Has serum CA 125 assay at the time of relapse a prognostic relevance for patients with recurrent ovarian carcinoma after primary cytoreduction and platinum- and paclitaxel-based chemotherapy?

PURPOSE OF INVESTIGATION: To correlate serum CA125 at relapse with survival in ovarian cancer patients who achieved a complete response after primary cytoreduction and paclitaxel- and platinum-based chemotherapy. MATERIALS AND METHODS: The study was conducted in 104 patients. RESULTS: The 25%, 50%, and 75% quantiles of CA125 levels at relapse were 46, 118, and 190 U/ml. By log-rank test, survival after recurrence was related to consolidation treatment (p = 0.046), platinum-free interval (PFI) (p < 0.000005), number of recurrence sites (p = 0.03), treatment at recurrence (p = 0.002), and serum CA125 taking 118 U/ml as cut-off (p = 0.013). On multivariate analysis, consolidation treatment (p = 0.007), PFI (p = 0.0001), treatment at recurrence (p = 0.01), and serum CA125 taking 118 U/ml as cut-off (p = 0.04) were independent prognostic variables for survival. CONCLUSIONS: Serum CA125 at relapse was an independent prognostic variable. Patients with serum CA125 > 118 U/m had 1.943 higher risk of death than those with lower antigen value.

Prognostic value of lymph node status and number of removed nodes in patients with squamous cell carcinoma of the vulva treated with modified radical vulvectomy and inguinal-femoral lymphadenectomy.

PURPOSE OF INVESTIGATION: To assess the outcome of patients with squamous cell vulvar carcinoma treated with deep partial or total vulvectomy and inguinal-femoral lymphadenectomy. MATERIALS AND METHODS: The authors assessed 87 patients who underwent primary surgery. RESULTS: Tumor recurred in 34 patients, and the first relapse was local in 19, inguinal in ten, and distant in five. Five-year disease-free survival was 56.7% and was related to Stage (p < 0.0001), grade (p = 0.023), and node status (p < 0.0001). Groin failure occurred in 4.9% of node-negative patients compared with 29.6% of node-positive patients (p = 0.0096). Distant recurrences only developed in women with positive nodes. Among the 47 patients who underwent bilateral lymphadenectomy and who had negative nodes, groin recurrence occurred in 12% of those who had < or = 15 nodes removed and 0% of those who had > 15 nodes removed. CONCLUSIONS: Stage and node status were the most important prognostic variables. There was a trend favoring a better groin control in patients with node-negative disease who underwent extensive lymphadenectomy.

Analysis of the pattern of hypersensitivity reactions in patients receiving carboplatin retreatment for recurrent ovarian cancer.

The aim of this paper was to assess hypersensitivity reactions in 69 patients who received carboplatin (CBDCA) retreatment for recurrent ovarian cancer. Hypersensitivity reactions developed in 15 (21.7%) patients and occurred during the second cycle of retreatment in 13 (86.7%) of them. Reactions consisted of skin rash, flushing, itching, or abdominal cramping in eight (53.3%) and severe respiratory or cardiovascular events in seven patients (46.7%). One patient had a chest pain, without any other symptoms suggestive of hypersensitivity, followed by cardiac arrest unresponsive to standard resuscitative measures. All the other cases promptly recovered from symptoms. Logistic regression analysis showed that allergy history and CBDCA retreatment interval (interval time between the last cycle of first-line chemotherapy and CBDCA retreatment) were independent predictive variables for the risk of hypersensitivity, whereas patient age, first-line chemotherapy, total CBDCA dose given during first-line treatment, recurrence treated with CBDCA (first versus other), and CBDCA regimen at recurrence had no predictive value. Hypersensitivity reaction rate was higher in patients with CBDCA retreatment interval longer than 23.4 months compared to those with a shorter interval (36.3% versus 8.3%, P = 0.0132). Nine patients were subsequently treated with cisplatin, and two (22.2%) still developed allergic reactions. In conclusion, hypersensitivity reactions to CBDCA retreatment can occur in approximately one fifth of the cases, and a CBDCA retreatment interval longer than 2 years appears to be the strongest predictive variable for the development of allergic reactions.

Epidoxorubicin and lonidamine in refractory or recurrent epithelial ovarian cancer.

Lonidamine (150 mg x 3 day orally, days 1-5) plus high dose epidoxorubicin (120 mg/m2 intravenously, day 3) was tested in 26 patients with refractory or recurrent epithelial ovarian cancer, to assess the anti-tumour activity and the toxicity of this combination of drugs. All patients were evaluable for toxicity and 24 for tumour response. Two complete responses (8.3%) and six partial responses (25.0%) were recorded for a total response rate of 33.3%. 6 of 8 responding patients were pretreated with anthracyclines. Stable disease was obtained in 7 patients (29.2%). Toxicity was acceptable; only 1 (3.8%) patient stopped chemotherapy because of a left ventricular ejection rate reduction > 20%. The most relevant side-effect was leucopenia (grade 3-4, 34.6%). In conclusion, the association of lonidamine and high-dose epidoxorubicin has promising activity as second-line treatment in patients with refractory or recurrent epithelial ovarian cancer.

Serum CA125 assay in the early diagnosis of recurrent epithelial ovarian cancer.

The present retrospective study included 59 patients who had been treated for epithelial ovarian cancer and who showed an elevation of serum CA125 above 35 U/ml without clinical evidence of disease. Eight patients underwent chemotherapy at the time of serum antigen elevation (group A). The other 51 patients (group B) were periodically checked, and received further chemotherapy only when recurrent disease was detected. Forty-four of the 59 patients relapsed. One patient of group B experienced two different recurrences with two distinct time intervals. The median follow-up of survivors from CA125 elevation was 10 months (range 2-92 months). Of the 8 patients of group A, 3 (37.5%) developed recurrent disease after 14, 17 and 22 months, respectively, from antigen elevation. Of the 51 patients of group B, 41 (80.4%) relapsed. The overall recurrence rate was 82.4% (42/51). The median lead time between CA125 increase and clinical detection of relapse was 3 months (range 2-27 months). The present data confirmed the reliability of serum CA125 assay as predictor of clinical relapse in epithelial ovarian cancer. The recurrence rate seemed to be lower for patients who received chemotherapy at the time of CA125 elevation (37.5% versus 82.4%, p=0.02). However, the small number of patients, the short follow-up, and the non-randomized design of the study do not allow to draw any conclusion on the appropriate timing for second-line chemotherapy.

Serum levels of soluble receptors for tumor necrosis factor (p55 and p75 sTNFr) in patients with cervical cancer.

The pretreatment serum levels of the soluble receptors for tumor necrosis factor (p55 and p75 sTNFr) were retrospectively measured in 54 patients with cervical cancer and in 55 patients with benign uterine disease as controls. Serum mean (+/- standard deviation) concentrations of both p55 and p75 sTNFr were higher in patients with cervical cancer than in controls (2.6+/-1.3 vs 1.9+/-0.7 ng/ml, p<0.0001, and, respectively, 7.8+/-4.3 vs 5.9+/-3.0 ng/ml, p=0.009). Both receptor levels were significantly higher in patients with stage IIb-IV than in those with stage I-IIa cervical cancer or with cervical intraepithelial neoplasia (CIN) 3. Among the 31 patients with stage I-IIa disease who underwent initial surgery, the preoperative serum p55 and p75 sTNFr values correlated neither with the common prognostic variables nor with the clinical outcome. In conclusion, serum p55 and p75 sTNFr levels are significantly elevated in patients with cervical cancer. However, the serum measurement of these soluble receptors seems to be of limited clinical value for the management of patients with this malignancy.

Times and sites of relapses after negative second look in advanced epithelial ovarian cancer.

In the present work the Authors analyse the role and significance of second-look laparotomy in the management of epithelial ovarian cancer. Eighty-three patients with advanced epithelial ovarian cancer (stage III-IV) followed at the Institutes of Gynecology of Padua and Pisa Universities underwent a negative second-look at a median time of 6 months after first surgery. The incidence of relapse after the negative second-look was 45.8% and the mean time of relapse was 27.4 months. Advanced surgical-pathological stage, the presence of residual disease > 2 cm and serous histotype are risk factors for relapse after a negative second-look. These data confirm that, because of its poor prognostic value, a negative second-look laparotomy should not influence the choice of second line chemotherapy and, consequently is no longer routinely performed in our Institutes.

Malnutrition and cachexia in ovarian cancer patients: pathophysiology and management.

In ovarian cancer patients the poor nutritional status and cachexia are caused by the metabolic effects of the enlarging tumor masses and bowel obstruction. These patients may have a high resting energy expenditure due to increase in Cori cycle activity, glucose and triglyceride-fatty acid cycling and gluconeogenesis. Biochemical mediators of cachexia include cytokines, such as tumor necrosis factor and interleukin-6, and tumor-produced catabolic factors, such as lipid-mobilizing factor, proteolysis-inducing factor, and anemia-inducing factor. Mechanisms involved in the pathogenesis of obstruction may include extrinsic occlusion of the bowel due to pelvic, mesenteric omental masses, or intestinal motility disorders due to infilor tration of the mesentery or bowel muscle and nerves. The relief of malnutrition and cachexia may be attempted through nutritional support, pharmacological approach (megestrol acetate, cyclooxygenase inhibitors) and palliative treatment of bowel obstruction. Very few agents have been demonstrated to have true anticachectic activity, so future research should be addressed to the identification of drugs able to block the activity of tumor-produced catabolic factors. The decision regarding optimum management of bowel obstruction should be individualized. Krebs' and Goplerud's score (based on age, nutritional status, tumor status, ascites, previous chemotherapy and irradiation) seems to offer reliable eligibility criteria for those patients who can benefit from surgery.

An intensive follow-up does not change survival of patients with clinical stage I endometrial cancer.

BACKGROUND: The clinical benefit of an intensive follow-up protocol in endometrial cancer patients is still uncertain. MATERIALS AND METHODS: One hundred and thirty-three patients underwent initial abdominal surgery for clinical stage I endometrial cancer between January 1988 and December 1997 and were periodically followed-up until April 1999 or until death. After surgery, 89 patients received postoperative adjuvant treatment. Periodical surveillance included physical examination, vaginal smear and abdominal-pelvic ultrasound every 3-4 months for the first 2 years from surgery, every 6 months for the next 3 years and yearly thereafter. Chest X-ray was performed every 6 months for the first 2 years, every year for the next 3 years and with individually increasing intervals afterwards. An abdominal-pelvic CT scan was carried out yearly for 5 years. RESULTS: Twenty-four patients (18.0%) developed recurrent disease after a median time of 17.5 months (range, 6-64 months). Five-year actuarial disease-free survival was 81.2% in the whole series, and in detail 94.2% in the subset of patients with FIGO stage Ia grade 1 or 2 endometrioid adenocarcinoma, or stage Ib grade 1 endometrioid adenocarcinoma (low-risk), compared to 76.0 (p = 0.0472) in the subset of patients with stage Ia grade 3 endometrioid adenocarcinoma, stage Ib grade 2 or 3 endometrioid adenocarcinoma, stage equal to or greater than Ic endometrioid adenocarcinoma any grade, and aggressive histologies (high-risk). The site of recurrent disease was local in 6 (25.0%) patients, distant in 17 (70.8%), and local plus distant in 1 (4.2%). Eleven (45.8%) recurrences were symptomatic and 13 (54.2%) were asymptomatic. The median survival after recurrence was 10 months. Survival was longer in patients who relapsed after 17.5 months from initial surgery when compared to those who relapsed earlier (p = 0.02). Conversely, survival after recurrence was not related to patient characteristics at diagnosis, such as FIGO stage, tumor grade, myometrial invasion, histologic type, and lymph node status. It is noteworthy that survival was similar in asymptomatic women, in whom the relapse was occasionally detected by follow-up examinations, and in symptomatic ones. CONCLUSION: An intensive surveillance protocol seems to have no significant impact on the outcome of patients with clinical stage I endometrial cancer. Simplified follow-up programs tailored for patient subsets with different recurrence risk are required.

Hormone replacement therapy and gynecological cancer.

The aim of this paper is to report the risk of development of gynecological cancer in women receiving hormone replacement therapy and to review the current knowledge on the administration of hormone replacement therapy following treatment of gynecological cancer. Estrogens alone may act as promoting factors for endometrial carcinogenesis. However, the addition of progestins reduces the risk of endometrial cancer to that of nonusers. Hormone replacement therapy could be given to selected patients following treatment for endometrial cancer. However, we think that this therapy should be reserved only for patients enrolled in controlled clinical trials. Ovarian cancer does not seem to be sensitive to estrogens, even if current literature does not allow firm conclusions to be drawn. Hormone replacement therapy should be offered to patients previously treated for ovarian cancer and cervical cancer.

Serum interleukin-6 levels in uterine malignancies. Preliminary data.

Interleukin-6 (IL-6) was measured with an enzyme-immunoassay in blood samples drawn at diagnosis from 37 patients with endometrial cancer, 36 with cervical cancer, 9 with cervical intraepithelial neoplasia (CIN) and 68 with benign uterine disease. The minimal detectable dose of IL-6 was 3 pg/mL. Detectable serum IL-6 levels were found in 9% of patients with benign uterine diseases, 11% of patients with CIN, 44% of patients with cervical cancer and 11% of patients with endometrial cancer. As regards cervical cancer, serum IL-6 levels > 3 pg/mL were found in 36.0% of 25 patients with stage Ib-IIa disease and in 64% of 11 patients with stage IIb-IV disease. As regards endometrial cancer, serum detectable IL-6 levels were observed in 0% of 30 patients with stage I-II disease and in 57% of 7 patients with stage III-IV disease (p = 0.0005). These preliminary data suggest that IL-6 may be involved in the progression of uterine malignancies.

Surgical cytoreduction during second-look laparotomy in patients with advanced ovarian cancer.

BACKGROUND: The impact of surgical cytoreduction performed during second-look on the survival of patients with advanced ovarian cancer is still under debate. MATERIALS AND METHODS: The present retrospective investigation assessed 81 patients with advanced ovarian cancer who underwent second-look laparotomy after initial cytoreductive surgery and six cycles of platinum-based chemotherapy. RESULTS: At the beginning of the second-look, 31 patients were in pathological complete response, 7 had microscopic residual disease, 22 had macroscopic residual disease < 2 cm and 21 had a larger residuum. Twenty-two (51.2%) of the 43 patients with macroscopic disease underwent surgical cytoreduction during the second-look: 11 patients were completely cytoreduced, whereas 11 were debulked from residuum > 2 cm to macroscopic residuum < 2 cm. Patients with microscopic residual disease after the second-look had improved survival when compared to those with macroscopic residuum after re-exploration (median survival, 43 months versus 19.5 months, p = 0.002). Among the former, no difference in survival was detected between the patients who had microscopic disease at the beginning of the second-look and those who were surgically cytoreduced to microscopic disease. Moreover, the patients with macroscopic residuum < 2 cm after the second-look had a better survival than those with a larger residuum after re-exploration (median survival, 24 months versus 10 months, p = 0.0001). Among the former, no difference in survival was seen between the patients who had residual disease < 2 cm at the beginning of the second-look and those who were surgically cytoreduced to < 2 cm. However, ovarian cancer recurred in all the patients with small or large macroscopic residuum after the second-look. CONCLUSION: The complete resection of macroscopic persistent tumor seems to give ovarian cancer patients the highest likelihood of long-term survival and should represent the goal of surgical cytoreduction during second-look laparotomy.

Combination of cisplatin, epirubicin, and cyclophosphamide (PEC regimen) in advanced or recurrent endometrial cancer: a retrospective clinical study.

The combination of cisplatin (50 mg/m2), epirubicin (60 mg/m2), and cyclophosphamide (600 mg/m2) (PEC regimen) was given every 4 weeks to 19 patients with advanced (n.2) or recurrent (n.17) endometrial cancer. The median number of cycles delivered to each patient was 5 (range, 2-8). All patients were evaluable for toxicity and 16 for response. Two (12.5%) patients experienced a complete response and 5 (31.2%) had a partial response, for an overall objective response rate of 43.7%. The median duration of objective response was 10 months (range, 3-28 months). Median survival was 10 months (range, 3-68+ months) in the whole series. According to response to chemotherapy, median survival was 12 months (range, 3-68+ months) for responders and 9 months (range, 6-17 months) for nonresponders. Hematologic toxicity was relatively frequent but it could be easily managed, and significant nonhematologic toxicities were not found except for nausea and vomiting. In conclusion, PEC regimen has a good activity in advanced or recurrent endometrial cancer, but the short duration of responses limits the impact of the treatment on survival time.

Recurrence patterns in patients with early stage cervical cancer treated with radical hysterectomy and external pelvic irradiation.

One hundred and fifty patients with clinical FIGO stage IB-II cervical cancer who underwent radical surgery followed by external pelvic irradiation between 1978 and 1991 were reviewed. Until June 1994, 28 (18.7%) patients developed recurrent disease. Seventeen (60.7%) of them experienced a pelvic failure, 7 (25.0%) an extrapelvic failure and 4 (14.3%) both a pelvic and an extrapelvic failure. The median time to recurrence was 16 months for patients with pelvic failure (range = 4-50 months), 27 months for those with extrapelvic failure (range = 6-49 months), and 21 months for those with both pelvic and extrapelvic failure (range u 8-56 months). Recurrence rates were significantly related to surgical-pathologic stage, tumor size and lymph node status, but not to histologic type. An extrapelvic recurrence, alone or associated with a pelvic failure, was found in 0.9% of 117 patients with negative lymph nodes, 6.2% of 16 patients with one or two positive lymph nodes, and 52.9% of 17 patients with three or more positive lymph nodes, (p = 0.0001). It is worth noting that 9 (81.8%) out of the 11 patients who developed extrapelvic recurrences had three or more involved lymph nodes. The number of positive lymph nodes (p = 0.0001) and the tumor size (p = 0.0046) were independent prognostic variables for disease-free survival.

Analysis of treatment failures in patients with early-stage uterine leiomyosarcoma.

This retrospective study aimed to investigate the treatment failures in 26 patients with stages I-II uterine leiomyosarcoma (> or = 10 mitoses per 10 high-power field (HPF) who underwent total abdominal hysterectomy and bilateral salpingo-oophorectomy +/- adjuvant external pelvic irradiation. Thirteen (50%) patients developed recurrent disease, after a median time of 10 months from surgery (range = 4-72 months). Recurrence was pelvic in 3 (23%) patients, extrapelvic in 9 (69%) patients, and both pelvic and extrapelvic in 1 (8%) patient. Disease-free survival was better for premenopausal than for postmenopausal patients (p = 0.002) and for patients with < 20 mitoses per 10 HPF than for those with > or = 20 mitoses per 10 HPF (p = 0.006). In conclusion, patients with early-stage disease who had undergone locoregional treatment experienced a high recurrence rate. Most of the treatment failures were extrapelvic. Multicentric randomized trials on the role of adjuvant chemotherapy are advocated.

Serum anti-p53 antibodies in the follow-up of patients with advanced ovarian carcinoma.

Serum anti-p53 antibodies have been detected in different human malignancies, including ovarian carcinoma. In the present investigation these autoantibodies were retrospectively measured with a new ELISA (Immunotech, Marseilles, France) before first surgery and subsequently at different times during the course of disease from 40 patients with advanced ovarian carcinoma. Anti-p53 antibodies were preoperatively found in 15 (37.5%) patients. With regard to the follow-up of these 15 patients, anti-p53 antibodies were detected in 87.8% of the 41 samples drawn when there was clinical evidence of disease compared to 57.1% of the 14 samples collected when there was no clinical evidence of tumor (p = 0.037). As for the 25 patients whose serum originally scored negative, the autoantibodies were found only in 1.8% of the 113 samples obtained during the follow-up, independently of the status of disease. In conclusion, anti-p53 antibodies are often detected in serum from patients with advanced ovarian carcinoma. However, the serial measurement of these autoantibodies does not seem to give useful clinical information for the follow-up of these patients.

The measurement of plasma D-dimer (DD) levels in the follow-up of patients with ovarian cancer.

Peripheral blood samples for the measurement of DD and CA 125 were drawn from 39 patients with ovarian cancer at different times from first surgery. Both median DD and CA 125 levels were significantly higher in the 20 samples drawn from patients with clinically evident disease than in the 37 samples from patients without clinical evidence of disease (477 vs 300 ng/ml p = 0.006, and 66 vs 11 U/ml, p < 0.0001, respectively). DD levels did not correlate with CA 125 levels. The sensitivity, specificity and diagnostic accuracy of the tests in the assessment of clinical disease status were as follows: 65%, 62% and 63% for DD (cut-off = 416 ng/ml); 70%, 92% and 84% for CA 125 (cut-off = 35 U/ml); 90%, 59% and 70% for DD "or" CA 125; and 45%, 95% and 77% for DD "and" CA 125. DD levels correlated with the clinical course of disease in ovarian cancer patients. However, the concomitant determination of DD and CA 125 did not improve the reliability of CA 125 assay alone in the follow-up of these patients.

Serum preoperative vascular endothelial growth factor (VEGF) in epithelial ovarian cancer: relationship with prognostic variables and clinical outcome.

Substantial experimental and clinical evidence links tumor growth, progression and metastatic potential with neoangiogenesis. This process is modulated by several angiogenic growth factors, such as vascular endothelial growth factor (VEGF). Little data are currently available on serum VEGF levels in cancer patients. In the present retrospective investigation preoperative serum VEGF was higher in 53 patients with epithelial ovarian cancer than in 25 patients with benign ovarian disease as controls (median, range: 229.7, 23.5-1807.5 pg/ml versus 140.3, 14.7-1038.7 pg/ml, p = 0.034). With regard to FIGO stage, antigen values were significantly elevated in stage III-IV (p = 0.027) but not in stage I-II ovarian cancer patients when compared to controls. In patients with advanced disease preoperative serum VEGE was significantly related to the presence of ascites (p = 0.013), but not to common prognostic variables, response to chemotherapy and survival. In conclusion, preoperative serum VEGF assay reflects tumor progression and ascites generation in epithelial ovarian cancer, but it seems to have a limited predictive and prognostic value in patients with advanced disease.

p53 status is neither a predictive nor a prognostic variable in patients with advanced ovarian cancer treated with a paclitaxel-based regimen.

BACKGROUND: The aim of this study was to assess the relationship between p53 status and the clinical outcome of patients with advanced ovarian cancer treated with a paclitaxel-based regimen. PATIENTS AND METHODS: The investigation was conducted on 38 patients with FIGO stage III-IV ovarian cancer from whom tumor tissue samples for p53 protein immunostaining were obtained during initial cytoreductive surgery. All these patients subsequently received six cycles of first-line combination chemotherapy with paclitaxel 175 mg/m2 (3-hour infusion) plus carboplatin AUC 6 with or without epidoxorubicin 75 mg/m2. RESULTS: Positive p53 immunostaining was detected in tissue samples collected from 24 (63.2%) ovarian cancers. A clinical complete response was obtained in 14 (58.3%) of the 24 patients with positive p53 immunostaining compared to 9 (64.3%) of the 14 patients with negative p53 immunostaining (p = 0.717). A pathological complete response was found in 6 (25.0%) of the former compared to 4 (28.6%) of the latter (p = 0.956). Similarly, survival did not correlate with p53 status (p = 0.1271). DISCUSSION: p53 status seems to be neither a predictive nor a prognostic variable in patients with advanced ovarian cancer treated with a paclitaxel-based regimen. These results are consistent with experimental data showing that paclitaxel cytotoxicity in ovarian cancer is likely to be mediated by a p53-independent pathway.

Preoperative serum antibodies against the p53 protein in patients with ovarian and endometrial cancer.

Antibodies against the p53 protein were measured with a sandwich-type enzyme-linked immunosorbent assay in blood samples preoperatively collected from 30 patients with ovarian cancer and 30 patients with endometrial cancer. Anti-p53 antibodies were detected in 33.3% of patients with ovarian cancer, comprising 22.2% of the 9 patients with stage I-II disease, 30.8% of the 13 patients with stage III disease, and 50.0% of the 8 patients with stage IV disease. Anti-p53 antibodies were found in none of the 4 patients with well differentiated tumors and in 38.5% of the 26 patients with moderately or poorly differentiated tumors. Among the 21 patients with stage III-IV disease, a complete clinical response to front-line platinum-based chemotherapy was obtained by 46.2% of the 13 patients without anti-p53 antibodies and 25.0% of the 8 patients with anti-p53 antibodies. Antibodies against the p53 protein were detected in only 6.7% of patients with endometrial cancer. The low incidence of anti-p53 antibodies in patients with endometrial cancer seems to suggest that the serum assay of these autoantibodies has a limited clinical relevance in the management of this malignancy. On the other hand in patients with ovarian cancer the incidence of serum anti-p53 antibodies is relatively high, and, moreover, it seems to increase with tumor stage and grade and seems to be associated with a lower response rate to chemotherapy. However, the small number of patients did not allow us to obtain statistically significant differences.