Pignatta, S., Tazzari, M., Indio, V., Tumedei, M. M., Limarzi, F., Tauceri, F., Tebaldi, M., Fanelli, D., & Bulgarelli, J. (2025). Comparative analysis of library preparation approaches for FFPE gene expression profiling and related recommendations. Scientific Reports, 15, 12992.

Pignatta et al. (2025) — Comparative Analysis of Library Preparation for FFPE Gene Expression

Bibliographic Reference

Pignatta, S., Tazzari, M., Indio, V., Tumedei, M. M., Limarzi, F., Tauceri, F., Tebaldi, M., Fanelli, D., & Bulgarelli, J. (2025). Comparative analysis of library preparation approaches for FFPE gene expression profiling and related recommendations. Scientific Reports, 15, 12992. DOI: 10.1038/s41598-025-12992-7.

Core Argument

RNA-seq from FFPE tissue faces two compounding challenges: RNA degradation from formalin fixation AND limited input material from small clinical biopsies. Library preparation kit selection directly determines whether sufficient transcriptomic data can be extracted from scarce FFPE specimens. The authors present a direct comparison of two stranded RNA-seq kits — TaKaRa SMARTer Stranded Total RNA-Seq Kit v2 (Kit A, requiring 1 ng input) vs. Illumina Stranded Total RNA Prep Ligation with Ribo-Zero Plus (Kit B, requiring 20 ng input) — using FFPE melanoma samples. Both kits generated high-quality data, but Kit A matched Kit B’s performance with 20-fold less RNA input, at the cost of increased sequencing depth. The study includes a practical pathologist-assisted microdissection workflow for maximizing tumor-content RNA extraction from FFPE blocks.

Methods

• Specimens: 6 FFPE melanoma samples (lymph node metastases) from patients treated with nivolumab
• Macrodissection: Pathologist-assisted marking of ROI (tumor microenvironment regions for RNA, high-tumor-content regions for DNA)
• RNA quality: DV200 metric (percentage of RNA fragments >200 nt) used for FFPE RNA assessment
• Library prep: Kit A (TaKaRa SMARTer v2, 1 ng input, random priming + template switching) vs. Kit B (Illumina Ribo-Zero Plus, 20 ng input, ligation-based)
• Sequencing: NovaSeq 6000, paired-end
• Analysis: Gene expression quantification, correlation analysis, differential expression comparison, cost/time assessment

Key Findings

• Kit A achieves comparable gene expression with 20-fold less RNA. Gene expression quantification from Kit A (1 ng input) was highly concordant with Kit B (20 ng input), with strong correlation in gene-level counts. This is a critical advantage for small FFPE biopsies where RNA yield is limiting — a common scenario in clinical archives where only a few 5–10 µm sections are available.

• Kit B provides broader transcript coverage at lower sequencing depth. The ligation-based Kit B captured a larger number of unique genes at equivalent sequencing depth, reflecting its more comprehensive rRNA depletion and broader transcript representation. Kit A required deeper sequencing to achieve comparable gene detection breadth — a cost-quality trade-off.

• Pathologist-assisted microdissection is essential for FFPE RNA-seq. The authors’ workflow includes explicit ROI marking by a pathologist to separate tumor microenvironment regions (for RNA expression) from high-tumor-content regions (for DNA). Bulk scraping without microdissection would mix lymph node parenchyma with tumor, diluting the tumor-specific transcriptomic signal. This is the RNA-level analogue of TCA for DNA: selecting the right region is prerequisite to meaningful data.

• Practical trade-offs documented. Kit A: lower input (1 ng), faster workflow (∼6h), higher per-sample sequencing cost. Kit B: higher input (20 ng), longer workflow (∼10h, including overnight rRNA depletion), lower per-sample sequencing cost for equivalent gene detection breadth. The choice depends on sample availability vs. budget.

Concepts Introduced or Used

FFPE RNA-seq, library preparation, DV200, macrodissection, template switching, rRNA depletion, stranded RNA-seq, gene expression quantification, input RNA quantity, pathologist-assisted ROI selection

Entities Referenced

• TaKaRa SMARTer Stranded Total RNA-Seq Kit v2
• Illumina Stranded Total RNA Prep Ligation with Ribo-Zero Plus
• NovaSeq 6000
• Melanoma lymph node metastases
• Nivolumab (anti-PD1 immunotherapy)

Limitations (as stated by authors)

• Small sample size (n=6 FFPE specimens, single tumor type)
• Single-institution study; results may vary with different FFPE processing protocols
• DV200 values were relatively high for FFPE specimens; results may not generalize to more degraded samples
• Only two kits compared; other commercial kits (e.g., NEB, Qiagen) not evaluated

Relevance to Clonal Evolution

This paper closes the FFPE series by addressing the most practical question: given that FFPE introduces artifacts (Greytak, Do, Steiert) and that analyte quality follows DNA > RNA >> protein (Zhu), how do we actually extract usable transcriptomic data from the limited FFPE material available in clinical archives? The answer is operational: (1) pathologist-guided macrodissection to enrich the ROI, (2) selection of a library prep kit matched to the available RNA quantity, and (3) acceptance of a cost-quality trade-off (deeper sequencing for lower-input kits).

For clonal evolution, transcriptomic ITH — the non-genetic dimension of intratumor-heterogeneity — requires RNA-seq data. Most archival specimens with long-term clinical follow-up are FFPE, not fresh-frozen. This paper demonstrates that FFPE-derived transcriptomic data can be of sufficient quality for gene expression quantification, provided that the library preparation method is matched to the input constraints. The practical implication: retrospective studies of transcriptional ITH from FFPE archives are feasible but require careful documentation of the library preparation method, RNA input quantity, and macrodissection procedure — extending the ERROR-FFPE-DNA checklist (steiert2023-ffpe-dna-ngs-paradigms) to the RNA domain.