- Email: [email protected]
Evaluation of cervical biopsy samples by proton magnetic resonance spectroscopy
174
SOCIETY OF GYNECOLOGISTS
LN is the natural logarithm, i is the y-axis intercept and is a function of tumor burden at initial presentation, and s, the slope of the curve, measures the rapidity of decline of CA-125 measurements following surgery and during chemotherapy. Sixty-four patients with CA-125 positive advanced stage epithelial ovarian cancers have been followed for a mean of 480 days. Twenty patients have expired: 17 from their cancer, 3 without evidence of disease. Second look coeliotomy (SLC) was performed on 39 patients. The difference between s for the I9 with negative SLC was significant over those with positive SLC (P < 0.005). The mean serum CA-125 level was ~30 U/ml prior to SLC for both groups. The difference in s but not i is significant (P < 0.01) for the 6 patients with negative SLC who recur vs those with negative SLC who do not. For the 64 patients as a whole, life table analysis gives a 42% 42-month survival for the group with s < 0.045, but an 80% 42-month survival for the group with s > 0.045. We conclude that the value of s from the equation above accurately predicts response to therapy as evaluated clinically, by second look laparotomy, and by subsequent survival. s can be determined within 2 months of surgery and thus provides valuable prognostic information long before it would otherwise be available. 26. The Prognostic implications of Low Serum CA-125 Levels Prior to Second-Look Operation for Stage III and IV Epithelial Ovarian Cancer. H. GALLION,H. AVERETTE,*E. Pmmmx,t L. COPELAND,+
J. CAIN,# N. HUSSEINZADEH,~ W. NAHHAS,~~ S. PLJRSELL,’R. HIGGINS,J. VAN NAGELL,P. DEPRIEST, A. MAGGARD, AND R. KRYSCIO, UKMC, *UM, -FUAB, +OSU, OUWA, IUC. IIWSU, ‘UL From July 1986 to June 1990, 75 patients (pts) with Stage 111or IV epithelial ovarian cancer underwent second-look laparotomy. All pts had elevated (>35 u/ml) serum CA-125 prior to therapy. At the completion of platinum-based chemotherapy, all pts were clinically and radiologically free of disease and had serum CA-125 levels ~35 u/ml. The mean age of the pts was 54 years (range 22-79) and the mean gravidity was 2 (range O-6). Sixty-seven pts had Stage 111disease and 8 pts had Stage IV. Cell types were serous. 57; endometrioid, 9; mucinous, 5; clear cell. 3; and malignant Brenner tumor, I. Ten tumors were grade I,21 grade 2, and 44 grade 3. Second look findings according to preoperative CA-125: CA 125
Negative
Microscopic
Gross
o-7 8-35
23 (68%) 10 (24%)
1 (3%)
IO (29%)
9 (22%)
22 (54%)
Sixty-eight percent of pts with CA-125 s7 u/ml had no residual disease. In contrast, only 24% of pts with CA-125 between 8 and 35 u/ml had no disease at second-look laparotomy (P < 0.001). Ninety percent of pts with microscopic residual disease and 69% of pts with gross residual disease had CA-125 levels between 8 and 35 u/ml. The negative predictive value for a CA-125 ~7 u/ml was 0.68 vs 0.44 for CA-125 levels ~35 u/ml. These data indicate that a serum CA-125 level of <7 u/ml is a more accurate predictor of negative second-look findings than serum levels of ~35 u/ml in pts with Stage III and IV epithelial ovarian cancer. 27. Patterns of Failure in Patients with Malignant Ovarian Germ Cell Tumors. M. MESSING,D. GERSHENSON, M. MORRIS,T. BURKE, J. KAVANAGH, AND J. WHARTON, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030.
Between 1970 and 1990, 160 patients with malignant nondysgerminomatous ovarian germ cell tumors have been treated at our institution. Forty-two of these patients have failed primary therapy and constitute the basis for this study. The median age was 17 years (range, 6-33 years). FIG0 stage distribution included 17 stage I, 5 stage II, 17 stage III, and 3 stage IV. Histologic type was as follows: 13 immature teratomas, 8 endomermal sinus tumors, and 21 mixed germ cell tumors. Primary therapy consisted of surgery alone in 14 patients, surgery plus chemotherapy in 23 patients, surgery plus radiotherapy in 2 patients,
ONCOLOGISTS-ABSTRACTS and all three modalities in 3 patients. Initial chemotherapy included a combination of vincristine, actinomycin-D, and cyclophosphamide (VAC) in 17 patients, cisplatin combinations in 5 patients, and other drugs in 4 patients. Median progression-free survival from initial diagnosis was 6.8 months (range, 0.9-24 months). Of 34 patients who received chemotherapy as part of salvage, 5/11 (45%) who received VAC are NED; 6/ll (55%) of patients who received cisplatin combinations are NED. VAC failures were salvaged with cisplatin combinations in 3/7 (43%) cases. Of 5 cisplatin combination failures, two (40%) were salvaged. Twelve of the 42 patients (29%) are currently alive NED. Reasons for primary treatment failure included surgery alone in I4 patients (7 because of misdiagnosis), radiotherapy in 5 patients, and toxicity in I patient. Of the 22 patients who failed chemotherapy, I2 did so because of suboptimal regimen, 4 because of possible dose-intensity problems, and 6 for unexplained reasons. Patients with ovarian germ cell tumors have an excellent probability of cure with meticulous, aggressive therapy, but successful salvage may be difficult in patients who fail primary treatment. 28. Evaluation of Cervical Biopsy Samples by Proton Magnetic Resonance Spectroscopy. M. PR~FONTAINE,* I. C. P. SMITH,? T. KRoi=r,t
J. K. SAuNDERs.tANDA. KusEL,t *Ottawa General Hospital, University of Ottawa, Ottawa. Canada; and tlnstitute for Biological Sciences, National Research Council, Ottawa, Canada. The pathological evaluation of premalignant cervical biopsies involves some subjectivity. The risk of progression for dysplasia is difficult to evaluate. Recently Mountford and co-workers reported promising highresolution ‘H MRS data from punch biopsies of the uterine cervix (Magn. Reson. Medi. 13, 324, 1990). We have undertaken to confirm and extend these accomplishments; our hypothesis is that MRS can provide more objective and reproducible criteria for diagnosis and grading of cervical biopsies. Colposcopically directed biopsies from 50 patients with malignant and premalignant disease have been investigated by MRS. The malignant specimens yielded intense narrow-lined lipid spectra with numerous contributions from metabolites, long Tz relaxation times from the composite resonance at 1.3 ppm, and several methyl to methine couplings (1.3 to 4.1 ppm) in their 2D COSY spectra. Specimens which were HPV-infected with and without dysplasia gave weak and broad-lined ID spectra, 2D COSY spectra with crosspeaks for only a few of the metabolites observed for the malignant biopsies, but a long TL for the composite resonance at 1.3 ppm. Healthy cervix or specimens containing only inflammatory cells are instantly recognizable by the absence of resonances other than that of water. Premalignant biopsy specimens (CIN l-3) generate a number of resonances which correlate with mild, moderate, or severe dysplasia. It appears likely that a combination of the above MRS parameters could characterize the abnormality of cervical biopsies precisely and in particular distinguish which cervical atypia will regress from those which are committed to becoming malignant. 29. Possible Clinical Usefulness of DNA Ploidy Measurement by Flow Cytometry in Stage IB-IIA Squamous Cell Cervical Cancer (CC). G. ZANETTA,* G. KEENEY,? S. CHA,$ J. K.&TzMANN,t H. S. WIEAND,~ W. KINNEY,* AND K. PODRATZ,* Mayo Clinic, *Depart-
ment of Gynecology, tDepartment of Path. & Lab. Medicine, $Cancer Center Statistics, Rochester, Minnesota 55905. The utility of DNA ploidy measured by flow cytometry in CC is unclear. This study evaluated its prognostic significance in a large, homogeneous population: 394 patients (pts) with stage IB-IIA squamous cell CC were analyzed. All pts underwent a Wertheim hysterectomy plus lymphadenectomy at the Mayo Clinic from 1956 to 1985. Paraffinembedded samples containing 20% or more of tumor cells were obtained in 367 pts. Samples were dewaxed, rehydrated, stained with propidium iodide, and analyzed. We observed 137 (39%) diploid, 207 (61%) nondiploid (26 tetraploid, 157aneuploid, 24 polyploid), and 13 unclassifiable
SOCIETY OF GYNECOLOGISTS
LN is the natural logarithm, i is the y-axis intercept and is a function of tumor burden at initial presentation, and s, the slope of the curve, measures the rapidity of decline of CA-125 measurements following surgery and during chemotherapy. Sixty-four patients with CA-125 positive advanced stage epithelial ovarian cancers have been followed for a mean of 480 days. Twenty patients have expired: 17 from their cancer, 3 without evidence of disease. Second look coeliotomy (SLC) was performed on 39 patients. The difference between s for the I9 with negative SLC was significant over those with positive SLC (P < 0.005). The mean serum CA-125 level was ~30 U/ml prior to SLC for both groups. The difference in s but not i is significant (P < 0.01) for the 6 patients with negative SLC who recur vs those with negative SLC who do not. For the 64 patients as a whole, life table analysis gives a 42% 42-month survival for the group with s < 0.045, but an 80% 42-month survival for the group with s > 0.045. We conclude that the value of s from the equation above accurately predicts response to therapy as evaluated clinically, by second look laparotomy, and by subsequent survival. s can be determined within 2 months of surgery and thus provides valuable prognostic information long before it would otherwise be available. 26. The Prognostic implications of Low Serum CA-125 Levels Prior to Second-Look Operation for Stage III and IV Epithelial Ovarian Cancer. H. GALLION,H. AVERETTE,*E. Pmmmx,t L. COPELAND,+
J. CAIN,# N. HUSSEINZADEH,~ W. NAHHAS,~~ S. PLJRSELL,’R. HIGGINS,J. VAN NAGELL,P. DEPRIEST, A. MAGGARD, AND R. KRYSCIO, UKMC, *UM, -FUAB, +OSU, OUWA, IUC. IIWSU, ‘UL From July 1986 to June 1990, 75 patients (pts) with Stage 111or IV epithelial ovarian cancer underwent second-look laparotomy. All pts had elevated (>35 u/ml) serum CA-125 prior to therapy. At the completion of platinum-based chemotherapy, all pts were clinically and radiologically free of disease and had serum CA-125 levels ~35 u/ml. The mean age of the pts was 54 years (range 22-79) and the mean gravidity was 2 (range O-6). Sixty-seven pts had Stage 111disease and 8 pts had Stage IV. Cell types were serous. 57; endometrioid, 9; mucinous, 5; clear cell. 3; and malignant Brenner tumor, I. Ten tumors were grade I,21 grade 2, and 44 grade 3. Second look findings according to preoperative CA-125: CA 125
Negative
Microscopic
Gross
o-7 8-35
23 (68%) 10 (24%)
1 (3%)
IO (29%)
9 (22%)
22 (54%)
Sixty-eight percent of pts with CA-125 s7 u/ml had no residual disease. In contrast, only 24% of pts with CA-125 between 8 and 35 u/ml had no disease at second-look laparotomy (P < 0.001). Ninety percent of pts with microscopic residual disease and 69% of pts with gross residual disease had CA-125 levels between 8 and 35 u/ml. The negative predictive value for a CA-125 ~7 u/ml was 0.68 vs 0.44 for CA-125 levels ~35 u/ml. These data indicate that a serum CA-125 level of <7 u/ml is a more accurate predictor of negative second-look findings than serum levels of ~35 u/ml in pts with Stage III and IV epithelial ovarian cancer. 27. Patterns of Failure in Patients with Malignant Ovarian Germ Cell Tumors. M. MESSING,D. GERSHENSON, M. MORRIS,T. BURKE, J. KAVANAGH, AND J. WHARTON, University of Texas M.D. Anderson Cancer Center, Houston, Texas 77030.
Between 1970 and 1990, 160 patients with malignant nondysgerminomatous ovarian germ cell tumors have been treated at our institution. Forty-two of these patients have failed primary therapy and constitute the basis for this study. The median age was 17 years (range, 6-33 years). FIG0 stage distribution included 17 stage I, 5 stage II, 17 stage III, and 3 stage IV. Histologic type was as follows: 13 immature teratomas, 8 endomermal sinus tumors, and 21 mixed germ cell tumors. Primary therapy consisted of surgery alone in 14 patients, surgery plus chemotherapy in 23 patients, surgery plus radiotherapy in 2 patients,
ONCOLOGISTS-ABSTRACTS and all three modalities in 3 patients. Initial chemotherapy included a combination of vincristine, actinomycin-D, and cyclophosphamide (VAC) in 17 patients, cisplatin combinations in 5 patients, and other drugs in 4 patients. Median progression-free survival from initial diagnosis was 6.8 months (range, 0.9-24 months). Of 34 patients who received chemotherapy as part of salvage, 5/11 (45%) who received VAC are NED; 6/ll (55%) of patients who received cisplatin combinations are NED. VAC failures were salvaged with cisplatin combinations in 3/7 (43%) cases. Of 5 cisplatin combination failures, two (40%) were salvaged. Twelve of the 42 patients (29%) are currently alive NED. Reasons for primary treatment failure included surgery alone in I4 patients (7 because of misdiagnosis), radiotherapy in 5 patients, and toxicity in I patient. Of the 22 patients who failed chemotherapy, I2 did so because of suboptimal regimen, 4 because of possible dose-intensity problems, and 6 for unexplained reasons. Patients with ovarian germ cell tumors have an excellent probability of cure with meticulous, aggressive therapy, but successful salvage may be difficult in patients who fail primary treatment. 28. Evaluation of Cervical Biopsy Samples by Proton Magnetic Resonance Spectroscopy. M. PR~FONTAINE,* I. C. P. SMITH,? T. KRoi=r,t
J. K. SAuNDERs.tANDA. KusEL,t *Ottawa General Hospital, University of Ottawa, Ottawa. Canada; and tlnstitute for Biological Sciences, National Research Council, Ottawa, Canada. The pathological evaluation of premalignant cervical biopsies involves some subjectivity. The risk of progression for dysplasia is difficult to evaluate. Recently Mountford and co-workers reported promising highresolution ‘H MRS data from punch biopsies of the uterine cervix (Magn. Reson. Medi. 13, 324, 1990). We have undertaken to confirm and extend these accomplishments; our hypothesis is that MRS can provide more objective and reproducible criteria for diagnosis and grading of cervical biopsies. Colposcopically directed biopsies from 50 patients with malignant and premalignant disease have been investigated by MRS. The malignant specimens yielded intense narrow-lined lipid spectra with numerous contributions from metabolites, long Tz relaxation times from the composite resonance at 1.3 ppm, and several methyl to methine couplings (1.3 to 4.1 ppm) in their 2D COSY spectra. Specimens which were HPV-infected with and without dysplasia gave weak and broad-lined ID spectra, 2D COSY spectra with crosspeaks for only a few of the metabolites observed for the malignant biopsies, but a long TL for the composite resonance at 1.3 ppm. Healthy cervix or specimens containing only inflammatory cells are instantly recognizable by the absence of resonances other than that of water. Premalignant biopsy specimens (CIN l-3) generate a number of resonances which correlate with mild, moderate, or severe dysplasia. It appears likely that a combination of the above MRS parameters could characterize the abnormality of cervical biopsies precisely and in particular distinguish which cervical atypia will regress from those which are committed to becoming malignant. 29. Possible Clinical Usefulness of DNA Ploidy Measurement by Flow Cytometry in Stage IB-IIA Squamous Cell Cervical Cancer (CC). G. ZANETTA,* G. KEENEY,? S. CHA,$ J. K.&TzMANN,t H. S. WIEAND,~ W. KINNEY,* AND K. PODRATZ,* Mayo Clinic, *Depart-
ment of Gynecology, tDepartment of Path. & Lab. Medicine, $Cancer Center Statistics, Rochester, Minnesota 55905. The utility of DNA ploidy measured by flow cytometry in CC is unclear. This study evaluated its prognostic significance in a large, homogeneous population: 394 patients (pts) with stage IB-IIA squamous cell CC were analyzed. All pts underwent a Wertheim hysterectomy plus lymphadenectomy at the Mayo Clinic from 1956 to 1985. Paraffinembedded samples containing 20% or more of tumor cells were obtained in 367 pts. Samples were dewaxed, rehydrated, stained with propidium iodide, and analyzed. We observed 137 (39%) diploid, 207 (61%) nondiploid (26 tetraploid, 157aneuploid, 24 polyploid), and 13 unclassifiable