Efficacy and safety of adjuvant treatment with Wobe-Mugos E in breast cancer
A multicentric retrolective cohort analysis of therapeutic data in parallel groups (RetrospectTM)

Beuth J. (1); Bock P.R. (2); Hanisch J. (2)

1 Institute of Medical Microbiology and Hygiene, University of Cologne, Germany
2 IFAG Basel, Basel, Switzerland

Presented at the satellite symposium "Meet the experts", The European cancer Conference, 12.-16. September 1999, Vienna.
Abstrakt z přednášky: Enzyme-Forum - MUCOS - Scientific Posters, pp. 17-18

599 KA (19-07-2)

Part 1: sub-sample-analysis (1): Wobe-Mugos E as the only adjuvant treatment compared to no adjuvant treatment (N=840)

Introduction

The objective of this study is to prove the clinical efficacy and safety of the oral enzyme preparation Wobe-Mugos® E as adjuvant therapy in breast cancer. The study was designed as a multicentric retrolective cohort analysis with parallel groups (Retrospect ™) using standardized medical record data according to the written study protocol. All patient data from the participating centres which fits with the inclusion and exclusion criteria laid down in the study protocol were collected and analyzed without any further selection.

Methods

Primary test criterion of therapeutic efficacy
The primary test criterion of efficacy in this study is according to the study protocol the multivariate, directional simultaneous test according to Wei and Lachin et al. on relevant symptom-scores (change from baseline), using the generalized multivariate Mann-Whitney statistic with its 95% CI as the measure of effect size.

Secondary test criteria of therapeutic efficacy

• single relevant symptom-scores (change from baseline)
• sum-score (change from baseline), a total of all single symptom scores
• doctor's judgement on therapeutic efficacy (score on an ordinal scale)
• adverse reactions (ADRs) of the oncological basic therapy (no. patients with ADRs, ADR-frequency, severity, reduction of ADR-occurrence)
• physical performance index; Karnofsky index (change from baseline)
• time to recurrence, metastasis and death (survival analysis)

Primary test criteria of safety
doctor's judgement of safety (score on an ordinal scale)

Secondary test criteria of safety

• no. of patients with adverse reactions attributed to the enzyme therapy
• frequency of enzyme-attributed adverse reactions per patient
• kind, severity, duration, measures to treat enzyme-attributed adverse reactions

Disease stage assessment
The disease stage and severity was assessed according to the UICC-stage

Baseline characteristic
To assure the comparability of the test and the control groups, patient's demographic data as well as the baseline disease characteristic and the symptoms severity degree were compared between the treatment groups before evaluating the efficacy and safety results.

Analysis of interactions ("sensitivity analysis")
To evaluate the reliability of the test results on efficacy, an analysis of possible interactions with potential confounding factors like UICC stage, patient's age, oncologic basic therapy (e.g. radiation, cytostatic drug, and hormone-antagonists), as well as the quality of data documentation in the medical records were analyzed by multivariate analysis of variance (ANOVA) and multivariate logistic regression , respectively.

Results

Sample size
The present analysis represents a sub-sample (N=840) from the whole study sample of breast cancer data with a total of 2339 patients, with the test group of Wobe-Mugos® E as the only adjuvant treatment (N=300, 35.7%) compared to a parallel matched control group without any adjuvant treatment (N=540, 64.3%)

Baseline comparability

• The patients demographic data was not significantly different between the therapy groups with the exception of age (mean 59.3 vs 61.2 years; P=0.021) and body weight (67.3 vs 69.9 kg; P=0.001). These differences are, however, not clinically relevant.
• The UICC disease stage distribution was not significantly different between the therapy groups (P(exact) = 0.384).
• The disease risk- and prognostic criteria were not significantly different between the therapy groups at the baseline, with the exception of tumor multiplicity (15.4 vs 8.0%; P=0.001). This difference is not regarded as clinically relevant in this analysis.
• The baseline values of disease symptom severity as well as the calculated multivariate combined symptom effect size (multivariate Mann-Whitney-statistic) were not significantly different between the therapy groups.
• The frequency of oncological basic treatment with radiation (62.3 vs 64.6%), cytostatic drugs (34.0 vs 37.4 %) and hormone-antagonists (65.9 vs 69.1%) were not statistically different between the therapy groups.

Conclusion: The patients demographic parameter, the disease risk- and prognostic criteria, the baseline symptom severity and the basic oncological therapy are comparable between the treatment groups.

Primary endpoint results of efficacy
The primary endpoint of this study is the multivariate effect size (Mann-Whitney-statistic, MWS) based on the change from the baseline of the relevant disease symptoms tested with the multivariate directional test (Wilcoxon-Mann-Whitney, WMW) according to Wai and Lachin. The analysis confirmed a statistically significant and clinically relevant superiority of the Wobe-Mugos® E treatment over "no adjuvant treatment" in this breast cancer study (MWS=0.60, 95% CI= 0.56-0.65); P<0.0001).
Conclusion: According to the study protocol, the clinical efficacy of Wobe-Mugos® E in breast cancer is proven by this hypothesis test result on the primary efficacy criterion.

Secondary endpoint results of efficacy

Single symptoms
Among the single disease-relevant symptoms, following proved statistically significant and clinically relevant superiority of Wobe-Mugos® E, even after the conservative Bonferroni a-adjustment for multiple criteria:

• Nausea (MWS=0.64, 95% CI=0.56-0.71; P(WMW) = 0.0003)
• Reduced appetite (MWS=0.62, 95% CI=0.57-0.72; P(WMW) < 0.0001)
• Epigastric (stomach) pain/disorders (MWS=0.65, 95% CI=0.57-0.73; P(WMW) = 0.0002)
• Headache (MWS=0.65, 95% CI=0.56-0.73; P(WMW) = 0.0005)
• Weakness (MWS=0.60, 95% CI=0.55-0.66; P(WMW) = 0.0003)
• Memory/concentration disorders (MWS=0.70, 95% CI=0.62-0.78; P(WMW) < 0.0001)
• Restlessness/nervous (MWS=0.61, 95% CI=0.55-0.68; P(WMW) = 0.0009)

Symptom-sum-score
The symptom sum-score proved a statistically significant superiority of Wobe-Mugos® E (mean difference = -2.43; P(WMW) = 0.009).

Physician's judgement of efficacy
The physician's judgement of efficacy proved a superiority of Wobe-Mugos® E. Judgement "excellent" = 71.9 vs 57.7%; MWS=0.57, 95% CI=0.54-0.61; P(WMW) < 0.0001).

No. of patients with ADRs
The number of patients with ADRs was significantly lower in the Wobe-Mugos® E group (27.0 vs 58.5%; MWS=0.66, 95% CI=0.62-0.69; P(WMW) < 0.0001).

Degree of ADR severity
The degree of ADR severity was significantly lower in the Wobe-Mugos® E group. (median 1.0 vs 2.0; MWS=0.62, 95% CI=0.56-0.68; P(WMW) = 0.0003).

Reduction of ADRs induced by oncologic basic therapy (doctor's judgement)
The reduction of ADRs was significantly more expressed in the Wobe-Mugos® E group (Judgement "large reduction" =75.8 vs 57.2%; (MWS=0.59, 95% CI=0.55-0.63; P(WMW) < 0.0001).

Karnofsky index (change from baseline)
The Karnofsky index change was significantly but only borderline superior in Wobe-Mugos® E group (MWS=0.54, 95% CI=0.50-0.59; P(WMW) = 0.0457). The "performance index" was not statistically different between the treatment groups.
Conclusion: The hypotheses tests on the secondary criteria of efficacy confirmed the superiority of Wobe-Mugos® E over "no adjuvant therapy" in breast cancer.

Time to recurrence
Only in the UICC-stage IIa + IIb sufficient number of patients were available for evaluation. In the WME-group, one recurrence among 153 evaluable patients and in the control group 12 recurrences among 250 evaluable patients were observed. While the median time to recurrence could not be estimated, the estimated mean time to recurrence was 189 versus 119 months in the WME and the control group, respectively, P(log rank) = 0.0109.
Conclusion: in the UICC stage IIA+IIB patients, the WME therapy has shown a significant trend to longer time to recurrence than the control group.

Time to metastasis
Only in the UICC-stage IIa + IIb sufficient number of patients were available for evaluation. In the WME-group, 8 recurrences among 152 evaluable patients and in the control group 23 recurrences among 249 evaluable patients were observed. While the median time to recurrence could not be estimated, the estimated mean time to recurrence was 156 versus 96 months in the WME and the control group, respectively, P(log rank) = 0.0471.
Conclusion: in the UICC stage IIA+IIB patients, the WME therapy has shown a significant trend to longer time metastasis than the control group.

Time to death (survival)
Only in the UICC-stage IIa + IIb sufficient number of patients were available for evaluation. In the WME-group, no mortalities among 154 evaluable patients and in the control group 7 mortalities among 251 evaluable patients were observed. While the median time to recurrence could not be estimated, the estimated mean time to recurrence could not be estimated in the WME group and was estimated with 125 months in the control group, P(log rank) = 0.0229
Conclusion: in the UICC stage IIA+IIB patients, the WME therapy has shown a significant trend to longer survival than the control group.

Primary endpoint results of safety (tolerance)
The primary endpoint of safety in this study, the physician's judgement of safety (tolerance), confirmed a statistically significant and clinically relevant tolerance superiority of the Wobe-Mugos® E treatment over "no adjuvant treatment" in this breast cancer study ("excellent" results: 70.7 vs 37.8%; MWS=0.69, 95% CI= 0.66-0.73); P<0.0001).
The analysis of single ADRs was not performed, because in this sub-sample no enzyme-attributed side-effects were reported.
Conclusion: According to the study protocol, this test result proved the overall safety of the Wobe-Mugos® E therapy in breast cancer patients.

Potential effect (interaction) of data quality on the efficacy results (symptoms)
In this multivariate analysis of variance (ANOVA), the comparative treatment results of efficacy on individual symptoms were adjusted for potential effects of data source quality (e.g. values fully or only partly documented in the medical record) and of the UICC disease stage. In most cases, the general effect of data source quality was not significant except for nausea (mean score 3.12 vs 3.76) and restlessness (means score 3.07 vs 3.46), e.g. in both cases better results were found with fully documented data. The comparison of the multivariate adjusted treatment results with Wobe-Mugos® E versus control group resulted in the same conclusion on efficacy as from the original (unadjusted) data, except that after adjustment, also for the symptom "depression" the Wobe-Mugos® E treatment became significantly superior.
Conclusion: the data source quality is not influencing significantly (confounding) the comparison of the treatment results between Wobe-Mugos® E and the control group. The only exception is the symptom "depression" for which, after adjustment, Wobe-Mugos E became also significantly superior.

Potential effect (interaction) of oncologic basic therapy, age, and UICC disease stage on the efficacy results (symptoms)
In this multivariate analysis of variance (ANOVA), the comparative treatment results of efficacy on individual symptoms were adjusted for potential effects of the oncologic basic treatment (e.g. radio- chemo- or hormone therapy, respectively), of the UICC disease stage and patient's age. In most cases, the general effect of radiotherapy and hormone therapy on the symptoms was not statistically significant, except for some, not well defined influence of hormone therapy on the occurrence of infections.
The general effect of the cytostatic therapy, however, resulted in significant increase of symptom score (e.g. "worsening") in the symptoms: reduced appetite, depression/asthenia, memory/concentration disorders, dyspnea at work, skin disorders/mucositis, and the occurrence of infections. A general significant effect of patients age was found only in case of reduced appetite, which was more pronounced with higher age.
The comparison of the multivariate adjusted treatment results with Wobe-Mugos® E versus control group resulted for all symptoms in the same conclusion on efficacy as with the original (unadjusted) data.
Conclusion: after multivariate adjustment of the treatment results for the potential influencing (confounding) effect of the oncologic basic treatment, the UICC disease stage and the patients age, the comparison of the treatment results between Wobe-Mugos E and the control group resulted in an identical conclusion as from the original unadjusted data.

Potential effect (interaction) of oncologic basic therapy, age, and UICC disease stage on the secondary endpoints of efficacy
The comparison of the multivariate adjusted treatment results with Wobe-Mugos® E versus control group resulted for all secondary efficacy criteria in the same conclusion on efficacy as with the original (unadjusted) data, e.g. the significant superiority of the Wobe-Mugos® E therapy.
However, generally, in all criteria, except the number of patients with side-effects, the higher UICC stage resulted in worse results (e.g. worse judgement on efficacy and safety and higher risk for ADRs). In general terms, the overall results with the secondary criteria of efficacy were significantly adversely influenced by radiotherapy (physician's judgement on efficacy, number- and severity of side-effects), by cytostatic treatment (physician's judgement on efficacy, number of patients with side-effects, number- and severity of side-effects), but only exceptionally by the therapy with hormone-antagonists (total number of side-effects).
The analysis of the interactions using the multivariate ANOVA and, alternatively, the multivariate logistic regression resulted in similar conclusions, proving the reliability of these results.
Conclusion: the multivariate adjustment of the treatment results, as expressed by the secondary criteria of efficacy (physician's judgement and adverse drug reactions), for the potential influencing (confounding) effect of the oncologic basic treatment, the UICC disease stage and the patients age, resulted in an identical conclusion on the superiority of Wobe-Mugos E as from the original unadjusted data.