Surgical vs Progestin Treatment for Severe Deep Dyspareunia
Surgical vs Progestin Treatment for Severe Deep Dyspareunia
Our findings suggest that, overall, surgery and low-dose progestin therapy are both effective in improving sexual functioning, psychological well-being and health-related quality of life in women with endometriosis-associated deep dyspareunia, although with a different chronological pattern. Excision of lesions is followed by a rapid and substantial benefit, with a greater immediate effect with respect to progestin treatment. However, the gradual recurrence of pain at intercourse after 6 months since surgery, combined with progressive amelioration of clinical conditions in women using NETA, lead to a final practical equipoise in the considered functional outcomes at 1-year follow-up.
We were unable to discriminate between patients who underwent incomplete first-line surgery and those in whom new lesions developed after radical excision. In fact, the majority of the study participants underwent primary surgery elsewhere. This limits a reliable evaluation of the actual degree of radicality at primary surgery and, thus, the proper identification of persistent versus recurrent lesions. Therefore, we are unable to demonstrate that our findings may apply to both subgroups of patients, i.e. those who underwent incomplete surgery, and those in whom recurrences developed de novo after actual radical excision.
Given the complete lesion excision obtained at second-line surgery, it appears reasonable to ascribe pain relapse in women in the surgery group to endometriosis recurrence, as removal of lesions does not interfere with the pathogenesis of the disease. However, a psychogenic component cannot be excluded, as after surgery, patients may expect worsening of symptoms over time anyway. On the other hand, the progressive effect observed in progestin users is probably related to a gradual decrease in pelvic inflammatory status (Tokushige et al., 2009).
In both groups, the mean total FSFI score never exceeded the 'normality' cut-off limit of 26.55 at any time point (Wiegel et al., 2005). Therefore, the sexual dysfunction observed at baseline was completely corrected neither by surgery nor by medical treatment. In particular, performance in the surgery group improved substantially in the short term, then gradually deteriorated. However, at 12-month of evaluation, desire, arousal and lubrication scores were still significantly higher than in the NETA group, resulting in a tendency towards better total FSFI score after surgery. This could also reflect the adverse impact of progestins on libido (Vercellini et al., 2002, 2005). In other words, women using NETA probably experience a drug-induced decrease in desire, arousal and lubrication, but once intercourse initiates, orgasm capacity and overall satisfaction are not worse than in the surgical group. Conversely, 1 year after surgery, women maintained an unaffected sexual predisposition, but experienced slightly more pain during intercourse.
We evaluated possible treatment effect modifications by selected variables by including product terms in the model. Despite a few statistically significant differences, we looked at mean effect over time in stratified analyses and observed results that, with the exception of the effect of NETA on women without rectovaginal lesions (Vercellini et al., 2012), are probably of modest clinical relevance or potentially due to random fluctuation of data (e.g. a slightly higher benefit of surgery on desire in women without ovarian endometriomas). Moreover, we feel that the generalizability of this subgroup analyses is somewhat uncertain.
Anxiety and depression constitute important factors in the evaluation of sexual problems in women with chronic pelvic pain. In fact, it has been demonstrated that both psychological conditions mediate the effect of pain on sexual disorders (ter Kuile et al., 2010). We observed a remarkable improvement in overall psychological status and in anxiety and depression subscales in women in both study groups. In the present study, only women suffering from severe limitation in sexual activity were included. Living with this symptom may result in more frustrating in comparison with dysmenorrhoea that is usually associated with less important psychological consequences (Vercellini et al., 2002, 2005). This could explain the high HADS scale scores reported by our patients at baseline. However, it is not possible to exclude that other forms of pelvic pain could have influenced the anxiety and depression scores throughout the study period.
Endometriosis is associated with debilitating pelvic pain, impairment of psychological as well as social functioning and reduction in quality of life as well as sexual satisfaction (Tripoli et al., 2011). The use of a disease-specific instrument to assess the health-related quality of life burden of endometriosis has been repeatedly suggested (Khong et al., 2010; Vincent et al., 2010). The EHP-30 questionnaire is a user-friendly self-report tool suitable for use in endometriosis-related clinical research (Khong et al., 2010). According to literature data, both surgical and pharmacological treatments improve patient's physical and psychological functioning, vitality, pain level and general health (Gao et al., 2006). In our study, variations of the EHP-30 questionnaire scores in both study groups followed a substantially similar temporal pattern with respect to the other considered study outcomes. This further emphasizes the importance of being relieved from dyspareunia and the possibility to resume a better sex life in the overall quality of a woman's life.
We recruited only women with severe pain at intercourse. In this regard, the findings of our study are difficult to compare with literature data, as in most reports variation in dyspareunia was a secondary study outcome, and in the few in which dyspareunia was a selection criterion, women with moderate pain were also recruited. Overall, we confirm the results observed by Ferrero et al. (2007a) on the effect of radical surgery for endometriosis in women with moderate to severe deep dyspareunia. Also in their study, reduction in pain at intercourse and improvement in quality of sex life resulted more marked in patients with deep lesions.
We used NETA in a low dose. Other investigators have suggested that higher doses are required to obtain pain relief (Muneyyirci-Delale and Karacan, 1998; Kaser et al., 2012). Therefore, it is certainly possible that had higher doses been used, a larger effect could have resulted. Other medical therapies have been demonstrated to benefit women with endometriosis-associated deep dyspareunia. In a randomized controlled trial, a GnRH agonist was more effective than an estrogen–progestin on pain at intercourse (Vercellini et al., 1993), and the combination of NETA with an aromatase inhibitor (letrozole) reduced deep dyspareunia to a larger extent than NETA alone in some women with rectovaginal endometriosis (Ferrero et al., 2009). Moreover, Ferrero et al. (2007b) suggested that the best results can be obtained with surgery, followed by post-operative medical treatment. The findings of our study appear to support this view, and the combination of surgical and long-term adjuvant pharmacological therapy deserves further research.
The most important methodological drawback of our study is the non-random allocation of treatments. The choice of undergoing surgery or using a progestin was based on patients' preference, and the unbalanced distribution between the two study arms is almost certainly due to the refusal of most women to undergo further surgery after an unsuccessful first-line procedure. In theory, a study based completely on patient preference introduces a major selection bias, thus limiting interpretation of the results. In fact, the profile of patients selecting such different options may vary in several and undefined aspects, including psychological pattern, attitude towards sexual activity and ability to cope with a stressful, painful and chronic condition. The observed similarity in baseline clinical characteristics of subjects in the two study groups may reduce, but not eliminate this potential bias. In addition, lack of statistically significant differences in the considered demographic aspects may be due to lack of power relative to the small sample size.
Moreover, the number of dropouts was unbalanced, as substantially more subjects withdrew from the progestin arm. This is partly expected, as women may easily decide to interrupt drug use even for minor problems (e.g. erratic bleeding or weight gain), whereas withdrawal from the surgery arm is usually dictated only by important reasons (e.g. development of an ovarian cyst and recurrence of severe symptoms). This could result in overestimation of the progestin effect in a 'per protocol' analysis.
In spite of the above methodological implications, the results of two systematic literature reviews demonstrated that estimates of treatment effects in well-designed cohort studies are not consistently larger than or qualitatively different from those obtained in RCT on the same topic (Benson and Hartz, 2000; Concato et al., 2000). In addition, we consider that in case of very diverse treatment alternatives associated with major differences in terms of risks and morbidity, the study design adopted by us may reveal more practical and feasible, and potentially more representative, 'real world' conditions than a formal RCT (Brocklehurst, 1997). In fact, it has been repeatedly demonstrated that setting up a RCT in these circumstances may easily lead to insurmountable recruiting difficulties (McCulloch et al., 2002; Carragee, 2006; Weinstein et al., 2006a, 2006b; Crowther et al., 2012).
Women are more and more willing to be informed on the therapeutic options and to decide which one best suits their needs. The alternative health outcome paradigm is becoming the 'goal-oriented patient care' (Reuben and Tinetti, 2012), the 'patient-centred medicine' (Bardes, 2012) and the 'shared decision making' (Barry and Edgman-Levitan, 2012). When the study outcomes are functional, as in our case, a comparison between two groups of patients who have chosen their treatment may represent the maximum effect size of the intervention (Schmoor et al., 1996).
In conclusion, surgery and low-dose oral NETA demonstrated a similar final beneficial outcome in women with endometriosis-associated deep dyspareunia in terms of improvement of sexual functioning, psychological well-being and health-related quality of life at 1-year follow-up. The findings should be considered with caution owing to lack of randomization, potential between-group heterogeneity and difference in dropout rates. Moreover, in light of the observed differences in the temporal pattern of the effect, further comparative studies with longer follow-up are warranted.
Discussion
Our findings suggest that, overall, surgery and low-dose progestin therapy are both effective in improving sexual functioning, psychological well-being and health-related quality of life in women with endometriosis-associated deep dyspareunia, although with a different chronological pattern. Excision of lesions is followed by a rapid and substantial benefit, with a greater immediate effect with respect to progestin treatment. However, the gradual recurrence of pain at intercourse after 6 months since surgery, combined with progressive amelioration of clinical conditions in women using NETA, lead to a final practical equipoise in the considered functional outcomes at 1-year follow-up.
We were unable to discriminate between patients who underwent incomplete first-line surgery and those in whom new lesions developed after radical excision. In fact, the majority of the study participants underwent primary surgery elsewhere. This limits a reliable evaluation of the actual degree of radicality at primary surgery and, thus, the proper identification of persistent versus recurrent lesions. Therefore, we are unable to demonstrate that our findings may apply to both subgroups of patients, i.e. those who underwent incomplete surgery, and those in whom recurrences developed de novo after actual radical excision.
Given the complete lesion excision obtained at second-line surgery, it appears reasonable to ascribe pain relapse in women in the surgery group to endometriosis recurrence, as removal of lesions does not interfere with the pathogenesis of the disease. However, a psychogenic component cannot be excluded, as after surgery, patients may expect worsening of symptoms over time anyway. On the other hand, the progressive effect observed in progestin users is probably related to a gradual decrease in pelvic inflammatory status (Tokushige et al., 2009).
In both groups, the mean total FSFI score never exceeded the 'normality' cut-off limit of 26.55 at any time point (Wiegel et al., 2005). Therefore, the sexual dysfunction observed at baseline was completely corrected neither by surgery nor by medical treatment. In particular, performance in the surgery group improved substantially in the short term, then gradually deteriorated. However, at 12-month of evaluation, desire, arousal and lubrication scores were still significantly higher than in the NETA group, resulting in a tendency towards better total FSFI score after surgery. This could also reflect the adverse impact of progestins on libido (Vercellini et al., 2002, 2005). In other words, women using NETA probably experience a drug-induced decrease in desire, arousal and lubrication, but once intercourse initiates, orgasm capacity and overall satisfaction are not worse than in the surgical group. Conversely, 1 year after surgery, women maintained an unaffected sexual predisposition, but experienced slightly more pain during intercourse.
We evaluated possible treatment effect modifications by selected variables by including product terms in the model. Despite a few statistically significant differences, we looked at mean effect over time in stratified analyses and observed results that, with the exception of the effect of NETA on women without rectovaginal lesions (Vercellini et al., 2012), are probably of modest clinical relevance or potentially due to random fluctuation of data (e.g. a slightly higher benefit of surgery on desire in women without ovarian endometriomas). Moreover, we feel that the generalizability of this subgroup analyses is somewhat uncertain.
Anxiety and depression constitute important factors in the evaluation of sexual problems in women with chronic pelvic pain. In fact, it has been demonstrated that both psychological conditions mediate the effect of pain on sexual disorders (ter Kuile et al., 2010). We observed a remarkable improvement in overall psychological status and in anxiety and depression subscales in women in both study groups. In the present study, only women suffering from severe limitation in sexual activity were included. Living with this symptom may result in more frustrating in comparison with dysmenorrhoea that is usually associated with less important psychological consequences (Vercellini et al., 2002, 2005). This could explain the high HADS scale scores reported by our patients at baseline. However, it is not possible to exclude that other forms of pelvic pain could have influenced the anxiety and depression scores throughout the study period.
Endometriosis is associated with debilitating pelvic pain, impairment of psychological as well as social functioning and reduction in quality of life as well as sexual satisfaction (Tripoli et al., 2011). The use of a disease-specific instrument to assess the health-related quality of life burden of endometriosis has been repeatedly suggested (Khong et al., 2010; Vincent et al., 2010). The EHP-30 questionnaire is a user-friendly self-report tool suitable for use in endometriosis-related clinical research (Khong et al., 2010). According to literature data, both surgical and pharmacological treatments improve patient's physical and psychological functioning, vitality, pain level and general health (Gao et al., 2006). In our study, variations of the EHP-30 questionnaire scores in both study groups followed a substantially similar temporal pattern with respect to the other considered study outcomes. This further emphasizes the importance of being relieved from dyspareunia and the possibility to resume a better sex life in the overall quality of a woman's life.
We recruited only women with severe pain at intercourse. In this regard, the findings of our study are difficult to compare with literature data, as in most reports variation in dyspareunia was a secondary study outcome, and in the few in which dyspareunia was a selection criterion, women with moderate pain were also recruited. Overall, we confirm the results observed by Ferrero et al. (2007a) on the effect of radical surgery for endometriosis in women with moderate to severe deep dyspareunia. Also in their study, reduction in pain at intercourse and improvement in quality of sex life resulted more marked in patients with deep lesions.
We used NETA in a low dose. Other investigators have suggested that higher doses are required to obtain pain relief (Muneyyirci-Delale and Karacan, 1998; Kaser et al., 2012). Therefore, it is certainly possible that had higher doses been used, a larger effect could have resulted. Other medical therapies have been demonstrated to benefit women with endometriosis-associated deep dyspareunia. In a randomized controlled trial, a GnRH agonist was more effective than an estrogen–progestin on pain at intercourse (Vercellini et al., 1993), and the combination of NETA with an aromatase inhibitor (letrozole) reduced deep dyspareunia to a larger extent than NETA alone in some women with rectovaginal endometriosis (Ferrero et al., 2009). Moreover, Ferrero et al. (2007b) suggested that the best results can be obtained with surgery, followed by post-operative medical treatment. The findings of our study appear to support this view, and the combination of surgical and long-term adjuvant pharmacological therapy deserves further research.
The most important methodological drawback of our study is the non-random allocation of treatments. The choice of undergoing surgery or using a progestin was based on patients' preference, and the unbalanced distribution between the two study arms is almost certainly due to the refusal of most women to undergo further surgery after an unsuccessful first-line procedure. In theory, a study based completely on patient preference introduces a major selection bias, thus limiting interpretation of the results. In fact, the profile of patients selecting such different options may vary in several and undefined aspects, including psychological pattern, attitude towards sexual activity and ability to cope with a stressful, painful and chronic condition. The observed similarity in baseline clinical characteristics of subjects in the two study groups may reduce, but not eliminate this potential bias. In addition, lack of statistically significant differences in the considered demographic aspects may be due to lack of power relative to the small sample size.
Moreover, the number of dropouts was unbalanced, as substantially more subjects withdrew from the progestin arm. This is partly expected, as women may easily decide to interrupt drug use even for minor problems (e.g. erratic bleeding or weight gain), whereas withdrawal from the surgery arm is usually dictated only by important reasons (e.g. development of an ovarian cyst and recurrence of severe symptoms). This could result in overestimation of the progestin effect in a 'per protocol' analysis.
In spite of the above methodological implications, the results of two systematic literature reviews demonstrated that estimates of treatment effects in well-designed cohort studies are not consistently larger than or qualitatively different from those obtained in RCT on the same topic (Benson and Hartz, 2000; Concato et al., 2000). In addition, we consider that in case of very diverse treatment alternatives associated with major differences in terms of risks and morbidity, the study design adopted by us may reveal more practical and feasible, and potentially more representative, 'real world' conditions than a formal RCT (Brocklehurst, 1997). In fact, it has been repeatedly demonstrated that setting up a RCT in these circumstances may easily lead to insurmountable recruiting difficulties (McCulloch et al., 2002; Carragee, 2006; Weinstein et al., 2006a, 2006b; Crowther et al., 2012).
Women are more and more willing to be informed on the therapeutic options and to decide which one best suits their needs. The alternative health outcome paradigm is becoming the 'goal-oriented patient care' (Reuben and Tinetti, 2012), the 'patient-centred medicine' (Bardes, 2012) and the 'shared decision making' (Barry and Edgman-Levitan, 2012). When the study outcomes are functional, as in our case, a comparison between two groups of patients who have chosen their treatment may represent the maximum effect size of the intervention (Schmoor et al., 1996).
In conclusion, surgery and low-dose oral NETA demonstrated a similar final beneficial outcome in women with endometriosis-associated deep dyspareunia in terms of improvement of sexual functioning, psychological well-being and health-related quality of life at 1-year follow-up. The findings should be considered with caution owing to lack of randomization, potential between-group heterogeneity and difference in dropout rates. Moreover, in light of the observed differences in the temporal pattern of the effect, further comparative studies with longer follow-up are warranted.
