Increase in the incidence of peritoneal collagen balls over a 10-year period.

OBJECTIVE: To describe the detection of collagen balls in peritoneal washings over a 10-year period, revealing an unexpected and unexplained higher incidence than in the past. STUDY DESIGN: Reports of routinely processed, Papanicolaou-stained smears and cytospins and hematoxylineosin-stained cell blocks from peritoneal washes and ascitic fluids seen over an 8-year period (1995-2002) were reviewed, and the percentage of specimens in which collagen balls were noted was determined. To rule out a learning curve phenomenon, the first 100 consecutive peritoneal washings and ascitic specimens from years 1993-2001 plus 2002 were rescreened, and the percentage of specimens containing collagen balls was determined. RESULTS: The percentages of cases with collagen balls increased steadily beginning with 8 of 467 cases (3.2%) in 1995 to 185 of 650 (28.5%) in 2002, as reported in the case records. Rescreened cases also showed a similar increase, 4.0% in 1993 to 29% in 2002. No procedural modifications were made in specimen processing during this interval, nor are we aware of any procedural change in obtaining cytologic samples. CONCLUSION: An unexplained 7-fold increase in the percentage of peritoneal samples with collagen balls occurred over a 10-year period. This increase cannot be attributed to changes in specimen handling or to a learning curve phenomenon. This finding is of unknown significance and may warrant further investigation.