Distinct Gene Expression Profile in Patients With Poor Postoperative Outcomes After Rotator Cuff Repair: A Case-Control Study

Abstract

Background:

Impaired healing after rotator cuff repair is a major concern, with retear rates as high as 94%. A method to predict whether patients are likely to experience poor surgical outcomes would change clinical practice. While various patient factors, such as age and tear size, have been linked to poor functional outcomes, it is currently very challenging to predict outcomes before surgery.

Purpose:

To evaluate gene expression differences in tissue collected during surgery between patients who ultimately went on to have good outcomes and those who experienced a retear, in an effort to determine if surgical outcomes can be predicted.

Study Design:

Case-control study; Level of evidence, 3.

Methods:

Rotator cuff tissue was collected at the time of surgery from 140 patients. Patients were tracked for a minimum of 6 months to identify those with good or poor outcomes, using clinical functional scores and follow-up magnetic resonance imaging to confirm failure to heal or retear. Gene expression differences between 8 patients with poor outcomes and 28 patients with good outcomes were assessed using a multiplex gene expression analysis via NanoString and a custom-curated panel of 145 genes related to various stages of rotator cuff healing.

Results:

Although significant differences in the expression of individual genes were not observed, gene set enrichment analysis highlighted major differences in gene sets. Patients who had poor healing outcomes showed greater expression of gene sets related to extracellular matrix production (P < .0001) and cellular biosynthetic pathways (P < .001), while patients who had good healing outcomes showed greater expression of genes associated with the proinflammatory (M1) macrophage phenotype (P < .05).

Conclusion:

These results suggest that a more proinflammatory, fibrotic environment before repair may play a role in poor healing outcome. With validation in a larger cohort, these results may ultimately lead to diagnostic methods to preoperatively predict those at risk for poor surgical outcomes.