Cancer Aware · Patient & Family Guide

Immunotherapy in Cervical Cancer

Immune checkpoint inhibitors release some of the brakes that tumours use to evade immune attack. They can help in particular settings, but are not suitable for every cervical cancer.

Educational information for patients and families. This guide supports, but does not replace, individualized medical advice. It cannot diagnose, stage cancer or prescribe treatment.

All 25 chapters

Chapter 18 of 25

At a glance: Immunotherapy in Cervical Cancer
  1. Identify curative or metastatic setting
  2. Check the relevant biomarker and label
  3. Discuss immune toxicity and prior treatment
  4. Assess response and safety over time

What you will learn

Understand PD-1/PD-L1 and CPS; separate KEYNOTE-826 from KEYNOTE-A18; recognize immune-related adverse effects.

How eligibility is assessed

Pembrolizumab blocks PD-1, a signalling pathway involved in limiting immune responses. A PD-L1 Combined Positive Score, or CPS, counts appropriately staining tumour and immune cells relative to viable tumour cells. It is not a percentage chance of response and is not interchangeable with every other PD-L1 score. Ask which assay and cutoff apply to the proposed treatment.

In persistent, recurrent or metastatic disease, pembrolizumab with platinum/taxane chemotherapy, with or without bevacizumab, has a verified US indication for CPS at least 1. Pembrolizumab alone also has a cervical indication after chemotherapy for appropriate PD-L1-positive disease. Tumour-agnostic indications, such as MSI-high or mismatch-repair-deficient disease, are separate and have their own criteria. A biomarker result alone does not establish the safest option.

KEYNOTE-826: advanced disease

This randomized phase III trial enrolled 617 patients with persistent, recurrent or metastatic cervical cancer not amenable to curative treatment and no prior systemic chemotherapy for that setting; prior radiosensitizing chemotherapy was allowed. Pembrolizumab was compared with placebo, both with platinum/paclitaxel and optional bevacizumab.

In the final analysis published in Journal of Clinical Oncology in 2023, median overall survival in the CPS-at-least-1 population was 28.6 versus 16.5 months, hazard ratio 0.60. These are group results, not an individual lifespan prediction. The small CPS-negative subgroup does not justify assuming the same benefit irrespective of biomarker or label, and evidence after prior checkpoint therapy is a different question.

KEYNOTE-A18: curative-intent treatment

This randomized trial enrolled 1,060 previously untreated patients with FIGO 2014 IB2–IIB node-positive or III–IVA cervical cancer. Pembrolizumab or placebo accompanied chemoradiation, followed by continued pembrolizumab or placebo. The 2024 Lancet overall-survival analysis reported 36-month survival of 82.6% versus 74.8%, hazard ratio 0.67, in the trial population.

The verified US approval applies to FIGO 2014 III–IVA disease with chemoradiotherapy and does not require PD-L1 selection. It must not be described as a universal indication for every locally advanced stage. This regimen includes appropriate external-beam radiation and brachytherapy; immunotherapy does not replace either. Regional approvals and reimbursement can differ.

Immune toxicity requires early action

Checkpoint therapy can inflame healthy organs, including lungs, bowel, liver, thyroid, pituitary, kidneys and, rarely, the heart or nervous system. New breathlessness, chest pain, severe diarrhoea, jaundice, marked weakness, severe headache, confusion or reduced urine output need urgent contact. Toxicities can arise after treatment ends. Carry the drug name and an immunotherapy alert card when available.

Tell the team about autoimmune disease, an organ transplant, immune-suppressing medicines and pregnancy possibility. Treatment may need to pause and steroids or other therapies may be required; do not self-treat or abruptly stop prescribed steroids. Blood tests and symptom review continue even when you feel well. Scan changes need clinical interpretation; apparent progression should not automatically be called “pseudoprogression.”

Misconception and common question

“Immunotherapy is natural and has no serious side effects” is false. Does CPS 10 mean a 10% chance of benefit? No: CPS is a staining score, not a prediction percentage.

Take-home messages

Identify the disease setting, biomarker requirement and local approval. Discuss prior immunotherapy exposure. Report new symptoms early, including after the last dose.

References and review status

Evidence/editorial check: 28 October 2026. AI-assisted preparation is separate from physician sign-off. Individual physician review is not recorded. Local approvals, resources and clinical guidance must be confirmed by the treating team.

Evidence scope and update policy