Abstract
Objective
We aimed to 1) evaluate the efficacy of immune checkpoint inhibitors (ICIs) in cervical cancer patients according to the site of disease, 2) investigate the mechanism responsible for differential ICIs sensitivities with focuses on CD8+ T lymphocytes and programmed death-ligand 1 (PD-L1) expression.
Methods
We retrospectively reviewed clinical data from patients with recurrent or metastatic cervical cancer treated with pembrolizumab or cemiplimab between January 2019 and January 2024 (clinical cohort). Target diseases were classified according to the site of diseases: within previously irradiated field (in-field diseases), out-of-field diseases, and both. Immunohistochemical investigations were performed using paired tumor samples (i.e. initial cervical tumor and locally-recurrent tumor developed after definitive radiotherapy: Immunohistochemical cohort). Survival rates were estimated using the Kaplan–Meier method and compared using the log-rank test.
Results
Fifty patients treated with pembrolizumab-containing chemotherapies (n=39) or cemiplimab (n=11) were assessed. Of these, six patients (12.0%) had in-field diseases alone, twenty-eight patients (56.0%) had out-of-field diseases, and the remaining sixteen (32%) patients had both types of diseases. In-field diseases demonstrated a significantly lower response rate compared to out-of-field diseases (36.3% vs. 72.7%, p=0.004). Patients with in-field diseases demonstrated significantly shorter progression-free survival (p=0.003) and overall survival (p=0.003) than those with out-of-field diseases. In-field diseases were associated with decreased tumor-infiltrating CD8+ T lymphocytes and PD-L1 expression.
Conclusion
In-field cervical cancer recurrence was associated with decreased sensitivity to ICIs-containing chemotherapies when compared to out-of-field diseases. Decreased tumor-infiltrating CD8+ T lymphocytes and PD-L1 expression are possible reasons for this differential sensitivity to ICI-containing chemotherapies.
Keywords: Uterine Cervical Neoplasms; Neoplasm Recurrence, Local; Radiotherapy; Immune Checkpoint Inhibitors
Synopsis
The activity of immune checkpoint inhibitor (ICI)-containing chemotherapies against in-field cervical cancer recurrence and underlying mechanism were investigated. In-field recurrence was associated with decreased sensitivity to ICIs-containing chemotherapies. Decreased tumor-infiltrating CD8+ T lymphocytes and programmed death-ligand 1 expression may explain its ICI-insensitive nature.
INTORODUCTION
Approximately one-third of patients with invasive cervical cancer will develop recurrent disease after primary treatment, usually within 3 years [1], and the treatment strategy for recurrent cervical cancer is generally determined by factors such as the site of recurrence (local vs. distant), prior radiation therapy, age, or performance status [2]. Previous studies have shown that recurrence tends to be localized to the pelvis in approximately 40% of patients treated with definitive radiotherapy (RT) [3]. Pelvic exenteration or salvage hysterectomy which can improve long-term survival by 30%–60%, is considered in these cases as a curative treatment option. However, they are highly invasive, associated with significantly high complication rate when compared with surgeries in patients with newly-diagnosed cervical cancer, and thus can only be performed by well-trained surgeons [4,5,6].
Chemotherapy has been a major treatment option for recurrent cervical cancer in the radiation field, but it is primarily intended to palliate symptoms and improve quality of life, and the response rate is reported to be 33%, which is lower compared to the 60%–75% response rate for recurrences outside the radiation field [7]. Recently, two large randomized controlled trials (RCTs) of immune checkpoint inhibitors (ICIs) targeting programmed cell death protein-1 (PD-1)/programmed death-ligand 1 (PD-L1) pathway have shown survival benefits of ICIs in advanced or recurrent cervical cancer and ICI alone or in combination with platinum-based chemotherapy have become a standard treatment in patients with recurrent, persistent and advanced cervical cancer [8,9]. One is anti-PD-1 monoclonal antibody pembrolizumab. In a phase III randomized clinical trial (KEYNOTE-826) designed to assess the clinical benefit of adding pembrolizumab with or without bevacizumab as first-line treatment for patients with persistent, recurrent, or metastatic cervical cancer, addition of pembrolizumab to platinum-based combination chemotherapy demonstrated statistically significant and clinically meaningful survival benefits: median overall survival (OS) of 24.4 months vs. 16.5 months; 95% confidence interval (CI)=0.54–0.84; p<0.001 and median progression free survival (PFS) of 10.4 months vs. 8.2 months, hazard ratio (HR)=0.65; 95% CI=0.53–0.79; p<0.001 [8]. Another ICI approved for patients with cervical cancer is the anti-PD-1 monoclonal antibody cemiplimab. In a randomized, phase III, EMPOWER-Cervical 1/GOG-3016/ENGOT-cx9 trial investigating the activity of single agent cemiplimab compared with standard chemotherapy in previously treated, recurrent cervical cancer patients, cemiplimab achieved promising anti-tumor efficacies: overall response rate of 16.4%, median duration of response of 16.4 months, and a significantly longer OS compared with standard chemotherapy (median OS of 12 months vs. 8.5 months; HR=0.69; 95% CI=0.56–0.84) [9]. However, in both RCTs, the clinical activity of ICIs has not been investigated according to the locations of diseases: i.e. within previously irradiated field (in-field diseases) or outside the irradiated field (out-of-field diseases).
ICIs were first found to reinvigorate CD8+ T lymphocytes in tumors. Accordingly, tumor immunological “heat” has recently gained attentions, as it predicts ICI sensitivity: so called “cold (immune-desert/immune-excluded)” tumors characterized by a lack of CD8+ TILs and an abundance of immunosuppressive cells are immunotherapy-insensitive. Whereas “hot (immune-inflamed)” tumors characterized by abundant CD8+ T cells infiltration are immunotherapy-sensitive [10]. In addition, as ICIs approved for cervical cancer treatment target PD-1/PD-L1 signaling pathway, PD-L1 expression on tumor or immune cells emerged as a potential predictive biomarker [8].
In this study, using clinical data, we retrospectively investigated the efficacy of ICIs in recurrent cervical cancer patients according to the site of diseases: in-field diseases versus out-of-field diseases. We also investigated the background mechanism responsible for the differential sensitivity of in-field diseases and out-of-field diseases to ICIs, with focuses on CD8+ T lymphocytes and PD-L1 expression in tumor.
MATERIALS AND METHODS
1. Patients (clinical cohort)
Ethics approval for this retrospective study was obtained by central ethics approval (IRB at Osaka International Cancer Institute, approval No. 23251). Patients with persistent, recurrent, or metastatic cervical cancer treated with at least two cycles of ICIs (pembrolizumab or cemiplimab) between January 2019 and January 2024 were identified and retrospectively reviewed (clinical cohort). Patients who received ICIs only once were excluded from this study. For all patients, clinical data on the following characteristics were collected: age, initial clinical stage, histology, primary treatment, prior recurrence treatment, previous chemotherapy regimen, dose of ICIs administered, site of recurrent disease, adverse events associated with ICIs, treatment response, OS and PFS. Target lesions were classified according to the site of diseases: within the previously irradiated field (in-field diseases), outside the previously irradiated field (out-of-field diseases), and both. In the survival analyses, patients with in-field diseases with or without out-of-field diseases are defined as “patients with in-field diseases.” Patients with out-of-field diseases alone are defined as “patients with out-of-field diseases.” An overview of the current study population and analyses for the clinical and immunohistochemical investigations is provided in Fig. S1.
2. Administration of ICI
Patients with newly-diagnosed stage IVB cervical cancer or those with recurrent cervical cancer who had not previously received platinum-based chemotherapy against recurrent lesion were treated with pembrolizumab (200 mg) in combination with platinum-based chemotherapy with or without bevacizumab every three weeks as reported previously [8]. Patients who had received chemotherapies for recurrent diseases were treated with cemiplimab (350 mg) every three weeks as reported previously [9].
3. Response evaluation
Patients were encouraged to undergo regular follow-up in the outpatient unit by gynecological oncologists during and after treatment, as reported previously [11,12]. Computed tomography or magnetic resonance imaging was performed approximately every three cycles from the start of treatment until disease progression, death or treatment discontinuation. We evaluated the tumor response to ICIs after every three cycles of each regimen. Response assessments were primarily based on the Revised Guidelines for Evaluating Treatment Effectiveness in Solid Tumors Guidelines version 1.1 [13].
4. Toxicity assessment
The patients were assessed for toxicities every two weeks during treatment. The severity of acute complications was classified according to the Common Terminology Criteria for Adverse Events (CTCAE) v5.0.
5. Immunohistochemical investigations
Immunohistochemical investigations were performed using paired tumor samples (i.e. biopsy from newly-diagnosed cervical tumor and matched locally-recurrent tumor resected from the same patients) obtained from 23 locally-recurrent cervical cancer patients who had been treated with definitive RT consisting with external-beam RT (total of 45–50 Gy) plus intracavitary brachytherapy (ICBT) (total of 12–24 Gy) and subsequently treated with salvage hysterectomy (Immunohistochemical cohort). The staining was performed using a Ventana BenchMark GX IHC/ISH Staining Module®. Primary antibodies used are anti-CD8 monoclonal antibody (SP239, pre-diluted, Ventana Medical System Inc., Tucson, AZ, USA) and anti-PD-L1 monoclonal antibody (374517, Biolegend, San Diego, CA, USA). The slides were examined using a bright-field microscope. Under Olympus BX51 microscope using high power field (HPF) (400 × magnification), CD8 immunopositive cells was semi-quantitatively scored from 0 to 3 according to the proportion of positively stained cells as follows: (Score 0; 0 cell/HPF, Score 1; 1–10 cells/HPF, Score 2; 10–100 cells/HPF, Score 3; 100– cells/HPF) (Fig. S2A). PD-L1 expression score on tumor cells or stromal immune cells was counted by scoring the proportion of membranous positive cells over the total number of cells, and was scored as follows: 0: <1%; 1: 1%–9%; 2: 10%–49%; 3: 50%–100% (Fig. S2B). For both CD8 and PD-L1, the mean scores of three separate filed was used for analyses. If only one or two score was available, which was mainly due to the small tumor size, these cases were not evaluated and were left out of all following analyses. The evaluations of pathological specimens were performed by H.N. and S.M. without knowledge of the clinical outcomes of patients.
6. Statistical analysis
PFS was defined as the time from the date of the administration of ICI to the date of the first physical or radiographic evidence of disease progression. OS was defined as the time from the administration of ICI to the date of death or the last follow-up visit. Frequency counts and proportions were compared between groups using chi-squared or Fisher’s exact tests, as applicable. Survival analysis was based on the Kaplan–Meier method and the results were compared using log-rank tests. All analyses were conducted using JMP version 17.0 (SAS Institute, Cary, NC, USA), and a p-value of <0.05 was considered statistically significant.
RESULTS
1. Treatment outcomes in overall population
A total of fifty patients with recurrent or persistent cervical cancer treated with ICI were identified. The patient characteristics are shown in Table 1. Thirty-nine patients (78.0%) received pembrolizumab-containing chemotherapy, while eleven patients (22.0%) were treated with cemiplimab. Eight patients had newly-diagnosed stage IVB cervical cancer, and 42 had recurrent cervical cancer developed after initial treatments. Thirty-six patients (72.0%) had been previously treated with RT in the definitive (external beam radiotherapy [EBRT] + ICBT, n=20) or adjuvant settings (EBRT alone, n=16). Of these, six patients (12.0%) had diseases solely within previously irradiated field (in-field diseases), twenty-eight patients (56.0%) had diseases outside the irradiated field (out-of-field diseases), and the remaining sixteen (32.0%) patients had both in-field and out-of-field diseases. The characteristics of patients with in-field and out-of-field diseases are shown in Table S1 and S2.
Table 1. Clinicopathological characteristics of the patients (n=50).
| Variables | Values | |
|---|---|---|
| Age (yr) | ||
| Median (range) | 54 (25–79) | |
| ≤39 | 3 (6.0) | |
| 40–64 | 41 (82.0) | |
| ≥65 | 6 (12.0) | |
| Types of ICI | ||
| Pembrolizumab | 39 (78.0) | |
| Cemiplimab | 11 (22.0) | |
| Number of the administration of ICI | ||
| 2–5 | 16 (32.0) | |
| 6–9 | 20 (40.0) | |
| ≥10 | 14 (28.0) | |
| Disease status | ||
| Advanced disease | 8 (16.0) | |
| Recurrent disease | 42 (84.0) | |
| Initial clinical stage* | ||
| IB1-IIA | 8 (16.0) | |
| IIB-IIIA | 7 (14.0) | |
| IIIB-IVA | 21 (42.0) | |
| IVB | 14 (28.0) | |
| Histology | ||
| SCC | 31 (62.0) | |
| AC | 17 (34.0) | |
| Others | 2 (4.0) | |
| History of radiotherapy | ||
| Yes | 36 (72.0) | |
| No | 14 (28.0) | |
| Number of regimens before ICI | ||
| 0 | 10 (20.0) | |
| 1 | 21 (42.0) | |
| ≥2 | 19 (38.0) | |
| Prior bevacizumab use | ||
| Yes | 15 (30.0) | |
| No | 35 (70.0) | |
| Location of recurrence | ||
| In-field† | 6 (12.0) | |
| Out-of-field‡ | 28 (56.0) | |
| Both | 16 (32.0) | |
Values are presented as number (%) unless otherwise indicated.
AC, adenocarcinoma; ICI, immune checkpoint inhibitor; SCC, squamous cell carcinoma.
*FIGO 2008 staging system.
†Within irradiated field.
‡Outside the irradiated field.
Overall, the treatment with ICI-containing chemotherapy was well tolerated, and no treatment related deaths occurred. Grade 3–4 acute toxicities were observed in 7 patients (14.0%). The observed grade 3–4 acute toxicities were rectovaginal fistula, pulmonary embolism, cytopenia, hepatic dysfunction, proteinuria, and skin rash. A median follow-up period was 7.4 months. As shown in Fig. 1A, in the overall population, the median PFS was 8.3 months, with an estimated 1-year OS rate of 71.1%. When examined in detail, as shown in Fig. 1B, the median PFS was 9.6 months and the estimated 1-year survival rate was 78.7% in patients treated with pembrolizumab. Moreover, among those treated with cemiplimab, the median PFS was 2.9 months and an estimated 1-year survival rate was 36.5%.
Fig. 1. Kaplan-Meier estimates of PFS and OS according to the types of ICIs and the location of diseases (all patients). (A) (i) Kaplan–Meier estimates of PFS in the overall population. (ii) Kaplan–Meier estimates of OS in the overall population. (B) (i) Kaplan–Meier estimates of PFS according to the types of ICIs. (ii) Kaplan–Meier estimates of OS according to the types of ICIs (i.e. pembrolizumab-containing chemotherapies and cemiplimab). (C) (i) Kaplan–Meier estimates of PFS according to the location of diseases (out-of-field diseases alone [n=28] vs. in-field diseases with or without out-of-field diseases [n=22]; p=0.003). (ii) Kaplan–Meier estimates of OS according to the location of recurrence (out-of-field diseases alone [n=28] vs. in-field diseases with or without out-of-field diseases [n=22]; p=0.003).
ICI, immune checkpoint inhibitor; OS, overall survival; PFS, progression free survival.
Among the 22 in-field diseases, as shown (Table 2, upper panel), 3 (13.6%) achieved complete response (CR), 5 (22.7%) achieved partial response (PR), resulting in the overall response rate of 36.3% (8/22). Conversely, among the 44 out-of-field diseases (Table 2, upper panel), 10 (22.7%) achieved CR, 22 (50.0%) achieved PR, resulting in an overall response rate of 72.7% (32/44). When the response rates of the two groups were compared, patients with in-field diseases exhibited significantly lower response rate than those with out-of-field diseases (p=0.004). In the survival analyses, in the overall population, patients with in-field diseases with or without out-of-field diseases demonstrated significantly shorter PFS (p=0.003) and OS (p=0.003) than those with out-of-field diseases alone (Fig. 1C).
Table 2. Tumor responses to immune checkpoint inhibitors according to the site of diseases.
| Variables | Total number | CR (%) | PR (%) | SD (%) | PD (%) | Response rate (%) | |
|---|---|---|---|---|---|---|---|
| All patients | |||||||
| In-field diseases* | 22 | 3 (13.6) | 5 (22.7) | 3 (13.6) | 11 (50.0) | 36.3 | |
| Out-of field diseases† | 44 | 10 (22.7) | 22 (50.0) | 4 (9.0) | 8 (18.2) | 72.7 | |
| Recurrent cervical cancer patients | |||||||
| In-field diseases* | 22 | 3 (13.6) | 5 (22.7) | 3 (13.6) | 11 (50.0) | 36.3 | |
| Out-of-field diseases† | 36 | 9 (25.0) | 15 (41.7) | 4 (11.1) | 8 (22.2) | 66.7 | |
CR, complete response; PD, progressive disease; PR, partial response; SD, stable disease.
*All in-field diseases (patients with in-field diseases alone and those with in-field- plus out-of-field diseases).
†All out-of-field diseases (patients with out-of-field diseases alone and those with out-of-field plus in-field diseases).
2. Treatment outcomes in patients with recurrent cervical cancer
To further investigate the potential impact of previous RT on the efficacy of ICI-containing chemotherapies, after excluding patients with advanced cervical cancer, a total of 42 patients with recurrent cervical cancer were analyzed. As shown in Fig. 2A, patients with in-field diseases with or without out-of-field diseases demonstrated significantly shorter PFS (p=0.046) and OS (p=0.024) than those with out-of-field diseases alone. When the response rate of the two groups were compared (Table 2, lower panel), patients with in-field diseases exhibited significantly lower response rate than those with out-of-field diseases (36.3% vs. 66.7%, p=0.024).
Fig. 2. Kaplan-Meier estimates of PFS and OS according to the location of diseases in patients with recurrent cervical cancer. (A) (i) Kaplan–Meier estimates of PFS according to the location of diseases (out-of-field diseases alone [n=22] vs. in-field diseases with or without out-of-field diseases [n=20]; p=0.046). (ii) Kaplan–Meier estimates of OS according to the location of diseases (out-of-field diseases alone [n=22] vs. in-field diseases with or without out-of-field diseases [n=20]; p=0.024). (B) (i) Kaplan–Meier estimates of PFS according to the location of diseases in patients treated with pembrolizumab-containing chemotherapies (out-of-field diseases alone [n=15] vs. in-field diseases with or without out-of-field diseases [n=16]; p=0.038). (ii) Kaplan–Meier estimates of OS according to the location of diseases (out-of-field diseases alone [n=15] vs. in-field recurrence with or without out-of-field diseases [n=16]; p=0.059). (C) (i) Kaplan–Meier estimates of PFS according to the location of diseases in patients treated with cemiplimab (out-of-field diseases alone [n=5] vs. others [in-field diseases with or without out-of-field diseases] [n=6]; p=0.075). (ii) Kaplan–Meier estimates of OS according to the location of diseases in patients with cemiplimab (out-of-field diseases alone [n=5] vs. others [in-field diseases with or without out-of-field diseases] [n=6]; p=0.056).
OS, overall survival; PFS, progression free survival.
We next examined the treatment outcomes according to the type of ICIs. Among the 31 patients treated with pembrolizumab-containing chemotherapy, patients with in-field diseases with or without out-of-field diseases demonstrated significantly shorter PFS (p=0.038) and OS (p=0.059) than those with out-of-field diseases alone (Fig. 2B). Moreover, patients with in-field diseases exhibited significantly lower response rate than those with out-of-field diseases (43.8% vs. 80.8%, p=0.014). In line with this, among the 11 patients treated with cemiplimab, patients with in-field diseases with or without out-of-field diseases demonstrated marginally shorter PFS (p=0.075) and OS (p=0.056) than those with out-of-field diseases alone (Fig. 2C). On the other hand, patients with in-field diseases exhibited similar response compared with those with out-of-field diseases (33.3% vs. 30.0%, p=0.889). Correctively, although retrospective, these results strongly indicate that in-field diseases are relatively insensitive to pembrolizumab-containing chemotherapies in comparison to out-of-field diseases.
3. Effect of RT on the number of tumor-infiltrating CD8+ T lymphocytes
To explain the background mechanism responsible for the relatively ICI-insensitive nature of in-field diseases, we investigated the number and the location of CD8+ lymphocytes within the tumor microenvironment using immunohistochemistry. For this purpose, we identified 23 locally-recurrent cervical cancer developed after definitive RT and subsequently treated with salvage hysterectomy. Using paired tumor samples (i.e. biopsy from newly-diagnosed cervical tumor and resected locally-recurrent tumor (in-field diseases) from the same patients), immunostaining for CD8 were performed and CD8 immunoreactivity in tumor was scored semiquantitatively. Representative photomicrographs of tumor-infiltrating CD8+ T lymphocytes in in-field diseases and newly-diagnosed cervical cancer are shown in Fig. 3A (i). Among the 23 in-field diseases developed after definitive RT, 4 (17.4%) were scored as 0, 12 (52.2%) were scored as +1, 7 (30.4%) were scored as +2. In contrast, among the 23 newly-diagnosed cervical cancer, 11 (47.8%) were scored as +1, 11 (47.8%) were scored as +2, and 1 (4.3%) were scored as +3 (Table 3). When compared, immunoreactivity for CD8 in tumor nest was greater in newly-diagnosed cervical cancer than in locally-recurrent cervical cancer (in-field diseases) (p=0.027), indicating that CD8+ lymphocytes do not frequently infiltrate into tumor nest in in-field diseases when compared to newly-diagnosed cervical cancer.
Fig. 3. Immunoreactivities of locally-recurrent cervical cancer developed after definitive radiotherapy (in-field diseases) and matched initial cervical tumor (newly-diagnosed tumor). (A) Representative images of immunohistochemistry for CD8 + T lymphocytes in patients with newly-diagnosed cervical cancer and in-field diseases (magnification: ×400). (i) tumor-infiltrating CD8+ T lymphocytes. Left: newly-diagnosed tumor containing abundant tumor-infiltrating CD8 + T lymphocytes. Right: locally-recurrent tumor containing a few tumor-infiltrating CD8 + T lymphocytes. (ii) stromal CD8 + T lymphocytes. Left: newly-diagnosed tumor containing abundant CD8 + T lymphocytes in both tumor nest and stroma. Right: locally-recurrent tumor containing abundant stromal CD8 + T lymphocytes and a few tumor-infiltrating CD8+ T lymphocytes. (B) Representative images of immunohistochemistry for tumor programmed death-ligand 1 in patients with newly-diagnosed cervical cancer and in-field recurrence (magnification: ×400).
Table 3. Immunohistochemical analyses for CD8 and PD-L1.
| Variables | Primary cancer (n=23) | Recurrent cancer (n=23) | p-value | |
|---|---|---|---|---|
| CD8+ T cells in tumor | 0.027 | |||
| Mean | 1.57 | 1.13 | ||
| 0–1 | 11 (47.8) | 16 (69.6) | ||
| 2–3 | 12 (52.2) | 7 (30.4) | ||
| CD8+ T cells in stroma | 0.796 | |||
| Mean | 2.35 | 2.30 | ||
| 0–1 | 1 (4.4) | 1 (4.4) | ||
| 2–3 | 22 (95.7) | 22 (95.7) | ||
| PD-L1 in tumor* | <0.001 | |||
| Mean | 1.55 | 0.36 | ||
| 0–1 | 5 (45.5) | 11 (86.7) | ||
| 2–3 | 6 (54.5) | 0 | ||
Values are presented as number (%).
PD-L1, programmed death-ligand 1.
*Tumor PD-L1 expression was assessed in 11 cases of each primary and recurrent cancer.
We next investigated the immunoreactivity for CD8 in tumor stroma. Representative photomicrographs of stromal CD8+ T lymphocytes in in-field diseases and newly-diagnosed cervical cancer are shown in Fig. 3A (ii). Among the 23 in-field diseases developed after definitive RT, 1 (4.3%) were scored as +1, 14 (60.9%) were scored as +2, and 8 (34.8%) were scored as +3. In contrast, among the 23 newly-diagnosed cervical cancer, 1 (4.3%) were scored as +1, 13 (56.5%) were scored as +2, and 9 (39.1%) were scored as +3. When compared (Table 3), CD8 immunoreactivity in the stroma of in-field diseases was almost identical to that observed in newly-diagnosed cervical cancer (p=0.796). Collectively, these results indicate that previous RT may affect the infiltration of CD8+ T lymphocytes into cervical cancer, but the effect is largely within the tumor nest, with very little effect on the tumor stroma.
Finally, to further explain the relatively ICI-insensitive nature of in-field disease, we investigated the tumor PD-L1 expression using same paired tumor samples (Fig. 3B). As shown in Table 3, 11 paired tumor samples were used for PD-L1 immunohistochemical analysis, because insufficient tumor remained after immunostaining for CD8 in the remaining cases. Representative photomicrographs of tumor PD-L1 expression in in-field diseases and newly-diagnosed cervical cancer are shown in Fig. 3B. Among the 11 in-field diseases developed after definitive RT, 7 (63.6%) were scored as 0 and 4 (36.4%) were scored as +1. In contrast, among the 11 newly-diagnosed cervical cancer, 1 (9.1%) were scored as 0, 4 (36.4%) were scored as +1, 5 (45.5%) were scored as +2, and 1 (9.1%) were scored as +3 (Table 3). When the two groups were compared, tumor PD-L1 immunoreactivity was greater in newly-diagnosed cervical cancer than in in-field diseases (p<0.001).
DISCUSSION
In our clinical investigation, we have shown, for the first time, that ICI-containing chemotherapies are less effective against in-field diseases than out-of-field diseases in patients with cervical cancer. Our immunohistochemical investigations showed tumor-infiltrating CD8+ T lymphocytes were less frequently observed in in-field diseases than in newly-diagnosed cervical cancer. Based on these, we think immunologically-cold nature of in-field diseases may be a possible reason for the in-field diseases being insensitive to ICI-containing chemotherapies.
Due to the lack of highly effective chemotherapy in this patient population as well as the previously reported promising anti-tumor effects of ICI alone or ICI-containing chemotherapies [14], physicians have very high hopes for ICIs in the treatment of recurrent cervical cancers. Moreover, due to the difficulty in operating in the previously irradiated field, most in-field recurrences have recently been treated with ICI-containing chemotherapies. The median PFS of 9.6 months and 2.9 months observed in patients treated with pembrolizumab-containing regimen and cemiplimab in the current study are almost exactly equivalent to the median PFS of 10.4 months and 2.8 months reported in the previous landmark RCTs of pembrolizumab and cemiplimab [8,9], indicating that our study population is composed of very common recurrent or advanced cervical cancer patients. Although it is necessary to verify whether ICI-containing chemotherapy is really less effective for in-field recurrence than for out-of-field recurrence in larger clinical studies, we believe that our findings, in-field diseases being insensitive to ICI-containing chemotherapies, will provide physician an alarm or an opportunity to reconsider the universal use of ICI-containing chemotherapies in patients with recurrent and advanced cervical cancer. When ICI-containing chemotherapy is employed for in-field recurrence, it is recommended that biopsy samples be obtained from locally-recurrent tumor to confirm whether tumor-infiltrating CD8+ T cells and tumor PD-L1 expression are sufficient. It is not recommended to perform such examinations using specimens obtained at the initial presentation (before RT), as the immune microenvironment differs between newly-diagnosed tumor and locally-recurrent tumor developed after RT. Until more effective anticancer agents including ICIs become clinically available, surgical salvage by pelvic exenteration or radical hysterectomy may be the only curable treatment option in this patient population, with a long-term survival rate of roughly 30%–60% [5,6,15,16]. Re-irradiation may also be an alternative as it may provide a long-term disease control in this patient population [17,18].
Recently, concurrent and adjuvant pembrolizumab with definitive RT has gained attention as a promising treatment strategy against locally-advanced cervical cancer. The rationale of combination of anti-PD-1/PD-L1 antibody therapy and radiotherapy has been explained by the preclinical evidences suggesting the immune activation by irradiation [19]. The results obtained from current investigation, in-field diseases being insensitive to ICI-containing chemotherapies or having decreased tumor PD-L1 expression, may suggest the possibility that radiation-induced immune activation is a temporal effect (i.e. during or just after radiotherapy) and reverts back to normal after a period of time following RT.
The reason for the in-field disease being less sensitive to ICI-containing chemotherapies than out-of-field diseases remains unknown. However, based on the results from our immunohistochemical analyses, we think immunologically-cold nature of in-field diseases may be a possible explanation. One is a decreased tumor PD-L1 expression (Fig. 3B), the extent of which has been shown to predict the sensitivity to ICIs in various cancers [8]. The other is associated with the tumor-infiltrating CD8+ T lymphocytes, which has long been a marker to predict the sensitivity of cancers to anticancer treatments including ICIs [20,21,22]. It has been reported that the production of interferons by tumor-infiltrating CD8+ T lymphocytes induce PD-L1 expression on tumor cells [23]. Moreover, in previous studies of cervical cancer, tumor-infiltrating CD8+ T lymphocytes density is positively associated with tumor PD-L1 expression [24]. Therefore, it is reasonable that in-field diseases exhibit decreased PD-L1 expression due to decreased tumor-infiltrating CD8+ T lymphocytes, thereby show lower sensitivity to ICI-containing chemotherapies than out-of-field diseases.
Precise mechanisms responsible for the “immunologically cold” nature of in-field disease, developed after radiotherapy, remains unknown. However, there are several possibilities. The first is a lack of tumor immunogenicity of previously-irradiated tumor. Neoantigens, present on tumor cells for T cell receptor recognition through human leukocyte antigen (HLA) molecules, are the main elicitors of T cell-mediated immune responses [25]. It has been recently reported that HLA-1 molecule can be epigenetically downregulated in tumor cells [26]. Thus, it is possible that previous RT might have affected the amount of neoantigen or HLA-1 expression in recurrent tumor, and impaired the infiltration of CD8+ T lymphocytes. The second is an impaired CD8+ T lymphocytes infiltration. CD8+ T lymphocytes are recruited and infiltrated into tumors in responses to tumor-derived chemokines [27]. Therefore, previous RT might have inhibited the production of chemokine from the in-field diseases. The third is the T cell suppression or T cell death in locally-recurrent tumor. It has been reported that radiotherapy can lead to an immunosuppressive milieu: radiation create the DNA damage in tumor cells, causing them to release immunosuppressive cytokines like transforming growth factor-β, and induce immune suppressive cells (ISCs) such as tumor-associated macrophages, myeloid-derived suppressor cells, regulatory T cells. Thus, these ISCs in in-field diseases might have inhibited the activity of CD8+ T lymphocytes, and suppressed anti-tumor immunity and promote cancer cell survival in the in-field diseases [28,29]. In addition to these, tumor stroma, mainly composed of fibroblasts and extracellular matrix, may further explain the differential sensitivities of in-field diseases or out-of-field diseases to ICIs, as tumor stroma can act as physical barriers for CD8+ T lymphocytes infiltration and survival [30]. Previous investigations have shown that CD8+ T lymphocytes exclusion occurred in patients with fibroblast- and collagen rich tumors, mainly due to the restriction of CD8+ T lymphocytes extravasation [31,32]. It is also known that radiation can cause so-called radiation-induced fibrosis (RIF) in the irradiated field, and that excess collagen reduces vascularity in the irradiated field [33], leading to a reduction in blood supply to the tumor [34]. Collectively, these radiation-induced alterations including RIF or decreased vascularity can result in an inability of CD8+ T lymphocytes to infiltrate the tumor or to hinder the effective delivery of anti-cancer agents into tumor, thereby compromise the efficacies of ICIs or ICI-containing chemotherapies against in-field disease. As the number of stromal CD8+ T lymphocytes in in-field diseases is equivalent to that observed in newly-diagnosed cervical cancer (Table 2), RIF that usually occur 4 to 12 months after radiotherapy exposure [35] as well as the reduced vascularity in the irradiated field, might have affected CD8+ T lymphocytes infiltration into tumor nest in locally-recurrent cervical cancer. To clarify abovementioned mechanism, well designed future muti-institutional investigation, enrolling only patients in which paired clinical samples from initial tumor and recurrent tumor (in-field or out-of-field) are available, is needed.
A main strength of our study is, although retrospective, this is the first study that investigated the efficacy of ICI-containing chemotherapies according to the site of diseases: in-field diseases versus out-of-field diseases. Previously published two landmark RCTs did not include such investigations [8,9]. The second strength is that we have included mechanistic investigation to provide an explanation for differential sensitivity to ICI-containing chemotherapies according to the site of disease (in-field diseases versus. out-of-field diseases).
Our study had several limitations. First is the small sample size. The second is its retrospective design and the potential for selection bias from the choice of surgery as salvage treatment (i.e. surgery or ICI-containing chemotherapies) being at the physicians’ discretion. The third is that the study population was heterogeneous—ICIs employed in our patients include both pembrolizumab and cemiplimab, which may have slightly different mechanisms of action. Forth, we could not investigate the immunoreactivities for CD8 and PD-L1 in out-of-field diseases, because tumor samples were not available as most of these are diagnosed via imaging studies without performing biopsy.
In conclusion, the present study demonstrates for the first time that ICI-containing chemotherapies are less effective against in-field cervical cancer recurrences developed after radiotherapy. Decreased CD8+ T lymphocytes and decreased PD-L1 expression in in-field diseases may explain the immunotherapy-insensitive nature of this type of tumors.
Footnotes
Conflict of Interest: No potential conflict of interest relevant to this article was reported.
- Conceptualization: M.S.
- Data curation: M.S., K.N., N.H., M.M., Y.T., K.S., U.T.
- Formal analysis: M.S., K.N.
- Investigation: M.S., K.N., N.H., M.M., Y.T., K.S., U.T.
- Methodology: M.S., K.N.
- Project administration: M.S., K.N.
- Resources: N.H.
- Validation: N.H.
- Visualization: K.N.
- Writing - original draft: M.S., K.N.
- Writing - review & editing: M.S.
SUPPLEMENTARY MATERIALS
Clinicopathological characteristics and outcomes of the patients with in-field diseases
Clinicopathological characteristics and outcomes of the patients with out-of-field field diseases
Overview of the current study population (clinical investigation and Immunohistochemical investigation).
Scoring of immunoreactivity in locally recurrent cervical cancer developed after definitive radiotherapy (in-field recurrence) and in matched primary cervical tumors (newly diagnosed) for CD8 and PD-L1. (A) Representative photomicrographs illustrating Scores 0, 1, 2, and 3 for CD8 (magnification, ×400). (B) Representative photomicrographs illustrating Scores 0, 1, 2, and 3 for PD-L1 (magnification, ×400).
References
- 1.Cohen PA, Jhingran A, Oaknin A, Denny L. Cervical cancer. Lancet. 2019;393:169–182. doi: 10.1016/S0140-6736(18)32470-X. [DOI] [PubMed] [Google Scholar]
- 2.Peiretti M, Zapardiel I, Zanagnolo V, Landoni F, Morrow CP, Maggioni A. Management of recurrent cervical cancer: a review of the literature. Surg Oncol. 2012;21:e59–e66. doi: 10.1016/j.suronc.2011.12.008. [DOI] [PubMed] [Google Scholar]
- 3.Hong JH, Tsai CS, Lai CH, Chang TC, Wang CC, Chou HH, et al. Recurrent squamous cell carcinoma of cervix after definitive radiotherapy. Int J Radiat Oncol Biol Phys. 2004;60:249–257. doi: 10.1016/j.ijrobp.2004.02.044. [DOI] [PubMed] [Google Scholar]
- 4.Ubinha ACF, Pedrão PG, Tadini AC, Schmidt RL, Santos MHD, Andrade CEMD, et al. The role of pelvic exenteration in cervical cancer: a review of the literature. Cancers (Basel) 2024;16:817. doi: 10.3390/cancers16040817. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Mabuchi S, Matsumoto Y, Komura N, Sawada M, Tanaka M, Yokoi E, et al. The efficacy of surgical treatment of recurrent or persistent cervical cancer that develops in a previously irradiated field: a monoinstitutional experience. Int J Clin Oncol. 2017;22:927–936. doi: 10.1007/s10147-017-1134-x. [DOI] [PubMed] [Google Scholar]
- 6.Mabuchi S, Kozasa K, Kimura T. Radical hysterectomy after radiotherapy for recurrent or persistent cervical cancer. Int J Gynaecol Obstet. 2017;139:185–191. doi: 10.1002/ijgo.12284. [DOI] [PubMed] [Google Scholar]
- 7.Pectasides D, Fountzilas G, Papaxoinis G, Pectasides E, Xiros N, Sykiotis C, et al. Carboplatin and paclitaxel in metastatic or recurrent cervical cancer. Int J Gynecol Cancer. 2009;19:777–781. doi: 10.1111/IGC.0b013e3181a40a8b. [DOI] [PubMed] [Google Scholar]
- 8.Colombo N, Dubot C, Lorusso D, Caceres MV, Hasegawa K, Shapira-Frommer R, et al. Pembrolizumab for persistent, recurrent, or metastatic cervical cancer. N Engl J Med. 2021;385:1856–1867. doi: 10.1056/NEJMoa2112435. [DOI] [PubMed] [Google Scholar]
- 9.Tewari KS, Monk BJ, Vergote I, Miller A, de Melo AC, Kim HS, et al. Survival with cemiplimab in recurrent cervical cancer. N Engl J Med. 2022;386:544–555. doi: 10.1056/NEJMoa2112187. [DOI] [PubMed] [Google Scholar]
- 10.Zhang J, Huang D, Saw PE, Song E. Turning cold tumors hot: from molecular mechanisms to clinical applications. Trends Immunol. 2022;43:523–545. doi: 10.1016/j.it.2022.04.010. [DOI] [PubMed] [Google Scholar]
- 11.Matsumoto Y, Mabuchi S, Isohashi F, Komura N, Ogawa K, Kimura T. Impact of a reduction in follow-up frequency on life expectancy in uterine cervical cancer patients. Int J Clin Oncol. 2020;25:1170–1177. doi: 10.1007/s10147-020-01641-w. [DOI] [PubMed] [Google Scholar]
- 12.Mabuchi S, Isohashi F, Maruoka S, Hisamatsu T, Takiuchi T, Yoshioka Y, et al. Post-treatment follow-up procedures in cervical cancer patients previously treated with radiotherapy. Arch Gynecol Obstet. 2012;286:179–185. doi: 10.1007/s00404-012-2235-4. [DOI] [PubMed] [Google Scholar]
- 13.Armato SG, 3rd, Nowak AK. Revised modified response evaluation criteria in solid tumors for assessment of response in malignant pleural mesothelioma (version 1.1) J Thorac Oncol. 2018;13:1012–1021. doi: 10.1016/j.jtho.2018.04.034. [DOI] [PubMed] [Google Scholar]
- 14.Makker V, Colombo N, Casado Herráez A, Santin AD, Colomba E, Miller DS, et al. Lenvatinib plus pembrolizumab for advanced endometrial cancer. N Engl J Med. 2022;386:437–448. doi: 10.1056/NEJMoa2108330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mabuchi S, Komura N, Kodama M, Matsuzaki S, Matsumoto Y, Kamiura S, et al. Impact of lymphadenectomy in patients with locally recurrent or persistent cervical cancer treated with salvage hysterectomy. J Obstet Gynaecol Res. 2023;49:717–724. doi: 10.1111/jog.15495. [DOI] [PubMed] [Google Scholar]
- 16.Mabuchi S, Shimura K, Matsumoto Y. The significance of post-radiotherapy parametrial involvement and the necessity of parametrial resection in locally-recurrent or persistent cervical cancer developed after radiotherapy. J Gynecol Obstet Hum Reprod. 2021;50:102190. doi: 10.1016/j.jogoh.2021.102190. [DOI] [PubMed] [Google Scholar]
- 17.Isohashi F, Yoshida K, Murakami N, Masui K, Ishihara S, Ohkubo Y, et al. Reirradiation for recurrent gynecologic cancer using high-dose-rate brachytherapy in Japan: a multicenter survey on practice patterns and outcomes. Radiother Oncol. 2024;195:110269. doi: 10.1016/j.radonc.2024.110269. [DOI] [PubMed] [Google Scholar]
- 18.Mabuchi S, Takahashi R, Isohashi F, Yokoi T, Okazawa M, Sasano T, et al. Reirradiation using high-dose-rate interstitial brachytherapy for locally recurrent cervical cancer: a single institutional experience. Int J Gynecol Cancer. 2014;24:141–148. doi: 10.1097/IGC.0000000000000028. [DOI] [PubMed] [Google Scholar]
- 19.Sato H, Okonogi N, Nakano T. Rationale of combination of anti-PD-1/PD-L1 antibody therapy and radiotherapy for cancer treatment. Int J Clin Oncol. 2020;25:801–809. doi: 10.1007/s10147-020-01666-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Brummel K, Eerkens AL, de Bruyn M, Nijman HW. Tumour-infiltrating lymphocytes: from prognosis to treatment selection. Br J Cancer. 2023;128:451–458. doi: 10.1038/s41416-022-02119-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ohno A, Iwata T, Katoh Y, Taniguchi S, Tanaka K, Nishio H, et al. Tumor-infiltrating lymphocytes predict survival outcomes in patients with cervical cancer treated with concurrent chemoradiotherapy. Gynecol Oncol. 2020;159:329–334. doi: 10.1016/j.ygyno.2020.07.106. [DOI] [PubMed] [Google Scholar]
- 22.Tumeh PC, Harview CL, Yearley JH, Shintaku IP, Taylor EJ, Robert L, et al. PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature. 2014;515:568–571. doi: 10.1038/nature13954. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bald T, Landsberg J, Lopez-Ramos D, Renn M, Glodde N, Jansen P, et al. Immune cell-poor melanomas benefit from PD-1 blockade after targeted type I IFN activation. Cancer Discov. 2014;4:674–687. doi: 10.1158/2159-8290.CD-13-0458. [DOI] [PubMed] [Google Scholar]
- 24.Rivera-Colon G, Chen H, Molberg K, Niu S, Strickland AL, Castrillon DH, et al. PD-L1 expression in endocervical adenocarcinoma: correlation with patterns of tumor invasion, CD8+ tumor-infiltrating lymphocytes, and clinical outcomes. Am J Surg Pathol. 2021;45:742–752. doi: 10.1097/PAS.0000000000001633. [DOI] [PubMed] [Google Scholar]
- 25.Kim K, Kim HS, Kim JY, Jung H, Sun JM, Ahn JS, et al. Predicting clinical benefit of immunotherapy by antigenic or functional mutations affecting tumour immunogenicity. Nat Commun. 2020;11:951. doi: 10.1038/s41467-020-14562-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yamamoto K, Venida A, Yano J, Biancur DE, Kakiuchi M, Gupta S, et al. Autophagy promotes immune evasion of pancreatic cancer by degrading MHC-I. Nature. 2020;581:100–105. doi: 10.1038/s41586-020-2229-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Nagarsheth N, Wicha MS, Zou W. Chemokines in the cancer microenvironment and their relevance in cancer immunotherapy. Nat Rev Immunol. 2017;17:559–572. doi: 10.1038/nri.2017.49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Mabuchi S, Matsumoto Y, Kawano M, Minami K, Seo Y, Sasano T, et al. Uterine cervical cancer displaying tumor-related leukocytosis: a distinct clinical entity with radioresistant feature. J Natl Cancer Inst. 2014;106:dju147. doi: 10.1093/jnci/dju147. [DOI] [PubMed] [Google Scholar]
- 29.Komura N, Mabuchi S, Shimura K, Yokoi E, Kozasa K, Kuroda H, et al. The role of myeloid-derived suppressor cells in increasing cancer stem-like cells and promoting PD-L1 expression in epithelial ovarian cancer. Cancer Immunol Immunother. 2020;69:2477–2499. doi: 10.1007/s00262-020-02628-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Sun X, Wu B, Chiang HC, Deng H, Zhang X, Xiong W, et al. Tumour DDR1 promotes collagen fibre alignment to instigate immune exclusion. Nature. 2021;599:673–678. doi: 10.1038/s41586-021-04057-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Tauriello DVF, Palomo-Ponce S, Stork D, Berenguer-Llergo A, Badia-Ramentol J, Iglesias M, et al. TGFβ drives immune evasion in genetically reconstituted colon cancer metastasis. Nature. 2018;554:538–543. doi: 10.1038/nature25492. [DOI] [PubMed] [Google Scholar]
- 32.Xiao Z, Todd L, Huang L, Noguera-Ortega E, Lu Z, Huang L, et al. Desmoplastic stroma restricts T cell extravasation and mediates immune exclusion and immunosuppression in solid tumors. Nat Commun. 2023;14:5110. doi: 10.1038/s41467-023-40850-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Lefaix JL, Daburon F. Diagnosis of acute localized irradiation lesions: review of the French experimental experience. Health Phys. 1998;75:375–384. doi: 10.1097/00004032-199810000-00003. [DOI] [PubMed] [Google Scholar]
- 34.Paris F, Fuks Z, Kang A, Capodieci P, Juan G, Ehleiter D, et al. Endothelial apoptosis as the primary lesion initiating intestinal radiation damage in mice. Science. 2001;293:293–297. doi: 10.1126/science.1060191. [DOI] [PubMed] [Google Scholar]
- 35.Zheng M, Liu Z, He Y. Radiation-induced fibrosis: Mechanisms and therapeutic strategies from an immune microenvironment perspective. Immunology. 2024;172:533–546. doi: 10.1111/imm.13788. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Clinicopathological characteristics and outcomes of the patients with in-field diseases
Clinicopathological characteristics and outcomes of the patients with out-of-field field diseases
Overview of the current study population (clinical investigation and Immunohistochemical investigation).
Scoring of immunoreactivity in locally recurrent cervical cancer developed after definitive radiotherapy (in-field recurrence) and in matched primary cervical tumors (newly diagnosed) for CD8 and PD-L1. (A) Representative photomicrographs illustrating Scores 0, 1, 2, and 3 for CD8 (magnification, ×400). (B) Representative photomicrographs illustrating Scores 0, 1, 2, and 3 for PD-L1 (magnification, ×400).