy were both 100% (12/12 pN + patients identified), despite the higher risk of shine-through phenomenon and false negative results of this location. Comparable findings were reported by Marcinow et al. in 20 patients with oral cavity squamous cell carcinoma undergoing elective neck dissection [16]. All 12 pN + patients were correctly staged by 99mTc-tilmanocept, resulting in a false negative rate of 0% and a negative predictive value of 100%. Finally, the SPECT/CT images allowed to identify additional SNs in 11 of 20 (55%) cases, despite not reaching statistical significance compared to planar scintigraphy.

Regarding the comparison between 99mTc-tilmanocept and 99mTc-nanocolloid, den Toom et al. [25] described a faster injection site clearance rate for 99mTc-tilmanocept in 20 OSCC patients (remaining activity 29.9% vs 60.9%, p < 0.001) despite a lower SN uptake (1.95% vs 3.16%, p = 0.01), thus resulting in no significant difference in SN/injection site ratio (0.066 vs 0.054) and median number of detected SNs (3.0 vs 2.5) and higher echelon nodes (2.0 vs 2.5, respectively). The prospective cohort of den Toom et al. [25] was later combined by Mahieu et al. with a retrospective cohort (n = 10 patients investigated only with 99mTc-tilmanocept), showing a 99mTc-tilmanocept sensitivity of 83.3% and a negative predictive value of 93.3% (average follow-up: 22.6 months), but a staging accuracy comparable to 99mTc-nanocolloid [29].

Meta-analysis

The estimated pooled per-patient preoperative and intraoperative DRs of 99mTc-tilmanocept were 0.94 (95%CI, 0.88–1.01) and 0.99 (0.98–1.00) for breast cancer (Figs. 2 and 3), 0.98 (0.96–0.99) and 1.00 (0.99–1.00) for melanoma (Figs. 4 and 5), and 0.97 (0.93–1.02) and 0.99 (0.96–1.01) for head and neck carcinoma (Figs. 6 and 7). In breast cancer, the per-lesion pooled intraoperative DR of 99mTc-tilmanocept was 0.98 (0.97–0.99) (Fig. 8), while the per-patient intraoperative DR of 99mTc-sulfur colloid was 1.00 (95% CI 0.98–1.02) (Fig. 9). Finally, the staging accuracy of 99mTc-tilmanocept SN biopsy in melanoma patients was also estimated by means of the pooled pN + sensitivity (i.e., SN + /pN + patients ratio), which was 0.97 (95% CI, 0.92–1.03) (Fig. 10).

Fig. 2
figure 2

Breast cancer—99mTc-tilmanocept preoperative sentinel node detection rate (per patient)

Full size image
Fig. 3
figure 3

Breast cancer—99mTc-tilmanocept intraoperative sentinel node detection rate (per patient)

Full size image
Fig. 4
figure 4

Melanoma—99mTc-tilmanocept preoperative sentinel node detection rate

Full size image
Fig. 5
figure 5

Melanoma—99mTc-tilmanocept intraoperative sentinel node detection rate

Full size image
Fig. 6
figure 6

Head and neck cancer—99mTc-tilmanocept preoperative sentinel node detection rate

Full size image
Fig. 7
figure 7

Head and neck cancer—99mTc-tilmanocept intraoperative sentinel node detection rate

Full size image
Fig. 8
figure 8

Breast cancer—99mTc-tilmanocept intraoperative sentinel node detection rate (per lesion)

Full size image
Fig. 9
figure 9

Breast cancer—99mTc-sulfur colloid intraoperative sentinel node detection rate (per patient)

Full size image
Fig. 10
figure 10

Melanoma—sensitivity of 99mTc-tilmanocept sentinel node (SN) biopsy for pathology-positive (pN +) patients (SN + / pN + patients ratio)

Full size image

Discussion

Although multiple radiopharmaceuticals are available for SN mapping, there is currently a particular interest in 99mTc-tilmanocept due to its favorable characteristics, including a low molecular weight and a specific binding capability for the mannose receptors (CD 206) located on the surface of lymphatic reticuloendothelial cells. Today, 99mTc-tilmanocept is approved by the FDA and EMA for the use in breast cancer, melanoma, and head and neck cancer [33, 34]. On the basis of the published evidence and a consensus of European experts, an updated overview on 99mTc-tilmanocept performance for SN mapping in breast cancer, melanoma, and head and neck cancer is provided.

In breast cancer and melanoma, 99mTc-tilmanocept and 99mTc-sulfur colloid have shown comparable pre- and/or intraoperative DRs [6, 10,11,12, 17, 19, 21, 23, 24, 30, 32]. There are, however, clinically relevant differences to consider.

First, in a study by Baker et al. on breast cancer, 99mTc-tilmanocept allowed to remove fewer SNs per patient (mean 1.85 vs 3.24) while still detecting a higher proportion of metastatic SNs among pN + patients than 99mTc-sulfur colloid [17]. This could possibly lead to a clinical benefit through the combination of improved oncological outcomes and potentially lower morbidity rates. Similar results were obtained in melanoma patients by Wallace et al. [10] and Silvestri et al. [23], who reported a lower number of SNs visualized at lymphoscintigraphy and fewer excised SNs during surgery, with comparable clinical outcomes. However, some of the later studies in breast cancer and melanoma were not able to confirm these results [21, 24, 30].

One possible reason behind the removal of fewer SNs is 99mTc-tilmanocept specificity for CD206 receptors, which allows a prolonged binding of the tracer to the first LN, thus limiting its migration to distal LNs. Furthermore, the potential benefits of removing fewer SNs include not only a decreased patient morbidity, but also a reduction in the operative time and the pathology costs of the procedure.

Multiple studies in breast cancer and melanoma also reported a faster injection site clearance rate for 99mTc-tilmanocept compared to 99mTc-sulfur colloid [6, 10,11,12]. The faster clearance of 99mTc-tilmanocept from the injection site may be explained by its smaller molecular size (diameter of 7 nm) compared to 99mTc-sulfur colloid (between 100 and 200 nm). This may benefit the preoperative SN visualization, particularly when the primary tumor (the injection site) is close to the SN(s); moreover, we strongly advise the use of 99mTc-tilmanocept in conjunction with single photon emission computed tomography/computed tomography (SPECT/CT) [35, 36].

Another potential advantage of 99mTc-tilmanocept over 99mTc-sulfur colloid could be the reduction of the injection site pain. However, currently available evidence is limited and discordant: while Murphy et al. [24] reported comparable pain scores between tracers, Unkart et al. [21] found 99mTc-sulfur colloid to be correlated with significantly more pain during the first 3 min post-injection.

A possible explanation of the reduced pain associated with 99mTc-tilmanocept injection could be the small particle size that allows for a fast clearance from the injection site and a reduced stretch on the nociceptive pain receptors in the dermis, as suggested by Unkart et al. [19]. Although a correlation between the post-injection site pain and the pH of the preparations could also represent a viable hypothesis, the study by Unkart et al. [19] found the pH values to be comparable between 99mTc-tilmanocept and filtered 99mTc-sulfur colloid. Moreover, a randomized double-blinded trial by Stojadinovic et al. reported that changing the pH of the 99mTc-sulfur colloid preparation by adding sodium bicarbonate did not significantly alter the post-injection pain scores [37].

Finally, the comparative performance of 99mTc-tilmanocept and 99mTc-sulfur colloid did not appear to be significantly affected by the injection-to-surgery time (1-day vs 2-day protocol) in both breast cancer [6, 11, 12] and melanoma [13]. This feature allows for greater flexibility in patient scheduling. In our opinion, the 2-day protocol is more suitable for centers with a high patient throughput as it allows surgeons to start the surgical procedures earlier in the morning.

Similarly to breast cancer and melanoma, 99mTc-tilmanocept showed a preoperative and intraoperative DR comparable to other radiocolloids (i.e., 99mTc-nanocolloidal albumin) in head and neck cancer [25, 29], as well as a significantly faster injection site clearance rate that could be particularly useful in floor of mouth tumors [25]. However, considerations regarding the benefits of one tracer over the other are less well described for head and neck cancer. In addition, currently there are no studies that compare 99mTc-tilmanocept with a hybrid tracer such as indocyanine (ICG)-99mTc-nanocolloid. Indeed, ICG-99mTc-nanocolloid combines radioactive and fluorescent guidance in a single injection, allowing both a preoperative lymphatic mapping and a better intraoperative visualization of SNs compared to blue dye [38,39,40]. These favorable characteristics of ICG-99mTc-nanocolloid could be especially advantageous in localizing SNs in case of a more complex drainage pattern and/or surgical anatomy such as in head and neck cancer. Despite the promising performance of 99mTc-tilmanocept compared to other tracers, currently the cost of this new radiotracer is significantly higher than that of 99mTc-nanocolloidal albumin or 99mTc-sulfur colloid, and proper cost-effectiveness analyses are still needed.

When assessing the performance of 99mTc-tilmanocept for SN biopsy, it is also critical to consider the pathological staging accuracy in terms of sensitivity for nodal metastasis, false negative rate, and negative predictive value. In breast cancer, the reported pN + sensitivity ranged between 89 and 99% in two studies, although the estimations were based on either a limited sample size or a short follow-up [13, 26]. In melanoma, the staging accuracy was consistently higher than 90% [13, 14, 23]. In head and neck cancer, the negative predictive value ranged between 93 and 100% [16, 18, 27, 29].

The reported variation of false negative rates in head and neck cancer (range 0–16.7%) is possibly due to the use of a different reference standard (elective neck dissection vs follow-up) [16, 18, 29]. Indeed, as pointed by Mahieu et al., neck dissection specimens can still harbor undetected micro-metastases in 15% of cases, and a composite reference standard (including also long-term follow-up) can improve the estimation accuracy by identifying more false negative cases [29].

Regarding the comparison of 99mTc-tilmanocept staging performance with other tracers, no significant differences (in terms of pN + sensitivity or false negative rate) were reported between 99mTc-tilmanocept and 99mTc-sulfur colloid in two melanoma studies; however, the shorter follow-up of the 99mTc-tilmanocept cohorts could have affected the false negative rate estimation accuracy [23, 30]. 99mTc-Tilmanocept was also compared with blue dye in another study on melanoma, showing a superior staging performance of the former [14]. Finally, one study on head and neck cancer compared 99mTc-tilmanocept with 99mTc-nanocolloidal albumin, showing a comparable staging accuracy [29]. Since the evidence on the comparative pathological staging accuracy is still limited, larger prospective randomized studies are still needed to better clarify this topic.

Our study has both strengths and limitations. To our knowledge, this is the first systematic review and meta-analysis on the performance of 99mTc-tilmanocept SN biopsy in staging LN involvement in breast cancer, melanoma, and head and neck cancer. In 2012, Tokin et al. published a meta-analysis based on 148 breast cancer patients injected with 99mTc-tilmanocept to compare its efficacy to that of 99mTc-nanocolloidal albumin [41].

The main limitations of the current study include the limited number, sample sizes, limited follow-up (possibly affecting the false negative rates), and heterogeneous endpoints of the available original studies. Consequently, many studies did not have enough clinical outcome data to be pooled in meta-analysis. Moreover, some studies did not include 99mTc-sulfur colloid or 99mTc-nanocolloidal albumin and/or blue dye as reference standard in their design, hindering the possibility to perform comparative analyses and potentially causing bias.

In conclusion, 99mTc-tilmanocept is a promising radiotracer for SN mapping in breast cancer, melanoma, or head and neck cancer. However, we strongly believe that further multicentric trials are still needed to investigate whether 99mTc-tilmanocept is superior to other radiotracers used in clinical routine. Finally, we support the use of 99mTc-tilmanocept in conjunction with SPECT/CT imaging, particularly for preoperative mapping of SNs close to the primary tumor (i.e., floor of mouth tumors).