- Academic Editor
†These authors contributed equally.
Background: Re-irradiation of patients with previously irradiated
gynecological tumors represents one of the hot topics of modern oncology. It is
generally performed using Brachytherapy (alone or after external beam radiation
therapy (EBRT) re-treatment) or Stereotactic Body Radiotherapy (SBRT). Literature
provides few data concerning SBRT re-irradiation (re-RT). Here we provided a statistical
and comparative analysis of the studies to make a general assessment of the
efficacy and reliability of SBRT, considering the potential benefits achievable
in terms of local control, overall survival and toxicity. Methods: A
computerized literature search was performed in 3 electronic databases (MEDLINE,
EMBASE, and Cochrane) from 1996 to 2020. Only studies analysing outcomes of
re-irradiated (re-I) patients were taken into consideration. Quality assessment
score and risk of bias were assessed for each article. Random-effects models were
used due to great subjectivity given the lack of related control groups in the
non-comparative studies and a tendency towards high heterogeneity (examined by
the Cochran Q chi-square test and the I
Gynecological tumors represent 7.7% of all malignancies [1] with 2.3% and 3.3% of them respectively represented by endometrial cancer and cervical cancer.
The management of primary tumors requires a multidisciplinary approach with surgery, radiation therapy, and chemotherapy being the different kinds of treatment we can employ to achieve local control and reduce the risk of death.
Recurrence rates for endometrial and cervical cancer after surgery and adjuvant radiotherapy (RT) for endometrial cancer and after radio-chemotherapy treatment for cervical cancer is respectively about 5% and 10% with differences according to the International Federation of Gynecology and Obstetrics (FIGO) stage to they belong [2, 3].
The main treatments for local recurrences are surgery (i.e., surgical exenteration) and radiotherapy.
Literature provides few data concerning gynecological tumors re-irradiation. It must be planned and performed considering already delivered doses to both targets and organs at risk (OARs) and normal tissues’ tolerance in previous RT course. It may be performed using Brachytherapy (BT) (alone or after external beam re-irradiation) or Stereotactic Body Radiotherapy (SBRT).
To date, the treatment of pelvic recurrences, due to their proximity to radiosensitive anatomical structures (urinary tracts and the bowel), is controversial and associated with disappointing outcomes [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24].
According to recent systematic reviews, brachytherapy may be used in the management of women with recurrent gynaecologic tumors with negligible toxicities [2]. When brachytherapy is not available, SBRT represents a feasible alternative [25, 26, 27].
Stereotactic radiation therapy, in combination with Intensity Modulated Radiation Therapy/Volumetric Modulated Arc Therapy (IMRT/VMAT), permit to perform a dose escalated treatment delivering high dose within bulky tumors to intra- and extracranial lesion, minimizing side effect and obtaining better clinical outcome [28, 29, 30, 31].
SBRT may be described as a form of external beam radiation therapy (EBRT) that
accurately delivers high and conformed doses to an extracranial target through
hypofractionated radiation schemes (usually
Several reviews reported promising results in terms of local control and survival regarding use of SBRT in recurrent gynaecological cancer [38, 39].
Here we provided a statistical and comparative analysis of the studies to make a general assessment of the efficacy and reliability of SBRT, considering the potential benefits achievable in terms of local control, overall survival and toxicity.
A systematic review was conducted employing the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The computerized literature search was performed in 3 electronic databases (MEDLINE, EMBASE, and Cochrane) from 1996 to 2020.
The search strategy included terms related to “stereotactic radiation therapy”, “re-irradiation”, “reirradiation” and “gynaecological cancer”, “recurrent cervical cancer”, “recurrent endometrial cancer”.
Criteria for inclusion and exclusion:
Inclusion and exclusion criteria were predetermined. Full-text screening without
duplicate citations was carried out. Articles, including
Two studies with outcomes limited to re-irradiation patients were screened.
Included studies were retrospective, analysing more than 5 patients. Only studies analysing outcomes of re-irradiated (re-I) patients, where re-I field overlaps with that from primary RT course, were taken into consideration. Discrepancies among reviewers were solved by discussion (Fig. 1).
PRISMA flow diagram detailing how articles were selected for inclusion in the systematic review.
Review of the trials:
Studies were first reviewed using a list of predefined, pertinent issues concerning the characteristics of patients and treatments. Quality assessment score of included studies was assessed according to checklist for quality appraisal of case series studies produced by Institute of Health Economics (IHE) and modified to improve applicability (Table 1, Ref. [17, 18, 19, 20, 21, 23, 27]). Risk of bias was assessed for each article using quality assessment tool for domain of hypothesis stated, multicentric studies, prospective study, patients’ selection, consecutive patients, described characteristics of patients, clearly stated eligibility criteria, described intervention, reported losses to follow-up and adverse events. Risk-of-bias assessment resulted in a low risk of bias in all included papers (Fig. 2a,b, Ref. [17, 18, 19, 20, 21, 23, 27]).
Risk of bias assessment. (a) Author’s judgments about each risk of bias items. Risk of bias was assessed for each article using quality assessment tool for domain of hypothesis stated, multicentric studies, prospective study, patients’ selection, consecutive patients, described characteristics of patients, clearly stated eligibility criteria, described intervention, reported losses to follow-up and adverse events. (b) Risk of bias summary. Risk-of-bias assessment resulted in a low risk of bias in all included papers.
Reference | Was hypothesis stated? | Was the study prospective? | Is multicentric? | Consecutive patients | Described characteristics of patients | Clearly stated eligibility criteria | Described intervention | Reported losses to follow-up and adverse events | Conclusions of the study supported by the results | Patients’ selection | Final score | ||||||||||||||||||||
Yes | Partial | No | Yes | Partial | No | Yes | Partial | No | Yes | Partial | No | Yes | Partial | No | Yes | Partial | No | Yes | Partial | No | Yes | Partial | No | Yes | Partial | No | Yes | Partial | No | ||
Deodato et al. [23] | 2 | - | - | 2 | - | - | - | 1 | - | - | - | 0 | 2 | - | - | 2 | - | - | 2 | - | - | 2 | - | - | - | 1 | - | - | 1 | - | 15 |
Yazici et al. [21] | 2 | - | - | 2 | - | - | - | - | 0 | - | 1 | - | 2 | - | - | 2 | - | - | 2 | - | - | - | 1 | - | 2 | - | - | - | 1 | - | 15 |
Seo et al. [19] | 2 | - | - | 2 | - | - | - | - | 0 | 2 | - | - | 2 | - | - | 2 | - | - | 2 | - | - | - | 1 | - | - | 1 | - | 2 | - | - | 16 |
Park et al. [20] | 2 | - | - | 2 | - | - | 2 | - | - | 2 | - | - | - | 1 | - | 2 | - | - | 2 | - | - | - | 1 | - | 2 | - | - | 2 | - | - | 18 |
Hasan et al. [17] | 2 | - | - | - | - | 0 | - | - | 0 | - | 1 | - | 2 | - | - | 2 | - | - | 2 | - | - | 2 | - | - | 2 | - | - | - | 1 | - | 14 |
Pontoriero et al. [18] | 2 | - | - | 1 | - | - | - | - | 0 | 2 | - | - | 2 | - | - | 2 | - | - | 2 | - | - | 2 | - | - | 2 | - | - | 2 | - | - | 15 |
Hsin-Yi Cheng et al. [27] | 2 | - | - | 2 | - | - | - | 1 | - | - | - | 0 | 2 | - | - | 2 | - | - | - | 1 | - | - | 1 | - | - | 1 | - | - | 1 | - | 14 |
Overall Survival (OS) and Local Control (LC) were analysed and to improve the comparability of the different re-irradiation studies and to assess the relationship between re-irradiation and 2-year OS and 2-year LC, we calculated the estimated sub-population of Grade 3 (G3) and Grade 4 (G4) treatment-related toxicities.
To determine the pooled OS and local control at 2-year and G3–G4 toxicity rate,
an established meta-analysis technique over single and multi-arm studies was
performed. We calculated the estimated population proportion of toxicity, 2 years
overall survival, 1-year locoregional control with 95% confidence interval (CI) for each separate
study [40]. Pooled effect size aided the general evaluation of re-irradiation
risk and effect. Heterogeneity across studies was examined by the Cochran Q
chi-square test and the I
21 articles, focusing on the role of SBRT in recurrent gynaecological cancers, were identified. Only those with outcomes limited to re-I patients and specific radiotherapy techniques were included, and 7 articles fulfilled the inclusion criteria.
The selected studies, published between 2009 and 2020, counted for a total of 195 patients, 140 of whom were previously irradiated.
Patient characteristics:
The study populations were heterogeneous. Six studies included patients with primary cervical [17, 19, 20, 21, 23, 27] and endometrial cancer [17, 18, 19, 21, 23, 27]. Two studies reported a combination of patients with primary vulvar cancer [23, 27], vaginal cancer [23] and ovarian cancer [17, 21].
Primary treatment:
Patients were mainly treated with a combination of surgery, neoadjuvant/adjuvant chemotherapy, and external beam radiation therapy (range 37.5 Gy-70 Gy in 25–28 fx) with or without brachytherapy (range 15–21 Gy in 3 fractions). Data about primary radiotherapy treatment schedule were not available in two studies alone [17, 27].
Recurrent Site and treatment schedule:
The recurrence site had different locations and the main site were pelvic recurrence in two studies [18, 28], pelvic lymph node (para-iliac nodes, low para-aortic and Common iliac) [17, 20, 21]. Hasan et al. [17] and Hsin-Yi Cheng et al. [27] reported a case of cervical and vulvar recurrence, respectively.
Two cases of local progression [19, 23] and only one of distant recurrence (but not specified) [23] were observed.
The recurrence treatment was generally a radiotherapeutic approach with SBRT. Only one studies reported an association of EBRT [27]. Reirradiation dose and schedule were heterogeneous (range 15 Gy–66 Gy in 3–6 fractions).
Cumulative Biologically effective doses (Cum BED) were calculated according to the linear-quadratic model assuming an alpha/beta value of 10 Gy. In only three studies, dose to OAR were reported [18, 19, 27].
Patient characteristics and details of radiation delivery are resumed in Table 2 (Ref. [17, 18, 19, 20, 21, 23, 27]).
Author and year | Study type | Pts with P-RT (total) | Pts with P-RT (ReIrradiation) | Primary site | Treatment at primary diagnosis | P-RT technique, dose (range), franctionation | Recurrent site | Re-RT technique, dose (range), fractionation | Toxicities |
Deodato et al. [23] | R | 11 | 6 | Cervix/Endometrium/Vagina/Vulva | Surgery: CT: CTRT: CT + CTRT: Surgery + CT: Surgery + CT + RT: RT + BT | 37.5 Gy, 50 Gy, 65 Gy, 50.4 Gy, 63 Gy, 45 Gy (Median dose 51.82 Gy) | Local and distant recurrence (not specified) | SBRT 30 Gy/6 fx or 25 Gy/5 fx or 20 Gy/4 fx | 0 |
Yazici et al. [21] | R | 16 | 11 | Cervix, Endometrium, Ovary | S: S + EBRT/S + CTRT/CTRT + BRT/S + CT | EBRT 50.4 Gy (range 45–60 Gy) with BT doses until to a 85–90 Gy) | pelvic wall, para-iliac nodes, low para-aortic | SBRT 15–40 in 3–5 fx | G3 toxicities: 19% (3) proctitis; G4 toxicities: fistula 19% (3) |
Seo et al. [19] | R | 23 | 23 | Uterus, Cervix | Surgery: Surgery + EBRT: Surgery + EBRT + BT: EBRT + BT | EBRT 40–70 Gy | Local progression | SBRT 36–45/3 fx | G4 toxicities: retto vaginal fistula 13% (3) |
Park et al. [20] | R | 85 | 68 | Cervix | Surgery + RT/CTRT | IMRT, 50,4 Gy (28 fx) | Paraortic, Common iliac, external and internal iliac nodes | SBRT 39 Gy in 3 fx (median) | G3 toxicities: 3% (1) urethral stricture, 3% (1) ileus, 3%(1) enterocolitis; G4 toxicities: 6% (2) rectovaginal fistula |
Hasan et al. [17] | R | 30 | 13 | Cervix, Endometrium, Ovary | OP |
NR | Cervix, vaginal vault, pelvic nodes, PAN, distant metastases | SBRT median dose 27.5 Gy (15–40)/3–5 fx | 7.7% (1) rectovaginal/vesicovaginal fistulas |
Pontoriero et al. [18] | R | 5 | 5 | Endometrium | EBRT + BT | EBRT 45 Gy in 25 fx + BT 15 Gy in 3 fx | pelvic recurrence | SBRT 18 Gy in 3–4 fx | 0 |
Cheng et al. [27] | R | 25 | 14 | Endomethrium/Cervix/Vulva | OP/NCT + OP/OP + ACT/RT | NR | Central (3), Central extending to pelvic side wall (6), Pelvic wall (1), vulva (1) | 45 Gy in 25 fx + SBRT 25 Gy in 5 fx | Diarrhea (7%), Hemorrhage (7%), Fistula (14%), Sigmoid Perforation (7%) |
R, retrospective; P-RT, previous radiotherapy; re-RT, re-irradiation; pts patients; fx, fractions; adj, adjuvant; OP, operation; RT, radiotherapy; CT, chemotherapy; CTRT, chemoradiotherapy; NCT, Neoadjuvant chemotherapy, ACT, Adjuvant chemotherapy, BRT, external beam radiotherapy; BT, brachytherapy; SBRT, stereotactic body radiotherapy; NR, Not reported; IMRT, Intensity Modulated Radiation Therapy; EBRT, external beam radiation therapy; S: Surgery; G3, Grade 3; G4, Grade 4; EBRT, external beam radiation therapy; PAN, Para-aortic nodes.
*Related to re-irradiated patients.
Outcomes were evaluable for 140 patients.
4 studies were analysed for 2-year OS [17, 19, 20, 21] and 3 for 2-year LC [17, 19, 20] respectively. The median 2-year OS was of 46.5% in 115 patients evaluated; 2-year LC was analysed in 104 patients with a median of 65%.
Data on G3 toxicity were available in 129 patients [17, 18, 19, 20, 23, 28]. The most frequent G3 toxicities were urethral stricture (3%), enterocolitis (3%), fistula (14%), sigmoid perforation (7%), recto-vaginal fistula (3%), enterovaginal/vescicovaginal fistula (7.7%), genitourinary (not specified). Only two cases of G4 toxicity (recto-vaginal fistula) was registered (19%) [19, 20].
The toxicities related to doses to the OARs are tabulated in Table 3 (Ref. [18, 19, 27]).
References | Rectum | Bladder | Bowel loop | Late toxicities Grade |
Seo et al. [19] | D5cc |
Dmax 3–35 Gy | Dmax 7–30 Gy | retto vaginal fistula (13%) |
V40 |
(median 13 Gy) | (median 20 Gy) | ||
GTV |
||||
Pontoriero et al. [18] | Dmax 13 Gy | Dmax 5 Gy | Dmax 5 Gy | 0 |
Cheng, et al. [27] | D2cc-EQD2 68.82 (median, Gy) | D2cc-EQD2 73.1 (median, Gy) | D2cc-EQD2 66.86 (median, Gy) | Diarrhea (7%), Hemorrhage (7%), Fistula (14%), Sigmoid Perforation (7%) |
EQD2, Equivalent dose in 2; D2cc, dose received by 2 cc of OAR; D5cc, dose received by 5 cc of OAR; GTV, gross tumor volume; V40, volume of rectum receiving 40 Gy.
With a median follow up time of 14.5 months, using SBRT re-I technique, the pooled 2-year OS of 52.7% (95% CI: 0.372 to 0.651) and 2-year LC of 75.7% (95% CI: 0.614 to 0.852) were observed (Fig. 3a,b, Ref. [17, 18, 19, 20, 21, 23, 27]).
Pooled 2-year OS. (a) Pooled objective 2 years overall survival in re-irradiated patients for included studies: pooled 2-year OS of 52.7% (95% confidence interval (CI): 0.372 to 0.651) was observed; (b) Pooled objective 2 years local control in re-irradiated patients for included studies: pooled 2-year LC was of 75.7% (95% CI: 0.614 to 0.852). OS, Overall Survival; CI, Confidence interval.
SBRT re-irradiation technique does not affect toxicities with pooled
Pooled
Re-irradiation for gynaecological tumors has traditionally been associated with high morbidity. With the development of image-guided techniques as brachytherapy and stereotactic radiotherapy, it’s now possible to deliver higher and conformal dose, reducing exposure of normal tissues.
SBRT may be considered as an alternative to brachytherapy for local symptom control [42].
Re-irradiation with brachytherapy results in relatively reasonable local control and toxicities. Local control ranged from 44% to 88% over 1 and 5 years; OS from 39.5 e 82% at 2 and 5 years; late G3 and 4 toxicity varied very broadly from 0% to 42.9% [2].
Cyberknife system, with prescription to specific isodose lines, precise targeting and delivery of radiation, allow to deliver inhomogeneous dose and dose escalation respecting the normal tissue tolerance of OAR, improving local control and toxicity profiles [25]. On the other hand, Hsieh et al. [26] suggested the possibility of replacing brachytherapy with SBRT administered through helical tomotherapy.
Brachytherapy is supposed to be used for recurrences involving the primary tumor site and near ones (such as cervix, parametria and vagina) thanks to high dose gradients with major dose distribution to the site of interest and minor spread to the surrounding structures.
SBRT may be used for both local and distant recurrences, being the preferrable alternative especially in case of pelvic ones with high tumor burden and nodal ones, considering the possibility of target’s intra-/interfraction motion. Fiducial tracking system contributes for accurate patient positioning and targeting ensuring tracking of target translation and rotational movements [43]. Some authors suggest use of SBRT with endovaginal device as “fiducials” when patients have clinical conditions that do not allow to deliver BT. Thus, SBRT represent an alternative to BT delivering high doses of external radiation resembling BT dose distribution [18].
The evidence available about SBRT re-irradiation of pelvic tumors are based on small single institutions and some studies were not considered for statistical analysis due to the high heterogeneity.
Considering adjacent normal organs, the dose that can be administered may vary depending on the site of recurrence. In selected studies recurrence site and relative re-irradiation doses were different and heterogeneous. Distant progression (not specified) were retreated with doses ranged between 20–30 Gy in 4–5 fx in [23] and 36–45 Gy/3 fx [19]. In different lymph nodes metastases were delivered doses of 15–40 Gy/3–5 fx or 39 Gy/3 fx [20, 21]. Cervix, vaginal vault, pelvic wall and vulva were retreated with a range of 15 Gy–40 Gy [17, 28]. Only Pontoriero et al. [18] define recurrence site in pelvic recurrence re-irradiated delivering 18 Gy in 3–4 fractions. More detailed data on cumulative radiation doses delivered on OAR, which can be translated in retreatment setting, derives from studies on radiation boost after external beam irradiation with curative purpose [28, 30].
According to these studies, many factors may affect the probability of local control after re-I, such as SBRT re-irradiation dose, volume recurrence [19, 42] and histology [22, 44].
It has been shown that increasing the dose delivered to the tumor may improve
local control. In Park et al. [20] achieving a BED
Volume recurrence (cc) is related to better prognosis, both in term of local control and overall survival [19, 42]. A longer 2-year OS and 2 year local progression free survival (LPFS) in patients with small volume (89%, p = 0.0001 and 85%, p = 0.0199, respectively) were observed [19]. In Reshko et al. [44] a local control resulted slightly inferior in larger tumors (hazard ratio (HR) 1.01, 95% CI: 1.00–1.01, p = 0.02). When average tumor was 18 cc, a complete response rate was observed, than higher volume that are associated with progression of disease [42]. Dewas et al. [22] reported 1-year LC rates of 51.4% in 15/16 patients who receiving SBRT re-irradiation for pelvic recurrence of different histology of the primary lesion.
Adenocarcinoma are associated with higher chances of recurrence and worse OS if compared with other histology (HR 1.66, 95% CI: 1.03–2.68, p = 0.038) [42]. Conversely, Dewas [22] showed a better LC of adenocarcinoma than squamous carcinoma (p = 0.09).
Deodato et al. [23] reported acceptable outcomes of SBRT in 11 patients with locally recurrent gynaecological cancer (uterine cervix [n = 4], ovary [n = 4] and endometrial [n = 3]) with 2-OS of 63.6% and 1-year LC of 93%.
Two Korean groups reported results in patients treated with SBRT for recurrent or metastatic uterine cervical cancer with OS at 2 years of 43% and 57.5% respectively [19, 20].
In Park, 68 out of 85 patients were treated with SBRT for recurrent or
metastatic uterine cervical cancer. With a median follow up of 20.4 months a 5y
OS 32.9% were reported and only 5 patients showed toxicities
In our analysis, pooled 2-year OS and 2-year LC were of 52.7% and 75.7%.
The main concern of salvage SBRT is the potential severe late toxicity.
Retrospective studies demonstrate that despite high cumulative dose were delivered with SBRT after a previous external beam radiation with or without brachytherapy, the toxicity was rather low.
In our analysis, few studies reported late toxicities related to re-irradiated patients.
In Pontoriero et al. [18] 60% showed Grade 1 (G1) toxicities (two patients genitourinary (GU) and only one gastrointestinal (GI)); 20% showed Grade 2 (G2) Cystitis and diarrhea; Hsin-Yi Cheng et al. [27] reported 14% GU late toxicities (Grade 1) and 3 patients genitourinary gastrointestinal and legs edema (Grade 2).
In Yazici et al. [21], sixteen patients with recurrent
gynaecological cancer treated with mixed approach, showed a grade
In Pontoriero et al. [18], patients were retreated with median
cumulative Equivalent dose in 2 (EQD2) Gy fractions of 85 Gy; after a median
follow-up of 12 months (range, 8–34 months), no severe (
More recently, Reshko et al. [44] and Hsin-Yi Cheng et al.
[27] showed that SBRT re-irradiation in feasible with acceptable toxicities in
patients unable to have brachytherapy. Our pooled analysis showed
According to our review, SBRT re-irradiation technique seems to be feasible and safe, when brachytherapy re-RT technique is not available.
The data underlined high heterogeneity and further studies, randomized clinical trials, prospective studies and analysis of large real-world high-quality datasets are warranted to standardize the best radiation therapy in recurrent gynaecological cancer.
AP and Sper designed and coordinated the study and the analysis. ABos, CC, GFerr, FC, VZ performed the literature search and selected included studies. PC revised the literature search and performed data extraction. GFeri, SPar and CS checked data extraction. AP, PC and ABro performed statistical data analysis and provided graphics. AP and PC drafted the manuscript. AP and SPer critically revised the manuscript. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work to take public responsibility for appropriate portions of the content and agreed to be accountable for all aspects of the work in ensuring that questions related to its accuracy or integrity.
Not applicable.
We would like to express our gratitude to all those who helped us during the writing of this manuscript. Thanks to all the peer reviewers for their opinions and suggestions.
This research received no external funding.
The authors declare no conflict of interest. Antonio Pontoriero is serving as one of the Guest editors of this journal. We declare that Antonio Pontoriero had no involvement in the peer review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to Christos Iavazzo.
Publisher’s Note: IMR Press stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.