Immunology · Year 2 · from Immunology

Case 3: MSI-High Colorectal Cancer Responding to Immunotherapy

Patient Presentation

Demographics: 52-year-old female

Chief Complaint: Follow-up for metastatic colorectal cancer treatment

History of Present Illness: The patient was diagnosed with metastatic colorectal cancer 6 months ago. Molecular testing revealed the tumor is microsatellite instability-high (MSI-H) with deficient mismatch repair (dMMR). She was started on pembrolizumab (anti-PD-1) as first-line therapy. After 4 cycles, she reports feeling well with improved energy and appetite.

Past Medical History:

  • Colorectal cancer with liver metastases
  • No family history of Lynch syndrome
  • No prior chemotherapy

Physical Examination:

  • General: Well-appearing, improved from prior visits
  • Abdomen: Soft, non-tender, no hepatomegaly
  • No lymphadenopathy

Workup and Results

Laboratory Studies:

  • CEA: 12 ng/mL (down from 156 ng/mL at diagnosis)
  • LFTs: Normal
  • CBC: Normal

Imaging:

  • CT chest/abdomen/pelvis: 70% reduction in liver metastases

Molecular Testing (at diagnosis):

  • MSI-H (microsatellite instability-high)
  • dMMR (deficient mismatch repair: MLH1 loss)
  • Tumor mutational burden: 42 mutations/Mb (high)

Clinical Image

CT comparison showing significant tumor regression in MSI-H colorectal cancer liver metastases after pembrolizumab treatment, demonstrating the excellent response of high tumor mutational burden tumors to checkpoint inhibitor therapy.

Diagnosis

MSI-High Metastatic Colorectal Cancer with Excellent Response to Checkpoint Inhibitor

Discussion

This case illustrates predictors of immunotherapy response:

  • MSI-H/dMMR Predicts Response: The lecture describes how MSI-H tumors have deficient mismatch repair, leading to accumulation of mutations and high tumor mutational burden (TMB). This generates numerous neoantigens that make tumors highly immunogenic.
  • TMB Correlates with Response: The lecture notes that tumor mutational burden correlates with neoantigen load and checkpoint inhibitor response. MSI-H colorectal cancers have 10-100x higher TMB than microsatellite stable tumors.
  • Tissue-Agnostic Approval: The lecture explains that pembrolizumab has tissue-agnostic FDA approval for MSI-H/dMMR tumors regardless of site of origin - the first such approval based on a biomarker rather than tumor type.
  • "Hot" vs "Cold" Tumors: MSI-H tumors are typically "hot" (T-cell infiltrated) due to their high neoantigen load, making them excellent candidates for checkpoint inhibitors.

Treatment Plan

  1. Continue Pembrolizumab:
  • Standard dosing every 3 weeks
  • Continue for up to 2 years if responding
  1. Monitoring:
  • CEA every 6 weeks
  • CT imaging every 3 months
  • Monitor for irAEs
  1. Genetic Counseling:
  • Consider Lynch syndrome evaluation
  • MLH1 promoter methylation testing (sporadic vs germline)
  • Family history assessment
  1. Future Considerations:
  • If complete response: consider treatment discontinuation at 2 years
  • Surveillance protocol after completion

Teaching Points

  1. MSI-H/dMMR tumors have high TMB and are highly responsive to checkpoint inhibitors
  2. Pembrolizumab has tissue-agnostic approval for MSI-H/dMMR tumors
  3. Hot tumors (T-cell infiltrated) respond better to immunotherapy than cold tumors
  4. Testing for MSI/MMR status should be routine in colorectal cancer
  5. Lynch syndrome should be considered in patients with dMMR tumors

Image Reference

For visual reference of tumor immunology concepts, see:


Learning Points

  1. irAEs Mechanism: Result from T cell activation against normal tissues when checkpoints are blocked
  1. CRS Treatment: Tocilizumab (anti-IL-6R) is first-line; corticosteroids for ICANS
  1. MSI-H Predicts Response: High tumor mutational burden generates neoantigens making tumors immunogenic
  1. Hot vs Cold Tumors: Hot (T-cell infiltrated) tumors respond better to checkpoint inhibitors
  1. Tissue-Agnostic Approval: Pembrolizumab approved for MSI-H/dMMR tumors regardless of site of origin

All cases for this lecture as Markdown