Advertisement

T-Cell Repertoire in Combination with T-Cell Density Predicts Clinical Outcomes in Patients with Merkel Cell Carcinoma

Open ArchivePublished:April 15, 2020DOI:https://doi.org/10.1016/j.jid.2020.02.031
      The integrity of the immune system represents a pivotal risk factor and prognostic biomarker for Merkel cell carcinoma. A higher density of tumor-associated T cells correlates with improved Merkel cell carcinoma–specific survival, but the prognostic importance of the T-cell infiltrate reactivity is unknown. We evaluated the T-cell receptor repertoire associated with 72 primary Merkel cell carcinomas and correlated metrics of the T-cell receptor repertoire with clinicopathologic characteristics and patient outcomes. We showed that a high Simpson’s Dominance index (SDom) was significantly associated with fewer metastases (P = 0.01), lower stage at presentation (P = 0.02), lower final stage at last follow-up (P = 0.05), and longer time to first lymph node metastasis (P = 0.04). These correlations were mostly preserved in the Merkel cell polyomavirus–negative subgroup. Combining SDom with CD3+ or CD8+ T-cell density revealed three distinct prognostic groups with respect to disease-specific survival. Patients with both high SDom and high CD3+ or CD8+ T-cell density had markedly improved disease-specific survival compared with patients with low SDom and low CD3+ or CD8+ T-cell density (P = 0.002 and P = 0.03, respectively). Patients with either high SDom or high CD3+ or CD8+ had intermediate disease-specific survival. Our findings demonstrate that the quality of the tumor-associated T-cell infiltrate informs patient prognosis in primary Merkel cell carcinoma beyond the T-cell density.

      Abbreviations:

      DSS (disease-specific survival), LN (lymph node), MCC (Merkel cell carcinoma), MCPyV (Merkel cell polyomavirus), MCPyV(+) (Merkel cell polyomavirus T-antigen–positive), MCPyV(−) (Merkel cell polyomavirus T-antigen–negative), MDACC (The University of Texas MD Anderson Cancer Center), OS (overall survival), SDom (Simpson’s Dominance index), TCR (T-cell receptor), UDE (University of Duisburg-Essen)

      Introduction

      Merkel cell carcinoma (MCC) is an aggressive cutaneous neuroendocrine malignancy with an increasing incidence (
      • Coggshall K.
      • Tello T.L.
      • North J.P.
      • Yu S.S.
      Merkel cell carcinoma: An update and review: pathogenesis, diagnosis, and staging.
      ,
      • Paulson K.G.
      • Park S.Y.
      • Vandeven N.A.
      • Lachance K.
      • Thomas H.
      • Chapuis A.G.
      • et al.
      Merkel cell carcinoma: Current US incidence and projected increases based on changing demographics.
      ,
      • Stang A.
      • Becker J.C.
      • Nghiem P.
      • Ferlay J.
      The association between geographic location and incidence of Merkel cell carcinoma in comparison to melanoma: an international assessment.
      ,
      • Tetzlaff M.T.
      • Nagarajan P.
      Update on Merkel cell carcinoma.
      ). Risk factors include advanced age, fair skin, male sex, immunosuppression, and chronic sun exposure (
      • Albores-Saavedra J.
      • Batich K.
      • Chable-Montero F.
      • Sagy N.
      • Schwartz A.M.
      • Henson D.E.
      Merkel cell carcinoma demographics, morphology, and survival based on 3870 cases: a population based study.
      ,
      • Becker J.C.
      • Stang A.
      • DeCaprio J.A.
      • Cerroni L.
      • Lebbé C.
      • Veness M.
      • et al.
      Merkel cell carcinoma.
      ,
      • Engels E.A.
      • Frisch M.
      • Goedert J.J.
      • Biggar R.J.
      • Miller R.W.
      Merkel cell carcinoma and HIV infection.
      ,
      • Tadmor T.
      • Liphshitz I.
      • Aviv A.
      • Landgren O.
      • Barchana M.
      • Polliack A.
      Increased incidence of chronic lymphocytic leukaemia and lymphomas in patients with Merkel cell carcinoma - a population based study of 335 cases with neuroendocrine skin tumour.
      ). One of the following two pathways drives MCC pathogenesis: (i) expression of viral oncoproteins (large and small T antigens) in tumors with genomic integration of the Merkel cell polyomavirus (MCPyV) (
      • Feng H.
      • Shuda M.
      • Chang Y.
      • Moore P.S.
      Clonal integration of a polyomavirus in human Merkel cell carcinoma.
      ,
      • Sihto H.
      • Kukko H.
      • Koljonen V.
      • Sankila R.
      • Böhling T.
      • Joensuu H.
      Merkel cell polyomavirus infection, large T antigen, retinoblastoma protein and outcome in Merkel cell carcinoma.
      ) or (ii) progressive accumulation of UV light–induced somatic mutations in MCCs lacking MCPyV (
      • Goh G.
      • Walradt T.
      • Markarov V.
      • Blom A.
      • Riaz N.
      • Doumani R.
      • et al.
      Mutational landscape of MCPyV-positive and MCPyV-negative Merkel cell carcinomas with implications for immunotherapy.
      ,
      • Harms P.W.
      • Vats P.
      • Verhaegen M.E.
      • Robinson D.R.
      • Wu Y.M.
      • Dhanasekaran S.M.
      • et al.
      The distinctive mutational spectra of polyomavirus-negative Merkel cell carcinoma.
      ,
      • Wong S.Q.
      • Waldeck K.
      • Vergara I.A.
      • Schröder J.
      • Madore J.
      • Wilmott J.S.
      • et al.
      UV-associated mutations underlie the etiology of MCV-negative Merkel cell carcinomas.
      ).
      Survival in MCC correlates with the extent of disease burden at presentation; 5-year overall survival (OS) rates in patients presenting with localized disease, regional metastases, and distant metastases are 50.6%, 35.4%, and 13.5%, respectively (
      • Harms K.L.
      • Healy M.A.
      • Nghiem P.
      • Sober A.J.
      • Johnson T.M.
      • Bichakjian C.K.
      • et al.
      Analysis of prognostic factors from 9387 Merkel cell carcinoma cases forms the basis for the new 8th edition AJCC staging system.
      ,
      • Santos J.S.
      • Teixeira L.S.
      • dos Santos W.N.
      • Lemos V.A.
      • Godoy J.M.
      • Ferreira S.L.
      Uranium determination using atomic spectrometric techniques: an overview.
      ). Patient prognosis is determined according to the American Joint Committee on Cancer stage (
      • Harms K.L.
      • Healy M.A.
      • Nghiem P.
      • Sober A.J.
      • Johnson T.M.
      • Bichakjian C.K.
      • et al.
      Analysis of prognostic factors from 9387 Merkel cell carcinoma cases forms the basis for the new 8th edition AJCC staging system.
      ,
      • Santos J.S.
      • Teixeira L.S.
      • dos Santos W.N.
      • Lemos V.A.
      • Godoy J.M.
      • Ferreira S.L.
      Uranium determination using atomic spectrometric techniques: an overview.
      ). However, the 5-year mortality rate in patients with stage Ia disease is approximately 20%, underscoring a critical unmet need to identify novel, stage-independent prognostic biomarkers (
      • Fitzgerald T.L.
      • Dennis S.
      • Kachare S.D.
      • Vohra N.A.
      • Wong J.H.
      • Zervos E.E.
      Dramatic increase in the incidence and mortality from Merkel cell carcinoma in the United States.
      ,
      • Tarantola T.I.
      • Vallow L.A.
      • Halyard M.Y.
      • Weenig R.H.
      • Warschaw K.E.
      • Grotz T.E.
      • et al.
      Prognostic factors in Merkel cell carcinoma: analysis of 240 cases.
      ).
      Although various clinical and pathologic variables have been investigated as prognostic biomarkers in MCC (
      • Albores-Saavedra J.
      • Batich K.
      • Chable-Montero F.
      • Sagy N.
      • Schwartz A.M.
      • Henson D.E.
      Merkel cell carcinoma demographics, morphology, and survival based on 3870 cases: a population based study.
      ,
      • Andea A.A.
      • Coit D.G.
      • Amin B.
      • Busam K.J.
      Merkel cell carcinoma: histologic features and prognosis.
      ,
      • Asioli S.
      • Righi A.
      • de Biase D.
      • Morandi L.
      • Caliendo V.
      • Picciotto F.
      • et al.
      Expression of p63 is the sole independent marker of aggressiveness in localised (stage I-II) Merkel cell carcinomas.
      ,
      • Feldmeyer L.
      • Hudgens C.W.
      • Ray-Lyons G.
      • Nagarajan P.
      • Aung P.P.
      • Curry J.L.
      • et al.
      Density, distribution, and composition of immune infiltrates correlate with survival in Merkel cell carcinoma.
      ,
      • Fleming K.E.
      • Ly T.Y.
      • Pasternak S.
      • Godlewski M.
      • Doucette S.
      • Walsh N.M.
      Support for p63 expression as an adverse prognostic marker in Merkel cell carcinoma: report on a Canadian cohort.
      ,
      • Güler-Nizam E.
      • Leiter U.
      • Metzler G.
      • Breuninger H.
      • Garbe C.
      • Eigentler T.K.
      Clinical course and prognostic factors of Merkel cell carcinoma of the skin.
      ,
      • Skelton H.G.
      • Smith K.J.
      • Hitchcock C.L.
      • McCarthy W.F.
      • Lupton G.P.
      • Graham J.H.
      Merkel cell carcinoma: analysis of clinical, histologic, and immunohistologic features of 132 cases with relation to survival.
      ,
      • Tarantola T.I.
      • Vallow L.A.
      • Halyard M.Y.
      • Weenig R.H.
      • Warschaw K.E.
      • Grotz T.E.
      • et al.
      Prognostic factors in Merkel cell carcinoma: analysis of 240 cases.
      ), integrity of the host immune system has emerged as the most significant. Immunosuppressed patients exhibit an increased propensity to develop MCC (
      • Engels E.A.
      • Frisch M.
      • Goedert J.J.
      • Biggar R.J.
      • Miller R.W.
      Merkel cell carcinoma and HIV infection.
      ,
      • Penn I.
      • First M.R.
      Merkel's cell carcinoma in organ recipients: report of 41 cases.
      ,
      • Tadmor T.
      • Liphshitz I.
      • Aviv A.
      • Landgren O.
      • Barchana M.
      • Polliack A.
      Increased incidence of chronic lymphocytic leukaemia and lymphomas in patients with Merkel cell carcinoma - a population based study of 335 cases with neuroendocrine skin tumour.
      ). For each disease stage, immunosuppressed patients have worse disease-specific survival (DSS) than immunocompetent individuals (
      • Paulson K.G.
      • Iyer J.G.
      • Blom A.
      • Warton E.M.
      • Sokil M.
      • Yelistratova L.
      • et al.
      Systemic immune suppression predicts diminished Merkel cell carcinoma-specific survival independent of stage.
      ). Gene expression profiling in primary MCC revealed a correlation between increased expression of genes reflective of immune activation (including CD8A) and improved prognosis (
      • Paulson K.G.
      • Iyer J.G.
      • Tegeder A.R.
      • Thibodeau R.
      • Schelter J.
      • Koba S.
      • et al.
      Transcriptome-wide studies of merkel cell carcinoma and validation of intratumoral CD8+ lymphocyte invasion as an independent predictor of survival.
      ). Consistent with this, higher densities of MCC-associated CD8+ T cells correlated with improved MCC-specific survival (
      • Feldmeyer L.
      • Hudgens C.W.
      • Ray-Lyons G.
      • Nagarajan P.
      • Aung P.P.
      • Curry J.L.
      • et al.
      Density, distribution, and composition of immune infiltrates correlate with survival in Merkel cell carcinoma.
      ,
      • Paulson K.G.
      • Iyer J.G.
      • Simonson W.T.
      • Blom A.
      • Thibodeau R.M.
      • Schmidt M.
      • et al.
      CD8+ lymphocyte intratumoral infiltration as a stage-independent predictor of Merkel cell carcinoma survival: a population-based study.
      ,
      • Paulson K.G.
      • Iyer J.G.
      • Tegeder A.R.
      • Thibodeau R.
      • Schelter J.
      • Koba S.
      • et al.
      Transcriptome-wide studies of merkel cell carcinoma and validation of intratumoral CD8+ lymphocyte invasion as an independent predictor of survival.
      ,
      • Sihto H.
      • Böhling T.
      • Kavola H.
      • Koljonen V.
      • Salmi M.
      • Jalkanen S.
      • et al.
      Tumor infiltrating immune cells and outcome of Merkel cell carcinoma: a population-based study.
      ).
      T-cell receptor (TCR) sequencing characterizes the population dynamics of the TCR repertoire within a tumor microenvironment. Each T cell possesses a unique variable CDR3β region generated randomly during thymic development. Engagement of a TCR with its cognate antigen (a tumor-associated antigen or mutation-derived neoantigen) elicits clonal expansion of that T cell and shifts the composition of the TCR repertoire to be enriched for T cells reactive to tumor antigen(s) (
      • Li B.
      • Li T.
      • Pignon J.C.
      • Wang B.
      • Wang J.
      • Shukla S.A.
      • et al.
      Landscape of tumor-infiltrating T cell repertoire of human cancers.
      ). Metrics of the quality of the TCR repertoire include richness and Simpson’s Dominance index (SDom). Richness measures the number of independent T-cell clonotypes in a sample (diversity). SDom measures the relative fractional abundance of T-cell clones in a sample and reflects antigen-driven T-cell expansion. Attributes of the TCR repertoire are prognostic in lung adenocarcinoma (
      • Reuben A.
      • Gittelman R.
      • Gao J.
      • Zhang J.
      • Yusko E.C.
      • Wu C.J.
      • et al.
      TCR repertoire intratumor heterogeneity in localized lung adenocarcinomas: an association with predicted neoantigen heterogeneity and postsurgical recurrence.
      ). In melanoma and prostate carcinoma, maintenance of high-frequency TCR clones during α-CTLA-4 therapy correlated with improved OS (
      • Cha E.
      • Klinger M.
      • Hou Y.
      • Cummings C.
      • Ribas A.
      • Faham M.
      • et al.
      Improved survival with T cell clonotype stability after anti-CTLA-4 treatment in cancer patients.
      ). Moreover, a higher proportion of preexisting dominant TCR clones correlated with improved response to PD-1 blockade in melanoma (
      • Roh W.
      • Chen P.L.
      • Reuben A.
      • Spencer C.N.
      • Prieto P.A.
      • Miller J.P.
      • et al.
      Integrated molecular analysis of tumor biopsies on sequential CTLA-4 and PD-1 blockade reveals markers of response and resistance.
      ,
      • Tumeh P.C.
      • Harview C.L.
      • Yearley J.H.
      • Shintaku I.P.
      • Taylor E.J.
      • Robert L.
      • et al.
      PD-1 blockade induces responses by inhibiting adaptive immune resistance.
      ). Initial studies suggest that the TCR repertoire may be prognostic in MCC (
      • Cui J.H.
      • Lin K.R.
      • Yuan S.H.
      • Jin Y.B.
      • Chen X.P.
      • Su X.K.
      • et al.
      TCR repertoire as a novel indicator for immune monitoring and prognosis assessment of patients with cervical cancer.
      ,
      • Schrama D.
      • Ritter C.
      • Becker J.C.
      T cell receptor repertoire usage in cancer as a surrogate marker for immune responses.
      ). Here, we performed TCR sequencing to determine associations between metrics of the TCR repertoire (richness and SDom), clinicopathologic attributes, and survival in patients with primary MCC.

      Results

      Clinical and pathologic characteristics

      The clinical and pathologic characteristics of the patients are summarized in Table 1 and Figure 1. Overall, 72 patients (46 men and 26 women) with primary MCC were included. We compared the clinical and pathologic variables of patients from The University of Texas MD Anderson Cancer Center (MDACC) and the University of Duisburg-Essen (UDE) (Supplementary Table S1). The only significant difference was the frequency of patients alive and disease-free (higher at MDACC). The median age was 72 years. Six patients (8%) had chronic lymphocytic leukemia, and six (8%) were iatrogenically immunosuppressed. A total of 41 tumors (59%) were MCPyV T-antigen–positive (MCPyV(+)). We did not observe significant differences in the density of the tumor-associated immune infiltrate or T-cell repertoire metrics comparing immunocompetent versus immunosuppressed patients (Supplementary Table S2). Demographic differences between patients with MCPyV(+) and MCPyV T-antigen–negative (MCPyV(−)) MCC were not apparent except that median tumor thickness was higher for MCPyV(+) tumors, 9 mm versus 8 mm (Table 1). Most primary MCCs (26 of 72) arose on the head and neck. The median tumor size was 17 mm, and the median thickness was 9 mm. Given the similarities among the patients from each institution, results were analyzed as a single cohort.
      Table 1Clinicopathologic Characteristics and Outcome Variables
      CharacteristicAll patients (n = 72
      MCPyV status unknown for 3 cases.
      )
      MCPyV(+) (n = 41)MCPyV(−) (n = 28)P
      Age, n72 (100)41 (100)28 (100)
       Median (range), years72 (32, 91)74 (32, 91)73 (53, 91)0.43
      Sex, n (%)72 (100)41 (100)28 (100)0.13
       Male46 (64)23 (56)21 (75)
       Female26 (36)18 (44)7 (25)
      Primary tumor site, n (%)68 (94)38 (93)27 (96)0.17
       Head and neck26 (38)10 (26)13 (48)
       Trunk12 (18)6 (16)6 (22)
       Upper extremity16 (24)12 (32)4 (15)
       Lower extremity14 (21)10 (26)4 (15)
      Tumor size, n69 (96)38 (93)28 (100)
       Median (range), mm17 (3, 93)18 (4, 93)17 (3, 45)0.08
      Tumor thickness,69 (96)38 (93)28 (100)
       Median (range), mm9 (1.6, 22.5)9 (1.6, 22.5)8 (2.4, 20)0.004
      Any metastasis, n (%)67 (93)38 (93)26 (93)1.00
       Yes44 (66)25 (66)17 (65)
       No23 (34)13 (34)9 (35)
      Metastasis to SLN, n (%)72 (100)41 (100)28 (100)0.41
       Yes20 (28)9 (22)9 (32)
       No52 (72)32 (78)19 (68)
      Metastasis beyond SLN, n (%)65 (90)36 (88)26 (93)0.45
       Yes34 (52)17 (47)15 (58)
       No31 (48)19 (53)11 (42)
      Distant metastasis, n (%)64 (89)35 (85)26 (93)0.60
       Yes23 (36)11 (31)10 (38)
       No41 (64)24 (69)16 (62)
      Time to first LN metastasis, n (%)38 (53)19 (46)17 (61)0.18
       Median (range), months0 (0, 27)0 (0, 27)0 (0, 17)
      Time to first distant metastasis, n (%)22 (31)10 (24)10 (36)
       Median (range), months12 (0, 41)12 (0, 35)12 (0, 41)0.44
      AJCC stage at presentation, n (%)72 (100)41 (100)28 (100)1.00
       I–II34 (47)19 (46)14 (50)
       III33 (46)19 (46)12 (43)
       IV5 (7)3 (8)2 (7)
      Final AJCC stage, n (%)64 (89)35 (85)26 (93)0.90
       I–II23 (36)13 (37)9 (35)
       III17 (26)10 (29)7 (27)
       IV24 (38)12 (34)10 (38)
      Duration of follow-up, n (%)64 (89)35 (85)26 (93)
       Median (range), months29 (0, 176)32 (4, 176)24 (0, 114)0.35
      Vital status at last follow-up, n (%)63 (88)35 (85)25 (89)0.58
       Died of disease17 (24)7 (17)8 (30)
       Died of other/unknown cause20 (28)10 (24)9 (33)
       Alive with disease7 (10)5 (12)2 (7)
       Alive and disease free18 (25)13 (32)5 (19)
       Lost to follow-up10 (14)6 (15)3 (11)
      Abbreviations: AJCC, American Joint Committee on Cancer; LN, lymph node; MCPyV, Merkel cell polyomavirus; MCPyV(+), Merkel cell polyomavirus T-antigen–positive; MCPyV(−), Merkel cell polyomavirus T-antigen–negative; SLN, sentinel lymph node.
      Bold value designates statistical significance at P ≤ 0.05.
      1 MCPyV status unknown for 3 cases.
      Figure thumbnail gr1
      Figure 1TCR sequencing in Merkel cell carcinoma. (a) Heat map showing the distribution of age, sex, MCPyV status, tumor site, tumor size, stage at presentation, final stage, richness, and SDom in MDACC and UDE cohorts. (b) When TCR on tumor-infiltrating lymphocytes binds to a TAA, it undergoes a TAA-specific clonal expansion resulting in a shift of the intratumoral TCR repertoire. The degree of clonal expansion, reflected by CDR3 clonal enrichment, can be measured by SDom, which offers information about the size-frequency distribution of the T-cell response. High diversity with enrichment of discrete clones in the T-cell population corresponds to a high SDom, whereas a low SDom corresponds to a low diversity with no enrichment of discrete clones. Bar = 60 μm (left) and 200 μm (right). F, female; M, male; Max, maximum; MCPyV, Merkel cell polyomavirus; MDACC, The University of Texas MD Anderson Cancer Center; Min, minimum; SDom, Simpson’s Dominance index; TAA, tumor-associated antigen; TCR, T-cell receptor; UDE, University of Duisburg-Essen.
      Median follow-up period was 29 months (range, 0–176 months) (Table 1). Among patients with metastasis (n = 44), the median time to first distant metastasis was 12 months. A total of 38 patients presented with regional (stage III) or distant (stage IV) metastases, and six patients developed metastases after wide local excision of the primary MCC. At last follow-up, 25 patients were alive, including 18 who were disease-free; 17 patients died of disease; 20 died of an unknown cause; and 10 were lost to follow-up.

      Increased SDom is associated with improved outcome in the overall cohort

      Relationships between TCR metrics and clinical-pathologic variables were assessed. In the overall cohort, no significant correlations were identified between richness or SDom and patient age, sex, immune status, primary tumor site, tumor size, or tumor thickness. When potential associations were evaluated among MCPyV(+) (n = 41) or MCPyV(−) (n = 28) MCCs, a significant correlation was detected between greater richness and greater tumor thickness in the MCPyV(−) group (P = 0.04) (Table 2). No significant correlations were detected between richness or SDom and MCPyV status (Table 2).
      Table 2Correlations Between TCR Repertoire and Clinicopathologic Characteristics and Outcome Variables in All, MCPyV(−), and MCPyV(+) MCCs
      CharacteristicAllMCPyV(−)MCPyV(+)
      SDomRichnessSDomRichnessSDomRichness
      Age0.30.610.350.110.160.54
      Sex0.480.650.750.550.430.78
      Immunosuppression0.130.80.650.960.070.77
      MCPyV status0.680.53
      Primary tumor site0.340.430.340.580.150.77
      Tumor size0.270.860.190.210.540.47
      Tumor thickness0.430.710.070.040.950.16
      Any metastasis0.010.80.050.110.060.29
      Metastasis to SLN0.020.940.070.640.160.83
      Metastasis beyond SLN0.0480.510.220.060.170.44
      Metastasis to an LN0.020.610.050.110.180.5
      Distant metastasis0.040.530.160.210.240.1
      Time to first LN metastasis0.050.730.080.230.240.62
      Time to first distant metastasis0.10.920.250.290.310.18
      AJCC stage at presentation0.020.010.040.0040.370.09
      Final AJCC stage0.050.090.06<0.0010.660.81
      OS0.730.280.760.180.390.51
      DSS0.130.710.230.310.490.58
      Abbreviations: AJCC, American Joint Committee on Cancer; DSS, disease-specific survival; LN, lymph node; MCPyV, Merkel cell polyomavirus; MCPyV(+), Merkel cell polyomavirus T-antigen–positive; MCPyV(−), Merkel cell polyomavirus T-antigen–negative; OS, overall survival; SDom, Simpson’s Dominance index; SLN, sentinel lymph node.
      All values are P-values.
      Bold values designate statistical significance at P ≤ 0.05.
      Next, we analyzed correlations between TCR quality metrics (SDom and richness; see Materials and Methods) and indices of disease progression. Higher SDom significantly correlated with lower risk of metastasis to any site (regional or distant; Figure 2a), any regional lymph node metastasis (Figure 2b), sentinel lymph node (LN) metastasis (Figure 2c), regional LN metastasis beyond a sentinel LN (Figure 2d), and distant metastasis (Figure 2e). Furthermore, SDom was higher among primary MCCs from patients presenting with stage ≤II (stage I or II) disease compared with primary MCCs from patients presenting with stage >II (stage III or IV) disease (Figure 2f). Consistent with this, SDom was higher in primary MCCs from patients with a final American Joint Committee on Cancer stage ≤II compared with those with a final stage >II (Figure 2g). Even among patients who did develop a metastasis, patients whose tumors had a higher SDom had a longer time to first LN metastasis (Figure 2h). SDom showed similar correlations when patients were separated according to MCPyV status. Higher SDom correlated with reduced propensity to develop any metastasis (P = 0.05), metastasis to any LN (P = 0.05), and lower stage at presentation (P = 0.04) in MCPyV(−) MCCs (Table 2).
      Figure thumbnail gr2
      Figure 2T-cell receptor repertoire correlates with clinical outcome in primary Merkel cell carcinoma. Correlation between SDom and (a) metastasis in general, (b) metastasis to an LN in general, (c) metastasis to an SLN, (d) metastasis beyond an SLN, (e) distant metastasis, (f) stage at presentation, (g) final stage, (h) and time to first LN metastasis. LN, lymph node; MCPyV(+), Merkel cell polyomavirus T-antigen–positive; MCPyV(−), Merkel cell polyomavirus T-antigen–negative; SDom, Simpson’s Dominance index; SLN, sentinel lymph node.
      Richness only correlated significantly with stage at presentation; patients who presented with stage ≤II disease had a less diverse T-cell repertoire than patients presenting with stage >II disease (P = 0.01). No significant correlations were observed between richness and other clinical endpoints (Table 2). Considering MCPyV(−) separate from MCPyV(+) primary MCCs, increased richness was associated with more advanced stage (III–IV) at presentation and at last follow-up (P = 0.004 and P < 0.001, respectively) among primary MCPyV(−) MCCs (Table 2).

      Higher SDom is associated with a lower prevalence of virus-related TCRs

      Populations of T cells unrelated to antitumor responses (bystander T cells) have recently been described in many tumor types, and to the extent that they represent an unfocused antitumor T-cell response, their presence correlates with worse patient outcomes (
      • Reuben A.
      • Gittelman R.
      • Zhang J.
      • Quek K.
      • Vence L.M.
      • Behrens C.
      • et al.
      Association of the T-cell receptor landscape with survival in non-small cell lung cancer.
      ,
      • Simoni Y.
      • Becht E.
      • Fehlings M.
      • Loh C.Y.
      • Koo S.L.
      • Teng K.W.W.
      • et al.
      Bystander CD8+ T cells are abundant and phenotypically distinct in human tumour infiltrates.
      ). To categorize T-cell clonotypes associated with MCC and to evaluate for the presence of nontumor antigen–related T cells in MCCs, TCR sequences were compared against a database of previously characterized CDR3 sequences known to recognize viral antigens (
      • Simoni Y.
      • Becht E.
      • Fehlings M.
      • Loh C.Y.
      • Koo S.L.
      • Teng K.W.W.
      • et al.
      Bystander CD8+ T cells are abundant and phenotypically distinct in human tumour infiltrates.
      ). This revealed that MCCs with a low SDom harbored consistently higher frequencies of Epstein-Barr virus–, cytomegalovirus–, HIV-, and hepatitis C virus–related TCRs compared with MCCs with high SDom (Figure 3a and b; see Supplementary Table S3 for a comprehensive list of the TCR clonotypes and their known viral epitopes). No correlation between MCPyV positivity and presence of viral-directed TCRs was observed (Figure 3c), arguing against cross-reactivity between these viral epitopes and MCPyV epitopes.
      Figure thumbnail gr3
      Figure 3Higher prevalence of virus-associated TCRs is seen in patients with low Simpson’s D. (a) Heat map showing the distribution of viral-associated (bystander) TCR sequences in Simpson’s D high versus low tumors grouped according to MCPyV status. From left to right are the viral-associated TCRs grouped according to relative frequency in the patient population in the MDACC and UDE cohorts (red box indicates presence, white box indicates absence). (b) Frequency of viral-associated (bystander) TCR sequences is higher in Simpson’s D low versus high patient tumors but (c) is not related to MCPyV status. CMV, cytomegalovirus; EBV, Epstein-Barr virus; HBV, hepatitis B virus; HCV, hepatitis C virus; MCPyV, Merkel cell polyomavirus; MDACC, The University of Texas MD Anderson Cancer Center; Simpson’s D, Simpson’s Dominance index; UDE, University of Duisburg-Essen.

      Combination of metrics of T-cell repertoire quality with T-cell quantity is a powerful prognostic indicator

      Building on biomarker work from our group and others (
      • Feldmeyer L.
      • Hudgens C.W.
      • Ray-Lyons G.
      • Nagarajan P.
      • Aung P.P.
      • Curry J.L.
      • et al.
      Density, distribution, and composition of immune infiltrates correlate with survival in Merkel cell carcinoma.
      ,
      • Paulson K.G.
      • Iyer J.G.
      • Simonson W.T.
      • Blom A.
      • Thibodeau R.M.
      • Schmidt M.
      • et al.
      CD8+ lymphocyte intratumoral infiltration as a stage-independent predictor of Merkel cell carcinoma survival: a population-based study.
      ,
      • Paulson K.G.
      • Iyer J.G.
      • Tegeder A.R.
      • Thibodeau R.
      • Schelter J.
      • Koba S.
      • et al.
      Transcriptome-wide studies of merkel cell carcinoma and validation of intratumoral CD8+ lymphocyte invasion as an independent predictor of survival.
      ,
      • Sihto H.
      • Böhling T.
      • Kavola H.
      • Koljonen V.
      • Salmi M.
      • Jalkanen S.
      • et al.
      Tumor infiltrating immune cells and outcome of Merkel cell carcinoma: a population-based study.
      ), we sought to integrate CD3+ and CD8+ T-cell density with measures of the T-cell repertoire to determine whether both T-cell quantity and T-cell quality impact MCC survival. Higher SDom did not significantly associate with longer DSS (Figure 4a and b, Supplementary Figure S1a and b). We previously showed that an increased density of CD3+ and CD8+ T cells at the periphery of primary MCC correlated with improved OS (
      • Feldmeyer L.
      • Hudgens C.W.
      • Ray-Lyons G.
      • Nagarajan P.
      • Aung P.P.
      • Curry J.L.
      • et al.
      Density, distribution, and composition of immune infiltrates correlate with survival in Merkel cell carcinoma.
      ). In this expanded patient cohort, higher peripheral CD3+ or CD8+ T-cell density also significantly correlated with longer DSS (Figure 4c and d and Supplementary Figure S1c and d).
      Figure thumbnail gr4
      Figure 4Quantity and quality of the tumor-associated T-cell infiltrate correlates with survival in primary Merkel cell carcinoma. (a) SDom in patients alive (blue) or dead from disease (red). (b) KMA of DSS in patients with SDom ≥ median (solid line) versus SDom < median (dashed line). (c) CD3+ T cells/mm2 in patients alive (blue) or dead from disease (red). (d) KMA of DSS in patients with CD3+ T cells/mm2 ≥ median (solid line) versus CD3+ T cells/mm2 < median (dashed line). (e) Correlation between SDom and CD3 density in patients alive (blue) or dead from disease (red) (horizontal and vertical lines correspond to median values). (f) KMA of DSS in patients with SDom ≥ median + CD3+ T cells/mm2 ≥ median (solid line) versus SDom < median + CD3+ T cells/mm2 < median) (dashed line). DSS, disease-specific survival; KMA, Kaplan-Meier analysis; SDom, Simpson’s Dominance index.
      To determine whether SDom could further refine prognostic models among the subsets of patients with either high or low CD3+ or CD8+ T-cell densities, we grouped patients into one of four groups: (i) high CD3+ or CD8+ T-cell density and high SDom, (ii) low CD3+ or CD8+ T-cell density and high SDom, (iii) high CD3+ or CD8+ T-cell density and low SDom, and (iv) low CD3+ or CD8+ T-cell density and low SDom. We used the median value of each variable as a cutoff between low and high groups, with a median SDom = 0.00273878, median CD3+ T-cell density = 1,827.37 cells/mm2, and median CD8+ T-cell density = 766.25 cells/mm2. Analysis of these groups revealed that no patient in the high CD3+ or CD8+ T-cell density and high SDom group died during follow-up, whereas patients with only one high parameter (groups ii and iii) or low values for both parameters (group iv) fared more poorly (P = 0.002 and P = 0.03, respectively; Figure 4e and f, Supplementary Figure S1e and f). When we grouped patients according to stage at presentation (I/II versus III/IV), identical patterns of DSS were observed among patients presenting at higher stage (III/IV) when comparing those with high CD3+/CD8+ T-cell density and high SDom with those with low CD3+/CD8+ T-cell density and low SDom, although this did not achieve statistical significance (Supplementary Figure S3b and d). T-cell metrics did not appear to impact DSS of patients presenting at lower disease stages (Supplementary Figure S3a and c).

      Discussion

      In this study, we applied TCR sequencing to 72 primary MCCs to determine the prognostic impact of the quality of the MCC-associated T-cell repertoire. Primary MCCs with a higher SDom produced fewer metastases and had lower disease stage at presentation. As a reflection of this, high SDom MCCs contained fewer bystander T cells. Moreover, MCCs with both a high density of either CD3+ or CD8+ T cells and a high SDom were associated with longer MCC-specific survival compared with MCCs with a low density of either CD3+ or CD8+ T cells and low SDom. SDom modified the survival defined according to the quantity of the T-cell infiltrate; high SDom improved survival of MCCs with low density CD3+ or CD8+ T-cell infiltrates, whereas a low SDom adversely affected survival in MCCs with high density of CD3+ or CD8+ T cells. Together, these findings demonstrate that not only the quantity but also the quality (i.e., the extent to which that T-cell infiltrate represents an antigen-modified immune response) impacts survival in primary MCC.
      By recognizing tumor antigens through their TCR, T cells play a crucial role in regulating tumor growth and progression (
      • Boon T.
      • Coulie P.G.
      • Van den Eynde B.
      Tumor antigens recognized by T cells.
      ,
      • Echchakir H.
      • Vergnon I.
      • Dorothée G.
      • Grunenwald D.
      • Chouaib S.
      • Mami-Chouaib F.
      Evidence for in situ expansion of diverse antitumor-specific cytotoxic T lymphocyte clones in a human large cell carcinoma of the lung.
      ,
      • Karanikas V.
      • Colau D.
      • Baurain J.F.
      • Chiari R.
      • Thonnard J.
      • Gutierrez-Roelens I.
      • et al.
      High frequency of cytolytic T lymphocytes directed against a tumor-specific mutated antigen detectable with HLA tetramers in the blood of a lung carcinoma patient with long survival.
      ,
      • Pagès F.
      • Berger A.
      • Camus M.
      • Sanchez-Cabo F.
      • Costes A.
      • Molidor R.
      • et al.
      Effector memory T cells, early metastasis, and survival in colorectal cancer.
      ). As such, an analysis of the T-cell repertoire associated with a tumor sample provides insight into antigen-specific immune responses (
      • Cui J.H.
      • Lin K.R.
      • Yuan S.H.
      • Jin Y.B.
      • Chen X.P.
      • Su X.K.
      • et al.
      TCR repertoire as a novel indicator for immune monitoring and prognosis assessment of patients with cervical cancer.
      ,
      • Hodi F.S.
      • Dranoff G.
      The biologic importance of tumor-infiltrating lymphocytes.
      ,
      • Hou D.
      • Chen C.
      • Seely E.J.
      • Chen S.
      • Song Y.
      High-throughput sequencing-based immune repertoire study during infectious disease.
      ,
      • Reuben A.
      • Gittelman R.
      • Gao J.
      • Zhang J.
      • Yusko E.C.
      • Wu C.J.
      • et al.
      TCR repertoire intratumor heterogeneity in localized lung adenocarcinomas: an association with predicted neoantigen heterogeneity and postsurgical recurrence.
      ). The T-cell repertoire lacks dominant clones in physiologic conditions, whereas in disease states, disease-specific antigens drive T-cell expansion resulting in dominance of those reactive clones (
      • Hodges E.
      • Krishna M.T.
      • Pickard C.
      • Smith J.L.
      Diagnostic role of tests for T cell receptor (TCR) genes.
      ). Thus, we hypothesized that an effective antitumor T-cell response would contain a T-cell repertoire dominated by specific clones primed and expanded because of exposure to immunogenic MCC antigens. In this context, a more focused (less rich) T-cell repertoire enriched for certain clonotypes would be predicted to correlate with a better prognosis (higher SDom).
      Our results align with this hypothesis. Primary MCCs with a higher SDom exhibited a reduced propensity to metastasize and were of lower disease stage at presentation and last follow-up. Even among patients who did develop metastases, higher SDom correlated with longer time to develop first metastasis. One reflection of the apparent robustness of high SDom T-cell infiltrates was the apparent reduction in prevalence of T cells with TCRs directed against presumed nontumor (viral) antigens. Such T cells have been termed bystander T cells, as they are considered unrelated to the tumor and therefore likely do not contribute to antitumor responses. Their enrichment in low compared with high SDom MCCs further argues that a high SDom reflects an effective, tumor antigen–focused immune response, whereas a low SDom MCC contains a higher frequency of TCRs directed against tumor-unrelated antigens. Taken together, the quality of the T-cell repertoire in MCC strongly impacts a patient’s disease course. Consistent with this,
      • Miller N.J.
      • Church C.D.
      • Dong L.
      • Crispin D.
      • Fitzgibbon M.P.
      • Lachance K.
      • et al.
      Tumor-infiltrating Merkel cell polyomavirus-specific T cells are diverse and associated with improved patient survival.
      found that enrichment of specific T-cell clones against a single epitope of MCPyV(+) MCC correlated with improved survival. These studies only focused on KLL-specific T cells and not population-based metrics of the complete TCR repertoire. We did not observe any significant overlap between our CDR3 TCR sequences and those identified by Miller et al., who showed mostly private, distinct KLL-specific TCRs and, as such, suggested that epitope recognition can be achieved through a variety of different TCR sequences. The finding that our sequences did not overlap with theirs further supports this contention.
      Increased intratumoral CD8+ T-cell infiltration was associated with improved MCC-specific survival independent of stage at presentation (
      • Paulson K.G.
      • Iyer J.G.
      • Tegeder A.R.
      • Thibodeau R.
      • Schelter J.
      • Koba S.
      • et al.
      Transcriptome-wide studies of merkel cell carcinoma and validation of intratumoral CD8+ lymphocyte invasion as an independent predictor of survival.
      ). Increased CD8+ T-cell infiltration was also associated with improved DSS in 137 MCCs (
      • Paulson K.G.
      • Iyer J.G.
      • Simonson W.T.
      • Blom A.
      • Thibodeau R.M.
      • Schmidt M.
      • et al.
      CD8+ lymphocyte intratumoral infiltration as a stage-independent predictor of Merkel cell carcinoma survival: a population-based study.
      ). Increased intratumoral CD3+ and CD8+ T-cell densities correlated with improved OS in 116 MCCs (
      • Sihto H.
      • Böhling T.
      • Kavola H.
      • Koljonen V.
      • Salmi M.
      • Jalkanen S.
      • et al.
      Tumor infiltrating immune cells and outcome of Merkel cell carcinoma: a population-based study.
      ). Similarly, we showed that increased CD3+ and CD8+ T-cell densities at the tumor periphery correlated with improved OS and DSS (
      • Feldmeyer L.
      • Hudgens C.W.
      • Ray-Lyons G.
      • Nagarajan P.
      • Aung P.P.
      • Curry J.L.
      • et al.
      Density, distribution, and composition of immune infiltrates correlate with survival in Merkel cell carcinoma.
      ). Together, these studies underscored the prognostic significance of the quantity of the MCC-associated T-cell infiltrate. Further, they provided a strong rationale to explore the antigen specificity of the MCC-associated T-cell infiltrate and the relative contribution of the quality of those T cells to patient survival. In this context, the degree of clonal expansion of the T-cell repertoire in response to tumor-specific antigens is envisioned to reflect that quality. Although we confirmed that a higher density of CD3+ and CD8+ T cells correlates with improved DSS, incorporating SDom further stratified patient survival. Our data shows that high CD3+ and CD8+ T-cell densities alone identified a cohort of patients with differential survival (∼82% vs. ∼42% for CD3 and ∼82% vs. ∼45% for CD8 at 50 months). The integration of SDom with CD3+ and CD8+ T-cell density instead identified a cohort of patients with 100% DSS (high CD3/CD8 and high SDom) after 50 months compared with 20% for those with low CD3 and low SDom and 32% for those with low CD8 and low SDom. Therefore, patients whose tumors had high density CD3+ or CD8+ T cells had better survival with a high SDom. Similarly, patients whose tumors had low CD3+ or CD8+ T-cell density had a survival advantage if the tumor had a high SDom, and patients with a high CD3+ or CD8+ T-cell infiltrate had slightly worse survival if their SDom was low. Patients whose tumors had both low CD3+ or CD8+ T-cell density and low SDom had the worst survival. Furthermore, because higher disease stage at presentation was associated with lower SDom, and stage is a well-established prognostic factor in MCC, we thought to assess the impact of T-cell density and SDom on DSS in the context of stage at presentation. We found a slight survival advantage among patients presenting with stage III/IV when comparing those with high CD3/CD8 T-cell density and high SDom to those with low CD3/CD8 T-cell density and low SDom, although this did not achieve statistical significance, possibly related to the small cohort size. Therefore, the possibility of a confounding effect of disease stage and SDom on survival requires further investigation in a larger cohort.
      Our study has important limitations. As we relied on formalin-fixed, paraffin-embedded tissue resources, exhaustive functional studies aimed at identifying the antigens targeted by dominant TCR clones were not possible. Patients in our cohort were not subject to uniform treatment strategies; thus, the possible impact of these therapeutic interventions could not be systematically captured. The patients were treated in the preimmunotherapy era dominated by surgery, radiation, and cytotoxic chemotherapy; thus, the nature of their response in the present era of checkpoint inhibitors, cytokines, and vaccines is unknown. In addition, given the exploratory nature of the study and small sample size, we were not sufficiently powered to perform robust corrections for multiple hypothesis testing, and this may increase the likelihood of obtaining false positive results. Research on the clinical impact of the quality of the tumor-associated T-cell infiltrate, including its association with patients’ characteristics and clinical outcomes, is an emerging concept, further compounded by the rarity of MCC itself. Our study did not begin with a clear predefined primary hypothesis. Instead, we systematically evaluated associations between the tumor-associated T-cell repertoire and other clinical variables by first testing many comparisons. We recognized early in this analysis that SDom emerged as a consistent prognostic variable and therefore focused the remainder of the study on associations with SDom and clinical outcome. Some of the variables were indeed confounding (such as the association between high SDom and sentinel LN metastasis and stage at presentation). Although the number of cases included in this study is comparatively large for a rare tumor like MCC, it is statistically insufficient and exploratory in nature for robust multiple hypothesis testing (
      • Althouse A.D.
      Adjust for multiple comparisons? It's not that simple.
      ). A subsequent study with clearly defined preplanned hypotheses should be conducted to confirm our findings, which we believe identifies a bona fide group of patients with MCC that have a superior survival compared with those who do not. Finally, although the presence of bystander T cells was enriched in low SDom tumors, these tumors also generally showed a commensurate increase in TCR richness. Thus, the possibility that these viral clonotypes were enriched purely by chance cannot be formally excluded.
      Despite these limitations, our findings have important translational implications. First, the extent to which SDom impacts prognosis supports that both quantity and quality of the tumor-directed T-cell infiltrate are important in MCC. Additional studies will be critical to identify the discrete antigens against which efficacious T-cell infiltrates are directed and refine this prognostic model. Second, patients with primary MCC containing high CD3+ or CD8+ T-cell density and high SDom appear to have a distinctively superior survival advantage. Thus, as management strategies for MCC evolve, these patients might be considered for less aggressive or toxic interventions in future clinical trials. Third, analysis of the T-cell repertoire may also be leveraged as a biomarker predictive of response to MCC therapy (
      • Durgeau A.
      • Virk Y.
      • Corgnac S.
      • Mami-Chouaib F.
      Recent advances in targeting CD8 T-cell immunity for more effective cancer immunotherapy.
      ,
      • Kuang M.
      • Cheng J.
      • Zhang C.
      • Feng L.
      • Xu X.
      • Zhang Y.
      • et al.
      A novel signature for stratifying the molecular heterogeneity of the tissue-infiltrating T-cell receptor repertoire reflects gastric cancer prognosis.
      ,
      • Li B.
      • Li T.
      • Pignon J.C.
      • Wang B.
      • Wang J.
      • Shukla S.A.
      • et al.
      Landscape of tumor-infiltrating T cell repertoire of human cancers.
      ) or to sequester high-risk patients for future novel combinations of immune checkpoint inhibitor therapy, considering the association between more clonal pretreatment T-cell repertoire and improved response to pembrolizumab in metastatic melanoma (
      • Tumeh P.C.
      • Harview C.L.
      • Yearley J.H.
      • Shintaku I.P.
      • Taylor E.J.
      • Robert L.
      • et al.
      PD-1 blockade induces responses by inhibiting adaptive immune resistance.
      ). Although inhibitors of the PD-1/PD-L1 axis have recently become the standard of care for advanced MCC (
      • Kaufman H.L.
      • Russell J.
      • Hamid O.
      • Bhatia S.
      • Terheyden P.
      • D'Angelo S.P.
      • et al.
      Avelumab in patients with chemotherapy-refractory metastatic Merkel cell carcinoma: a multicentre, single-group, open-label, phase 2 trial.
      ,
      • Nghiem P.T.
      • Bhatia S.
      • Lipson E.J.
      • Kudchadkar R.R.
      • Miller N.J.
      • Annamalai L.
      • et al.
      PD-1 blockade with pembrolizumab in advanced Merkel-cell carcinoma.
      ), robust biomarkers predictive of response remain elusive. Additional studies are needed to determine if metrics of the T-cell repertoire in pre- or on-treatment biopsies also predict response to immune checkpoint inhibitors.
      In summary, our findings build upon emerging evidence that both the quantity as well as the quality of the tumor-associated T-cell infiltrate represents a critical prognostic variable in MCC. Further studies are required to validate these findings and investigate the relevance of T-cell repertoire metrics to stratify patients with MCC for more aggressive therapy versus observation. Our findings additionally provide the framework for further exploration of the T-cell repertoire in other tumor types to stratify patient outcomes according to similar metrics. Although our findings may not impact immediate practice, they do represent, to our knowledge, a previously unreported objective biomarker that more accurately predicts MCC DSS and delineates patients who may require more aggressive clinical interventions from those who would benefit from simple observation and be spared potentially toxic side effects. Such personalized surveillance and management strategies are critical ultimately to improve outcomes in this aggressive cutaneous malignancy and also to ensure best utilization of health care resources and minimize possible toxicity.

      Materials and Methods

      Case selection

      With approval from the institutional review boards of MDACC and UDE, records of patients diagnosed with primary MCC during 2002–2015 and 1999–2012 for MDACC and UDE, respectively, were reviewed. Only patients with sufficient pathologic material for histological assessment and DNA extraction were included. MCPyV status was determined by immunohistochemistry for the MCPyV T antigen for MDACC (Santa Cruz Biotechnology, Dallas, TX; sc-136172; CM2B4; 1:100) or real-time PCR for UDE cases (
      • Becker J.C.
      • Houben R.
      • Ugurel S.
      • Trefzer U.
      • Pföhler C.
      • Schrama D.
      MC polyomavirus is frequently present in Merkel cell carcinoma of European patients.
      ). MCPyV status could not be obtained for three tumors. Sufficient tissue was available to test 12 of the 29 primary MCCs from the UDE using CM2B4 mouse monoclonal antibody. The originally designated MCPyV status was confirmed in 11 of 12 MCCs from UDE when comparing the PCR result to CM2B4 immunohistochemistry; thus, we do not believe that the manner in which MCPyV status was designated impacts our conclusions. Clinical variables collected included age; sex; immune status; primary tumor site; site of metastases, if present (sentinel LN, other LN beyond the sentinel LN, skin, viscera, and central nervous system); time to first LN metastasis; time to first distant metastasis; stage at presentation; final stage; OS; and DSS. Disease was staged according to the 8th edition of the American Joint Committee on Cancer staging system (
      • Harms K.L.
      • Healy M.A.
      • Nghiem P.
      • Sober A.J.
      • Johnson T.M.
      • Bichakjian C.K.
      • et al.
      Analysis of prognostic factors from 9387 Merkel cell carcinoma cases forms the basis for the new 8th edition AJCC staging system.
      ). Histopathologic variables included tumor size, tumor thickness, and presence and density of tumor-infiltrating lymphocytes (
      • Andea A.A.
      • Coit D.G.
      • Amin B.
      • Busam K.J.
      Merkel cell carcinoma: histologic features and prognosis.
      ).

      TCR CDR3β sequencing

      DNA was extracted from 10–15 5-μm sections of formalin-fixed, paraffin-embedded tissue from primary MCCs using the QIAamp DNA FFPE tissue kit (Qiagen, Hilden, Germany), using a matched H&E-stained slide to direct microdissection of tumor plus a surrounding 5-mm area of the tumor-associated stroma/tumor-infiltrating lymphocytes. Sequencing of the variable CDR3β region of the TCR was performed using immunoSEQ (Adaptive Biotechnologies, Seattle, WA) (
      • Robins H.S.
      • Srivastava S.K.
      • Campregher P.V.
      • Turtle C.J.
      • Andriesen J.
      • Riddell S.R.
      • et al.
      Overlap and effective size of the human CD8+ T cell receptor repertoire.
      ). For each primary MCC, two principle metrics reflected the quality of the TCR repertoire, richness and SDom. Richness measures T-cell diversity and reflects the number of unique tumor-associated T cells possessing distinct V-J rearrangements. Richness is defined as the ratio between the number of observed rearrangements in a sample and the number of hypothetical rearrangements possible between V-J genes. SDom measures reactivity or the extent to which the TCR repertoire reflects an antigen-driven population. Over the total set of observed rearrangements, SDom represents the sum of the square fractional abundances of each TCR rearrangement present in a sample (
      • Simpson E.H.
      Measurement of diversity.
      ). Higher SDom is reflective of a more robust antitumor T-cell response, because a higher SDom is observed when a subset of T cells (presumably tumor-directed) predominates in a given tumor/T-cell microenvironment. Values for SDom range from 0 to 1. A score approaching 0 corresponds to a polyclonal, infinitely large, perfectly even repertoire, whereas a score approaching 1 corresponds to a nearly monoclonal sample in which one or a few clones dominate. In our series, richness ranged from 21 to 24,939 (median, 687) and SDom from 0 to 0.05 (median, 0.003). Because our initial exploratory analyses revealed consistent statistically significant relationships between SDom (but not richness or clonality) and measures of clinical outcome, we focused our subsequent analyses on SDom.
      For analysis of virus-related CDR3 sequences, the Sequence Search Tool included in Adaptive Biotechnologies' Analyzer was utilized to screen for discrete viral specific clonotypes in each sample.

      Immune profiling

      To determine CD8+ and CD3+ T-cell density, immunohistochemical studies were performed on a Leica Bond autostainer for CD3 (Dako, Carpinteria, CA; A0452; 1:100) or CD8 (Life Sciences Technologies, Waltham, MA; MS457s; 1:25) and 3,3'-diaminobenzidine chromogen and quantified using Aperio Scanscope AT Turbo, as described (
      • Feldmeyer L.
      • Hudgens C.W.
      • Ray-Lyons G.
      • Nagarajan P.
      • Aung P.P.
      • Curry J.L.
      • et al.
      Density, distribution, and composition of immune infiltrates correlate with survival in Merkel cell carcinoma.
      ) (See Supplementary Figure S2 for a schematic as to this methodology).

      Statistical analyses

      Clinical and pathologic variables were summarized using descriptive statistics. Pearson’s correlation coefficients assessed associations between continuous variables. T-tests and Wilcoxon tests compared means and medians of continuous variables between groups, respectively. Fisher’s exact tests tested distributions of categorical variables across groups. Kaplan-Meier method and log-rank tests compared distributions of survival outcomes. Cox proportional hazards models evaluated risk factors associated with the survival outcomes. Time to first LN or distant metastasis was calculated from the date of diagnosis of primary MCC. OS was defined as the time from diagnosis of MCC to death owing to any cause, whereas DSS was defined as the time from diagnosis of MCC to death owing to MCC. For survival analyses, we grouped patients according to their TCR repertoire metrics and/or peripheral CD3+ or CD8+ T-cell density into high and low groups using the median value of each metric as the cutoff. Statistical analyses were performed using R, version 3.4.3, and plots were generated using GraphPad Prism 8.0.

      Data availability statement

      Datasets related to this article will be available at Adaptive immuneACCESS database. DOI 10.21417/MF2020JID. https://clients.adaptivebiotech.com/pub/farah-2020-jid.

      Conflict of Interest

      SU declares research support from Bristol-Myers Squibb and Merck Serono; speakers and advisory board honoraria from Bristol-Myers Squibb, Merck Sharp & Dohme, Merck Serono, Novartis, and Roche; and travel support from Bristol-Myers Squibb and Merck Sharp & Dohme. DS reports related and unrelated advisory board honoraria from Roche, Novartis, Bristol-Myers Squibb, Merck Sharp & Dome, Amgen, Incyte, Merck Serono, and Pierre-Fabre as well as research funding from Novartis and Bristol-Myers Squibb. MKKW reports advisory board affiliation with EMD-Serono, Pfizer, and Bristol Meyers Squibb. JW is an inventor on a US patent application (PCT/US17/53.717) submitted by the University of Texas MD Anderson Cancer Center that covers methods to enhance immune checkpoint blockade responses by modulating the microbiome. JW reports compensation for speaker’s bureau and honoraria from Imedex, Dava Oncology, Omniprex, Illumina, Gilead, PeerView, Physician Education Resource, MedImmune, Exelixis, and Bristol-Myers Squibb and serves as a consultant/advisory board member for Roche/Genentech, Novartis, AstraZeneca, GlaxoSmithKline, Bristol-Myers Squibb, Merck, Biothera Pharmaceuticals, and Microbiome DX. JW also receives research support from GlaxoSmithKline, Roche/Genentech, Bristol-Myers Squibb, and Novartis. JCB reports related and unrelated speaker’s bureau honoraria from Amgen, Pfizer, MerckSerono, and Sanofi; is a paid consultant/advisory board member for eTheRNA, MerckSerono, Pfizer, 4SC, and Sanofi; and has received research funding from Alcedis, Amgen, Bristol-Myers Squibb, and Merck Serono. MTT reports unrelated advisory board relationships with Myriad Genetics, Seattle Genetics, and Novartis LLC.

      Acknowledgments

      Alexandre Reuben, Jürgen C. Becker, and Michael T. Tetzlaff jointly supervised this study and are all equal contributors for correspondence. The authors thank Ms Kim Vu for her assistance with figures and Ms Stephanie Deming for her editorial assistance.

      Author Contributions

      Conceptualization: MT, JB, AR; Data Curation: MF; Formal Analysis: MF, JN, IS, RY, WL, LK; Funding Acquisition: MT, JB; Investigation: MF, AR, JB, MT, CG, AF, JW; Methodology: MF, MT, AR; Project Administration: MT; Resources: MT, JB, AR; Supervision: MT, JB; Validation: MF, AR, MT; Visualization: MF, AR, MT; Writing - Original Draft Preparation: MF, AR; Writing - Review and Editing: MT, JB, IS, RY, LK, PN, JN, WL, PA, JC, CT, CH, SU, DS, CG, LL, AF, IW, VP, LW, MW, JW

      Supplementary Material

      Figure thumbnail fx1
      Supplementary Figure S1Quantity and quality of the tumor-associated T-cell infiltrate correlates with survival in primary Merkel cell carcinoma. (a) SDom in patients alive (blue) or dead from disease (red). (b) KMA of DSS in patients with SDom ≥ median (solid line) versus SDom < median (dashed line). (c) CD8+ T cells/mm2 in patients alive (blue) or dead from disease (red). (d) KMA of DSS in patients with CD8+ T cells/mm2 ≥ median (solid line) versus CD8+ T cells/mm2 < median (dashed line). (e) Correlation between SDom and CD8 density in patients alive (blue) or dead from disease (red) (horizontal and vertical lines correspond to median values). (f) KMA of DSS in patients with SDom ≥ median + CD8+ T cells/mm2 ≥ median (solid line) versus SDom < median + CD8+ T cells/mm2 < median (dashed line). DSS, disease-specific survival; KMA, Kaplan-Meier analysis; SDom, Simpson’s Dominance index.
      Figure thumbnail fx2
      Supplementary Figure S2Quantification of CD8 in primary Merkel cell carcinoma. CD8 high primary Merkel cell carcinoma. (a) Low power view of primary Merkel cell carcinoma with high CD8+ T-cell density. Green boxes indicate 1-mm2 areas for quantification (×40). Bar = 20 μm. (b) Higher power view of 1-mm2 area for CD8 quantification (×200). Bar = 100 μm. (c) Higher power view of 1-mm2 area for CD8 quantification with quantification algorithm applied. Positive cells are designated by yellow, brown, and orange, and negative cells are designated by blue. Densities reported as number of CD8+ cells/mm2 (×200). Bar = 100 μm. CD8 low primary Merkel cell carcinoma. (d) Low power view of primary Merkel cell carcinoma with low CD8+ T-cell density. Green boxes indicate 1-mm2 areas for quantification (×40). Bar = 20 μm. (e) Higher power view of 1-mm2 area for CD8 quantification (×200). Bar = 100 μm. (f) Higher power view of 1-mm2 area for CD8 quantification with quantification algorithm applied. Positive cells are designated by yellow, brown, and orange, and negative cells are designated by blue. Densities reported as number of CD8+ cells/mm2 (×200). Bar = 100 μm.
      Figure thumbnail fx3
      Supplementary Figure S3DSS according to stage at presentation, CD3/CD8 density, and SDom. (a) Stage I/II DSS plotted according to CD3 high and SDom high (dashed line; n = 5 patients), CD3 high/SDom low and CD3 low/SDom high (dotted line; n = 17 patients), or CD3 low and SDom low (n = 0 patients). (b) Stage III/IV DSS plotted according to CD3 high and SDom high (dashed line; n = 5 patients), CD3 high/SDom low and CD3 low/SDom high (dotted line; n = 15 patients), or CD3 low and SDom low (solid line; n = 9 patients). (c) Stage I/II DSS plotted according to CD8 high and SDom high (dashed line; n = 5 patients), CD8 high/SDom low and CD8 low/SDom high (dotted line; n = 17 patients), or CD8 low and SDom low (solid line; n = 1 patients). (d) Stage III/IV DSS plotted according to CD8 high and SDom high (dashed line; n = 6 patients), CD8 high/SDom low and CD8 low/SDom high (dotted line; n = 15 patients), or CD8 low and SDom low (solid line; n = 8 patients). DSS, disease-specific survival; SDom, Simpson’s Dominance index.
      Supplementary Table S1Clinicopathologic Characteristics and Outcome Variables in the MDACC and UDE Cohorts
      CharacteristicMDACC (n = 43)UDE (n = 29)P
      Age, n (%)43 (100)29 (100)0.46
       Median (range), years72 (32, 91)74 (55, 88)
      Sex, n (%)43 (100)29 (100)0.22
       Male30 (70)16 (55)
       Female13 (30)13 (45)
      Primary tumor site, n (%)43 (100)25 (86)0.09
       Head and neck14 (33)12 (48)
       Trunk8 (18)4 (16)
       Upper extremity14 (33)2 (8)
       Lower extremity7 (16)7 (28)
      MCPyV status, n (%)43 (100)26 (90)0.22
       MCPyV(+)23 (53)18 (69)
       MCPyV(−)20 (47)8 (31)
      Tumor size, n (%)43 (100)26 (90)0.49
       Median (range), mm16 (3, 93)19 (4, 20)
      Tumor thickness, n (%)43 (100)26 (90)0.56
       Median (range), mm8.9 (1.6, 22.5)8 (4, 20)
      Any metastasis, n (%)43 (100)24 (83)1.00
       Yes28 (65)16 (67)
       No15 (35)8 (33)
      Metastasis to SLN, n (%)43 (100)29 (100)0.60
       Yes13 (30)7 (24)
       No30 (70)22 (76)
      Metastasis beyond SLN, n (%)43 (100)22 (76)0.80
       Yes23 (53)11 (50)
       No20 (47)11 (50)
      Distant metastasis, n (%)43 (100)21 (72)0.58
       Yes14 (33)9 (43)
       No29 (67)12 (57)
      Time to first LN metastasis, n (%)27 (63)11 (38)0.19
       Median (range), months0 (0, 22)0 (0, 27)
      Time to first distant metastasis, n (%)14 (33)8 (28)0.49
       Median (range), months13 (0, 41)8.5 (0, 15)
      AJCC stage at presentation, n (%)43 (100)29 (100)0.53
       I–II19 (44)15 (52)
       III22 (51)11 (38)
       IV2 (5)3 (10)
      Final AJCC stage, n (%)43 (100)21 (72)0.84
       I–II15 (35)8 (38)
       III13 (30)4 (19)
       IV15 (35)9 (43)
      Duration of follow-up, n (%)43 (100)21 (72)0.78
       Median (range), months31 (0, 176)25 (8, 115)
      Vital status at last follow-up, n (%)43 (100)28 (100)<0.001
       Died of disease10 (23)7 (24)
       Died of other/unknown cause12 (28)8 (28)
       Alive with disease2 (5)5 (17)
       Alive and disease free18 (42)0 (0)
       Lost to follow-up1 (2)9 (31)
      Abbreviations: AJCC, American Joint Committee on Cancer; LN, lymph node; MCPyV, Merkel cell polyomavirus; MCPyV(+), Merkel cell polyomavirus T-antigen–positive; MCPyV(−), Merkel cell polyomavirus T-antigen–negative; MDACC, MD Anderson Cancer Center; SLN, sentinel lymph node; UDE, University of Duisburg-Essen.
      Bold value designates statistical significance at P < 0.05.
      Supplementary Table S2TCR Variables and Quantification of T-Cell Infiltrates in Immunocompetent Versus Immunosuppressed Patients
      VariableImmunocompetentImmunosuppressedP-value
      nMediannMedian
      SDom600.002748547120.0026901380.93
      Richness60692.512626.50.90
      CD8 (cells/mm2)43746.46921639971.00391810.81
      CD3 (cells/mm2)441,852.833992111,525.2540490.40
      SDom, Simpson’s Dominance index; TCR, T-cell receptor.

      References

        • Albores-Saavedra J.
        • Batich K.
        • Chable-Montero F.
        • Sagy N.
        • Schwartz A.M.
        • Henson D.E.
        Merkel cell carcinoma demographics, morphology, and survival based on 3870 cases: a population based study.
        J Cutan Pathol. 2010; 37: 20-27
        • Althouse A.D.
        Adjust for multiple comparisons? It's not that simple.
        Ann Thorac Surg. 2016; 101: 1644-1645
        • Andea A.A.
        • Coit D.G.
        • Amin B.
        • Busam K.J.
        Merkel cell carcinoma: histologic features and prognosis.
        Cancer. 2008; 113: 2549-2558
        • Asioli S.
        • Righi A.
        • de Biase D.
        • Morandi L.
        • Caliendo V.
        • Picciotto F.
        • et al.
        Expression of p63 is the sole independent marker of aggressiveness in localised (stage I-II) Merkel cell carcinomas.
        Mod Pathol. 2011; 24: 1451-1461
        • Becker J.C.
        • Houben R.
        • Ugurel S.
        • Trefzer U.
        • Pföhler C.
        • Schrama D.
        MC polyomavirus is frequently present in Merkel cell carcinoma of European patients.
        J Invest Dermatol. 2009; 129: 248-250
        • Becker J.C.
        • Stang A.
        • DeCaprio J.A.
        • Cerroni L.
        • Lebbé C.
        • Veness M.
        • et al.
        Merkel cell carcinoma.
        Nat Rev Dis Primers. 2017; 3: 17077
        • Boon T.
        • Coulie P.G.
        • Van den Eynde B.
        Tumor antigens recognized by T cells.
        Immunol Today. 1997; 18: 267-268
        • Cha E.
        • Klinger M.
        • Hou Y.
        • Cummings C.
        • Ribas A.
        • Faham M.
        • et al.
        Improved survival with T cell clonotype stability after anti-CTLA-4 treatment in cancer patients.
        Sci Transl Med. 2014; 6: 238ra70
        • Coggshall K.
        • Tello T.L.
        • North J.P.
        • Yu S.S.
        Merkel cell carcinoma: An update and review: pathogenesis, diagnosis, and staging.
        J Am Acad Dermatol. 2018; 78: 433-442
        • Cui J.H.
        • Lin K.R.
        • Yuan S.H.
        • Jin Y.B.
        • Chen X.P.
        • Su X.K.
        • et al.
        TCR repertoire as a novel indicator for immune monitoring and prognosis assessment of patients with cervical cancer.
        Front Immunol. 2018; 9: 2729
        • Durgeau A.
        • Virk Y.
        • Corgnac S.
        • Mami-Chouaib F.
        Recent advances in targeting CD8 T-cell immunity for more effective cancer immunotherapy.
        Front Immunol. 2018; 9: 14
        • Echchakir H.
        • Vergnon I.
        • Dorothée G.
        • Grunenwald D.
        • Chouaib S.
        • Mami-Chouaib F.
        Evidence for in situ expansion of diverse antitumor-specific cytotoxic T lymphocyte clones in a human large cell carcinoma of the lung.
        Int Immunol. 2000; 12: 537-546
        • Engels E.A.
        • Frisch M.
        • Goedert J.J.
        • Biggar R.J.
        • Miller R.W.
        Merkel cell carcinoma and HIV infection.
        Lancet. 2002; 359: 497-498
        • Feldmeyer L.
        • Hudgens C.W.
        • Ray-Lyons G.
        • Nagarajan P.
        • Aung P.P.
        • Curry J.L.
        • et al.
        Density, distribution, and composition of immune infiltrates correlate with survival in Merkel cell carcinoma.
        Clin Cancer Res. 2016; 22: 5553-5563
        • Feng H.
        • Shuda M.
        • Chang Y.
        • Moore P.S.
        Clonal integration of a polyomavirus in human Merkel cell carcinoma.
        Science. 2008; 319: 1096-1100
        • Fitzgerald T.L.
        • Dennis S.
        • Kachare S.D.
        • Vohra N.A.
        • Wong J.H.
        • Zervos E.E.
        Dramatic increase in the incidence and mortality from Merkel cell carcinoma in the United States.
        Am Surg. 2015; 81: 802-806
        • Fleming K.E.
        • Ly T.Y.
        • Pasternak S.
        • Godlewski M.
        • Doucette S.
        • Walsh N.M.
        Support for p63 expression as an adverse prognostic marker in Merkel cell carcinoma: report on a Canadian cohort.
        Hum Pathol. 2014; 45: 952-960
        • Goh G.
        • Walradt T.
        • Markarov V.
        • Blom A.
        • Riaz N.
        • Doumani R.
        • et al.
        Mutational landscape of MCPyV-positive and MCPyV-negative Merkel cell carcinomas with implications for immunotherapy.
        Oncotarget. 2016; 7: 3403-3415
        • Güler-Nizam E.
        • Leiter U.
        • Metzler G.
        • Breuninger H.
        • Garbe C.
        • Eigentler T.K.
        Clinical course and prognostic factors of Merkel cell carcinoma of the skin.
        Br J Dermatol. 2009; 161: 90-94
        • Harms K.L.
        • Healy M.A.
        • Nghiem P.
        • Sober A.J.
        • Johnson T.M.
        • Bichakjian C.K.
        • et al.
        Analysis of prognostic factors from 9387 Merkel cell carcinoma cases forms the basis for the new 8th edition AJCC staging system.
        Ann Surg Oncol. 2016; 23: 3564-3571
        • Harms P.W.
        • Vats P.
        • Verhaegen M.E.
        • Robinson D.R.
        • Wu Y.M.
        • Dhanasekaran S.M.
        • et al.
        The distinctive mutational spectra of polyomavirus-negative Merkel cell carcinoma.
        Cancer Res. 2015; 75: 3720-3727
        • Hodges E.
        • Krishna M.T.
        • Pickard C.
        • Smith J.L.
        Diagnostic role of tests for T cell receptor (TCR) genes.
        J Clin Pathol. 2003; 56: 1-11
        • Hodi F.S.
        • Dranoff G.
        The biologic importance of tumor-infiltrating lymphocytes.
        J Cutan Pathol. 2010; 37: 48-53
        • Hou D.
        • Chen C.
        • Seely E.J.
        • Chen S.
        • Song Y.
        High-throughput sequencing-based immune repertoire study during infectious disease.
        Front Immunol. 2016; 7: 336
        • Karanikas V.
        • Colau D.
        • Baurain J.F.
        • Chiari R.
        • Thonnard J.
        • Gutierrez-Roelens I.
        • et al.
        High frequency of cytolytic T lymphocytes directed against a tumor-specific mutated antigen detectable with HLA tetramers in the blood of a lung carcinoma patient with long survival.
        Cancer Res. 2001; 61: 3718-3724
        • Kaufman H.L.
        • Russell J.
        • Hamid O.
        • Bhatia S.
        • Terheyden P.
        • D'Angelo S.P.
        • et al.
        Avelumab in patients with chemotherapy-refractory metastatic Merkel cell carcinoma: a multicentre, single-group, open-label, phase 2 trial.
        Lancet Oncol. 2016; 17: 1374-1385
        • Kuang M.
        • Cheng J.
        • Zhang C.
        • Feng L.
        • Xu X.
        • Zhang Y.
        • et al.
        A novel signature for stratifying the molecular heterogeneity of the tissue-infiltrating T-cell receptor repertoire reflects gastric cancer prognosis.
        Sci Rep. 2017; 7: 7762
        • Li B.
        • Li T.
        • Pignon J.C.
        • Wang B.
        • Wang J.
        • Shukla S.A.
        • et al.
        Landscape of tumor-infiltrating T cell repertoire of human cancers.
        Nat Genet. 2016; 48: 725-732
        • Miller N.J.
        • Church C.D.
        • Dong L.
        • Crispin D.
        • Fitzgibbon M.P.
        • Lachance K.
        • et al.
        Tumor-infiltrating Merkel cell polyomavirus-specific T cells are diverse and associated with improved patient survival.
        Cancer Immunol Res. 2017; 5: 137-147
        • Nghiem P.T.
        • Bhatia S.
        • Lipson E.J.
        • Kudchadkar R.R.
        • Miller N.J.
        • Annamalai L.
        • et al.
        PD-1 blockade with pembrolizumab in advanced Merkel-cell carcinoma.
        N Engl J Med. 2016; 374: 2542-2552
        • Pagès F.
        • Berger A.
        • Camus M.
        • Sanchez-Cabo F.
        • Costes A.
        • Molidor R.
        • et al.
        Effector memory T cells, early metastasis, and survival in colorectal cancer.
        N Engl J Med. 2005; 353: 2654-2666
        • Paulson K.G.
        • Iyer J.G.
        • Blom A.
        • Warton E.M.
        • Sokil M.
        • Yelistratova L.
        • et al.
        Systemic immune suppression predicts diminished Merkel cell carcinoma-specific survival independent of stage.
        J Invest Dermatol. 2013; 133: 642-646
        • Paulson K.G.
        • Iyer J.G.
        • Simonson W.T.
        • Blom A.
        • Thibodeau R.M.
        • Schmidt M.
        • et al.
        CD8+ lymphocyte intratumoral infiltration as a stage-independent predictor of Merkel cell carcinoma survival: a population-based study.
        Am J Clin Pathol. 2014; 142: 452-458
        • Paulson K.G.
        • Iyer J.G.
        • Tegeder A.R.
        • Thibodeau R.
        • Schelter J.
        • Koba S.
        • et al.
        Transcriptome-wide studies of merkel cell carcinoma and validation of intratumoral CD8+ lymphocyte invasion as an independent predictor of survival.
        J Clin Oncol. 2011; 29: 1539-1546
        • Paulson K.G.
        • Park S.Y.
        • Vandeven N.A.
        • Lachance K.
        • Thomas H.
        • Chapuis A.G.
        • et al.
        Merkel cell carcinoma: Current US incidence and projected increases based on changing demographics.
        J Am Acad Dermatol. 2018; 78: 457-463.e2
        • Penn I.
        • First M.R.
        Merkel's cell carcinoma in organ recipients: report of 41 cases.
        Transplantation. 1999; 68: 1717-1721
        • Reuben A.
        • Gittelman R.
        • Gao J.
        • Zhang J.
        • Yusko E.C.
        • Wu C.J.
        • et al.
        TCR repertoire intratumor heterogeneity in localized lung adenocarcinomas: an association with predicted neoantigen heterogeneity and postsurgical recurrence.
        Cancer Discov. 2017; 7: 1088-1097
        • Reuben A.
        • Gittelman R.
        • Zhang J.
        • Quek K.
        • Vence L.M.
        • Behrens C.
        • et al.
        Association of the T-cell receptor landscape with survival in non-small cell lung cancer.
        J Clin Oncol. 2018; 36: 140
        • Robins H.S.
        • Srivastava S.K.
        • Campregher P.V.
        • Turtle C.J.
        • Andriesen J.
        • Riddell S.R.
        • et al.
        Overlap and effective size of the human CD8+ T cell receptor repertoire.
        Sci Transl Med. 2010; 2: 47ra64
        • Roh W.
        • Chen P.L.
        • Reuben A.
        • Spencer C.N.
        • Prieto P.A.
        • Miller J.P.
        • et al.
        Integrated molecular analysis of tumor biopsies on sequential CTLA-4 and PD-1 blockade reveals markers of response and resistance.
        Sci Transl Med. 2017; 9
        • Santos J.S.
        • Teixeira L.S.
        • dos Santos W.N.
        • Lemos V.A.
        • Godoy J.M.
        • Ferreira S.L.
        Uranium determination using atomic spectrometric techniques: an overview.
        Anal Chim Acta. 2010; 674: 143-156
        • Schrama D.
        • Ritter C.
        • Becker J.C.
        T cell receptor repertoire usage in cancer as a surrogate marker for immune responses.
        Semin Immunopathol. 2017; 39: 255-268
        • Sihto H.
        • Böhling T.
        • Kavola H.
        • Koljonen V.
        • Salmi M.
        • Jalkanen S.
        • et al.
        Tumor infiltrating immune cells and outcome of Merkel cell carcinoma: a population-based study.
        Clin Cancer Res. 2012; 18: 2872-2881
        • Sihto H.
        • Kukko H.
        • Koljonen V.
        • Sankila R.
        • Böhling T.
        • Joensuu H.
        Merkel cell polyomavirus infection, large T antigen, retinoblastoma protein and outcome in Merkel cell carcinoma.
        Clin Cancer Res. 2011; 17: 4806-4813
        • Simoni Y.
        • Becht E.
        • Fehlings M.
        • Loh C.Y.
        • Koo S.L.
        • Teng K.W.W.
        • et al.
        Bystander CD8+ T cells are abundant and phenotypically distinct in human tumour infiltrates.
        Nature. 2018; 557: 575-579
        • Simpson E.H.
        Measurement of diversity.
        Nature. 1949; 163: 688
        • Skelton H.G.
        • Smith K.J.
        • Hitchcock C.L.
        • McCarthy W.F.
        • Lupton G.P.
        • Graham J.H.
        Merkel cell carcinoma: analysis of clinical, histologic, and immunohistologic features of 132 cases with relation to survival.
        J Am Acad Dermatol. 1997; 37: 734-739
        • Stang A.
        • Becker J.C.
        • Nghiem P.
        • Ferlay J.
        The association between geographic location and incidence of Merkel cell carcinoma in comparison to melanoma: an international assessment.
        Eur J Cancer. 2018; 94: 47-60
        • Tadmor T.
        • Liphshitz I.
        • Aviv A.
        • Landgren O.
        • Barchana M.
        • Polliack A.
        Increased incidence of chronic lymphocytic leukaemia and lymphomas in patients with Merkel cell carcinoma - a population based study of 335 cases with neuroendocrine skin tumour.
        Br J Haematol. 2012; 157: 457-462
        • Tarantola T.I.
        • Vallow L.A.
        • Halyard M.Y.
        • Weenig R.H.
        • Warschaw K.E.
        • Grotz T.E.
        • et al.
        Prognostic factors in Merkel cell carcinoma: analysis of 240 cases.
        J Am Acad Dermatol. 2013; 68: 425-432
        • Tetzlaff M.T.
        • Nagarajan P.
        Update on Merkel cell carcinoma.
        Head Neck Pathol. 2018; 12: 31-43
        • Tumeh P.C.
        • Harview C.L.
        • Yearley J.H.
        • Shintaku I.P.
        • Taylor E.J.
        • Robert L.
        • et al.
        PD-1 blockade induces responses by inhibiting adaptive immune resistance.
        Nature. 2014; 515: 568-571
        • Wong S.Q.
        • Waldeck K.
        • Vergara I.A.
        • Schröder J.
        • Madore J.
        • Wilmott J.S.
        • et al.
        UV-associated mutations underlie the etiology of MCV-negative Merkel cell carcinomas.
        Cancer Res. 2015; 75: 5228-5234

      Linked Article

      • Erratum
        Journal of Investigative DermatologyVol. 141Issue 5
        • Preview
          T-Cell Repertoire in Combination with T-Cell Density Predicts Clinical Outcomes in Patients with Merkel Cell Carcinoma
        • Full-Text
        • PDF
        Open Archive