SAN ANTONIO – Lymphovascular invasion, not sentinel lymph node status, was strongly associated with recurrence and death from Merkel cell carcinoma in a series of 153 patients with clinically localized disease.
Patients without lymphatic vascular invasion (LVI) of their primary tumors did not develop recurrence or die, said Dr. Ryan Fields of the department of surgery at Memorial Sloan-Kettering Cancer Center, New York.
Although sentinel lymph node (SLN) status was not associated with recurrence or death from Merkel cell carcinoma (MCC), patients with a positive SLN were more likely to receive subsequent treatment including complete lymph node dissection, nodal radiation therapy, and/or chemotherapy.
Dr. Fields suggested that one explanation for the surprising lack of association with SLN status may be the additional treatment received by SLN-positive patients was so effective as to interrupt the metastatic cascade and prevent recurrence or death from MCC.
During a discussion of the study, Dr. Kelly McMasters, professor and chair of surgery at the University of Louisville in Kentucky, expressed doubt that adjuvant therapy was so effective as to skew the results and asked whether clinicians should biopsy their next MCC patient or whether this may simply lead them to overtreat their patients.
Dr. Fields acknowledged that the results are puzzling and said that all patients with MCC continue to be offered SLN biopsy at Memorial Sloan-Kettering, where the study was performed. He added that surgeons there are also trying to capture information on LVI to determine in which patients LVI could be used in lieu of SLN biopsy as a prognostic marker.
There is no universally accepted staging system for MCC, a rare cutaneous cancer that has a propensity for lymphatic spread. SLN biopsy has been used in an effort to provide more accurate staging and to guide subsequent treatment, although most studies have been small and did not look at outcome at the time of SLN status, he explained.
A previous analysis of 251 MCC patients treated from 1970 to 2002 at Memorial Sloan-Kettering reported that disease stage was the only independent predictor of survival and that stage-specific survival was decreased in patients with node-positive disease (J. Clin. Oncol. 2005;23:2300-9).
In the current analysis from 1996 to 2010, a total of 153 patients with stage I or II MCC underwent SLN biopsy, of which 45 (29%) were positive and 108 (71%) were negative.
The primary tumor was 2 cm or less (clinical stage I) in 122 patients and more than 2 cm (clinical stage II) in 31 patients. LVI was present in the primary tumor in 75 patients, absent in 69, and unknown in 9.
A positive SLN biopsy was significantly associated with stage II vs. stage I tumors (45% vs. 25%, P = .02), and with LVI vs. LVI absence (55% vs. 4%, P less than .01). Notably, 26% of positive SLNs occurred in tumors 1 cm or less, observed Dr. Fields.
After a median follow-up of 41 months, 16 nodal or distant recurrences occurred, 11 patients died of MCC, and 27 patients died of other causes.
"This highlights the fact that overall survival, which has been used quite a bit as an end point in Merkel cell cancers, is a poor end point measure in this population," Dr. Fields said.
Importantly, there were no nodal recurrences in SLN-positive patients who went on to receive adjuvant nodal radiation therapy after completion lymph node dissection or therapeutic radiotherapy alone.
Only two of the 32 SLN-positive patients (6%) who received no adjuvant chemotherapy developed a distant recurrence, suggesting that unproven adjuvant chemotherapy would be unlikely to benefit this patient population, Dr. Fields said.
Among the 108 patients with a negative SLN, 99 received no further treatment. Of these, 8% experienced a recurrence, which corresponds to a 15% false-negative rate for the SLNB procedure in MCC, he said.
When patients were stratified by tumor characteristics, there was no significant association between SLN-positive and SLN-negative patients in nodal/distant recurrence (P = .86) or death from MCC (P = .89). The association was significantly between stage II and stage I tumors for death (P =.05), but not for recurrence (P = .26).
In contrast, the presence of LVI was significantly associated with both nodal/distant recurrence and death from MCC (both P less than .001), Dr. Fields said. The 2-year confidence intervals of recurrence or death from MCC for LVI-positive patients were 30% and 15%, respectively.
Dr. Fields and his coauthors reported no study support or relevant conflicts of interest.