Does laparoscopic vs. conventional surgery increase exfoliated cancer cells in the peritoneal cavity during resection of colorectal cancer?

PURPOSE: Traumatic manipulation of cancer specimens during laparoscopic colectomy may increase exfoliation of malignant cells into the peritoneal cavity, causing an early occurrence of peritoneal carcinomatosis or port-sites recurrence. Because of this concern, the routine use of intraperitoneal chemotherapy after laparoscopic colectomy for cancer was suggested recently. We assessed if laparoscopic vs. conventional surgery increases exfoliated malignant cells in the peritoneal cavity during resection of colorectal cancer. METHODS: In a prospective, randomized fashion, 38 colorectal cancer patients undergoing an elective, curative operation were assigned to either a conventional or laparoscopic procedure between June 1996 and May 1997. In either group (n = 19), after the abdominal cavity was entered, saline was instilled into the peritoneal cavity, and the fluid was collected (Specimen 1). During surgery, all irrigating fluids were collected (Specimen 2). Both specimens were assessed for malignancy using four techniques: filtration process (ThinPrep), smear, cell block, and immunochemistry using Ber-EP4. The change in the amount of tumor cells in both specimens was compared between surgical groups. A pilot study was performed to validate the proposed cytologic method. RESULTS: In the pilot study of 20 consecutive patients with colorectal cancer, postresectional peritoneal cytology was positive in six patients, including two Stage II (T3, N0, M0) patients. The pilot study also validated that our semiquantitative scoring system can be reliably used to assess the amount of free peritoneal cancer cells. In the main study, 16 right colectomies, 3 extended right colectomies, 17 proctosigmoidectomies, and 1 left colectomy were performed. The T and N stages were T1 (n = 13, T2 (n = 5), T3 (n = 8), T4 (n = 11); N0 (n =22), N1 (n = 8), N2 (n = 7). Malignant cells were not detected in any Specimens 1 or, more importantly, in Specimens 2 in either surgical group. CONCLUSION: When performed according to strict oncologic surgical principles, laparoscopic techniques in curative colorectal cancer surgery did not have an increased risk of intraperitoneal cancer cell spillage, compared with conventional techniques. We hope that these results can decrease some of the concerns about tumor cell spillage and seeding during laparoscopy.

Endocervical adenocarcinoma in situ: an analysis of cellular features.

Cytologists increasingly encounter atypical endocervical cells, because of the increasing incidence of endocervical adenocarcinoma and the use of improved endocervical sampling devices. These atypical endocervical cells can cause diagnostic problems, especially in recognizing adenocarcinoma in situ (AIS) and distinguishing it from a variety of nonneoplastic changes. We analyzed 33 cervical smears from 22 patients with confirmed AIS and compared these to 19 cervical smears from 17 patients having atypical endocervical cells of undetermined significance and negative follow-up, including at least one tissue biopsy per case, to further investigate the cytologic features of AIS. The AIS smears typically had crowded three-dimensional cellular aggregates, with markedly hyperchromatic nuclei having altered polarity. Frequently, a minor component of AIS formed strips of distinctly columnar cells or sheets. Individual AIS cells occurred in 22 (67%) smears, but these were usually inconspicuous. The AIS smears also had increased nuclear to cytoplasmic ratios (100%), enlarged nuclei (94%), feathering (88%), rosettes (85%), nucleoli (76%), apoptosis (73%), mitoses (64%), multiple nucleoli (18%), and ciliated atypical cells (3%). Cytologic features occurring significantly (P < or = 0.001) more often in AIS cases were a predominance of three-dimensional crowded aggregates (79% vs. 32%), altered nuclear polarity in most groups (88% vs. 16%), marked hyperchromasia (91% vs. 16%), apoptosis (73% vs. 26%), an increased nuclear to cytoplasmic ratio (100% vs. 63%), feathering (88% vs. 26%), and individual atypical cells (67% vs. 16%). In summary, we identified a number of architectural and cellular features that occurred significantly more often in AIS cases than in cases having atypical endocervical cells of undetermined significance and negative follow-up.

Nongynecologic cytology utilizing the ThinPrep Processor.

OBJECTIVE: To assess the utility of the ThinPrep Processor (TP) for nongynecologic cytology. STUDY DESIGN: We reviewed the number of unsatisfactory specimens from the esophagus, common bile duct, hepatic duct, pancreas, gastric, bronchial wang, vertebra, submandibular area and neck over a one-year period, before and after TP implementation. For a one-year period after TP implementation, the cytologic diagnoses of selected TP specimens with corresponding surgical tissue diagnoses were compared, and the TP slides were reviewed in discrepant cases. RESULTS: The number of unsatisfactory specimens was reduced from 17% to 1% after TP implementations. The cytologic diagnoses of 145 TP specimens were in agreement with surgical tissue diagnoses. However, in 43 cases the cytology and tissue diagnoses were discordant. On review of 26 of the discrepant cases, the majority of TP slides were cellular, with good nuclear and cytoplasmic detail. Discrepancies resulted from sampling errors in 19 cases and TP interpretation errors in 7 cases. CONCLUSION: In our laboratory, TP is a reliable processor for nongynecologic specimens.

ThinPrep versus conventional smear cytologic preparations in the analysis of thyroid fine-needle aspiration specimens.

Paired fine-needle aspiration specimens were analyzed from 41 surgically resected thyroid nodules, to compare diagnostic accuracy, amount (absent, mild, moderate, or marked) and pattern (diffuse, droplets, or both) of colloid, nuclear detail (poor, satisfactory, or excellent) and cytoplasmic detail (intact or disrupted) in ThinPrep (TP) (Cytyc, Marlborough, MA) versus conventional smear (CS) cytologic preparations. The 41 surgical specimens included 25 colloid nodules, 6 papillary carcinomas, 4 follicular adenomas, 2 minimally invasive (encapsulated) follicular carcinomas, 3 Hashimoto's thyroiditis, and 1 Grave's disease. Both techniques identified seven of the eight carcinomas with the minimally invasive follicular carcinomas categorized as hypercellular follicular nodule, possibly malignant (HCFN). One papillary carcinoma was classified as a HCFN by both TP and CS techniques. The four follicular adenomas were classified as HCFN based on the TP slides. One oxyphilic follicular adenoma, associated with focal lymphocytic thyroiditis, was misinterpreted as Hashimoto's thyroiditis on a conventional smear. Three colloid nodules were interpreted as HCFN based on the TP slides. Two of these were similarly classified based on the conventional smear. ThinPrep slides contained less colloid and the colloid occurred as droplets rather than a diffuse pattern. TP slides had better nuclear detail but more often disrupted cytoplasm. In conclusion, the TP process does alter some cellular features; however, we experienced similar diagnostic accuracy with the TP and conventional smear preparations.