Hematologic Toxicity and Bone Marrow-Sparing Strategies in Chemoradiation for Locally Advanced Cervical Cancer: A Systematic Review

Simple Summary Chemoradiation as a standard treatment for locally advanced cervical cancer is known to induce severe hematologic toxicity. This systematic review aims to evaluate the relationship between pelvic bone marrow irradiation and hematologic toxicity in patients undergoing platin-based chemoradiation for locally advanced cervical cancer. We seek to summarize possible dose constraints for optimal bone marrow sparing and optimize clinical strategies to mitigate treatment-related toxicities. Abstract The standard treatment for locally advanced cervical cancer typically includes concomitant chemoradiation, a regimen known to induce severe hematologic toxicity (HT). Particularly, pelvic bone marrow dose exposure has been identified as a contributing factor to this hematologic toxicity. Chemotherapy further increases bone marrow suppression, often necessitating treatment interruptions or dose reductions. A systematic search for original articles published between 1 January 2006 and 7 January 2024 that reported on chemoradiotherapy for locally advanced cervical cancer and hematologic toxicities was conducted. Twenty-four articles comprising 1539 patients were included in the final analysis. HT of grade 2 and higher was observed across all studies and frequently exceeded 50%. When correlating active pelvic bone marrow and HT, significant correlations were found for volumes between 10 and 45 Gy and HT of grade 3 and higher. Several dose recommendations for pelvic bone and pelvic bone marrow sparing to reduce HT were established, including V10 < 90–95%, V20 < 65–86.6% and V40 < 22.8–40%. Applying dose constraints to the pelvic bone/bone marrow is a promising approach for reducing HT, and thus reliable implementation of therapy. However, prospective randomized controlled trials are needed to define precise dose constraints and optimize clinical strategies.


Introduction
Cervical cancer remains a significant global health concern, with substantial morbidity and mortality rates worldwide (an estimated 604,000 new cases and 342,000 deaths in 2020) [1,2].Although there is a downward trend in incidence due to the HPV vaccine, particularly in industrialized nations, it remains a leading cause of cancer-related deaths among women [1,3].Cervical cancer poses a significant health burden, necessitating effective preventive and therapeutic strategies.Human papillomavirus (HPV) infection (mainly HPV 16 and 18) stands as the primary trigger for cervical cancer development, emphasizing the importance of preventive measures such as HPV vaccination.The introduction of HPV vaccines has offered a significant reduction in the incidence of cervical cancer by targeting high-risk HPV strains [4,5].
The standard treatment for locally advanced cervical cancer typically involves a multimodal approach, consisting of external-beam radiation therapy (RT) with concomitant chemotherapy (CTx), followed by brachytherapy (BT) [6].However, while these treatments are essential for tumor control, they might lead to treatment-related side-effects in the surrounding healthy tissues.Particularly, gastrointestinal (GI) and genitourinary (GU) toxicities, as well as hematologic toxicities such as bone marrow deficiency, may have a significant impact [7].Prospective studies have revealed that the incidence of grade ≥ 3 HT in platin-based pelvic chemoradiation ranges between 20% and 25% [7].The application of extended-field para-aortic radiation with increased exposure of the pelvic and lower spine bones results in even broader exposure of the overall bone marrow, consequently leading to an elevated incidence of HT [8].
Bone marrow consists of hematopoietic stem cells, from which blood cells develop through cell division and cell differentiation.In adults, blood-forming bone marrow (red or active bone marrow (ABM)) is no longer found in all bones, but only in the sternum, ribs, skull bones, clavicles, vertebrae, pelvis, scapulae and the upper ends of humeri and femora.In a 40-year-old person, the ABM in the sacrum and pelvic bones makes up around 40% of the total active bone marrow, which therefore plays a pivotal role in replenishing blood cells [9].
Understanding the nuances of bone marrow irradiation is crucial, above all, the distinction between active and inactive bone marrow compartments.A study by Robinson et al. found that active and inactive pelvic bone marrow responded differently to concomitant chemoradiation: the volume of pelvic active bone marrow (PABM) shows a median absolute decline of −0.25 g/mL compared to −0.02 g/mL for inactive pelvic bone marrow, suggesting that defining PABM is central in radiation therapy.One approach to accomplish this is using an [ 18 F]FDG-PET in delineating the pelvic bone in addition to using the planning CT.ABM can be defined as having [ 18 F]FDG uptake greater than the mean of the whole structure [10].
The correlation between the volume of bone marrow irradiation and dose exposure with acute HT [11,12] underscores the need for precise treatment planning and delivery techniques to minimize these adverse effects.Moreover, chemotherapy further increases bone marrow suppression, often necessitating treatment interruptions or dose reductions, which can compromise oncologic outcomes [13,14].
While advancements in radiation techniques allow for bone marrow sparing [15,16], similar progress in chemotherapeutic regimens remains limited.However, emerging targeted therapies hold promise for future improvements in treatment tolerability and efficacy [17,18].This review article aims to consolidate the existing literature on the relationship between pelvic bone marrow irradiation and hematologic toxicity in patients undergoing primary platin-based chemoradiation for locally advanced cervical cancer.By elucidating these relationships, we seek to provide insights into bone marrow contouring methods and the optimization of clinical strategies to mitigate treatment-related toxicities.

Materials and Methods
This review was conducted in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [19].The systematic review followed the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA).The protocol has not been registered.

Search Strategy
A search of MEDLINE (via PubMed) and connected papers was conducted from 1 January 2006 to 7 January 2024.The search encompassed articles related to chemoradiation (CRT) in cervical cancer reporting on hematologic toxicities and bone marrow dose exposure.Following the removal of duplicates, two independent investigators (DK, SC) screened all records by title, abstract and full text.Any discrepancies were resolved through consensus-based discussion.

Eligibility Criteria
The eligibility criteria for inclusion in this review article encompassed studies focusing on cervical cancer patients who underwent (chemo)radiotherapy.Specifically, we considered studies reporting on the correlation between hematologic toxicity and the radiation dose received by the pelvic bone marrow.Included studies needed to have a minimum of 10 patients and were limited to those published from 1 January 2006 to 7 January 2024.Studies with pre-or postoperative treatment settings were eligible for inclusion.The exclusion criteria comprised letters, reviews, abstracts and editorials, as well as studies published in languages other than English or German.Additionally, studies involving mixed histologies that could not be analyzed separately were excluded.The detailed eligibility criteria are presented in Table 1.

Data Extraction and Analysis
Data extraction was performed independently by two authors (DK, SC) using a predefined Excel sheet (Office 365, Microsoft, Redmond, WA, USA).Any discrepancies were resolved through consensus-based discussion.Analyses, where applicable, were conducted using Excel.The review examined information regarding the study population, study design, RT technique, prescribed dose to planning target volume (PTV), inclusion of patients with extended-field RT (para-aortic), chemotherapeutic regimen and number of cycles administered, bone marrow definition (methods for contouring), whether the pelvic bone was split into subvolumes, distinction between active or inactive bone marrow, use of dose constraints for the pelvic bone, compromise of other organs at risk (OARs) for bone marrow sparing, assessment of hematologic toxicity (grading score used, frequency and timing of assessment, blood components analyzed, grade of HT measured), creation of more than one RT plan, dosimetric predictors and recommended dose limitations to the pelvic bone.
For every dosimetric factor identified, we assessed the number of studies examining its link to hematologic toxicity and the proportion of those studies reporting a meaningful association.Furthermore, we outlined the threshold doses for specific dose-volume parameters that forecast hematologic toxicity.

Quality Assessment
The quality of the included studies was evaluated independently by two examiners (DK, SC) using the Oxford CEBM Levels 2011 criteria [20].

Search Results
Our search terms identified 987 articles extracted from the included database.The search yield and selection process are shown in Figure 1.Thereafter, we identified 24 eligible studies.All the included studies are shown in Tables 2-4.A total of 17 studies were retrospective studies, and 7 studies had a prospective study design.One study had a prospective randomized controlled trial (RCT) study design and therefore reached level 2 on the Oxford CEBM Levels of evidence; however, the majority of studies could only be assigned to level 4 on the Oxford CEBM 2011 [20].
pelvic bone was split into subvolumes, distinction between active or inactive bone ma row, use of dose constraints for the pelvic bone, compromise of other organs at ris (OARs) for bone marrow sparing, assessment of hematologic toxicity (grading score use frequency and timing of assessment, blood components analyzed, grade of HT measured creation of more than one RT plan, dosimetric predictors and recommended dose limit tions to the pelvic bone.
For every dosimetric factor identified, we assessed the number of studies examinin its link to hematologic toxicity and the proportion of those studies reporting a meaningfu association.Furthermore, we outlined the threshold doses for specific dose-volume p rameters that forecast hematologic toxicity.

Quality Assessment
The quality of the included studies was evaluated independently by two examine (DK, SC) using the Oxford CEBM Levels 2011 criteria [20].

Search Results
Our search terms identified 987 articles extracted from the included database.Th search yield and selection process are shown in Figure 1.Thereafter, we identified 24 el gible studies.All the included studies are shown in Tables 2-4.A total of 17 studies wer retrospective studies, and 7 studies had a prospective study design.One study had a pro spective randomized controlled trial (RCT) study design and therefore reached level 2 o the Oxford CEBM Levels of evidence; however, the majority of studies could only be a signed to level 4 on the Oxford CEBM 2011 [20].

Population Characteristics
The incorporated studies spanned from 2006 to 2023, involving a cumulative cohort of 1539 patients.The cohort sizes varied between 17 and 164 patients across studies.Of these, 14 studies explicitly outlined a definitive treatment regimen, wherein patients were scheduled to undergo chemoradiation followed by brachytherapy.In contrast, four studies exclusively enrolled patients who underwent chemoradiotherapy in an adjuvant setting.Additionally, five articles included patients undergoing both definitive and adjuvant treatments, while one study did not specify the treatment intention.

Therapy Regimens
Significant disparities exist among the studies regarding both the RT technique employed and the dosages and cycles of chemotherapy administered.Variability was also observed in terms of the PTV, encompassing differences in both dose prescription and PTV definition.Furthermore, four studies incorporated patients necessitating para-aortic irradiation, thereby inherently expanding the irradiation volume in these cohorts.Specifically, one study exclusively enrolled patients requiring para-aortic irradiation [40].Of the included studies, 14 utilized intensity-modulated radiation therapy (IMRT) technology, with some employing IMRT techniques such as volumetric modulated arc therapy (VMAT), RapidArc or tomotherapy.Conversely, two studies exclusively employed threedimensional (3D) treatment planning, while in seven studies, either 3D plans or IMRT plans were utilized.One study employed the anterior-posterior/posterior-anterior (AP/PA) radiation technique in 77% of patients.
The majority of patients received concomitant chemotherapy following contemporary standards [44], with cisplatin administered at 40 mg/m 2 body surface area (BSA) in 15 out of 24 articles.In another three studies, concomitant chemotherapy also comprised cisplatin, albeit sometimes at lower doses (see Table 2).Moreover, three studies permitted the administration of carboplatin instead of cisplatin in the presence of contraindications.Induction chemotherapy was only applied in two studies: Zhang et al. [29] administered a single cycle of induction therapy (paclitaxel (175 mg/m 2 )/cisplatin (75 mg/m 2 ) to 78.6% of patients, while Li et al. [12] reported 47% of patients receiving unspecified induction chemotherapy and only 15% of patients receiving concomitant chemotherapy.In the remaining 22 studies, concomitant chemotherapy was administered, with modifications to cycles under circumstances of hematologic toxicity, patients' general condition or patient refusal.
Among the 12 articles administering chemotherapy with weekly cisplatin at 40 mg/m 2 BSA according to contemporary standards and providing precise information on the number of cycles administered, approximately 70.5% of patients received ≥ 5 cycles of concomitant chemotherapy.The lowest value is 37.7% of patients receiving ≥ 5 cycles [30], and the highest value is 100% of patients [38].
Furthermore, variations were observed in PTV dose prescription.The lowest administered dose to the PTV was 39.6 Gy (normofractionated) [21,22,25], while the maximum dose in the PTV was reported to be 68 Gy (normofractionated) in Zhou et al. [41], although it is presumed to be a simultaneous integrated boost dose rather than a dose to the entire PTV, albeit without explicit clarification.On average, doses ranging from 45 to 50.4 Gy (normofractionated) were most commonly administered to the PTV.

Hematologic Toxicity
HT was assessed using either the Common Terminology Criteria for Adverse Events (CTCAE) (45%) or the RTOG criteria (55%) [45,46].Nineteen studies performed a complete blood cell count weekly during CRT.In four articles, an additional blood cell count was conducted before the initiation of CRT [29,34,36,38], whereas in seven studies, blood values were also collected after the end of CRT, but no longer than 3 months after the end of CRT [31,38] (see Table 3).The evaluated blood components encompassed white blood cells, neutrophils, hemoglobin, thrombocytes and, in more recent investigations, lymphocytes.In instances where overall HT was not explicitly stated, Table 3 delineates the toxicities pertaining to individual blood components.It is noteworthy that hematologic toxicities of grade 2 or higher (HT 2+) were observed across all studies.HT 2+ was often found to be well over 50% in most studies that recorded this endpoint.The precise extent of HT, expressed as a percentage, along with specific information regarding the affected blood cells, if available, is documented in Table 3.

Pelvic Bone 3.5.1. Delineation Methods
The bone marrow (BM) was delineated by outlining the pelvic bone contour on the planning CT scans.In ten studies, beyond merely considering the entire pelvic bones collectively, specific subsites were delineated to investigate potential associations between these regions and HT.Furthermore, in five articles, regions of low density within the pelvic bones were delineated and interpreted as substitutes for bone marrow.Additionally, functional imaging was utilized in seven articles alongside the planning CT scans to delineate active bone marrow.Specifically, five studies utilized [ 18 F]FDG-PET-CT, while one study employed technetium-99m sulfur colloid single-photon emission tomography to quantify the standardized uptake values (SUVs).Two distinct methodologies were employed to identify active bone marrow: defining it as regions with SUVs greater than the mean SUV (SUV mean ) of the total body or the SUV mean of the entire bone.Additionally, in three studies, the pelvic bone was employed as an avoidance structure, ensuring coverage of the PTV while minimizing additional stress on other OARs.

Low-Density Bone Marrow Spaces
Despite their prior inclusion in the preceding results section, the findings for lowdensity bone marrow spaces are summarized as follows: Five articles defined BM as the regions of low density within the respective osseous structures [32][33][34][35][36].Among these investigations, two articles failed to establish significant correlations upon multivariate analysis [33,34].Mahantshetty et al. [32] discerned that a volumetric threshold of pelvic bone marrow V40 < 40% was notably linked with HT 2+.Similarly, Huang et al. [35] highlighted correlations, apart from V40, between lumbosacral spine marrow LSS-V10, V20 and V40, as well as Dmean, and HT 2+.Furthermore, Singareddy et al. [36] proposed the following dose constraints for the mitigation of adverse effects: whole pelvic bone marrow V20, V30 and V40 should be maintained below thresholds of 71.75%, 49.75% and 22.85%, respectively, with Dmean not exceeding 28.85 Gy.
In the included studies, various dose and/or volume parameters were examined to determine their correlation with hematologic toxicity.While several significant associations were identified, not every parameter currently has a recommended constraint established.

Discussion
The standard treatments for locally advanced cervical cancer, including pelvic chemoradiation, pose risks of hematologic toxicities due to bone marrow dose exposure.Specifically, the irradiation of bone marrow in the pelvic region has been recognized as a significant cause of hematologic toxicity, leading to efforts to reduce its effects.This is particularly important to ensure that the administration of simultaneous chemotherapy is not jeopardized.
One strategy involves implementing pelvic bone marrow dose constraints during the planning phase of radiotherapy with the aim of reducing the incidence and severity of HT during treatment.
An interruption of therapy and, thus, an extension of the overall treatment time has a negative impact on the oncological outcome and should always be avoided [47].For example, it was shown that stage III disease accounted for most adverse effects from the prolongation of OTT.A significant increase in the relative risk of local recurrence and death was found when the OTT was >52 days compared with a shorter duration [48].
This review article offers an analysis of research articles investigating the relationship between bone marrow dose exposure in patients with locally advanced cervical cancer and the occurrence of HT.In total, 24 articles were examined, revealing dosimetric predictors and recommending constraints to avoid/reduce the incidence of HT using three distinct methods for delineating bone marrow: whole pelvic bone, lower-density marrow spaces and functional imaging-based active bone marrow contouring.
Hematologic toxicity: The impact of the standard chemoradiotherapy regimen [44] for cervical carcinoma on the immune system shows substantial perturbations within the active bone marrow compartment [49].Elicin et al. emphasize the enduring impairment persisting posttreatment even after a three-month period [38], with some extent of marrow recovery noted only after a year-long interval [50].Endpoint analyses across the reviewed studies predominantly focused on HT grades 2+ or 3+, revealing a notable incidence exceeding 30% and 40%, respectively.However, the recommended dosimetric cut-off values were largely similar for HT 2+ and HT 3+.This can probably also be explained by the very unequal study settings: In some studies, chemotherapy was not administered if a certain degree of depletion of blood cells was seen [23,35] in order to avoid the progression of depletion.Indeed, the development of HT is influenced not only by RT but also by CTx interventions [51].However, discerning the precise contribution of each therapeutic modality, namely RT and CTx, to HT remains elusive.
Li et al. observed a significant increase in HT incidence when comparing concomitant CTx to RT alone (70.21% vs. 29.79%;p = 0.001).However, their study also permitted induction chemotherapy, revealing that patients subjected to this additional regimen experienced higher HT rates (63.83% vs. 36.17%;p = 0.01) [12].This disparity likely arises from the escalated dosage administered during induction CTx relative to concomitant CTx.Despite this variation, we opted to include both Li et al.'s study and that of Zhang et al. [29] in our review, notwithstanding the administration of induction CTx.Regardless, the chemotherapy protocols across the included studies were markedly heterogeneous (notably permitting carboplatin in three studies), with carboplatin being recognized for its more pronounced hematotoxicity.[52,53].Additionally, the diversity in the number of cycles administered across the studies, as well as the sparing documentation of the number of CTx cycles administered, further underlines the complexity of this comparison (see Table 2).
In contrast, Rose et al. reported a lack of significant correlation between dosimetric bone marrow parameters and the probability of having one or more CTx cycles held [22].
Chen et al.'s investigation elucidated that the effect of bone marrow irradiation on the nadir of neutrophils was masked by CTx when more than four cycles of concomitant CTx were administered.Additionally, Chen identified that the dose received by the ilium exerted a more pronounced influence on neutrophil decline, whereas the irradiation dose to the lower pelvis had a greater impact on hemoglobin, platelets and the neutrophil-to-lymphocyte ratio (NLR).Moreover, high-dose irradiation predominantly affected hemoglobin and platelet levels, whereas NLR was more significantly influenced by low-dose irradiation.Neutrophil counts were susceptible to both low-and high-dose irradiation [30].
Generally, the existing literature indicates that lymphocytes are most sensitive to irradiation [54,55].In four of the studies encompassed within this review, lymphocyte levels were assessed [29,30,34,43], revealing a significant reduction in lymphocyte counts (see Table 3).Thus, the studies examined in this review demonstrated a wide array of observations and experiences regarding the influence of the therapy on the individual blood components and regarding the interactions with chemotherapy.
Bone marrow contouring: Studies have already shown that bone marrow reacts very sensitively to radiation, although there are differences between the various progenitor cells.Notably, studies have elucidated the increased radiosensitivity of neutrophils and select lymphocytic subsets [56], with even minimal radiation dose exposure precipitating both acute and long-term bone marrow-related side effects [57,58].
Given that roughly 50% of active bone marrow is localized within the pelvic bones, encompassing the proximal femur and lumbosacral spine [59,60], it can be assumed that pelvic irradiation has an effect on hematopoietic stem cells, stromal components and microcirculatory networks.Although bone marrow has historically been ignored as an OAR in radiation planning, robust evidence suggests its vulnerability, particularly in cohorts necessitating extended-field radiotherapy, wherein the lumbosacral bone marrow may be substantially implicated [50,61].Only one of the included studies (Yan et al. [40]) exclusively examined patients subjected to para-aortic irradiation; in some of the remaining studies, patients with an indication for para-aortic irradiation were included but not analyzed separately (see Table 4).Yan et al.'s dosimetric predictors show agreement within the spectrum of the other recommended dose constraints and were in the medium range.Consequently, there is no conclusive evidence to suggest that HT would have significantly exceeded the values observed in the other studies analyzed.
It is essential to delineate the bone marrow, particularly its active regions, during radiation planning, and subsequently implement protective measures to spare its volume.One viable approach involves the integration of functional imaging modalities such as [ 18 F]FDG-PET alongside radiation planning to discern active bone marrow regions [62].The methodologies employed across the reviewed studies to identify ABM encompassed criteria such as ABM ≥ SUV mean -WB, ≥SUV mean -WPB and ≥SUV mean -WPBM.
An alternative method employed in select studies involved employing "freehand contouring" techniques to delineate lower-density marrow regions on planning CT scans, utilizing these areas as proxies for ABM.Notably, Mahantshetty et al. [32] demonstrated that using bone cavities as surrogates for BM allowed for the identification of dose parameters significantly correlated with HT 2+ (V40 < 40%).Conversely, in studies conducted by Kumar and Lewis, multivariate analyses failed to yield significant correlations for dose-volume histogram parameters pertaining to contoured low-density regions [33,34].Consequently, it is recommended that functional imaging techniques be utilized for the accurate identification of ABM.In the absence of functional imaging resources, employing the entire bone as a surrogate may be sufficient.
Rose et al. further investigated the potential association between dose parameters (V10, V20, V30, V40) reaching the inactive yellow bone marrow and HT.As anticipated, their findings revealed no discernible correlation [37].This outcome aligns with the established understanding that the yellow bone marrow houses mesenchymal stem cells crucial for cartilage, bone and adipose tissue generation [63].
Radiation technique: The radiation modality employed can influence both the extent of BM irradiation and the resultant dose exposure.IMRT offers superior conformality, facilitating the sparing of adjacent OARs [64].While extensively studied for organs at risk, such as the bladder and bowel [65], this principle extends to pelvic bone sparing when compared to conventional techniques like four-field box irradiation [66], AP/PA irradiation [16] or 3D irradiation [67].
However, despite evidence suggesting IMRT's efficacy in pelvic bone sparing, the translation of this reduction in radiation exposure to a significant decrease in HT is not consistently evident.Nonetheless, investigations comparing bone marrow-sparing IMRT (BMS-IMRT) with conventional IMRT consistently demonstrate fewer adverse effects regarding HT when BM considerations are taken into account [68,69].
Consequently, delineating the pelvic bone as an OAR and minimizing its irradiation should be standard practice, provided that it does not compromise PTV coverage and does not subject other OARs to undue stress [70].A recently published study has even shown that it is extremely useful to define ABM.Wang et al.'s prospective clinical trial demonstrated that ABMS VMAT significantly reduced grade 3+ HT and improved chemotherapy delivery in locally advanced cervical cancer patients undergoing chemoradiotherapy [71].
Among the studies encompassed within our analysis, 3D and IMRT techniques were predominantly used.The studies that used different irradiation techniques in their collective showed contradictory results.For instance, Yan et al. reported no statistically significant disparity in the risk of developing HT 3+ across irradiation techniques (3D or IMRT) [40].In contrast, Ajayakumar et al., who also employed either 3D or IMRT technique, presented conflicting findings [28].They demonstrated markedly reduced HT within the IMRT cohort.
Despite these inconsistencies, the literature predominantly shows the advantages of the IMRT technique.Thus, IMRT emerges as the preferred irradiation modality whenever feasible.

Dosimetric parameters:
The dosimetric parameters associated with HT were identified.Most authors considered the entire pelvic bone as a surrogate for bone marrow.The recommended threshold values encompass V10 < 90-95%, V20 < 65-86.6% and V40 < 22.85-40%.The notable breadth of these ranges may be attributed to substantial variations in study design, which include differences in RT techniques, dose prescriptions, chemotherapy regimens and the numbers of cycles applied, as well as the timing of HT assessments.Such multidimensional factors exert discernible influences on HT outcomes, such that different results were to be expected.Some studies investigated specific subsites within the pelvic bone for nuanced exploration.
Studies utilizing functional imaging to assess active bone marrow also evidenced significant correlations between certain dose parameters (V10, V20, V30, V40, V45) and HT.Nevertheless, the recommended doses do not differ significantly from those given for the whole pelvic bone.This may be due to the fact that the regions with active bone marrow are naturally also included in the delineated whole pelvic bone.However, in scenarios where sparing the entire pelvic bone proves unfeasible due to anatomical constraints, the precise localization of ABM regions becomes paramount.
Various studies have identified different relevant dose parameters associated with HT.Specifically, while some investigations have found associations within the V10 and V20 ranges, i.e., in the low-dose spectrum, others have exclusively revealed correlations within higher dose ranges, such as V40 (see Tables 4 and 5).One plausible explanation for this divergence could be multicollinearity, a phenomenon that yields wider confidence intervals and subsequently diminishes the reliability of probabilities regarding the impact of independent variables within a model.Another reason is certainly that not all dose levels were investigated in all studies.
Additionally, certain studies opted for volume constraints over dose constraints and considered bone marrow as a parallel organ.Indeed, the manifestation of HT likely arises from the interplay of both the dose and volume reaching the irradiated bone marrow.Ultimately, the discordant findings regarding dosimetric parameters correlated with HT highlight the complexity of this relationship, mandating the need for further comprehensive investigations.
Further research is needed to determine the optimal dose and volume constraints for bone marrow, an essential organ at risk, especially in the context of locally advanced cervical carcinoma.The current literature lacks prospective studies and long-term data, with only two studies examining post-treatment blood assessments three months after the end of treatment [31,38].Prolonged bone marrow impairment can hinder subsequent therapies, notably in the context of ongoing research exploring adjuvant immunotherapy [72].In conclusion, this review highlights the complex relationship between bone marrow irradiation and hematologic toxicity in locally advanced cervical cancer patients, emphasizing the need for continued research to improve treatment outcomes.

Conclusions
In conclusion, our systematic review emphasizes the significant impact of hematologic toxicity in locally advanced cervical cancer patients undergoing chemoradiotherapy, with HT of grade 2 and higher observed in over 50% of cases across all studies.The correlations between active pelvic bone marrow dose exposure and severe HT highlight the importance of dose optimization strategies.While dose recommendations for pelvic bone and bone marrow sparing show promise in reducing HT, further research, particularly through prospective randomized controlled trials, is crucial to establish precise dose constraints and optimize clinical approaches for improving treatment outcomes in this patient population.

Figure 1 .
Figure 1.PRISMA flowchart of the screening and selection process.

Figure 1 .
Figure 1.PRISMA flowchart of the screening and selection process.

Table 2 .
Included studies-general information on study design, radiotherapy (RT) and chemotherapy (CTx) details.

Table 3 .
Included studies-information regarding hematologic toxicity.

Table 4 .
Included studies: information regarding bone marrow sparing.

Table 5 .
Recommended dose constraints from the included studies.