Abstract
Purpose
The role of postoperative radiation therapy (PORT) in early stage Merkel cell carcinoma (MCC) is controversial. We analyzed the role of PORT in preventing local recurrences (LR) among patients with low-risk, pathologic stage I MCC based on the location of the primary tumors: head/neck (HN) versus non-HN sites.
Methods and Materials
One hundred forty-seven patients with MCC were identified that had “low risk” disease (pathologic T1 primary tumor, negative microscopic margins, negative pathologic node status, no immunosuppression or prior systemic therapy). LR was defined as tumor recurrence within 2 cm of the primary surgical bed, and its frequency was estimated with the cumulative incidence method.
Results
Seventy-nine patients received PORT (30 HN, 49 non-HN) with a median dose of 50 Gy (range, 8-64 Gy) and 68 patients were treated with surgery alone (30 HN, 38 non-HN). Overall, PORT was associated with a decreased risk of LR (5-year rate: 0% vs 9.5%; P = .004) with 6 LRs observed in the surgery alone group. Although the addition of PORT significantly reduced LR rates among patients with HN MCC (0% vs. 21%; P = .034), no LRs were observed in patients with non-HN MCC managed with surgery alone. There was no significant difference in MCC-specific survival comparing HN versus non-HN groups, with or without PORT.
Conclusions
For low-risk, pathologic stage I MCC of the extremities and trunk, excellent local control rates were achieved with surgery, and PORT is not indicated. However, PORT was associated with a significant reduction in LRs among low-risk MCC of the HN.
Introduction
Merkel cell carcinoma (MCC) is a rare, aggressive skin cancer with a recurrence risk of ∼40%.1 MCC primarily affects elderly Caucasian individuals,2,3 with 60% to 80% of MCC in the United States causally linked to the Merkel cell polyomavirus, whereas the remainder is attributed to ultraviolet exposure.1 Reported risk factors for MCC recurrence include stage and tumor size, close or positive resection margins, pathologically positive lymph nodes or lack of pathologic staging,4 head and neck (HN) location,5, 6, 7 and immunosuppression.8 Patients without clinically apparent disease in the lymph nodes are typically treated with primary resection and a sentinal lymph node biopsy (SLNB).9 In pathologically node-negative patients, adjuvant postoperative radiation therapy (PORT) to the primary site is recommended to prevent local recurrence (LR) in the presence of ≥1 baseline risk factors: large primary tumor (≥1 cm), chronic T cell immunosuppresion (human immunodeficiency virus, chronic lymphocytic leukemia, solid organ transplant), HN primary site, close or positive margins, or lymphovascular space invasion (LVSI).9 However, there is a paucity of data that is specific for stage I MCC. In patients with favorable stage I MCC without adverse features, observation without PORT is often recommended, and the need for PORT is controversial.4,10,11
Previous work has demonstrated that MCC is radiosensitive,12 and PORT has been associated with improved local control in several studies.13, 14, 15, 16, 17, 18 PORT is generally recommended for high-risk MCC cases, but studies are mixed on its role in early stage, low-risk disease.4,10,11,13,19, 20, 21 Because MCC is a rare disease, the literature is largely based upon retrospective studies, often with heterogenous patient populations. Although HN location is a recognized risk factor for LR, most studies that included early-stage MCC and suggested a low risk of LR without PORT did not specifically account for the location of the primary tumor.4,10,11,13,19, 20, 21 An earlier study using part of the same cohort as our investigation found that PORT effectively reduces LR rates in a homogenous series of patients with low-risk HN primary disease after demonstrating a 26% LR rate in patients with HN who had surgery alone versus no LRs in the surgery + PORT group.22 However, there are no such specific data evaluating the role of PORT in the non-HN setting for stage I, low-risk MCC.
To develop a more granular understanding of the role of PORT in low-risk MCC resected with pathologically negative margins, we retrospectively examined the impact of PORT on LR rates in patients with low-risk, pathologic stage I MCC stratified by their primary tumor site: HN versus non-HN.
Methods and Materials
We conducted an institutional review board–approved (Fred Hutchinson Cancer Center #6585) retrospective analysis of 147 patients with low-risk MCC from a Seattle-based repository of over 1500 patients with MCC treated between 2006 and 2020. This repository is a longitudinal database containing information on participant demographic characteristics, tumor features, treatment factors, and disease course. The database is updated annually with information from physician notes (electronic medical records), radiologic imaging reports, pathology reports, and communication with patients and their outside providers.
Inclusion criteria
Patients included in this study were enrolled in our repository ≤180 days from the date of surgical excision of the primary tumor to minimize ascertainment bias, had American Joint Committee on Cancer eight edition pathologic stage I disease (primary tumors ≤2 cm with a negative SLNB, and negative pathologic margins), no profound immunosuppression (including organ transplants, concurrent lymphoproliferative malignancies such as chronic lymphocytic leukemia, or human immunodeficiency virus infection), and no other prior systemic therapies. Before 2016, patients who fulfilled all these criteria were offered a choice of surveillance versus PORT. After 2016, based on findings reported by Takagishi et al, all patients with HN MCC were offered PORT to reduce risk of LR.22
Radiation therapy
Specific radiation techniques were at the discretion of the treating clinician. All patients received PORT to the primary resection bed. PORT was delivered using photon (IMRT or 3-dimensional conformal) or electron therapy (typically 6-10 MeV prescribed to the 90% isodose line) with a bolus to the entire surgical bed plus anatomically constrained 3 to 5 cm margins. Regional nodes were irradiated in a minority of cases as described further in the results section. Most patients were treated to a standard PORT dose of ∼50 Gy with conventional fractionation, and a minority were treated with a single fraction to 8 Gy as described further in the results section.
Endpoints
The primary endpoint was MCC LR, and the secondary endpoints were other MCC recurrences and MCC-specific mortality. For all endpoints, the date of initial surgery was considered time 0, and times were calculated until the date of recurrence, date of last follow-up (in the absence of an event), or date of death. Recurrences were categorized as local, in-transit, nodal, and distant. An LR was defined as a tumor recurrence within 2 cm of the primary surgical bed, in-transit as a cutaneous/subcutaneous lesion not involving region lymph nodes and arising >2 cm from the primary surgical bed, regional recurrence as occurring in the draining lymph node bed, and distant recurrence as arising beyond the draining lymph node bed. For all recurrence outcomes, only first recurrence events were considered in our statistical analyses.
Statistical analysis
Statistical analyses were performed using the Stata software version 14.0 (StataCorp, College Station, TX) and R (version 4.0.0; R Foundation for Statistical Computer, Vienna, Austria).
Patient and tumor characteristics were compared between groups using Fisher's exact test or the Wilcoxon rank-sum test. Event rates were estimated using cumulative incidence curves, accounting for competing risks. The competing risks were dependent on the endpoint. The competing risks for LRs and each other MCC recurrence type were any other MCC recurrence and death from any cause. Non-MCC death was a competing risk for MCC-specific survival. Five-year event rate estimates were extracted from these cumulative incidence curves. Differences in event rates between surgery alone and surgery + PORT groups were assessed using Cox regression models, censored by any competing risks and with PORT coded as a time-varying covariate. The time-varying covariate was used to minimize potential bias due to early recurrences or death not preventable by PORT due to later initiation of treatment. In these models, all patients start in the surgery alone group at time 0 (time of initial surgery) and enter the surgery + PORT group at the time PORT is initiated. Due to the small number of events in some groups, permutation tests were used to test for differences between surgery alone and surgery + PORT, using the Cox model partial likelihood ratio as the test statistic.
Results
Patient and pathologic tumor characteristics
The 147 individuals included in this cohort were patients with low-risk, pathologic stage I MCC (Fig. 1). Seventy-nine patients in this cohort received PORT (30 HN, 49 non-HN), while 68 patients were treated with surgery alone (30 HN, 38 non-HN). The median follow-up time was 5.1 years for the entire cohort (range, 7 days to 6 years). The surgery-alone group had a median follow-up time of 4.1 years (range, 7 days to 15 years), whereas the surgery + PORT group had a median follow-up of 5.2 years (range, 73 days to 16 years).
Figure 1.
Study cohort. Identification of 147 patients with resected, pathologic stage I, Merkel cell carcinoma included in this cohort. *All patients met the following criteria: primary tumor ≤2 cm, no profound immunosuppression, microscopically clear surgical margins, negative sentinel lymph node biopsy, no chemotherapy, and enrollment within 180 days from surgical excision of primary tumor. Abbreviations: HN = primary tumor of the head and neck; non-HN = primary tumor of the trunk, extremities, or buttocks; PORT = postoperative radiation therapy following primary surgical excision.
Patient and tumor characteristics are summarized in Table 1. Patients with non-HN had larger tumors than patients with HN in both the surgery-alone group (median, 1.00 cm vs 0.55 cm; P < .001) and surgery + PORT group (median, 1.10 vs 0.60 cm; P < .001; Table E1). Patients with HN had smaller intended surgical margins (margin width as reported by operative report) than patients with non-HN, although no significant differences in final pathologic margin status was observed in patients with HN versus non-HN (Table E1).
Table 1.
Clinical and tumor characteristics by treatment type (n = 147)
| All patients |
HN |
Non-HN |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Variable | Surgery (n = 68), n (%) | Surgery + PORT (n = 79), n (%) | P value* | Surgery (n = 30), n (%) | Surgery + PORT (n = 30), n (%) | P- value* | Surgery (n = 38), n (%) | Surgery + PORT (n = 49), n (%) | P- value* |
| Sex | .181 | .796 | .273 | ||||||
| Male | 36 (52.9) | 50 (63.3) | 15 (50.0) | 17 (56.7) | 21 (55.3) | 33 (67.4) | |||
| Female | 32 (47.1) | 29 (36.7) | 15 (50.0) | 13 (43.3) | 17 (44.7) | 16 (32.6) | |||
| Age at diagnosis (y) | .863 | .412 | .502 | ||||||
| <65 | 24 (35.3) | 28 (35.4) | 12 (40.0) | 8 (26.7) | 12 (31.6) | 20 (40.8) | |||
| ≥ 65 | 44 (64.7) | 51 (64.6) | 18 (60.0) | 22 (73.3) | 26 (68.4) | 29 (59.2) | |||
| Size of primary tumor (cm) | .035 | .254 | .129 | ||||||
| 0-1 | 52 (76.5) | 46 (58.2) | 28 (93.3) | 24 (80.0) | 24 (63.2) | 22 (44.9) | |||
| >1 | 16 (23.5) | 33 (41.8) | 2 (6.7) | 6 (20.0) | 14 (36.8) | 27 (55.1) | |||
| Median (cm) | 0.80 | 1.00 | .020 | 0.55 | 0.60 | .31 | 1.00 | 1.10 | .073 |
| Intended surgical margins (cm)† | .319 | .064 | .144 | ||||||
| ≤0.5 | 4 (6.3) | 3 (4.2) | 0 (0) | 2 (7.7) | 4 (11.4) | 1 (2.2) | |||
| >0.5-1.0 | 36 (57.1) | 33 (46.5) | 19 (67.8) | 18 (69.2) | 17 (48.6) | 15 (33.3) | |||
| >1.0-1.5 | 3 (4.8) | 5 (7.0) | 0 (0) | 1 (3.9) | 3 (8.6) | 4 (8.9) | |||
| ≥2.0 | 10 (15.9) | 22 (31.0) | 4 (14.3) | 5 (19.2) | 6 (17.1) | 17 (37.8) | |||
| Wide excision, not further specified | 10 (15.9) | 8 (11.3) | 5 (17.9) | 0 (0) | 5 (14.3) | 8 (17.8) | |||
| Microscopic margins (cm)† | .638 | .318 | .518 | ||||||
| <0.1 | 1 (1.5) | 5 (6.5) | 0 (0) | 0 (0) | 1 (2.6) | 5 (10.6) | |||
| 0.1-0.49 | 9 (13.4) | 12 (15.6) | 6 (20.6) | 5 (16.7) | 3 (7.9) | 7 (14.9) | |||
| 0.5-0.99 | 8 (11.9) | 7 (9.1) | 1 (3.5) | 2 (6.7) | 7 (18.4) | 5 (10.6) | |||
| ≥1 | 3 (4.5) | 4 (5.2) | 1 (3.5) | 1 (3.3) | 2 (5.3) | 3 (6.4) | |||
| No residual tumor detected | 44 (65.7) | 44 (57.1) | 21 (72.4) | 18 (60.0) | 23 (60.5) | 26 (55.4) | |||
| Margins negative | 2 (3.0) | 5 (6.5) | 0 (0) | 4 (13.3) | 2 (5.3) | 1 (2.1) | |||
| LVSI† | |||||||||
| Positive | 13 (21.3) | 20 (29.0) | .42 | 2 (7.4) | 5 (18.5) | .42 | 11 (32.4) | 15 (35.7) | .81 |
| Negative | 48 (78.7) | 49 (71.0) | 25 (92.6) | 22 (81.5) | 23 (67.6) | 27 (64.3) | |||
Abbreviations: HN = Merkel cell carcinoma of the head and neck; intended surgical margins = margin width as reported by operative report; LVSI = lymphovascular space invasion; non-HN = Merkel cell carcinoma of the trunk, extremities, or buttocks; PORT = postoperative radiation therapy.
Fisher's exact text or Wilcoxon rank-sum test.
Patients with missing values were excluded from the corresponding summary: intended surgical margins (n = 13), microscopic margins (n = 3), and LVSI (n = 17).
LVSI was more common in patients with non-HN compared with patients with HN (26/76 [34%] vs 7/54 [13%]; P = .008; Table E1). There was no significant difference in LVSI status between the surgery only and surgery + PORT groups overall (13/61 [21%] vs 20/69 [29%]; P = .42; Table 1). Similarly, after stratifying by primary tumor site, there were no significant differences in LVSI status by treatment modality for patients with HN (2/27 [7%] vs 5/27 [19%]; P = .42) or patients with non-HN (11/34 [32%] vs 15/42 [36%]; P = .81). In this low-risk cohort, there was not a significant difference in LR rates between LVSI-negative and LVSI-positive patients (5-year rate: 5.6% vs 0%; P = .18).
Among the 79 patients who received PORT, dose information was available for 72. The median dose was 50 Gy (range, 8-64 Gy), and the median dose per fraction was 2 Gy (range, 1.8-8.0 Gy/fraction). Sixty patients (83%) received 50 Gy in 25 fractions, 10 patients (14%) received a single fraction of 8 Gy, 1 patient (1%) received 50.4 Gy in 28 fractions (1.8 Gy/fraction), and 1 patient received 50 Gy in 24 fractions (2.1 Gy/fraction). PORT began a median of 41 days after surgery (range, 8-183 days). Given that this cohort had patients who were all SLNB negative, the majority of patients received PORT to the primary site alone. Eleven patients received PORT to the draining lymph node basin in addition to the primary tumor site (11/75, 15%), including 7/30 patients with HN (23%) and 4/45 patients with non-HN (9%); in 4 cases this information was missing.
Recurrence and survival
There were 6 total LRs, all in the surgery-alone group and none in the surgery + PORT group (5-year rate: 9.5% vs 0%; P = .004; Fig. 2A). Of those with an LR, the median time to LR was 10 months from surgical excision of the primary tumor (range, 14 days-27 months). When patients were further stratified by site of disease, the addition of PORT significantly reduced LR rates among patients with HN (5-year rate: 21% [6 LRs] vs 0%, P = 0.034; Fig. 2B), whereas no LRs were observed in the non-HN cohort (Fig. 2C).
Figure 2.
Local recurrence rate by primary tumor site and treatment modality. Cumulative incidence curves depicting probability of local recurrence is illustrated for 147 patients with resected, pathologic stage I, low-risk MCC stratified by (A) treatment type (surgery vs surgery + PORT), (B) MCC of the head and neck (HN), and (C) MCC of the trunk, extremities, or buttocks (non-HN). Across all disease sites, estimated local recurrence rates were significantly higher in patients treated with surgery alone than surgery + PORT (panel A, 5-year local recurrence rate: 9.5% vs 0.0%; P = .004). There was a significant difference in local recurrence rates among patients with HN treated with surgery alone versus surgery + PORT (panel B, 5-year rate: 21% vs 0%; P = .034), but not for patients with non-HN (panel C; 5-year rate: 0% vs 0%; P > .99). Nonlocalized MCC recurrences and death were treated as competing risks when estimating the cumulative incidence functions. Abbreviations: HN = primary tumor of the head and neck; non-HN = primary tumor of the trunk, extremities, or buttocks; MCC = Merkel cell carcinoma; PORT = postoperative radiation therapy following primary surgical excision.
As shown in Table 2, LRs occurred in patients with primary tumors on the forehead,2 cheek,3 and ear.1 The primary tumor size among patients with an LR was ≤0.5 cm in 5 patients and 0.7 cm in the sixth patient. Salvage treatment (as described in Table 2) was successful in obtaining local control in 4 of the 6 (67%) LR patients. Among the patients with LR, 1 patient died of MCC, and 2 patients died of non-MCC causes (Table 2).
Table 2.
Patient and tumor characteristics for local MCC recurrences (n = 6)
| Primary tumor site | Primary treatment modality | Age at dx (y) | Time to LR (d)* | Sex (M/F) | Primary tumor subsite | Primary tumor size (cm) | Primary tumor depth (cm) | Intended surgical margins (cm) | LVSI (Y/N) | Salvage treatment | Salvaged (Y/N) | Status |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| HN | Surgery | 59 | 404 | F | Forehead | 0.5 | 0.21 | 1.0 | N | RT | Y | Alive |
| HN | Surgery | 70 | 824 | F | Cheek | 0.3 | N/A | WE | N/A | WLE | N† | Alive |
| HN | Surgery | 76 | 14 | M | Forehead | 0.5 | 0.11 | 1.0 | N | RT | Y | Non-MCC death |
| HN | Surgery | 58 | 211 | M | Cheek | 0.5 | 0.40 | 1.0 | N | RT | Y | Alive |
| HN | Surgery | 84 | 82 | M | Cheek | 0.4 | 0.08 | 1.0 | N | RT | N‡ | MCC death |
| HN | Surgery | 80 | 497 | M | Ear | 0.7 | 0.30 | 1.0 | Y | RT | Y | Non-MCC death |
Abbreviations: dx = diagnosis; F = female; HN = MCC of the head and neck; intended surgical margins = margin width as reported by operative report; LR = local MCC recurrence as a first event; LVSI = lymphovascular space invasion; M = male; MCC = Merkel cell carcinoma; N = no; N/A = no data; RT = radiation therapy; WE = wide excision not further specified; WLE = wide local excision; Y = yes.
Days since surgical excision of the primary tumor.
Second LR 3 years after first LR; treated w/ WLE; now no evidence of disease.
Distant recurrence 1 year after LR; no cancer-directed therapy.
There was no significant difference in MCC-specific survival in the surgery versus surgery + PORT groups in the cohort as a whole (Fig. 3A) nor when stratified by HN (Fig. 3B) or non-HN (Fig. 3C). There were 9 patients with regional nodal recurrences, 5 patients with distant recurrences, and 2 patients with an in-transit recurrence as first events (Table 3). Both in-transit recurrences occurred in patients with non-HN. One patient had PORT and an in-transit recurrence outside of the radiation therapy field on the lower extremity. One patient was managed with surgery alone at the elbow and had an in-transit recurrence in the upper arm.
Figure 3.
MCC-specific survival based on primary tumor site and treatment type. Cumulative incidence curves illustrating MCC-specific survival for 147 patients with pathologically stage I low-risk disease stratified by (A) treatment type (surgery vs surgery + PORT), (B) disease of the head and neck (HN), and (C) MCC of the trunk, extremities, or buttocks (non-HN). There were no significant differences in disease-specific mortality between patients who received PORT and those with did not for the entire cohort (panel A, 5-year rate: 3.2% vs 3.9%; P = .94), nor for patients with HN (panel B, 5-year rate: 0% vs 7.9%; P = .48) and non-HN (panel C, 5-year rate: 4.9% vs 0%; P = .24) individually. Non-MCC-related death was treated as a competing risk when estimating the cumulative incidence functions. Abbreviations: HN = primary tumor of the head and neck; non-HN = primary tumor of the trunk, extremities, or buttocks; MCC = Merkel cell carcinoma; PORT = postoperative radiation therapy following primary surgical excision.
Table 3.
Patient and tumor characteristics for nonlocal MCC recurrences (n = 16)
| Type of first MCC recurrence | Primary tumor site | Primary treatment modality | Time to recurrence (d)* | Age at dx (y) | Sex (M/F) | Primary tumor subsite | Primary tumor size (cm) | Primary tumor depth (cm) | Surgical margins (cm) | LVSI (Y/N) | Salvage treatment | Salvaged (Y/N) | Status |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| In-transit | Non-HN | Surgery + PORT | 342 | 70 | F | Limb | 1.5 | 1.10 | 1.5 | N | RT | Y | Alive |
| In-transit | Non-HN | Surgery | 267 | 72 | M | Limb | 1.2 | 0.60 | 1.0 | N | SFRT | Y | Alive |
| Regional | Non-HN | Surgery + PORT | 349 | 72 | M | Limb | 1.2 | 0.20 | 1.0 | N | IMTX | Y | Alive |
| Regional | Non-HN | Surgery + PORT | 1240 | 80 | M | Limb | 1.1 | N/A | 2.0 | N/A | IMTX | Y | MCC death |
| Regional | Non-HN | Surgery + PORT | 267 | 76 | M | Limb | 1.2 | 0.62 | 1.0 | N | WLE | Y | Alive |
| Regional | Non-HN | Surgery + PORT | 320 | 71 | M | Limb | 1.6 | 0.60 | WE | Y | RT | Y | Alive |
| Regional | HN | Surgery + PORT | 304 | 67 | M | Forehead | 0.8 | 0.31 | 1.0 | N | IMTX | Y | Alive |
| Regional | Non-HN | Surgery + PORT | 349 | 59 | F | Limb | 1.0 | 0.40 | 1.1 | Y | RT | Y | Alive |
| Regional | HN | Surgery | 788 | 63 | F | Cheek | 0.5 | 0.08 | 1.0 | N | RT | Y | Alive |
| Regional | HN | Surgery + PORT | 358 | 55 | M | Forehead | 1.0 | 0.11 | 1.0 | Y | WLE | Y | Alive |
| Regional | HN | Surgery + PORT | 100 | 65 | M | Neck | 0.8 | 0.35 | 1.5 | N | WLE | Y | Alive |
| Distant | Non-HN | Surgery + PORT | 1157 | 62 | M | Limb | 1.5 | 0.32 | ME | N | IMTX | N | MCC death |
| Distant | Non-HN | Surgery + PORT | 834 | 68 | M | Limb | 2.0 | 1.30 | 2.0 | N | RT | Y | Alive |
| Distant | Non-HN | Surgery + PORT | 598 | 70 | F | Limb | 1.5 | N/A | WE | N | SFRT | Y | Alive |
| Distant | HN | Surgery | 769 | 71 | M | Neck | 1.2 | 0.14 | 2.0 | N | IMTX | N | MCC death |
| Distant | Non-HN | Surgery + PORT | 73 | 68 | M | Trunk | 1.9 | 0.21 | 2.0 | Y | None | N | MCC death |
Abbreviations: dx = diagnosis; HN = MCC of the head and neck; IMTX = immunotherapy; LR = local MCC recurrence as a first event; LVSI = lymphovascular space invasion; MCC = Merkel cell carcinoma; ME = Mohs excision; N/A = no data; NED = no evidence of disease (MCC); non-HN = MCC of the trunk, extremities, or buttocks; RT = radiation therapy; SFRT = single fraction radiation therapy; WE = wide excision not further specified; WLE = wide local excision.
Days since surgical excision of the primary tumor.
Discussion
This study aimed to develop a more granular understanding of the role of PORT in low-risk MCC and to identify those patients who may benefit from PORT versus cases where it may be safely excluded. Toward this goal, we retrospectively examined 147 patients with low-risk, pathologic stage I MCC who were treated with surgery versus surgery + PORT, stratified by primary tumor site (HN vs non-HN).
Among this entire cohort of patients with low-risk MCC, the 5-year LR rate of patients who had surgery alone was 9.5% compared with 0% for those who were treated with surgery + PORT (P = .004). However, when stratified by site of primary tumor, the addition of PORT significantly decreased the LR risk among patients with MCC of HN (21% for surgery alone vs 0% for surgery + PORT) whereas no LRs were observed in patients with MCC of non-HN locations. These findings confirm prior reports that PORT is associated with lower LR rates in patients with primary HN MCC, even among patients with the lowest-risk disease.13, 14, 15, 16, 17, 18,22 One possible explanation is that MCC in the HN may not always lend to wide surgical margins due to cosmetic and functional limitations. However, in this cohort there was no significant difference in the pathologic margin status between HN and non-HN locations. These data also indicate that patients with low-risk MCC of non-HN locations likely derive little or no benefit from PORT and may be safely observed after surgery. Thus, although PORT could be safely omitted for non-HN sites, it would be prudent to consider MCC of the HN as an indication for PORT, and our data support the current National Comprehensive Cancer Network guidelines that designate HN primary MCC as a baseline risk factor meriting consideration of PORT.9
To our knowledge, this is the largest homogeneous cohort of patients with pathologic stage I MCC and low-risk features (clear surgical margins, negative SLNB, and no immunosuppression). All primary tumors that had an LR among patients with HN were ≤0.7 cm. Five of the 6 patients who had a local failure had surgical margins of 1.0 cm, and 1 patient had a wide excision. LVSI was evenly distributed and was not associated with LR risk. Therefore, we recommend PORT to all patients with HN MCC given a risk of LR of approximately 20% to 25%. In contrast, excellent local control outcomes were achieved with surgery alone in patients with low-risk non-HN MCC. Indeed, no local failures were observed in the 38 patients with non-HN treated with surgery alone, nor in the 49 patients with non-HN treated with surgery + PORT. In nonimmunosuppressed patients with low-risk pathologic stage I MCC of the trunk or extremities, we recommend close observation. In addition, while treating HN sites, the regional lymph nodes and in-transit sites could also be comprehensively incorporated in the PORT volume, especially when there is a failed SLNB. This contrasts with treating limbs and trunk where there may be a dissociation between the primary site and regional nodes (eg, primary in the calf versus treating groin lymph nodes).
The efficacy of PORT in treating primary, stage I MCC tumors has been widely debated in the literature. Previous studies demonstrating the benefit of PORT to reduce LR rates and increase overall survival13, 14, 15, 16, 17, 18 included cohorts of patients with a mix of stages, risk factors, and primary tumor sites. Similarly, studies of node-negative MCC that suggest a low risk of LR without PORT4,10,11,13,19, 20, 21 also included patients with a mix of pathologic stages, risk factors, and primary tumor sites and do not account for the differential risk of LR based on location of the primary. It is not known why there are differences in biologic behavior and recurrence rates for HN versus non-HN MCC. We can hypothesize that sun exposure, even in virus-positive MCC, may play a role.23,24 It is interesting to speculate that the heterogeneity in skin may provide a different tumor microenvironment in sun-exposed versus non-sun-exposed areas.23, 24, 25
We recognize several limitations to this study including its retrospective nature and potential for selection bias, eg, larger tumors received PORT. Given that there were no events in the non-HN group, the study was not powered for multivariate analyses or to assess risk factors for LR, such as primary tumor size or surgical margin width. Although this is essentially a single-institution study, surgery and/or radiation may have occurred at other institutions. As such, there is some variability in patterns of practice including the dose and fractionation of radiation therapy while contributing to the generalizability of this data.
To our knowledge, this is the first study to characterize the efficacy of PORT stratified by disease site in low-risk, pathologic stage I MCC. Identifying patients with low-risk MCC of HN that may be spared PORT without increasing their LR risk will be the subject of future investigation.
Conclusion
One hundred forty-seven patients with low-risk, pathologic stage I MCC were stratified by the location of the primary (HN vs non-HN) and by treatment with surgery versus surgery + PORT. PORT did not benefit patients with non-HN MCC in whom no LR were observed. However, in keeping with multiple prior cohorts, our analysis identified the HN location as a significant risk factor for LR, and in our cohort, PORT was associated with a significant reduction in LRs among low-risk MCC in the HN (21% for surgery alone vs 0% for surgery + PORT).
Disclosures
Peter H. Goff reports research funding from Gilead Sciences, Inc, the Conquer Cancer Foundation of the American Society of Clinical Oncology, and the RSNA R&E Foundation not directly related to this work. Daniel S. Hippe reports research grants from GE Health care, Philips Health care, Canon Medical Systems USA, and Siemens Healthineers. Paul Nghiem reports research funding and personal fees from EMD-Serono and Merck & Co. All research funding and personal fees are unrelated to the submitted work.
Acknowledgments
We thank all of the patients with Merkel cell carcinoma and their families who participate in our research so we may better understand and treat this aggressive cancer.
Footnotes
Sources of support: This study was supported with P01 CA225517 (“Immunobiology and Immune Therapy for Merkel Cell Carcinoma”), Kelsey Dickson Team Science Courage Research Award: Advancing New Therapies for Merkel Cell Carcinoma (MCC), MCC Patient Gift Fund.
Research data are stored in an institutional repository and will be shared upon request to the corresponding author. Data generated and analyzed during this study are included in this published article (and its supplementary files).
Supplementary material associated with this article can be found in the online version at doi:10.1016/j.adro.2023.101364.
Appendix. Supplementary materials
Supplemental Table 1. Clinical and tumor characteristics by primary disease site (n=147).
References
- 1.Feng H, Shuda M, Chang Y, Moore PS. Clonal integration of a polyomavirus in human Merkel cell carcinoma. Science. 2008;319:1096–1100. doi: 10.1126/science.1152586. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Rockville Merkel Cell Carcinoma Group Merkel cell carcinoma: Recent progress and current priorities on etiology, pathogenesis, and clinical management. J Clin Oncol. 2009;27:4021–4026. doi: 10.1200/JCO.2009.22.6605. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Albores-Saavedra J, Batich K, Chable-Montero F, Sagy N, Schwartz AM, Henson DE. Merkel cell carcinoma demographics, morphology, and survival based on 3870 cases: A population based study. J Cutan Pathol. 2010;37:20–27. doi: 10.1111/j.1600-0560.2009.01370.x. [DOI] [PubMed] [Google Scholar]
- 4.Allen PJ, Bowne WB, Jaques DP, Brennan MF, Busam K, Coit DG. Merkel cell carcinoma: prognosis and treatment of patients from a single institution. J Clin Oncol. 2005;23:2300–2309. doi: 10.1200/JCO.2005.02.329. [DOI] [PubMed] [Google Scholar]
- 5.Bajetta E, Platania M. Pitfalls in neuroendocrine tumor diagnosis. Rare Tumors. 2009;1:96–97. doi: 10.4081/rt.2009.e32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.O'Connor WJ, Brodland DG. Merkel cell carcinoma. Dermatol Surg. 1996;22:262–267. doi: 10.1111/j.1524-4725.1996.tb00317.x. [DOI] [PubMed] [Google Scholar]
- 7.Morand GB, Madana J, Da Silva SD, Hier MP, Mlynarek AM, Black MJ. Merkel cell carcinoma of the head and neck: Poorer prognosis than non-head and neck sites. J Laryngol Otol. 2016;130:393–397. doi: 10.1017/S0022215116000153. [DOI] [PubMed] [Google Scholar]
- 8.Cook M, Baker K, Redman M, et al. Differential outcomes among immunosuppressed patients with Merkel cell carcinoma: Impact of immunosuppression type on cancer-specific and overall survival. Am J Clin Oncol. 2019;42:82–88. doi: 10.1097/COC.0000000000000482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.National Comprehensive Cancer Network. NCCN guidelines version 1.2023: Merkel cell carcinoma. Available at:https://www.nccn.org/guidelines/guidelines-process/transparency-process-and-recommendations/GetFileFromFileManagerGuid?FileManagerGuidId=692edc66-6685-4a91-a82d-ca8f40070aba. Accessed April 19, 2023.
- 10.Fields RC, Busam KJ, Chou JF, et al. Recurrence after complete resection and selective use of adjuvant therapy for stage I through III Merkel cell carcinoma. Cancer. 2012;118:3311–3320. doi: 10.1002/cncr.26626. [DOI] [PubMed] [Google Scholar]
- 11.Frohm ML, Griffith KA, Harms KL, et al. Recurrence and survival in patients with Merkel cell carcinoma undergoing surgery without adjuvant radiation therapy to the primary site. JAMA Dermatol. 2016;152:1001–1007. doi: 10.1001/jamadermatol.2016.1428. [DOI] [PubMed] [Google Scholar]
- 12.Iyer NG, Tan DSW, Tan VKM, et al. Randomized trial comparing surgery and adjuvant radiotherapy versus concurrent chemoradiotherapy in patients with advanced, nonmetastatic squamous cell carcinoma of the head and neck: 10-year update and subset analysis. Cancer. 2015;121:1599–1607. doi: 10.1002/cncr.29251. [DOI] [PubMed] [Google Scholar]
- 13.Bhatia S, Storer BE, Iyer JG, et al. Adjuvant radiation therapy and chemotherapy in Merkel cell carcinoma: Survival analyses of 6908 cases from the National Cancer Data base. J Natl Cancer Inst. 2016;108 doi: 10.1093/jnci/djw042. djw042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Iyer JG, Parvathaneni U, Gooley T, et al. Single-fraction radiation therapy in patients with metastatic Merkel cell carcinoma. Cancer Med. 2015;4:1161–1170. doi: 10.1002/cam4.458. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Gillenwater AM, Hessel AC, Morrison WH, et al. Merkel cell carcinoma of the head and neck: Effect of surgical excision and radiation on recurrence and survival. Arch Otolaryngol Head Neck Surg. 2001;127:149–154. doi: 10.1001/archotol.127.2.149. [DOI] [PubMed] [Google Scholar]
- 16.Medina-Franco H, Urist MM, Fiveash J, Heslin MJ, Bland KI, Beenken SW. Multimodality treatment of Merkel cell carcinoma: Case series and literature review of 1024 cases. Ann Surg Oncol. 2001;8:204–208. doi: 10.1007/s10434-001-0204-4. [DOI] [PubMed] [Google Scholar]
- 17.Petrelli F, Ghidini A, Torchio M, et al. Adjuvant radiotherapy for Merkel cell carcinoma: A systematic review and meta-analysis. Radiother Oncol. 2019;134:211–219. doi: 10.1016/j.radonc.2019.02.015. [DOI] [PubMed] [Google Scholar]
- 18.Veness MJ, Perera L, McCourt J, et al. Merkel cell carcinoma: Improved outcome with adjuvant radiotherapy. ANZ J Surg. 2005;75:275–281. doi: 10.1111/j.1445-2197.2005.03353.x. [DOI] [PubMed] [Google Scholar]
- 19.Strom T, Carr M, Zager JS, et al. Radiation therapy is associated with improved outcomes in Merkel Cell carcinoma. Ann Surg Oncol. 2016;23:3572–3578. doi: 10.1245/s10434-016-5293-1. [DOI] [PubMed] [Google Scholar]
- 20.Bishop AJ, Garden AS, Gunn B, et al. Merkel cell carcinoma of the head and neck: Favorable outcomes with radiotherapy. Head Neck. 2016;38:E452–E458. doi: 10.1002/hed.24017. [DOI] [PubMed] [Google Scholar]
- 21.Clark JR, Veness MJ, Gilbert R, O'Brien CJ, Gullane PJ. Merkel cell carcinoma of the head and neck: Is adjuvant radiotherapy necessary? Head Neck. 2007;29:249–257. doi: 10.1002/hed.20510. [DOI] [PubMed] [Google Scholar]
- 22.Takagishi S.R., Marx T.E., Lewis C. Postoperative radiation therapy is associated with a reduced risk of local recurrence among low risk Merkel cell carcinomas of the head and neck. Adv Radiat Oncol. 2016;1:244–251. doi: 10.1016/j.adro.2016.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Hernando B, Dietzen M, Parra G, et al. The effect of age on the acquisition and selection of cancer driver mutations in sun-exposed normal skin. Ann Oncol. 2021;32:412–421. doi: 10.1016/j.annonc.2020.11.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Wei L, Christensen SR, Fitzgerald ME, et al. Ultradeep sequencing differentiates patterns of skin clonal mutations associated with sun-exposure status and skin cancer burden. Science Adv. 2021;7:eabd7703. doi: 10.1126/sciadv.abd7703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Martincorena I, Roshan A, Gerstung M, et al. High burden and pervasive positive selection of somatic mutations in normal human skin. Science. 2015;348:880–886. doi: 10.1126/science.aaa6806. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Table 1. Clinical and tumor characteristics by primary disease site (n=147).