Cervical cancer

Overview and epidemiology

• Epidemiology (US 2025): ~14,000 new cases, ~4,300 deaths. Third most common gynecologic malignancy in the US. Globally, a leading cause of cancer death in resource-poor regions where screening and prevention programs are limited.
• HPV (esp. high-risk types 16, 18, 31, 33, 45) is causative in nearly all cervical cancer. HPV accounts for >90% of US cases; strains 16 or 18 account for ~70%.
• Pathogenesis: cervical cancer development is stepwise (persistent HR-HPV infection → precancer/CIN → invasive cancer). Progression from persistent infection to preinvasive disease often takes up to 5 years, and persistent infection/precancer arises in <10% of new infections. Invasive cancer arises over many years in only a minority with precancer, which is why screening and vaccination are effective.
• Risk factors: early intercourse, multiple partners, smoking (increases risk of persistent dysplasia), immunosuppression (HIV, transplant), other STDs, oral contraceptives (modest). Intrauterine device use has been associated with a lower risk of cervical cancer independent of HPV status.
• Histology: squamous cell (~70%), adenocarcinoma (~25%), adenosquamous, small cell/neuroendocrine (rare, aggressive), clear cell. Rare subtypes (clear cell, neuroendocrine) confer worse prognosis and need a specialized multidisciplinary approach.

Prevention and screening

• HPV vaccination: recombinant 9-valent vaccine (Gardasil 9, covers 9 types) routine at 11 to 12 (may begin at 9), catch-up through 26; ages 27 to 45 by shared clinical decision-making. Expanded from the original ≤26-year approval based on a randomized trial showing efficacy in people ≥26 and prevention of other HPV conditions (VIN, anal disease, oral HPV). Age-appropriate vaccination may still be given despite existing HPV infection or dysplasia (protects against not-yet-acquired vaccine types); it has no therapeutic effect on the existing infection or lesion.
• Screening (USPSTF 2024):
  • Ages 21 to 29: cervical cytology every 3 years.
  • Ages 30 to 65: cytology every 3 years OR primary HPV testing every 5 years OR co-test every 5 years.
  • Stop at age 65 if adequate prior screening.
  • After hysterectomy for benign reasons with no high-grade lesion: no screening.

Workup and staging (FIGO 2018)

• Most patients are asymptomatic early. Vaginal bleeding (intermenstrual, postcoital), pelvic pain, and vaginal discharge suggest more advanced disease.
• Pelvic exam, biopsy, exam under anesthesia (EUA) in advanced disease.
• Imaging: pelvic MRI for local extent; PET-CT for stage IB3 and above to assess pelvic/para-aortic nodes and distant disease.
• Surgical staging (para-aortic LN dissection) is an option for locally advanced disease.
• FIGO 2018 update: incorporates nodal involvement as stage IIIC (pelvic IIIC1 or para-aortic IIIC2), with clarification of whether detected radiologically (r) or pathologically (p). In US practice, MRI and PET are commonly used for staging, including nodal assessment.

Treatment by stage

• IA1 (≤3 mm depth, no LVSI): cervical conization (fertility-sparing) or simple hysterectomy.
• IA2 (3 to 5 mm): radical trachelectomy or radical hysterectomy + pelvic LND (or SLNB).

• Early invasive disease (stage IA or IB1) has excellent long-term survival, 5-year relative survival >90%. Primary treatment is surgical.
• Radical hysterectomy + pelvic LND OR definitive chemoRT.
• Open approach preferred: the LACC (Laparoscopic Approach to Cervical Cancer) trial randomized >600 patients with stage IA1/IA2/IB1 to minimally invasive vs open radical hysterectomy; minimally invasive surgery had worse DFS (HR 3.74) and OS (HR 6.0). Use open laparotomy, not minimally invasive, for radical hysterectomy.
• SHAPE trial: simple hysterectomy + pelvic LND non-inferior to radical for low-risk IA2/IB1 ≤2 cm.
• Fertility preservation: for lowest-risk disease (lesion <2 cm, node negative) desiring fertility, conization or trachelectomy (removal of cervix alone) are reasonable alternatives to radical hysterectomy.
• Adjuvant pelvic RT (with or without concurrent chemo) may be recommended based on final pathology: Intermediate risk (Sedlis: combinations of tumor size, LVSI, depth of stromal invasion): pelvic RT. High risk (positive nodes, positive margins, or parametrial involvement): concurrent cisplatin-based chemoradiation.

• Stage IB3 to IVA is locally advanced. Surgery should NOT be performed before chemoRT in locally advanced disease (high risk of short and long-term bowel/bladder toxicity). Backbone is definitive EBRT + brachytherapy with concurrent cisplatin. For FIGO 2014 stage III to IVA, add pembrolizumab (KEYNOTE-A18); INTERLACE induction chemo is another option for selected patients.
• Concurrent chemoRT: cisplatin 40 mg/m2 weekly (during external beam RT and brachytherapy) + EBRT (~45 Gy) + brachytherapy boost. Para-aortic node involvement (suspected or confirmed) warrants extended-field EBRT.
• INTERLACE (McCormack Lancet 2024): induction chemotherapy (carboplatin AUC2 + paclitaxel 80 mg/m2 weekly for 6 weeks) immediately before standard chemoRT in locally advanced cervical cancer. 5-year PFS 72% vs 64% (HR 0.65) and 5-year OS 80% vs 72% (HR ~0.60) favoring induction. Establishes short-course induction chemo before chemoRT as an option.
• KEYNOTE-A18 (Lorusso Lancet 2024; PMID: 38521086): pembrolizumab + concurrent chemoRT followed by adjuvant pembrolizumab (400 mg q6w) for 15 cycles. Trial enrolled FIGO 2014 IB2 to IIB node-positive or III to IVA; HRs are overall ITT. FDA indication (Jan 12, 2024) restricted to FIGO 2014 III to IVA. PFS HR 0.70 (24-mo PFS 67.8% vs 57.3%); OS endpoint also met. FDA Jan 12, 2024.
• KEYNOTE-A18 OS update (Lorusso Lancet 2024; PMID 39288779): mature 3-year OS 82.6% vs 74.8% (HR 0.67, p=0.0040), the first positive OS readout for IO + chemoRT in locally advanced cervical cancer. Establishes pembro + chemoRT as new standard for FIGO 2014 stage III to IVA (do not equate to FIGO 2018 ≥IIIC1: nodal upstaging alone is not equivalent, and IVB is excluded).
• CALLA (negative): durvalumab + chemoRT did not improve PFS in a similar setting.

• Isolated pelvic recurrence within a previously irradiated field: pelvic exenteration (en bloc resection with bowel and/or bladder) can be considered for curative intent in carefully selected patients; morbidity is substantial.
• PD-L1 negative, not a surgical candidate: first-line cisplatin + paclitaxel + bevacizumab (GOG 240 showed adding bevacizumab to chemotherapy increased OS by nearly 4 months), or atezolizumab + platinum/paclitaxel + bevacizumab (BEATcc; NCCN-listed regardless of PD-L1, not FDA-approved in cervical cancer). Carboplatin may substitute for cisplatin (more tolerable, especially after prior cisplatin/RT).
• Pembrolizumab + chemo with or without bevacizumab (KEYNOTE-826) for PD-L1 CPS ≥1:
  • Eligibility: PD-L1 CPS ≥1 (22C3 pharmDx assay; CPS = 100 x (PD-L1-staining tumor cells, lymphocytes, and macrophages) / (total viable tumor cells), capped at 100).
  • Efficacy (CPS ≥1): OS HR 0.64 (24-mo OS 53.0% vs 41.7%, median OS not reached vs ~16 mo); mPFS 10.4 vs 8.2 mo (HR 0.62). The 24.4 vs 16.5 mo, HR 0.67 values were the ITT population.
  • FDA approval: Oct 13, 2021 (Colombo NEJM 2021; PMID: 34534429).
  • KEYNOTE-826 final analysis (Monk JCO 2023): confirmed mOS 28.6 vs 16.5 mo (HR 0.60) in CPS ≥1 (all-comers 26.4 vs 16.8 mo, HR 0.63); 3-year OS 42.2% vs 26.7%. Confirms pembro + chemo with or without bev as 1L standard for PD-L1 CPS ≥1 metastatic/recurrent disease.
• BEATcc (Oaknin Lancet 2024): 1L metastatic/persistent/recurrent cervical cancer, atezolizumab + carboplatin/paclitaxel + bevacizumab vs chemo + bev (no PD-L1 selection). mPFS 13.7 vs 10.4 mo (HR 0.62); mOS 32.1 vs 22.8 mo (HR 0.68). Confirms class effect of IO + chemo + bev, positive regardless of PD-L1.
• Pembrolizumab monotherapy (2L) (KEYNOTE-158): FDA-approved after progression on chemo for patients not previously given pembrolizumab who have PD-L1-positive or MSI-H/dMMR tumors. ORR 12% overall (14.6% in PD-L1-positive); 6-mo PFS 25%; mOS 9.4 mo.
• Cemiplimab (EMPOWER-Cervical-1, Tewari NEJM 2022; PMID: 35139273): 2L recurrent/metastatic after platinum chemo, PD-L1 unselected, mOS 12.0 vs 8.5 mo (HR 0.69). EU/Canada/Japan approved 2022; FDA BLA WITHDRAWN 2022 (application pulled after FDA disagreement on post-marketing studies). Not FDA-approved for cervical cancer; pembrolizumab + chemo (KEYNOTE-826) remains the standard 1L IO option in the US.
• Tisotumab vedotin (innovaTV 301):
  • Mechanism: tissue factor (TF)-directed antibody-drug conjugate (microtubule inhibitor payload).
  • Efficacy: mOS 11.5 vs 9.5 mo (HR 0.70 vs chemo) in 2L/3L and beyond.
  • FDA approval: accelerated Sept 20, 2021 (innovaTV 204, ORR 24%, median response duration 8.3 mo); full approval Apr 29, 2024 based on innovaTV 301 (Vergote NEJM 2024), HR 0.70 vs chemo.
  • Toxicity: ocular AEs (mandate an eye-care plan, lubricating drops, vasoconstrictor drops before infusion), peripheral neuropathy, bleeding.
• Other 2L and beyond: topotecan, gemcitabine, vinorelbine, etc.
• Larotrectinib/entrectinib for NTRK fusion (rare).

Rare histologies

• Clear cell carcinoma (~4% of cervical adenocarcinomas, ≤1% of all cervical cancers) and neuroendocrine tumors (large- and small-cell, ~1 to 2% of cervical cancers) are rare. No randomized trials (only small prospective series, e.g., pembrolizumab phase II, n=7, minimal activity); treatment (platinum-based) is extrapolated from squamous/adenocarcinoma and small-cell lung cancer respectively. These patients are typically younger, more often metastatic, and have poorer survival at all stages.

Special situations and survivorship

• Fertility preservation: radical trachelectomy for selected IA2 to IB1 ≤2 cm.
• Pregnancy: individualize; consider delaying definitive treatment past the second trimester for very early disease.
• Small cell / neuroendocrine cervical cancer: aggressive, treat like SCLC.
• Renal insufficiency: rule out correctable causes (ureteral obstruction from tumor may need stent or nephrostomy) so cisplatin can be given; if cisplatin not feasible, use carboplatin, gemcitabine, or fluorouracil.
• Older patients (>65): offer curative-intent treatment but expect higher toxicity; carboplatin may be better tolerated than cisplatin.
• Survivorship: 5-year survival ~66% all stages, ~91% localized; >40% with regional nodal involvement die within 5 years. Follow with history and pelvic exam every 3 to 6 months for 2 years, every 6 to 12 months for years 3 to 5, then annually; imaging if clinically indicated. Watch for premature menopausal genitourinary symptoms (HRT may be offered in age-appropriate context), vaginal stenosis (vaginal dilators), and lower-extremity lymphedema after node dissection.

Vulvar note

• Differentiated VIN (dVIN) is NOT HPV-associated, is aggressive, and arises in lichen sclerosus (papery texture around the lesion). Adherence to topical steroids in lichen sclerosus lowers vulvar cancer risk: in a prospective cohort, no patients using steroids as directed over a mean 4.7 years developed vulvar SCC or VIN vs 4.7% of partially adherent patients (p<0.001).

High-yield cervical pearls

• HPV is causal (16 and 18 = ~70%): vaccinate routinely at 11 to 12 (from age 9), catch-up through 26; ages 27 to 45 by shared decision-making.
• Open radical hysterectomy, not minimally invasive (LACC).
• Concurrent cisplatin + RT + brachytherapy is the locally advanced backbone.
• INTERLACE: short-course induction chemo before chemoRT improves PFS/OS.
• KEYNOTE-A18: pembro + chemoRT then adjuvant pembro for stage III to IVA (positive OS).
• KEYNOTE-826: pembro + chemo with or without bev for PD-L1 CPS ≥1 metastatic.
• Tisotumab vedotin: 2L/3L and beyond; watch for ocular toxicity.
• SHAPE: simple hysterectomy OK for low-risk IB1 ≤2 cm.
• Stage IIIC (FIGO 2018) = nodal involvement.