Immunotherapy around the clock: impact of infusion timing on stage IV melanoma outcomes
Abstract
Although the impact of circadian timing on immunotherapy has yet to be integrated into clinical practice, chronoimmunotherapy is an emerging and promising field as circadian oscillations are observed in immune cell numbers as well as the expression of immunotherapy targets, e.g., programmed cell death protein-1 and its ligand programmed death ligand 1. Concurrent retrospective studies suggest that morning infusions may lead to higher effectiveness of immune checkpoint inhibitors in melanoma, non-small cell lung cancer, and kidney cancer. This paper discusses the results of a retrospective study (2016-2022) exploring the impact of infusion timing on the outcomes of all 73 patients with stage IV melanoma receiving immunotherapy at a particular medical center. While the median overall survival (OS) was 24.2 months (95% confidence interval [CI] 9.04-39.8), for a median follow-up of 15.3 months, our results show that having more than 75% of infusions in the afternoon results in shorter median OS (14.9 vs. 38.1 months; hazard ratio 0.45 [CI 0.23-0.86]; p < 0.01) with more expressive impacts on particular subgroups: women, older patients, and patients with a lower tumor burden at the outset of immunotherapy. Our findings highlight the potential benefits of follow-up validation in prospective and translational randomized studies.
Full text
Citation: Gonçalves, L.; Gonçalves, D.; Esteban-Casanelles, T.; Barroso, T.; Soares de Pinho, I.; Lopes-Brás, R.; Esperança-Martins, M.; Patel, V.; Torres, S.; Teixeira de Sousa, R.; et al. Immunotherapy around the Clock: Impact of Infusion Timing on Stage IV Melanoma Outcomes. Cells 2023, 12, 2068. https://doi.org/10.3390/ cells12162068 Academic Editors: Yona Keisari, Fabrizio Mattei, Anahid Jewett and Luyuan Li Received: 5 July 2023 Revised: 4 August 2023 Accepted: 9 August 2023 Published: 15 August 2023 Copyright: © 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). cells Article Immunotherapy around the Clock: Impact of Infusion Timing on Stage IV Melanoma Outcomes Lisa Gonçalves 1,* , Duarte Gonçalves 2, Teresa Esteban-Casanelles 3, Tiago Barroso 1, Inês Soares de Pinho 1, Raquel Lopes-Brás 1, Miguel Esperança-Martins 1,4 , Vanessa Patel 1, Sofia Torres 1, Rita Teixeira de Sousa 1, André Mansinho 5and Luís Costa 1,4 1Department of Oncology, Hospital de Santa Maria, Centro Hospitalar Universitário Lisboa Norte, 1649-035 Lisboa, Portugal; [email protected] (L.C.) 2Department of Economics, University College London, London WC1H 0AX, UK 3Department of Political Economy, King’s College London, London WC2B 4BG, UK; [email protected] 4Instituto de Medicina Molecular-João Lobo Antunes, Faculdade de Medicina de Lisboa, 1649-028 Lisboa, Portugal 5START Lisbon, Hospital de Santa Maria, Centro Hospitalar Universitário Lisboa Norte, 1649-035 Lisboa, Portugal *Correspondence: [email protected] Abstract: Although the impact of circadian timing on immunotherapy has yet to be integrated into clinical practice, chronoimmunotherapy is an emerging and promising field as circadian oscillations are observed in immune cell numbers as well as the expression of immunotherapy targets, e.g., programmed cell death protein-1 and its ligand programmed death ligand 1. Concurrent retrospective studies suggest that morning infusions may lead to higher effectiveness of immune checkpoint inhibitors in melanoma, non-small cell lung cancer, and kidney cancer. This paper discusses the results of a retrospective study (2016–2022) exploring the impact of infusion timing on the outcomes of all 73 patients with stage IV melanoma receiving immunotherapy at a particular medical center. While the median overall survival (OS) was 24.2 months (95% confidence interval [CI] 9.04–39.8), for a median follow-up of 15.3 months, our results show that having more than 75% of infusions in the afternoon results in shorter median OS (14.9 vs. 38.1 months; hazard ratio 0.45 [CI 0.23–0.86]; p< 0.01) with more expressive impacts on particular subgroups: women, older patients, and patients with a lower tumor burden at the outset of immunotherapy. Our findings highlight the potential benefits of follow-up validation in prospective and translational randomized studies. Keywords: cancer; chronobiology; circadian; immunotherapy; melanoma 1. Introduction Malignant melanoma affects all age groups and is the leading cause of death from cutaneous malignancies, accounting for more than 20,000 deaths annually in Europe [ 1 ]. The disease has markedly distinct prognoses according to the disease stage. In stages I–II, the 5-year overall survival (OS) ranges between 90 and 100%; stage III has a 5-year OS of approximately 75%; and stage IV (metastatic melanoma) has a 5-year OS of 9–35% according to the disease burden, lactate dehydrogenase (LDH) level, presence of central nervous system (CNS) metastases, and metastatic involvement of multiple organs. In addition, other factors have been shown to potentially influence prognosis, namely inflammatory markers such as peripheral blood neutrophil/lymphocyte ratio and platelet count. These factors independently contribute to the systemic inflammatory index that correlates with worse outcomes [2]. Immunotherapy is currently the standard of care in metastatic melanoma treatment [3] . Before 2011, therapeutic choices for metastatic melanoma were limited, and the median Cells 2023,12, 2068. https://doi.org/10.3390/cells12162068 https://www.mdpi.com/journal/cells
Cells 2023,12, 2068 2 of 16 OS was 9 months [ 4 ]. In 2011, the emergence of immune checkpoint inhibitors (ICIs) revolutionized the treatment landscape of metastatic melanoma, a malignancy documented as highly immunogenic due to high levels of tumor-infiltrating lymphocytes [ 5 ]. Immune checkpoints downregulate the immune system, preventing its overactivation, and are used by cancer cells as an escape mechanism from the immune system [ 6 , 7 ]. Despite frequently reported high-grade toxicity [ 8 ], ipilimumab—a fully humanized immunoglobulin G1 (IgG1) anti-cytotoxic lymphocyte antigen-4 (CTLA-4)-blocking antibody—was the first ICI approved for the first-line treatment of stage IV melanoma. In 2015, other ICIs emerged, targeting the negative regulation of T cell activation through the blockade of the programmed cell death protein-1 (PD-1)/programmed death ligand 1 (PD-L1) pathway. Inhibition of this pathway activates the host’s immune response. This led to a significant survival benefit in metastatic melanoma, extending the 12-month OS associated with the PD-1 inhibitors pembrolizumab and nivolumab to over 70% [ 9 – 12 ]. Recently, combined ICI with an anti-CTLA-4/anti-PD-1 induction phase followed by an anti-PD-1 maintenance phase (ipilimumab/nivolumab) was approved in metastatic melanoma, following results of 72.1 months of median OS versus 36.9 months with anti-PD-1 nivolumab, and 19.9 months with anti-CTL-4 ipilimumab monotherapies [ 13 ]. Notwithstanding the significant progress, the prognosis of these patients remains dismal, with a 5-year OS just over 50% with the antiCTLA-4/anti-PD-1 combination [ 13 , 14 ], with a considerable number of patients presenting as non-responders. The interplay between cancer cells and the tumor microenvironment affects cancer cell survival, local invasion, and metastatic dissemination [ 15 ]. The tumor microenvironment comprises several non-cancer cells, growth factors, cytokines, and the extracellular matrix. Tumor microenvironment cells predominantly consist of fibroblasts, epithelial cells, and immune cells such as lymphocytes, natural killer (NK) cells, tumor-associated macrophages (TAMs), myeloid-derived suppressor cells, dendritic cells (DCs), and tumor-associated neutrophils. Immune cells from the tumor microenvironment infiltrate the tumor and can elicit both tumorigenic and anti-tumorigenic effects, playing a decisive role in tumor growth and therapy response [ 15 ]. The tumor microenvironment has a high heterogeneity of cells, which vary according to tumor type, making each tumor microenvironment unique [ 15 , 16 ]. The immune landscape varies throughout the day, suggesting an impact on tumor proliferation. Immune cells, NKs, DCs, monocytes, and T and B lymphocytes exhibit circadian oscillations in the peripheral blood [ 17 – 20 ]. For instance, the number of circulating CD4+ and CD8+ T cells doubles between early morning and early night [ 17 ]. Wang et al. [21] recently showed more aggressive tumor behavior in mice engrafted with melanoma cells in the evening versus in the late afternoon, as well as variation in the anti-tumorigenic activity of CD8+ T cells according to the time of day. The authors also showed that mice with melanoma cells inoculated during the day had a higher DC count and better anti-tumor response, with greater tumor volume suppression, than those that were vaccinated during the night. CD8+ T cell clones from melanoma patients exhibited different T cell proliferation abilities according to the time of day [ 21 ]. Additionally, other studies demonstrated that peripheral blood immune cells and their migration to organs exhibit an oscillating off-phase pattern. These cells predominantly exit hematopoietic organs and enter the peripheral blood during the onset of the behavioral rest phase. Conversely, they migrate predominantly to peripheral organs during the onset of the behavioral active phase [22]. The PD-1/PD-L1 expression levels of ICIs’ targets fluctuate throughout the day, which may result in different ICI efficacy according to the timing of administration. The ICIs nivolumab and pembrolizumab, used in the treatment of metastatic melanoma, act by blocking the interaction of PD-1 with PD-L1/2, thereby inhibiting the PD-1 pathway and promoting immune system activation [ 7 ]. Besides being present in T cells, the PD-1 receptor is also expressed on TAMs and DCs from the tumor microenvironment, and its expression levels increase as cancer progresses [ 23 ]. Furthermore, the presence of PD-1 reduces phagocytosis and thus limits the ability of macrophages to eliminate cancer cells, which enhances the immune escape of tumor cells [ 24 ]. In a melanoma-bearing mouse
Cells 2023,12, 2068 3 of 16 model, circadian oscillations were reported in PD-1, Pdcd1, and in the PD-L1-encoding Cd274 gene [ 25 ]. Indeed, Tsuruta et al. [ 25 ] showed that circadian PD-1 expression on TAMs impacts the anti-tumor effect of the PD-1/PD-L1 inhibitor BMS-1 in melanoma-bearing mice, with tumor growth significantly suppressed by the administration of BMS-1 during the night. Tumor cells usually express PD-L1 on their surfaces; however, the expression of PD-1 in tumor cells has also been reported. The PD-1 mechanism in tumor cells is not yet clear, and it seems to differ between tumor types [23]. A few concurrent retrospective studies suggest an impact of circadian timing on treatment effectiveness in different metastatic settings. In metastatic melanoma, the MEMOIR retrospective study showed that administering more than 20% of ICI (ipilimumab, nivolumab, and pembrolizumab) infusions later than 4:30 p.m. was associated with worse outcomes [ 26 ]. In non-small cell lung cancer (NSCLC), Kabaroué et al. [ 27 ] report a retrospective study in metastatic NSCLC that indicates a major OS difference according to the immunotherapy infusion timing, with patients receiving morning infusions showing a four-fold OS increase compared to those receiving afternoon infusions. In metastatic renal cell carcinoma, two concurrent retrospective studies [ 28 , 29 ] showed that morning immunotherapy infusions were again associated with better objective response rates, time to treatment failure, and OS. A small-sample, pan-cancer retrospective study did not find significant OS differences between performing ICI infusions in the morning or late afternoon [ 30 ]. However, it is unclear if the allocation into morning and afternoon groups across patient characteristics and tumor types correlated with the effectiveness of the treatment, which could explain the lack of a significant effect. Additionally, when controlling for the number of infusions, Cortellini et al. [ 31 ] failed to find a significant association between evening infusions and aggravated OS. Because the number of infusions is also a result of treatment success itself, controlling for it entails an indirect selection on outcomes, which may explain the absence of statistically significant differences. These data stress the need for further studies. This paper reports on a retrospective study of patients with metastatic melanoma. We contribute to the growing body of evidence that suggests a significant beneficial impact of performing immunotherapy in the morning on patient outcomes. Further, we find suggestive evidence of more expressive positive impacts of morning infusions on particular subgroups: women, older patients, and patients with a lower tumor burden at the start of immunotherapy. Prospective studies and translational approaches are needed to further our understanding of the potential gains and mechanisms underlying these results. 2. Materials and Methods 2.1. Study Design We perform a retrospective cohort study of patients with stage IV melanoma receiving immunotherapy with either nivolumab, pembrolizumab, or ipilimumab plus nivolumab, in first or later lines of treatment, at the medical center ‘Centro Hospitalar Universitário Lisboa Norte’ in Portugal between July 2016 and March 2022. Inclusion in the sample was restricted to patients with an Eastern Cooperative Oncology Group (ECOG) performance status (PS) of 0–1 at the start of immunotherapy. 2.2. Participants Our dataset comprises 104 patients with metastatic melanoma, of which 78 received immunotherapy with either nivolumab, pembrolizumab, or nivolumab plus ipilimumab. Out of these, only 73 patients had an ECOG PS of 0–1 at the start of immunotherapy. This patient cohort completed a total of 1019 infusions between July 2016 and March 2022 (Figure 1). All data were pseudoanonymized.
Cells 2023,12, 2068 4 of 16 Metastatic melanoma patients July 2016–March 2022 (N=104) Treated with immunotherapy (nivolumab, pembrolizumab, nivolumab + ipilimumab) (N=78) ECOG performance status 0 or 1 (1019 immunotherapy infusions) (N=73) Excluded not treated with immunotherapy (N=26) Excluded ECOG performance status >1 (N=5) Figure 1. Patient Inclusion Criteria. ECOG: Eastern Cooperative Oncology Group. 2.3. Outcomes Data on demographic and clinical–pathological characteristics, infusion reactions, immune-related adverse events (irAEs), and timing of administration of each immunotherapy cycle were retrieved from the patients’ medical records. Adverse events were categorized according to the Common Terminology Criteria for Adverse Events (CTCAE-Version 5.0) [32] . Disease progression was determined based on RECIST 1.1 (Response Evaluation Criteria in Solid Tumors) criteria [33] or clinical assessment. 2.4. Treatment Groups and Allocation Infusion times were retrieved from medical records and split into two treatment groups: the AM (morning) treatment group (8 a.m.–2 p.m.) and the PM (afternoon) treatment group (2 p.m.–8 p.m.). The AM group included all patients who received less than 75% of infusions after 2 p.m., and the PM group included all patients who received at least 75% of infusions after 2 p.m. Although no explicit randomization device was used to allocate patients to study groups, we verify that neither patients’ baseline characteristics nor initial disease burden correlate with treatment allocation. On this basis, we argue that the estimates presented in this paper may be interpreted as causal. Note that the potential identification of a causal relationship is independent of the inference-related limitations posed by the relatively small sample used. 2.5. Statistical Analysis Tests of independence of treatment allocation from patient characteristics and initial disease burden relied on χ2 independence tests. Time-to-event outcomes were analyzed using Kaplan–Meier nonparametric estimators, and statistical inference tests were based on Cox regression analysis. All tests were conducted considering a two-sided 5% significance level. 3. Results 3.1. Patient Characteristics and Disease Burden by Treatment Group Patient characteristics at the start of immunotherapy are summarized in Table 1. A majority of patients in the study cohort were male (62%, N = 45). The median age at diagnosis was 64 years (range 25–89 years), and 70 years (range 29–91 years) at the start of immunotherapy. Most patients (85%; N = 62) presented a cutaneous melanoma subtype; the remaining presented rare melanoma subtypes—e.g., mucosal (12%; N = 9) and ocular melanoma (3%; N= 2).
Cells 2023,12, 2068 5 of 16 Table 1. Baseline Patient Characteristics by AM/PM Treatments at Start of Immunotherapy. All Patients Morning Group Afternoon Group χ2Statistic (N=73) (N=48) (N=25) (p-Value) Age (years) 0.30 (0.581) Median (range) 70 (29–91) 69 (35–91) 75 (29–86) <65 25 (34.2%) 18 (37.5%) 7 (28.0%) ≥65 48 (65.8%) 30 (62.5%) 18 (72.0%) Sex 0.31 (0.581) Female 28 (38.4%) 20 (41.7%) 8 (32.0%) Male 45 (61.6%) 28 (58.3%) 17 (68.0%) ECOG PS 0.00 (≥0.999) 0 48 (65.8%) 32 (66.7%) 16 (64.0%) 1 25 (34.2%) 16 (33.3%) 9 (36.0%) Melanoma subtype 1.84 (0.398) Cutaneous 62 (84.9%) 39 (81.2%) 23 (92.0%) Mucosal 2 (2.7%) 2 (4.2%) 0 (0.0%) Ocular 9 (12.3%) 7 (14.6%) 2 (8.0%) AM (Morning) treatment group: patients with < 75% of infusions after 2 p.m. PM (Afternoon) treatment group: patients with ≥ 75% of infusions after 2 p.m. χ2 test of independence between treatment allocation (AM/PM) and individual characteristics; p -values and degrees of freedom (df) reported in parentheses. ECOG PS: Eastern Cooperative Oncology Group performance status. Regarding the patients’ disease burden (Table 2), 32% ( N = 23) had lesions in a single metastatic site, 37% (N= 27) in two metastatic sites, 23% (N= 17) in three metastatic sites, and 8% ( N = 6) in four or more metastatic sites. CNS metastases were present in 19% ( N= 14 ) of cases. Patients’ median LDH level before the start of the immunotherapy was 410 U/L (range 117–4529 U/L), with 42% ( N = 31) showing a level above the upper limit of normal (ULN), and 15% (N= 11) twice above the ULN (set at 500 U/L). Table 2. Baseline Patient Disease Burden by AM/PM Treatments at Start of Immunotherapy. All Patients Morning Group Afternoon Group χ2Statistic (N=73) (N=48) (N=25) (p-Value) Metastatic Sites (N) 1.71 (0.634) 1 23 (31.5%) 17 (35.4%) 6 (24.0%) 2 27 (37.0%) 18 (37.5%) 9 (36.0%) 3 17 (23.3%) 10 (20.8%) 7 (28.0%) ≥4 6 (8.2%) 3 (6.2%) 3 (12.0%) CNS Metastases 0.66 (0.417) Yes 14 (19.2%) 11 (22.9%) 3 (12.0%) No 59 (80.8%) 37 (77.1%) 22 (88.0%) LDH (U/L) Median (range) 225 (117–4529) 223 (117–4529) 301 (135–1922) ≥250 U/L 31 (42.5%) 17 (35.4%) 14 (56.0%) 2.07 (0.150) ≥2 ULN 11 (15.1%) 5 (10.4%) 6 (24.0%) 1.43 (0.232) AM (Morning) treatment group: patients with < 75% of infusions after 2 p.m. PM (Afternoon) treatment group: patients with ≥ 75% of infusions after 2 p.m. χ2 test of independence between treatment allocation (AM/PM) and disease burden at the start of immunotherapy; p -values and degrees of freedom (df) reported in parentheses. CNS metastases: presence of metastases in the central nervous system. LDH: lactate dehydrogenase; U/L: units per liter; ULN: upper limit of normal. Of the 73 patients in the sample, 66% ( N = 48) were allocated to the AM treatment group ( < 75% of infusions in the afternoon) and 34% ( N = 25) to the PM group ( ≥ 75% of infusions in the afternoon). The two treatment groups were mainly composed of male patients who were 70 or older at the start of immunotherapy, and patients predominantly with an ECOG PS of 0, with no statistically significant differences between both regarding clinical–demographic features (Table 1). Also, no significant differences were found
Cells 2023,12, 2068 6 of 16 between the treatment groups regarding disease burden (Table 2), and number of immunotherapy infusion sessions (median of 15 [range 2–44] vs. 13 [range 2–59] in the AM and PM treatment groups, respectively). 3.2. Immunotherapy Toxicities The toxicities associated with immunotherapy and respective grades (CTCAE v5.0) are summarized in Table 3. The majority of patients in the study experienced at least one instance of irAEs (66%, N = 48). The most common irAE was fatigue ( N = 27, 37%), followed by cutaneous ( N = 24, 33%), endocrine ( N = 16, 22%), and renal ( N = 8, 11%) toxicities. Cutaneous toxicities consisted of rash, pruritus, and vitiligo; endocrine toxicities denotes hypothyroidism, hyperthyroidism, thyroiditis, adrenal insufficiency, and hypophysitis. Table 3. Toxicities Associated with Immunotherapy by AM/PM Treatments. All Patients Morning Group Afternoon Group χ2Statistic (N=73) (N=48) (N=25) (p-Value) Overall Toxicity 6.94 (0.225) No Toxicity 22 (30.1%) 15 (31.2%) 7 (28.0%) G1 12 (16.4%) 7 (14.6%) 5 (20.0%) G2 27 (37.0%) 16 (33.3%) 11 (44.0%) G3 7 (9.6%) 7 (14.6%) 0 (0.0%) G4 2 (2.7%) 2 (4.2%) 0 (0.0%) N/I 3 (4.1%) 1 (2.1%) 2 (8.0%) Fatigue 1.66 (0.435) G1/G2 27 (37.0%) 19 (39.6%) 8 (32.0%) G3/G4 0 (0%) 0 (0%) 0 (0%) Cutaneous 1.98 (0.577) G1/G2 23 (31.5%) 15 (31.2%) 8 (32.0%) G3/G4 1 (1.4%) 1 (2.1%) 0 (0.0%) Endocrine 3.65 (0.302) G1/G2 14 (19.2%) 11 (22.9%) 3 (12.0%) G3/G4 2 (2.7%) 2 (4.2%) 0 (0.0%) Hepatitis 3.21 (0.360) G1/G2 6 (8.2%) 3 (6.2%) 3 (12.0%) G3/G4 2 (2.7%) 2 (4.2%) 0 (0.0%) Pneumonitis 2.96 (0.397) G1/G2 2 (2.7%) 2 (4.2%) 0 (0.0%) G3/G4 1 (1.4%) 1 (2.1%) 0 (0.0%) Renal Insufficiency 1.55 (0.461) G1/G2 8 (11.0%) 5 (10.4%) 3 (12.0%) G3/G4 0 (0%) 0 (0%) 0 (0%) Uveitis 2.45 (0.485) G1/G2 1 (1.4%) 1 (2.1%) 0 (0.0%) G3/G4 1 (1.4%) 1 (2.1%) 0 (0.0%) Encephalitis 2.45 (0.485) G1/G2 1 (1.4%) 1 (2.1%) 0 (0.0%) G3/G4 1 (1.4%) 1 (2.1%) 0 (0.0%) Toxicities associated with immunotherapy and respective grades according to CTCAE v5.0 observed in the total study cohort and according to AM/PM treatment groups. AM (Morning) treatment group: patients with < 75% of infusions after 2 p.m. PM (Afternoon) treatment group: patients with ≥ 75% of infusions after 2 p.m. χ2 test of independence between treatment allocation (AM/PM) and toxicity grades; p -values and degrees of freedom (df) reported in parentheses. G x : grade x ; N/I: cases for which no information is available; overall toxicity corresponds to the maximum toxicity grade observed. Grade 3–4 irAEs were only reported in the AM treatment group. Despite this, no statistically significant differences were found in overall, G1/G2, or G3/G4 irAEs between both treatment groups. Severe irAEs included endocrine toxicity (hypothyroidism, N= 2 ), hep-
Cells 2023,12, 2068 7 of 16 atitis with transaminase elevation ( N = 2), cutaneous toxicity (rash, N = 1), uveitis (N= 1) , pneumonitis ( N = 1), and encephalitis ( N = 1). Three patients suspended immunotherapy due to toxicity. 3.3. Progression-Free and Overall Survival The median progression-free survival (PFS) of the study cohort was 10.7 months ( 95% CI 4.0–17.5 ) and the median OS was 24.2 months (95% CI 9.0–39.3), for a median follow-up of 15.3 months. There was a trend toward higher PFS in the AM treatment group (median 14.9 months [95% CI 6.1–20.4] vs. 6.6 months [95% CI 3.5–14.2] in the PM group), although not reaching statistical significance (p= 0.320). Overall survival (OS) was strikingly higher in the AM group: a median of 38.1 months (95% CI 18.9-not reached) versus 14.2 months (95% CI 4.7–31.4) in the afternoon group, with a hazard ratio (HR) of 0.45 (95% CI 0.23–0.86; p< 0.01; Figure 2). (a) (b) Figure 2. Kaplan–Meier Curves of Overall and Progression-Free Survival by AM/PM Treatments. ( a ) Overall Survival; ( b ) Progression-Free Survival. AM (Morning) group: patients with < 75% of infusions after 2 p.m.; PM (Afternoon) group: patients with ≥ 75% of infusions after 2 p.m. Lines: Kaplan–Meier nonparametric estimates of survival rates; shaded regions: 95% confidence intervals. 3.4. Subgroup Analysis We analyzed how the relative effectiveness of allocation to AM vs. PM treatment group varies across patient subgroups. Figures 3and 4respectively exhibit hazard ratio (HR) forest plots of PFS and OS across AM and PM treatment groups within different subgroups, as defined by patient characteristics and initial disease burden. In no case was allocation to the AM treatment group detrimental to the effectiveness of immunotherapy. The data suggest higher relative effectiveness of AM vs. PM treatment allocation on PFS for (i) female patients (median 32.6 vs. 5.7 months; p -value 0.056), and (ii) patients aged 65 and older (median 14.6 vs. 5.3 months; p -value 0.052). However, the HR is not significantly different from 1 at a two-sided 5% significance level in any of the subgroups (Figure 3).
Cells 2023,12, 2068 8 of 16 (a) (b) Figure 3. Forest Plot for Hazard Ratio of Progression-Free Survival. ( a ) By Patient Characteristics; ( b ) By Tumor Burden. AM (Morning) group: patients with < 75% of infusions after 2 p.m.; PM (Afternoon) group: patients with ≥ 75% of infusions after 2 p.m. Hazard ratio of AM vs. PM treatment groups estimated using Cox regressions; squares: estimated values (size proportional to sample size); whiskers: 95% confidence intervals. ECOG PS: Eastern Cooperative Oncology Group performance status. Mtx: number of metastatic sites. ( ∅ ) CNS Mtx: presence (absence) of metastases in the central nervous system. LDH: lactate dehydrogenase; ULN: upper limit of normal. (a) (b) Figure 4. Forest Plot for Hazard Ratio of Overall Survival. ( a ) By Patient Characteristics; ( b ) By Tumor Burden. AM (Morning) group: patients with < 75% of infusions after 2 p.m.; PM (Afternoon) group: patients with ≥ 75% of infusions after 2 p.m. Hazard ratio of AM vs. PM treatment groups estimated using Cox regressions; squares: estimated values (size proportional to sample size); whiskers: 95% confidence intervals. ECOG PS: Eastern Cooperative Oncology Group performance status. Mtx: number of metastatic sites. ( ∅ ) CNS Mtx: presence (absence) of metastases in the central nervous system. LDH: lactate dehydrogenase; ULN: upper limit of normal. Regarding the relative effectiveness of AM treatment allocation on OS, we detect meaningful differences across subgroups. Not only is AM vs. PM treatment allocation conducive to significant effects on HR for (i) female patients (HR 0.19 [95% CI 0.06–0.62]), and (ii) older patients (HR 0.31 [95% CI 0.14–0.72])—and not for male and younger patients—but we also find a significant effect for patients with an initial low tumor burden (fewer metastatic sites [HR 0.45; 95% CI 0.19–1.07]), less CNS involvement [HR 0.36; 95% CI 0.17–0.79], and LDH below 2ULN [HR 0.43; 95% CI 0.20–0.92]), but not for those with worse disease burden at the outset of immunotherapy. For completion, in Figure 5we report the Kaplan–Meier estimates of OS disaggregated by patient characteristics (age (a), sex (b), and ECOG PS (c)) and initial disease burden (number of metastatic sites (d), presence of metastases in the central nervous system (e), and LDH values (f)).
Cells 2023,12, 2068 9 of 16 (a) (b) (c) (d) (e) (f) Figure 5. Kaplan–Meier Curves for Overall Survival by AM/PM Treatments: Patient Characteristics and Disease Burden. ( a ) By Age; ( b ) By Sex; ( c ) By ECOG PS; ( d ) By Number of Metastatic Sites; (e) By Presence of CNS Metastases; ( f ) LDH. AM (Morning) group: patients with < 75% of infusions after 2 p.m.; PM (Afternoon) group: patients with ≥ 75% of infusions after 2 p.m. Lines: Kaplan–Meier nonparametric estimates of survival rates; shaded regions: 95% confidence intervals. ECOG PS: Eastern Cooperative Oncology Group performance status. Mtx: number of metastatic sites. ( ∅ ) CNS Mtx: presence (absence) of metastases in the central nervous system. LDH: lactate dehydrogenase; ULN: upper limit of normal.
Cells 2023,12, 2068 16 of 16 45. Zhu, X.; Suo, Y.; Fu, Y.; Zhang, F.; Ding, N.; Pang, K.; Xie, C.; Weng, X.; Tian, M.; He, H.; et al. In Vivo Flow Cytometry Reveals a Circadian Rhythm of Circulating Tumor Cells. Light Sci. Appl. 2021,10, 110. [CrossRef] [PubMed] 46. Hodi, F.S.; Chiarion Sileni, V.; Lewis, K.D.; Grob, J.J.; Rutkowski, P.; Lao, C.D.; Cowey, C.L.; Schadendorf, D.; Wagstaff, J.; Dummer, R.; et al. Long-Term Survival in Advanced Melanoma for Patients Treated with Nivolumab Plus Ipilimumab in CheckMate 067. J. Clin. Oncol. 2022,40, 9522. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.