GYNECOLOGIC
ONCOLOGY
38, 486-493 (1990)
Neoadjuvant Chemotherapy in Locally Advanced Carcinoma of the Cervix Uteri J. SARDI,M.D., C. SANANES,M.D., A. GIAROLI,M.D., G. MAYA, M.D., ANDG. DI PAOLA,M.D. Gynecology
Oncology
Unit,
Cdrdoba
2351,
C.P. 1120, Buenos Aires, Argentina
Received December 11, 1989
verse factors are responsible for that therapeutic failure, One hundred fifty-one patients (107 stage IIB and 44 stage IIIB) who had completed 2 years of follow-up after treatment but the most important perhaps is tumor size. The larger with neoadjuvant chemotherapy (“Quick” VBP scheme), prior to the tumor mass, the poorer the radiotherapy response standard therapy, are presented. Four groups of patients with and hence survival. Chemotherapy has generally been reserved for the different prognosis and operability status, according to tumor response, were identified. In the first group (excellent responders) treatment of metastases. But the possibility of diminish96% of the patients had no evidence of disease (NED), whereas ing the volume of the primary tumor with chemotherapy, in the fourth group (no response), only 33% had NED after 2 as a first approach, appears to be the most significant years of follow-up. In the first group, no pelvic failure was ob- achievement in recent years. served. On the other hand, of 21 patients in group 4, 11 (52%) It has been demonstrated that the use of preoperative had pelvic recurrences. The response to neoadjuvant therapy is adjuvant chemotherapy can increase disease-free surstrongly associated with the initial tumor volume ecographically vival in several kinds of nongynecologic cancer, such as evaluated. The critical pretreatment volume for response was 84 pediatric solid tumors [2,3], osteogenic sarcoma [4,5], cm3(4.85 cm in diameter). Surgery and radiotherapy were equally effective in those patients with initial tumor volume of 84 cm3 head and neck cancer [6,7], and breast cancer [8]. The theoretical advantages of such an approach inor less. But surgery appears to have a better outcome in those clude the following: (1) Drug delivery to the tumor is patients with tumors larger than the critical pretreatment tumor improved because vascularity has not been disturbed by volume. Radiotherapy in this controversial group was ineffective in avoiding pelvic recurrences, especially when the tumor volume surgery or radiotherapy, and drug dosage is not comwas not sufficiently reduced. The advantage of neoadjuvant che- promised in patients who have not undergone prior irmotherapy is evident in stage IIB, with 79% @S/107) of patients radiaton and myelosuppression. (2) The potential for rashowing no evidence of disease as compared with 47% (24/51) dioresponsiveness is improved by size reduction and in a nonrandomized control group (P < 0.01). Statistically sig- decrease in hypoxic cell traction. (3) An inoperable tunificant differences were also obtained in stage IIIB (50% vs 26% mor is rendered resectable. (4) Early treatment is proin the control group). 0 1990 Academic FVW, Inc. vided for occult micrometastatic disease. (5) Patient population that might benefit from similar chemotherapy administered as additional adjuvant treatment following INTRODUCTION surgery and/or radiotherapy is identified. In addition to theoretical advantages there are also It is generally accepted that radiotherapy and surgery are both effective techniques in the management of car- possible disadvantages to such an approach and these include (1) delay of potentially curable radiotherapy, (2) cinoma of the cervix uteri of small volume. Therapeutic results in patients with locally or region- prolongation of treatment, (3) increased toxicity of treatment, (4) possibility of tumor progression, and (5) the ally advanced cervical carcinoma (FIG0 stages II, III, potential for selecting out drug-resistant clones by priand IV) are frequently poor [ 11.Radiotherapy has always mary chemotherapy. been the elective treatment or those stages. Various adThis work has its starting point in a previous paper [9] that reported the introduction of chemotherapy as a Presented at the annual meeting of the Society of Gynecologic Onfirst-line treatment in cervical carcinoma. At the First cologists, San Francisco, CA, Februrary 4-7, 1990. 486 0090-8258190$ I so Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
NEOADJUVANT
CHEMOTHERAPY
Chair of Gynecology, University of Buenos Aires, a new treatment, which includes neoadjuvant chemotherapy as first-line treatment in regionally advanced carcinoma of the uterine cervix (stages II, III, and IV), has been used for the last 5 years. The chemotherapeutic scheme used was the modified vincristine-bleomvcin-cis-Dlatinum (VBP) scheme. The observed tumor responses’have already been reported, corresponding to phase I of this new management [ 101. Today, there is much evidence for the chemosensitivity of carcinoma of the cervix uteri when drugs are administered as neoadjuvant therapy, especially in cis-platinum-based and bleomycin-based schemes [ 1l-l 51. As there is an international consensus about tumor regression of cervical squamous carcinoma, an analysis of survival of patients treated with this therapeutic approach is urgently required. MATERIAL
IN LOCALLY
ADVANCED
CANCER
487
TABLE 1 “Quick” Scheme
,, cic-P,atinum 50 mg/m’ *, V&ristine I mg/m’
(push 15’) (push) 3. Bleomycin (continuous infusion 25 mg/m’ over 6 hr) Three courses repeated at IO-day intervals Day 1 by 1 Days 1,2,3
Toxicity of the drugs was evaluated as follows: 1. c&-Platinum: hemogram including platelet count, hepatoenzymogram, and creatinine clearance after each course of chemotherapy; audiometry at the beginning and end of treatments. 2. Bleomycin: functional respiratory study, chest roentgenogram, and blood gas at the end of each course. 3. Vincristine: complete neurologic study after each course.
To evaluate tumor response to the VBP scheme, careThis is a phase II trial started in 1984 in the Gynecologic Oncology Unit of the First Gynecological Chair ful objective parameters were considered. These paramof Buenos Aires University. Up to now, 328 patients eters were related to cervical mass, vaginal extension, with squamous cervical carcinoma have been included and parametrial compromise. To determine the cervical in this protocol. One hundred fifty-one patients have and the vaginal response, colposcopic measurements completed 2 years of follow-up and their results reported were done. in this paper. Of these 151, 107 patients are FIG0 stage With ecography the volume of the tumor was measured IIB and 44 are IIIB. The mean age was 44.3 (range 21- by multiplying the three spatial axes. The parametrial 80) years. response was assessed mainly by clinical examination, A nonrandomized control group of 101 patients treated taking into account its extension, elasticity, and cononly with radiotherapy (Fletcher Technique) was used sistency. Three experienced members of the department for comparison; 51 were stage IIB patients and 50 stage examined simultaneously the patients in all circumstanIIIB (mean age 42.4, range 23-75). This control group ces, to avoid the possibility of subjective error. The response was classified as complete tumor regreswas staged and treated in the same institution and by the same staff, immediately prior to the neoadjuvant trial. sion (CTR), partial tumor regression (PTR), stable disease (SD), or progressive disease (PD), in the central area and in the parametria. Parametrial response is of METHODS utmost importance because of its implication in manThe trial sequence of procedures was as follows: (1) agement and operability. Complete tumor regression in staging and tumor volume measurement, (2) neoadjuvant the central area was considered when colposcopy rechemotherapy and toxicity evaluation, (3) evaluation of vealed an almost reepithelialized cervix, with a marked tumor response, (4) conventional treatment (surgery plus volume reduction proved by ecography and clinical exradiotherapy or radiotherapy alone), (5) 2 years of follow- amination. In the parametrium this response criterion up and determination of sites of recurrences. was applied to those cases in which the clinical disapAll patients were classified according to the clinical pearance of infiltration was verified. Moderate tumor staging system proposed by the International Federation regression was diagnosed when reepithelialization was of Gynecology and Obstetrics with clinical examination observed at least in 50% of the ectocervical surface inunder anesthesia [l]. To assess tumor bulk, cervical ul- itially involved, with a moderate volume reduction trasound scanning (ecography) was performed. These proved by ecography. The parametrial extent should be studies were repeated when the chemotherapy was ended reduced to 50% or more. A change from irregular to to reevaluate each case and decide the subsequent smooth and indurated parametria was also considered a treatment. partial response. In the stable cases, there were no All cases received three courses of polychemotherapy changes or very subtle changes in the cervix, vagina, of the Quick VBP scheme (Table 1). and parametrium. In patients with progressive disease,
488
SARDI ET AL.
TABLE 2 Management and Operability Stage IIB No. Group 1
S+RT
28
Stage IIIB RT
No.
5
5
23
S+RT 2
85% Group 2 Group 3 Group 4 Total
41 29 9 107
36 1 14 3 76
RT 3
78%
5 9 91 15 18 7 6 12 0 31 44 18 S+RT = 94/151 = 62.2%
0 11 12 26
an increase in tumor bulk or in the vaginal or parametrial lesion was confirmed. In our experience with neoadjuvant chemotherapy, tumor response was always observed to follow a centripetal pattern. This means that the response occurs first in the parametria and vagina and afterward in the cervix. Cervical response was never found without parametrial response. Thus, four groups of tumor response were recognized: Group 1 Group 2
Excellent response Good response
Group 3 Group 4
Regular response No response
CTR in both areas CTR in the parametria, PTR in the central area PTR in both areas SD or PD in both areas
Conventional treatments were applied according to the results obtained after the use of chemotherapy as firstline treatment. When radiotherapy was chosen, irradiation with the Fletcher technique [ 161began 1 week after the last course of chemotherapy. When surgery was decided, a Wertheim-Meigs operation was performed with paraortic lymphadenectomy, 2 weeks after the last course of chemotherapy. After surgery, radiotherapy (5000 rad) to the whole pelvis was administered in all cases. Generally speaking, surgical treatment was considered preferable for the patients in whom parametrial infitration TABLE 3 Response to Neoadjuvant Chemotherapy According to Stages Stage IIB Parametrial response
Central response
Group 1
CTR
;;
Group 2 Group 3 Group 4 Total
CTR PTR SD
PTR SD
Stage IIIB
No. 1;
;6ju
29 9 107
27 8
N; 18 12 44
lj3l 40 47
was no longer clinically demonstrable, while radiotherapy was carried out in those patients who did not exhibit parametrial response or who exhibited partial tumor response. But, during the development of the protocol, it was discovered that many partial responders were also good candidates for surgery with free surgical margins. Wertheim-Meigs operation was performed in 76 stage IIB and 18 stage IIIB patients. After surgery all of them received 5000 rads to the whole pelvis. Radiotherapy according to Fletcher technique was carried out in 31 stage IIB and 26 stage IIIB patients. Thus, surgical treatment after neoadjuvant chemotherapy was performed in 62% (94/151) of all patients. Operability of stage IIB was 71% and of stage IIIB, 44% (Table 2). The number of patients with no evidence of disease 24 months after treatment was analyzed, and the results were compared with those obtained in the control group of patients subjected to the conventional strategy, immediately prior to the beginning of the neoadjuvant chemotherapy trial, following the same staging system, diagnostic, and radiotherapeutic criteria. For statistical analysis, the x2 test was used. A P value less than 0.05 was considered significant. RESULTS
The results are analyzed taking into account tumor response to neoadjuvant treatment, tumor volume, tumor extension. Also, toxicity to the Quick VBP scheme is considered. Results in Relation
to Tumor Response
ReTumor response to neoadjuvant chemotherapy. sponse was analyzed with respect to the four groups previously described. (Table 3). The complete parametrial response is greater in stage IIB (64%) than in stage IIIB (31%) and the percentage of patients with stable disease is higher in stage IIIB as compared with stage IIB. Perhaps the difference in results in stage IIIB patients is due to the initial tumor volume, which was higher in the last group of patients. Operability. Based on these findings, the management of the patients was as follows. Almost all group 1 and 2 patients (excellent and good responders) in stages IIB and IIIB were operated (59/69 stage IIB and 1l/14 stage IIIB). Almost all group 4 patients (nonresponders) were subjected to radiotherapy (6/9 stage IIB and 12/12 stage IIIB). Initially, group 3 patients (partial responders) were treated the same as group 4 patients (nonresponders), but later, during development of the protocol, it was discovered that many partial parametrial responders
NEOADJUVANTCHEMOTHERAPYIN
LOCALLY
TABLE 6 Pathological Findings
TABLE 4 Relationship between Tumor Response and Free Disease Interval after 2 Years of Follow-up Pelvic failures
NED Response Group Group Group Group
1 2 3 4
No. 33 50 47 21
No.
%
No.
%
0 6 12 12 25 11 52 x2 = 25.12
32 96 39 78 29 61 7 33 x2 = 28.44 P < 0.005
489
ADVANCEDCANCER
Distant metastases
Clinical response
No.
Group Group Group Group
%
1 3 5 10 6 12 14 3 x2 = 4.05
1 2 3 4
No.
Cervical lesion CO.5 cm (%o)
Parametrial residual lesion (%)
Lymph node metastases (%)
25 45 21 3
60 11 4 0
4 20 66 100
0 24 62 66
explore the effectiveness of conventional treatments in this trial with neoadjuvant chemotherapy. As in groups ’ No evidence disease; NS, not significant. 1, 2, and 4 the treatments were not harmonically distributed and no significant conclusion could be drawn. were also candidates for surgery with free surgical mar- In group 3 (partial responders) equal numbers of patients gins. These were the cases in which, besides the reduc- were subjected to both treatments; these results are pretion of infiltration, changes from nodular and irregular sented in Table 5. In the partial responders surgery appears to be more to smooth and indurated parametria were observed. effective than radiotherapy in avoiding pelvic failures, In these patients, microscopic residual tumor was not compared with a group of patients with similar mean prefound to be a significant obstacle to the surgical techand post-treatment volumes. nique , and in all cases free surgical margins were Pathological jindings of surgical specimens. Analysis determined. of the pathological findings of the 94 surgical specimens Thus, operability was found to be higher than what was considered of value in assessment of the objectivity the previous response criteria would have indicated (Taof the clinical response criteria. Some data of interest bles 2 ad 3). On the other hand, no stage IIIB patients are presented in Table 6. with stable disease could be excised, although some of A strong correlation between clinical response and them were subjected to staging laparatomy. Disease-free interval (DFI) according to tumor re- pathological findings of surgical specimens was verified. sponse. The DFI of the patients in relation to the re- An increasing residual tumor size in the central and parasponse to neoadjuvant chemotherapy according to the metrial areas according the worse response was observed. Also, an increase in the percentage of lymph groups previously described is presented in Table 4. node involvement after neoadjuvant therapy was verified The outcome after 2 years of follow-up for the four in the partial responders and nonresponders. groups of tumor response demonstrates a strong assocation between response level and percentage of diseaseResult in Relation to Tumor Volume free interval and site of recurrence. It is evident that pelvic failures occur more frequently in patients with Relationship between tumor response and tumor volstable disease than in complete responders. The per- ume. The above-mentioned different responses to neoadcentage of distant metastases remains constant in all juvant treatment between stage IIB and IIIB patients are groups, except the first one. probably due to the higher initial tumor volume in the Disease-free interval according to tumor response and later patients. The Gompertz equation establishes that conventional treatment. It was considered of interest to when tumor volume increases, the tumor response to P
ONCOLOGY
38, 486-493 (1990)
Neoadjuvant Chemotherapy in Locally Advanced Carcinoma of the Cervix Uteri J. SARDI,M.D., C. SANANES,M.D., A. GIAROLI,M.D., G. MAYA, M.D., ANDG. DI PAOLA,M.D. Gynecology
Oncology
Unit,
Cdrdoba
2351,
C.P. 1120, Buenos Aires, Argentina
Received December 11, 1989
verse factors are responsible for that therapeutic failure, One hundred fifty-one patients (107 stage IIB and 44 stage IIIB) who had completed 2 years of follow-up after treatment but the most important perhaps is tumor size. The larger with neoadjuvant chemotherapy (“Quick” VBP scheme), prior to the tumor mass, the poorer the radiotherapy response standard therapy, are presented. Four groups of patients with and hence survival. Chemotherapy has generally been reserved for the different prognosis and operability status, according to tumor response, were identified. In the first group (excellent responders) treatment of metastases. But the possibility of diminish96% of the patients had no evidence of disease (NED), whereas ing the volume of the primary tumor with chemotherapy, in the fourth group (no response), only 33% had NED after 2 as a first approach, appears to be the most significant years of follow-up. In the first group, no pelvic failure was ob- achievement in recent years. served. On the other hand, of 21 patients in group 4, 11 (52%) It has been demonstrated that the use of preoperative had pelvic recurrences. The response to neoadjuvant therapy is adjuvant chemotherapy can increase disease-free surstrongly associated with the initial tumor volume ecographically vival in several kinds of nongynecologic cancer, such as evaluated. The critical pretreatment volume for response was 84 pediatric solid tumors [2,3], osteogenic sarcoma [4,5], cm3(4.85 cm in diameter). Surgery and radiotherapy were equally effective in those patients with initial tumor volume of 84 cm3 head and neck cancer [6,7], and breast cancer [8]. The theoretical advantages of such an approach inor less. But surgery appears to have a better outcome in those clude the following: (1) Drug delivery to the tumor is patients with tumors larger than the critical pretreatment tumor improved because vascularity has not been disturbed by volume. Radiotherapy in this controversial group was ineffective in avoiding pelvic recurrences, especially when the tumor volume surgery or radiotherapy, and drug dosage is not comwas not sufficiently reduced. The advantage of neoadjuvant che- promised in patients who have not undergone prior irmotherapy is evident in stage IIB, with 79% @S/107) of patients radiaton and myelosuppression. (2) The potential for rashowing no evidence of disease as compared with 47% (24/51) dioresponsiveness is improved by size reduction and in a nonrandomized control group (P < 0.01). Statistically sig- decrease in hypoxic cell traction. (3) An inoperable tunificant differences were also obtained in stage IIIB (50% vs 26% mor is rendered resectable. (4) Early treatment is proin the control group). 0 1990 Academic FVW, Inc. vided for occult micrometastatic disease. (5) Patient population that might benefit from similar chemotherapy administered as additional adjuvant treatment following INTRODUCTION surgery and/or radiotherapy is identified. In addition to theoretical advantages there are also It is generally accepted that radiotherapy and surgery are both effective techniques in the management of car- possible disadvantages to such an approach and these include (1) delay of potentially curable radiotherapy, (2) cinoma of the cervix uteri of small volume. Therapeutic results in patients with locally or region- prolongation of treatment, (3) increased toxicity of treatment, (4) possibility of tumor progression, and (5) the ally advanced cervical carcinoma (FIG0 stages II, III, potential for selecting out drug-resistant clones by priand IV) are frequently poor [ 11.Radiotherapy has always mary chemotherapy. been the elective treatment or those stages. Various adThis work has its starting point in a previous paper [9] that reported the introduction of chemotherapy as a Presented at the annual meeting of the Society of Gynecologic Onfirst-line treatment in cervical carcinoma. At the First cologists, San Francisco, CA, Februrary 4-7, 1990. 486 0090-8258190$ I so Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
NEOADJUVANT
CHEMOTHERAPY
Chair of Gynecology, University of Buenos Aires, a new treatment, which includes neoadjuvant chemotherapy as first-line treatment in regionally advanced carcinoma of the uterine cervix (stages II, III, and IV), has been used for the last 5 years. The chemotherapeutic scheme used was the modified vincristine-bleomvcin-cis-Dlatinum (VBP) scheme. The observed tumor responses’have already been reported, corresponding to phase I of this new management [ 101. Today, there is much evidence for the chemosensitivity of carcinoma of the cervix uteri when drugs are administered as neoadjuvant therapy, especially in cis-platinum-based and bleomycin-based schemes [ 1l-l 51. As there is an international consensus about tumor regression of cervical squamous carcinoma, an analysis of survival of patients treated with this therapeutic approach is urgently required. MATERIAL
IN LOCALLY
ADVANCED
CANCER
487
TABLE 1 “Quick” Scheme
,, cic-P,atinum 50 mg/m’ *, V&ristine I mg/m’
(push 15’) (push) 3. Bleomycin (continuous infusion 25 mg/m’ over 6 hr) Three courses repeated at IO-day intervals Day 1 by 1 Days 1,2,3
Toxicity of the drugs was evaluated as follows: 1. c&-Platinum: hemogram including platelet count, hepatoenzymogram, and creatinine clearance after each course of chemotherapy; audiometry at the beginning and end of treatments. 2. Bleomycin: functional respiratory study, chest roentgenogram, and blood gas at the end of each course. 3. Vincristine: complete neurologic study after each course.
To evaluate tumor response to the VBP scheme, careThis is a phase II trial started in 1984 in the Gynecologic Oncology Unit of the First Gynecological Chair ful objective parameters were considered. These paramof Buenos Aires University. Up to now, 328 patients eters were related to cervical mass, vaginal extension, with squamous cervical carcinoma have been included and parametrial compromise. To determine the cervical in this protocol. One hundred fifty-one patients have and the vaginal response, colposcopic measurements completed 2 years of follow-up and their results reported were done. in this paper. Of these 151, 107 patients are FIG0 stage With ecography the volume of the tumor was measured IIB and 44 are IIIB. The mean age was 44.3 (range 21- by multiplying the three spatial axes. The parametrial 80) years. response was assessed mainly by clinical examination, A nonrandomized control group of 101 patients treated taking into account its extension, elasticity, and cononly with radiotherapy (Fletcher Technique) was used sistency. Three experienced members of the department for comparison; 51 were stage IIB patients and 50 stage examined simultaneously the patients in all circumstanIIIB (mean age 42.4, range 23-75). This control group ces, to avoid the possibility of subjective error. The response was classified as complete tumor regreswas staged and treated in the same institution and by the same staff, immediately prior to the neoadjuvant trial. sion (CTR), partial tumor regression (PTR), stable disease (SD), or progressive disease (PD), in the central area and in the parametria. Parametrial response is of METHODS utmost importance because of its implication in manThe trial sequence of procedures was as follows: (1) agement and operability. Complete tumor regression in staging and tumor volume measurement, (2) neoadjuvant the central area was considered when colposcopy rechemotherapy and toxicity evaluation, (3) evaluation of vealed an almost reepithelialized cervix, with a marked tumor response, (4) conventional treatment (surgery plus volume reduction proved by ecography and clinical exradiotherapy or radiotherapy alone), (5) 2 years of follow- amination. In the parametrium this response criterion up and determination of sites of recurrences. was applied to those cases in which the clinical disapAll patients were classified according to the clinical pearance of infiltration was verified. Moderate tumor staging system proposed by the International Federation regression was diagnosed when reepithelialization was of Gynecology and Obstetrics with clinical examination observed at least in 50% of the ectocervical surface inunder anesthesia [l]. To assess tumor bulk, cervical ul- itially involved, with a moderate volume reduction trasound scanning (ecography) was performed. These proved by ecography. The parametrial extent should be studies were repeated when the chemotherapy was ended reduced to 50% or more. A change from irregular to to reevaluate each case and decide the subsequent smooth and indurated parametria was also considered a treatment. partial response. In the stable cases, there were no All cases received three courses of polychemotherapy changes or very subtle changes in the cervix, vagina, of the Quick VBP scheme (Table 1). and parametrium. In patients with progressive disease,
488
SARDI ET AL.
TABLE 2 Management and Operability Stage IIB No. Group 1
S+RT
28
Stage IIIB RT
No.
5
5
23
S+RT 2
85% Group 2 Group 3 Group 4 Total
41 29 9 107
36 1 14 3 76
RT 3
78%
5 9 91 15 18 7 6 12 0 31 44 18 S+RT = 94/151 = 62.2%
0 11 12 26
an increase in tumor bulk or in the vaginal or parametrial lesion was confirmed. In our experience with neoadjuvant chemotherapy, tumor response was always observed to follow a centripetal pattern. This means that the response occurs first in the parametria and vagina and afterward in the cervix. Cervical response was never found without parametrial response. Thus, four groups of tumor response were recognized: Group 1 Group 2
Excellent response Good response
Group 3 Group 4
Regular response No response
CTR in both areas CTR in the parametria, PTR in the central area PTR in both areas SD or PD in both areas
Conventional treatments were applied according to the results obtained after the use of chemotherapy as firstline treatment. When radiotherapy was chosen, irradiation with the Fletcher technique [ 161began 1 week after the last course of chemotherapy. When surgery was decided, a Wertheim-Meigs operation was performed with paraortic lymphadenectomy, 2 weeks after the last course of chemotherapy. After surgery, radiotherapy (5000 rad) to the whole pelvis was administered in all cases. Generally speaking, surgical treatment was considered preferable for the patients in whom parametrial infitration TABLE 3 Response to Neoadjuvant Chemotherapy According to Stages Stage IIB Parametrial response
Central response
Group 1
CTR
;;
Group 2 Group 3 Group 4 Total
CTR PTR SD
PTR SD
Stage IIIB
No. 1;
;6ju
29 9 107
27 8
N; 18 12 44
lj3l 40 47
was no longer clinically demonstrable, while radiotherapy was carried out in those patients who did not exhibit parametrial response or who exhibited partial tumor response. But, during the development of the protocol, it was discovered that many partial responders were also good candidates for surgery with free surgical margins. Wertheim-Meigs operation was performed in 76 stage IIB and 18 stage IIIB patients. After surgery all of them received 5000 rads to the whole pelvis. Radiotherapy according to Fletcher technique was carried out in 31 stage IIB and 26 stage IIIB patients. Thus, surgical treatment after neoadjuvant chemotherapy was performed in 62% (94/151) of all patients. Operability of stage IIB was 71% and of stage IIIB, 44% (Table 2). The number of patients with no evidence of disease 24 months after treatment was analyzed, and the results were compared with those obtained in the control group of patients subjected to the conventional strategy, immediately prior to the beginning of the neoadjuvant chemotherapy trial, following the same staging system, diagnostic, and radiotherapeutic criteria. For statistical analysis, the x2 test was used. A P value less than 0.05 was considered significant. RESULTS
The results are analyzed taking into account tumor response to neoadjuvant treatment, tumor volume, tumor extension. Also, toxicity to the Quick VBP scheme is considered. Results in Relation
to Tumor Response
ReTumor response to neoadjuvant chemotherapy. sponse was analyzed with respect to the four groups previously described. (Table 3). The complete parametrial response is greater in stage IIB (64%) than in stage IIIB (31%) and the percentage of patients with stable disease is higher in stage IIIB as compared with stage IIB. Perhaps the difference in results in stage IIIB patients is due to the initial tumor volume, which was higher in the last group of patients. Operability. Based on these findings, the management of the patients was as follows. Almost all group 1 and 2 patients (excellent and good responders) in stages IIB and IIIB were operated (59/69 stage IIB and 1l/14 stage IIIB). Almost all group 4 patients (nonresponders) were subjected to radiotherapy (6/9 stage IIB and 12/12 stage IIIB). Initially, group 3 patients (partial responders) were treated the same as group 4 patients (nonresponders), but later, during development of the protocol, it was discovered that many partial parametrial responders
NEOADJUVANTCHEMOTHERAPYIN
LOCALLY
TABLE 6 Pathological Findings
TABLE 4 Relationship between Tumor Response and Free Disease Interval after 2 Years of Follow-up Pelvic failures
NED Response Group Group Group Group
1 2 3 4
No. 33 50 47 21
No.
%
No.
%
0 6 12 12 25 11 52 x2 = 25.12
32 96 39 78 29 61 7 33 x2 = 28.44 P < 0.005
489
ADVANCEDCANCER
Distant metastases
Clinical response
No.
Group Group Group Group
%
1 3 5 10 6 12 14 3 x2 = 4.05
1 2 3 4
No.
Cervical lesion CO.5 cm (%o)
Parametrial residual lesion (%)
Lymph node metastases (%)
25 45 21 3
60 11 4 0
4 20 66 100
0 24 62 66
explore the effectiveness of conventional treatments in this trial with neoadjuvant chemotherapy. As in groups ’ No evidence disease; NS, not significant. 1, 2, and 4 the treatments were not harmonically distributed and no significant conclusion could be drawn. were also candidates for surgery with free surgical mar- In group 3 (partial responders) equal numbers of patients gins. These were the cases in which, besides the reduc- were subjected to both treatments; these results are pretion of infiltration, changes from nodular and irregular sented in Table 5. In the partial responders surgery appears to be more to smooth and indurated parametria were observed. effective than radiotherapy in avoiding pelvic failures, In these patients, microscopic residual tumor was not compared with a group of patients with similar mean prefound to be a significant obstacle to the surgical techand post-treatment volumes. nique , and in all cases free surgical margins were Pathological jindings of surgical specimens. Analysis determined. of the pathological findings of the 94 surgical specimens Thus, operability was found to be higher than what was considered of value in assessment of the objectivity the previous response criteria would have indicated (Taof the clinical response criteria. Some data of interest bles 2 ad 3). On the other hand, no stage IIIB patients are presented in Table 6. with stable disease could be excised, although some of A strong correlation between clinical response and them were subjected to staging laparatomy. Disease-free interval (DFI) according to tumor re- pathological findings of surgical specimens was verified. sponse. The DFI of the patients in relation to the re- An increasing residual tumor size in the central and parasponse to neoadjuvant chemotherapy according to the metrial areas according the worse response was observed. Also, an increase in the percentage of lymph groups previously described is presented in Table 4. node involvement after neoadjuvant therapy was verified The outcome after 2 years of follow-up for the four in the partial responders and nonresponders. groups of tumor response demonstrates a strong assocation between response level and percentage of diseaseResult in Relation to Tumor Volume free interval and site of recurrence. It is evident that pelvic failures occur more frequently in patients with Relationship between tumor response and tumor volstable disease than in complete responders. The per- ume. The above-mentioned different responses to neoadcentage of distant metastases remains constant in all juvant treatment between stage IIB and IIIB patients are groups, except the first one. probably due to the higher initial tumor volume in the Disease-free interval according to tumor response and later patients. The Gompertz equation establishes that conventional treatment. It was considered of interest to when tumor volume increases, the tumor response to P
