Pelvic floor disorders in women with gynecologic malignancies: a systematic review

Introduction

Approximately 14.5 million cancer survivors in the USA were reported in 2014, and this number is expected to increase to 19 million by 2024 [1]. As the number of cancer survivors continues to increase, the quality of life (QOL) of these survivors is an important consideration for healthcare providers. In 2016, an estimated 105,890 women in the USA were diagnosed with a gynecologic malignancy, including endometrial, cervical, ovarian/fallopian tube/primary peritoneal, vaginal, and vulvar cancer [2]. Major advances in current radiotherapy, chemotherapy and surgical treatments offer the chance of enhanced long-term survival in these women [3, 4]. With such improvements in survival, the morbidity of therapy and its impact on QOL must increasingly be considered when evaluating treatment options. Gynecologic cancer therapy often includes multiple treatment modalities, which alone and in combination have the potential to affect the pelvic organs and lead to pelvic floor disorders (PFDs). Yet, the actual prevalence of PFDs in women with specific gynecologic malignancies has not been well described.

The overall prevalence of PFDs in healthy, nonpregnant women is reported to be up to 25% [5]. Specifically the prevalences of urinary incontinence (UI), fecal incontinence (FI), and symptoms of pelvic organ prolapse are 15.7%, 9%, and 2.9%, respectively [6]. The prevalence of PFDs is known to increase with age: 31.6% of women aged 50–59 years were noted to have at least one or more PFDs [5]. Despite PFDs being prevalent, especially in older women, with significant negative effects on QOL, the impact of gynecologic oncologic treatment on the prevalence of PFDs is not well understood and is likely infrequently discussed with oncologic patients [7].

There has not yet been a comprehensive systematic literature review investigating the rates of PFDs in women who have undergone different types of treatment for gynecologic malignancies. Our aims in this systematic review were: (1) to investigate the prevalence of PFDs in cancer survivors after surgical and nonsurgical treatments for cervical, uterine, ovarian, and vulvar cancer; and (2) to compare rates of PFDs among gynecologic cancer survivors and the general population. We hypothesized that the prevalence of PFDs in the gynecologic cancer population would exceed that in the general female population.

Materials and methods

The University of Maryland Institutional Review Board provided exemption for this research. An expert in systematic reviews conducted structured searches of the following databases: PubMed (1809 to present), the Cochrane Central Register of Controlled Trials (Wiley), and Embase (Embase.com, 1974 to present). All searches were updated in May 2017. The combined searches yielded 550 unique citations.

The Population, Intervention, Comparison, Outcomes and Study Design (PICOS) approach was used to set parameters for this systematic review. The population inclusion criteria were as follows: women with any gynecologic cancer (uterine, ovarian, fallopian tube, primary peritoneal, cervical, vulvar, and/or vaginal) before or after any oncologic treatment with data on prevalence of PFDs. Women without a gynecologic malignancy and those without any clinical data on specific gynecologic cancer were excluded. While our search included all gynecologic malignancies, this systematic review only includes data on cervical, uterine, ovarian and vulvar as the search did not yield any studies that met the inclusion criteria for fallopian tube, vaginal or peritoneal cancer. Our primary outcome of interest was to evaluate the prevalence of PFDs in those who received any intervention for any type of gynecologic malignancy. Based on the definitions established by the Standardization and Terminology Committees of the International Urogynecological Association (IUGA) and the International Continence Society (ICS) [8], a comprehensive list of approximately 100 PFD terms was compiled as search terms (Appendix A). Studies of any design were included. References retrieved had at least one term from each of three concepts: (1) PFD terms, (2) gynecologic cancer terms, and/or (3) prevalence terms. Search terms were adjusted slightly for each database, and search strategies incorporated both subject headings and text words. No date restriction was applied.

As shown in Fig. 1, 550 studies met the criteria for retrieval. Abstracts were doubly screened by two reviewers, with each team consisting of two independent reviewers; discrepancies were adjudicated by the principal investigator. An additional 20 studies were included based on review of references and expert opinion. After screening of abstracts, two reviewers evaluated the full text of 74 articles, and a total of 31 articles were included in the review. The methodologic quality of each study was assessed using predefined criteria from a three-category system modified from the Agency for Healthcare Research and Quality. Studies were graded as of good (A), fair (B), or poor (C) quality based on the likelihood of biases and the completeness of reporting [9]. This systematic review was registered with the PROSPERO international database of prospectively registered systematic reviews prior to data extraction. The MOOSE checklist and guidelines for reporting were used for this systematic review of observational studies [10].

Results

Of the 550 studies retrieved, 31 met the criteria for systematic review including 17 on cervical cancer, 10 on endometrial cancer, 4 on ovarian cancer, and 8 on vulvar cancer. The study selection flow chart is shown in Fig. 1. Most studies were cross-sectional in design, and the overall quality of the studies was poor, with the majority grade C. Treatment of gynecologic malignancies may include various combinations of surgery, external beam radiotherapy (EBRT), chemotherapy, and brachytherapy based on the extent of disease and other patient characteristics. The prevalence of major PFDs before and after treatment in survivors of major gynecologic malignancies is summarized in Table 1. The prevalence of PFDs according to cancer type is discussed in the following sections.

Discussion

This systematic review summarizes the prevalence of PFDs in survivors of gynecologic malignancies after treatment with various modalities. The prevalence of PFDs has broad range in cervical and endometrial cancer survivors due to differences in reported measures, type and timing of treatment. Data on ovarian and vulvar cancer are limited. Uterine cancer survivors have significantly higher rates of decreased sexual desire, and entry and deep dyspareunia 24 months after surgery than healthy women recruited in a gynecologic outpatient clinic. While our study suggests that all PFDs are highly prevalent in cancer survivors, the lack of comparative studies specific to the type of gynecologic cancer limits further comments on the differences in PFD prevalence between cancer survivors and the general gynecologic population..

The prevalence of UI in the general population ranged from 3.5% to 38.2% and FI from 2.6% to 21% [5]. In comparison, in patients with cervical cancer, the prevalence of SUI ranged from 4% to 76%, UUI from 3% to 59%, and FI from 2% to 34% after treatment. While it is well-established that the prevalence of PFDs increases with age in the general population [5], factors that affect rates of PFDs in the gynecologic cancer population are affected by the timing and type of treatment received. In patients who undergo surgical management of cervical cancer, UI rates tend to be highest in the early post-operative period. Ceccaroni et al. found a UI rate of 55% approximately 1 month after surgery [23], while Naik et al. found UI rates of 48% and 17% 1.5 months and after 12 months after surgery, respectively [14].

As advances in treatment improve survival in cancer survivors, the impact of specific treatment modalities on QOL is an important aspect of pretreatment counseling. While high-quality comparative data are limited, Ceccaroni et al. [23], Hazewinkel et al. [24], and Bergmark et al. [26] evaluated the impact of various treatment types on the prevalence of PFDs in cervical cancer survivors. Ceccaroni et al. suggest that the type of surgery affects the prevalence of PFDs. Nerve-sparing surgical techniques tended to be associated with lower rates of SUI, UUI and UR than the non-nerve-sparing approach [23]. In studies evaluating different modalities of treatment, patients with cervical cancer who underwent radiotherapy alone or in combination with other forms of treatment had higher rates of urinary, bowel and sexual dysfunction. In a cross-sectional study investigating PFDs up to 132 months after treatment, Hazewinkel et al. found that patients who underwent radiotherapy alone or in conjunction with surgery had higher rates of UUI, urinary frequency, FI, FU, and straining than those who underwent surgery alone [24]. Similarly, Bergmark et al. [26] found higher rates of hypoactive sexual disorder in patients who underwent radiotherapy (76%) than in those who underwent surgery only (35%). Pretreatment counseling by clinicians should consider changes over time, as radical surgical interventions result in immediate symptoms with the potential for improvement with time, whereas radiation tends to elicit late effects that manifest over many months.

In patients with endometrial cancer, long-term follow-up of the randomized controlled PORTEC trials has shed some light on the prevalence of PFDs. Specifically, in the follow-up study to PORTEC-2, a comparative study in which patients underwent surgery followed by either brachytherapy or radiotherapy; both treatment modalities were associated with high rates of urinary dysfunction, urinary frequency, dyspareunia and VD at 84 months after treatment [33]. Comparative data on sexual dysfunction between uterine cancer survivors and the general population are minimal. A national study on adult sexual behavior in the USA showed a 43% prevalence of sexual dysfunction in the general female population [47]. Aerts et al. [36] found that patients with endometrial cancer experienced higher rates of sexual dysfunction than the general population: entry dyspareunia in 36% and deep dyspareunia in 18% of patients with endometrial cancer 2 years after surgery compared with only 4% and 2% in the general population, respectively. Similarly, in a comparative study by Rutledge et al. [7], survivors of gynecologic malignancies reported significant problems with sexual dysfunction, including lower rates of desire and less ability to climax than patients with benign gynecologic conditions. However, rates of sexual dysfunction were not reported in relation to cancer type. While it is evident that sexual dysfunction is a significant problem in survivors of gynecologic malignancies, future studies should compare appropriately matched healthy subjects and cancer patients stratified by gynecologic cancer type.

The data on PFDs in ovarian cancer survivors are extremely limited, probably because the majority of patients present with stage III/IV disease which is associated with 5-year survival rates of only 20–35% [48]. There were no posttreatment studies identified by this review that reported on rates of urinary or bowel dysfunction. Carmack Taylor et al. [40] evaluated sexual dysfunction in ovarian cancer survivors. While the type of treatment received and time from treatment are unclear, extremely high rates of dyspareunia, VD and OD were noted [40].

The current literature on vulvar cancer focuses mostly on posttreatment urinary and bowel dysfunction with little or no mention of sexual dysfunction. The most commonly reported PFDs among patients with vulvar cancer were UI and FI, although several patients did experience spraying or splitting on urination. In general, the rates of PFDs were higher in patients who underwent surgery in addition to EBRT than in those who underwent surgery alone [44, 45]. But even varied surgical techniques alone led to differing rates of PFDs, with differences being seen mostly in rates of UI [42, 44, 45]. In the only study investigating sexual dysfunction, Höckel et al. [43] found body image and sexual life disturbances in 13% of patients with vulvar cancer after surgery.

While studies investigating the prevalence of PFDs after treatment predominate in the literature, it is important to consider baseline prevalence prior to any interventions. Bretschneider et al. [11] and Thomas et al. [12] determined the presence of urinary and bowel dysfunction in patients with cervical, endometrial, ovarian and vulvar cancer. While the studies had small sample sizes, the reported baseline prevalences were comparable to the prevalences of UI and FI in the general population, which further suggests the need to evaluate the impact of treatment and disease progression on the evolution of PFDs in this population.

We aimed to compare the prevalence of PFDs in gynecologic cancer survivors to the prevalence of PFDs in the general population. Aerts et al. [36] found that uterine cancer survivors 24 months after surgical treatment had decreased sexual desire and higher rates of entry dyspareunia than patients with benign gynecologic conditions and healthy controls, respectively. Women with uterine cancer also had higher rates of deep dyspareunia than healthy controls. There were no other studies that evaluated the prevalence of PFDs in gynecologic cancer survivors by cancer type in comparison with the rates in the general population. Although Rutledge et al. [7] did compare the prevalence of PFDs in women with ovarian, endometrial, cervical, and vulvar cancers, we were not able to separate the data according to each individual type of gynecologic cancer. Among cancer survivors in this study, the most prevalent gynecologic cancer was endometrial (45%), followed by ovarian (29%) and cervical (22%). The most common intervention was surgery (87%), followed by radiotherapy (35%) and chemotherapy (35%). The presence of PFD symptoms before the diagnosis of cancer was reported by 25% of the survivors of cancer. Most of the patients with cancer (62%) had completed their treatment within the previous 1–4 years. Of control patients and survivors, respectively, 56% and 70% reported any UI, and 26% and 42% reported moderate-to-severe UI (both p > 0.05). The authors reported a higher rate of FI (43%) in cancer survivors than in the general gynecologic population in the clinic (32%; p = 0.02).

Donovan et al. [49] estimated the prevalence of bladder and bowel symptoms in survivors of endometrial and cervical cancer in comparison with the prevalence in women without cancer. In a cohort of 104 survivors matched with the same number of controls, 97% underwent surgery, 30% underwent chemotherapy and 52% underwent radiotherapy. Significantly higher rates of SUI (73% vs. 50%), UUI (72% vs. 49%), nocturia (46% vs. 24%), and urgency (83% vs. 52%) were found in cancer survivors (p ≤ 0.02). Yeoh et al. [50] evaluated anorectal function using questionnaire and manometry data in 15 patients who had received radiotherapy 5 to 10 years previously for cervical and uterine cancer. Of these 15 patients, 67% had FU and 33% had FI, and these rates were higher than in normal control patients. Of the nine control patients, one (11%) had FU and none had FI.

The strengths of this review include broad search terms, transparency in data reporting and a systematic review of available literature. These factors contribute to its robust methodological approach. There are, however, several limitations. As with any systematic review, we were limited by the available literature and the overall poor quality of published data on this topic. While there are numerous published studies, there are no high-quality studies. Most studies were cross-sectional or retrospective in design, so it is difficult to comment on the prevalence of PFDs before and after treatment. Also, there was no consistency in the type of validated tools used for objective assessment of PFDs. Many studies used nonvalidated questionnaires, so that the reliability and reproducibility of the data published remain in question. Furthermore, studies were excluded if the reported prevalences of PFDs were not separated according to specific gynecologic cancer types. The lack of comparative studies evaluating the impact of specific treatment types on PFDs is another significant limitation. Because the type of treatment modality is dependent on cancer type and stage, many patients may receive multiple modalities of treatment, so it is difficult to comment on the impact of specific treatment modalities on the development of PFDs.

In conclusion, PFDs are widely prevalent in gynecologic cancer survivors and should be addressed with patients. Future studies should focus on prospective assessment of baseline prevalence and the impact of specific oncologic treatment modalities on PFDs using validated questionnaires. Our systematic review underscores the need for more comparative data on PFDs among gynecologic cancer survivors and the general population. Future research should focus on improving QOL in cancer survivors using oncologic treatments that cause less damage to the pelvic floor and lead to lower rates of PFDs. Gynecology oncologists should consider incorporating screening for PFDs at the time of diagnosis of cancer prior to initiation of therapy and should continue monitoring PFDs during treatment to determine the ideal time for intervention if clinically indicated. Early recognition and diagnosis of PFDs in gynecologic cancer survivors would allow comprehensive care in these women, which can help ameliorate morbidity from therapy and improve overall QOL.