Interobserver variability in histologic recognition, interpretation of KIT immunostaining, and determining MIB-1 labeling indices in gastrointestinal stromal tumors and other spindle cell tumors of the gastrointestinal tract.

The authors studied the concordance among seven pathologists for the histologic diagnosis, interpretation of KIT immunostaining, and determining MIB-1 labeling indices (LI) in 80 adult patients with primary spindle cell tumors, mainly of the gastrointestinal tract, mesentery, retroperitoneum, and pelvis, based on the review of tissue sections using an immunohistochemical panel of antibodies for KIT/CD117, CD34, desmin, smooth muscle actin (SMA), and S-100 protein. Tumors included 30 gastrointestinal stromal tumors (GISTs), 10 leiomyomas, 10 leiomyosarcomas, 10 schwannomas, 10 solitary fibrous tumors, and 10 desmoid-type fibromatoses. The overall concordance with the original diagnosis of each histologic type was 97.9%, the kappa value ranging from 0.95 to 1.00 (mean 0.97), indicating a perfect agreement. The overall interlaboratory concordance with the original interpretation of KIT immunostaining was 91.3%, the kappa value ranging from 0.77 to 0.90 (mean 0.86). The overall interlaboratory concordance with the original interpretation of KIT immunostaining was 91.9%, the kappa value ranging from 0.72 to 0.93 (mean 0.85). The overall concordance for determining MIB-1 LI was 90% with the original evaluation, and the overall kappa value ranged from 0.62 to 0.86 (mean 0.77). These results indicate that it is possible to reliably diagnose GIST and other spindle cell tumors of the gastrointestinal tract with the use of an immunohistochemical panel of antibodies for KIT, CD34, desmin, SMA, and S-100 protein. Although there is clearly unavoidable inter-observer and interlaboratory variability in the interpretation of KIT immunostained sections and interobserver variability in the determination of MIB-1 LI, the concordance between observes is very acceptable, and in most instances such variability can be eliminated by careful reviewing of the hematoxylin and eosin and immunostained sections.