Sentinel lymph node (SLN) biopsy is a minimally invasive procedure, but there is disagreement about its potential complications.

The American Society of Clinical Oncology (ASCO) and Society of Surgical Oncology (SSO) clinical practice guideline on the use of SLN biopsy in melanoma states that this technique may be considered for T1b melanomas (Breslow thickness, 0.8-1mm or <0.8mm with ulceration) after thorough discussion of risks and benefits with the patient. The guideline recommends the procedure for T2 and T3 melanomas (Breslow thickness, 1-4mm) and states that it may be recommended for T4 melanomas (Breslow thickness, >4mm), again after weighing up the risks and benefits with the patient. If the biopsy is positive, patients can be treated with complete lymph node dissection (CLND) or careful observation depending on the risk of micrometastatic disease, determined by an evaluation of clinical and pathologic factors.1

The European Society of Medical Oncology (ESMO) clinical practice guidelines recommend SLN biopsy for melanomas with a Breslow thickness of over 1mm or of over 0.75mm if there are additional risk factors, such as ulceration and a mitotic rate of 1 mitosis/mm2 (pT1b).2 In this case, the guidelines state that CLND must be considered if the biopsy is positive.

The above recommendation is made despite a lack of evidence showing the therapeutic value of SLN biopsy, followed or not by CLND, as the procedure has not been found to improve overall survival in patients with melanoma.3 Its usefulness therefore lies in its ability to provide prognostic information.

Although SLN biopsy causes less morbidity than CLND, its complication rates are still considerable. In addition, patients with a positive result were until very recently offered CLND and hence exposed to the risk of further complications.

Several studies in recent years have evaluated the therapeutic and prognostic value of SLN biopsy and its associated complications.4–13 Their results highlight variations in study design and complication rates ranging from 4.6%9 to 33%.6 There is thus no clear evidence on the true incidence of complications and sequelae associated with this technique.

The aim of this study was to describe the incidence of complications and sequelae in patients with melanoma who underwent SLN biopsy.

Our study shows that complications and sequelae are common in patients who undergo SLN biopsy.

Comparisons with other studies that have analyzed complications associated with SLN biopsy are difficult due to large differences. Most studies do not provide clear definitions of the complications analyzed, while others describe short-term or long-term complications or complications that require repeat surgery.

This heterogeneity was demonstrated in a meta-analysis of 21 studies by Moody et al,16 who reported an overall complication rate of 11.3% (95% CI, 8.1%-15.0%). The specific rates were 2.9% (95% CI, 1.5%-4.6%) for infections, 5.1% (95% CI, 2.5%-8.6%) for seroma, 0.5% (95% CI, 0.3%-0.9%) for hematoma, 1.3% (95% CI, 0.5%-2.6%) for lymphedema, and 0.3% (95% CI, 0.1%-0.6%) for nerve damage. No significant differences according to SLN biopsy site or geographic location (continent) were observed. The overall complication rate in our series was higher, at 37.9%, but this is partly because we included patients who underwent CLND after SLN biopsy. We believe that this design is important for studying overall SLN biopsy complications, as patients referred for SLN biopsy are likely to undergo CLND if the biopsy result is positive. Findings that CLND does not improve survival in micrometastatic SLN disease suggest that this procedure should no longer be indicated in this setting, but these findings emerged after our study started. Indeed a high proportion of patients with a positive SLN biopsy in our series did not undergo CLND. To enable comparison with future studies we have separated our findings for complications and sequelae according to patients who underwent SLN biopsy only and those who underwent SLN biopsy plus CLND (Table 4).

In the Multicenter Selective Lymphadenectomy Trial 1 (MSLT-1), Morton et al.17 described a complication rate of 10.1% for patients who underwent SLN biopsy only and 37.1% for those who underwent SLN biopsy followed by CLND. These lower rates could be due to the fact that only regional complications that occurred within 30 days of the procedure were considered. The complication rate associated with SLN biopsy rose to 12.6% when systemic complications (leg edema, thrombophlebitis, and urinary, pulmonary and other complications) were also considered. The article does not describe complications or sequelae that occurred after this initial period of 30 days.

In MSLT-2, Faries et al.18 only reported lymphedema rates, which were 6.3% in patients who underwent SLN biopsy and 24.1% in those who underwent SLN biopsy and CLND. In this case, the follow-up time was 3 years and the rates are similar to those observed in our study.

The authors of the DeCOG-SLT trial (CLND versus no dissection in patients with sentinel lymph biopsy–positive melanoma) only reported adverse events for patients who underwent CLND. These events were analyzed immediately after surgery and at 3 and 6 months. Follow-up was longer for moderate and severe complications. The complication rate was 24% (58/240); 6% (15/240) of the complications were moderate, while 8% (19/240) were severe. Moderate and severe complications included lymphedema, lymph fistula, seroma, infection, and delayed wound healing. Lymphedema occurred in 20 patients and persisted in 13 of these (5.4%).19 Rates for both lymphedema and persistent lymphedema in our series were more common, possibly because of the longer follow-up period.

It is difficult to compare complication rates for patients who underwent SLN biopsy only (30.9% in our series) with other studies, as most do not consider scarring complications, such as keloid or fibrous scars, or only report on complications in the immediate postoperative period. In one study that did include scarring, Hettiaratchy et al.6 reported a complication rate of 33%, which is similar to ours.

The only factor significantly associated with an increased likelihood of complications in our series was smoking, as 73.3% of smokers developed complications, compared with just 33.0% of nonsmokers. This result coincides with reports by Ascha et al.,12 who in their study of risk factors for lymphatic surgeries for melanoma found smoking to be the only significant predictor of complications associated with SLN biopsy, regardless of biopsy site (odds ratio, 1.8; 95% CI, 1.0-2.9; P=0.038).

Our study has some limitations, including the risk of information bias that might have affected the accuracy of our results. Patients, for example, may have been treated for complications at another center (although the medical records analyzed cover primary care) or may not have sought care. Likewise, complications may not have been accurately recorded. However, if this bias did exist, the true complication and sequela rates would be even higher. In our opinion, information bias is unlikely as similar rates have been reported in studies with a similar design.6

Another potential limitation is that diagnoses were based on data from medical records, which are less reliable than prospectively collected data based on pre-established definitions. Nevertheless, we do not believe that our retrospective design resulted in an overestimation of complications or sequelae.

Another consideration is that our results are from a single hospital and may not be representative. We did, however, analyze data for all consecutive patients in the database. At our hospital, SLN biopsy is performed by surgeons with extensive experience and we believe that the identification of the SLN in all cases attests to adequate performance of the procedure.

We used the ASCO/SSO clinical practice guideline to determine whether or not the SLN biopsies in our series were performed according to current indications, and while the indication rates may have varied slightly had we used another guideline, we do not believe that this would have altered our results.

The morbidity rates associated with SLN biopsy in our cohort and in previous studies have important implications for the routine use of this procedure as a staging tool. The use of any invasive technique should only be contemplated when the benefits for the patient outweigh the risks. Clinical trials to date have not demonstrated an overall survival benefit for SLN biopsy alone or for biopsy followed by CLND. Therefore, as this procedure provides purely prognostic information, the risk of complications and sequelae may be unacceptable. Techniques and methods with a lower risk, such as analysis of clinical data, imaging studies, and proteomic or genomic studies, could offer similarly useful information.

The indication for CLND in patients with micrometastatic disease was recently called into question. A change in practice in this regard would lead to a reduction in associated morbidity, but according to our findings, SLN biopsy is still associated with complication and sequela rates of 30.9% and 7.5% and rates this high may not be justifiable for a procedure that provides prognostic information only, even if the complications are of moderate severity.

The increased risk of complications in smokers is also noteworthy and suggests that the indications for SLN biopsy in this subgroup of patients should be even more restrictive.

The authors declare that they have no conflicts of interest.