Article Text

Download PDFPDF

Radiotherapy dose escalation on pelvic lymph node control in patients with cervical cancer
  1. Emile Gogineni1,
  2. Beatrice Bloom1,
  3. Ferney Diaz Molina1,
  4. Jeannine Villella2,2 and
  5. Anuj Goenka1
  1. 1 Department of Radiation Medicine, Northwell Health, Lake Success, New York, USA
  2. 2 Gynecologic Oncology, Northwell Health, New York, New York, USA
  1. Correspondence to Dr Anuj Goenka, Northwell Health, Lake Success, NY 11042, USA; agoenka{at}northwell.edu

Abstract

Objective Data supporting dose escalation for node-positive cervical cancer are currently limited to small retrospective studies. The goal of this study was to assess whether radiation dose was associated with lymph node control or gastrointestinal toxicity in patients with node-positive cervical cancer.

Methods A total of 390 patients with carcinoma of the uterine cervix were treated between October 1997 and October 2017. Patients included in our analysis were those with squamous cell carcinoma or adenocarcinoma who were node-positive, treated definitively, and with at least one follow-up visit and post-treatment imaging scan. We excluded those without follow-up and those treated with palliative intent. All patients were treated with external beam radiation to pelvic±para-aortic fields with concurrent weekly cisplatin. All lymph nodes present at the time of treatment were stratified by size as <2 cm or ≥2 cm. Acute and late gastrointestinal toxicity were recorded for all patients.

Results A total of 77 patients with 206 lymph nodes were identified. Median stage at presentation was FIGO IIB. Thirteen patients underwent definitive surgical resection followed by adjuvant radiation, of which 12 were treated to doses ≤5040 (range 2700–5940) cGy. Sixty-four patients were treated with definitive chemoradiation, of which 42 (66%) received ≤5040 (range 4500–5040) cGy and 22 (34%) received >5040 (range 5300–6640) cGy. Patients with pre-chemoradiation lymph nodes ≥2 cm had inferior lymph node control compared with patients with pre-chemoradiation lymph node <2 cm at 12 months (77% vs 100%, p=0.002). Radiation dose >5040 cGy was not significantly associated with improved lymph node control compared with ≤5040 cGy when analyzing all patients (12 months, 100% vs 89%, p=0.112). In patients with pre-chemoradiation lymph nodes ≥2 cm, radiation dose >5040 cGy was associated with improved lymph node control (12 months, 100% vs 60%, p=0.020). Acute grade ≥2 gastrointestinal toxicity was not associated with radiation dose >5040 cGy (20% vs 13%, p=0.424). Two patients developed grade ≥2 late gastrointestinal toxicity, both of whom were treated to ≤5040 cGy.

Conclusions This series supports the role of dose escalation for patients with lymph nodes ≥2 cm. Dose escalation is associated with improved control in patients with larger lymph nodes, and is not associated with greater gastrointestinal toxicity.

  • cervical cancer
  • radiation
  • radiotherapy, image-guided
  • radiotherapy, intensity-modulated
  • radiotherapy dosage

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

HIGHLIGHTS

  • Lymph nodes ≥2 cm have significantly lower control rates at 12 months than lymph nodes <2 cm (77% vs 100%, p=0.002).

  • Dose escalation >5040 cGy significantly improves control at 12 months for lymph nodes ≥2 cm (100% vs 60%, p=0.020).

  • Dose escalation >5040 cGy did not increase rates of acute or chronic gastrointestinal toxicity.

Introduction

Cervical cancer is the most commonly diagnosed gynecologic malignancy worldwide, with incidence approaching nearly 600 000 in 2018.1 Overall survival remains 70%–80% for the majority of patients with locally advanced cervical cancer;2 3 however, survival rates drop in patients with positive lymph nodes,4 with prognosis corresponding to the number and size of involved lymph nodes.5 The treatment of choice for locally advanced cervical cancer is radiation with concurrent platinum-based chemotherapy.2 3 Considerations for dose and field size are driven by the extent of disease, both locally and regionally, in addition to other prognostic factors such as performance status and tumor bulk.

Early trials which helped define the standard of care for treatment of locally advanced cervical cancer used low doses of external beam radiation.2 6 Escalation of radiation dose has shown a benefit for bulky nodes and primaries, albeit in small retrospective studies.7–9 Grigsby et al showed that patients with lymph nodes >2 cm had inferior overall and cause-specific survival than patients with lymph nodes ≤2 cm.

The sensitivity of the small bowel to radiation has historically limited the ability to dose escalate, particularly when treating pelvic and para-aortic fields.10 11 Intensity-modulated radiation therapy now provides the technology to limit dose to normal structures when treating extended fields, allowing consideration of dose escalation more routinely.12 13 We performed a retrospective analysis to review our institutional experience treating node-positive cervical cancer in order to assess whether radiation dose was associated with pelvic lymph node control or gastrointestinal toxicity. We hypothesized that patients with bulky lymph nodes ≥2 cm would have lower control rates than those without, and that higher doses of radiation would improve regional recurrence rates with acceptable gastrointestinal toxicity.

Methods

Following institutional review board approval, our institutional database was searched between October 1997 and October 2017 to identify all patients with cancer of the uterine cervix treated with radiation. A total of 390 patients were identified. Patients included in our analysis were those with squamous cell carcinoma or adenocarcinoma who were node-positive, treated definitively, and with at least one follow-up visit and post-treatment imaging scan. Excluded patients included those who were node-negative, those treated with palliative intent, and those without follow-up visits and imaging. All patients were staged using the 2018 International Federation of Gynecology and Obstetrics (FIGO) uterine cervical cancer staging system.

All patients were treated with external beam radiation to pelvic±para-aortic fields with concurrent weekly cisplatin. Dose, fractionation, and technique of radiation were captured. Radiation boost doses to the cervix and lymph nodes were documented separately. All pre-treatment and post-treatment imaging was reviewed. Size and positron emission tomography (PET) avidity were captured for both the primary cervical tumor and all lymph nodes. Lymph nodes were considered malignant if they measured at least 1 cm in largest dimension on computed tomography (CT) scan and/or had 18F-fluorodeoxyglucose (FDG) avidity with standard uptake values (SUV) >3 on PET scan and/or were identified by the reading radiologist as suspicious. Biopsy-proven lymph nodes were also included for analysis. All lymph nodes present at the time of treatment were stratified by size as <2 cm or ≥2 cm.

Local control, regional control, distant control, and survival data were captured. Local failure was defined as recurrence at the cervix, proven pathologically or via an increase in size and/or avidity on imaging. Individual lymph node failure was defined as recurrence of a pelvic or para-aortic lymph node that had been present at the time of initial radiation treatment. Regional failure was defined as the appearance of a new pathologic lymph node >1 cm on CT and/or an increase in size/avidity of a previously seen lymph node in any pelvic and/or para-aortic lymph node, irrespective of location of initial disease.

On post-treatment imaging, patients were grouped into those with complete response, partial response, and no response. Complete response was defined as no evidence of disease on CT and/or PET on follow-up imaging, with no pathologically enlarged or hypermetabolic lymph nodes seen. Patients with CT showing no lymph nodes >1 cm with a minimum reduction of 50% when compared with pre-treatment imaging were considered to have achieved a complete response. Partial response was defined as a decrease in size and/or avidity not meeting the above criteria for a complete response. No response included patients with any lymph node that was unchanged or increased in size or avidity. Patients who did not undergo follow-up imaging within 6 months from the completion of treatment were excluded from this portion of the analysis.

Gastrointestinal, gynecologic, and genitourinary toxicity data from all follow-up visits were scored using the Common Terminology Criteria for Adverse Events (CTCAE) version 5.0. Toxicities were categorized into acute for symptoms experienced during treatment or within 3 months of completion of chemoradiation and chronic for symptoms experienced later than 3 months following treatment.

Follow-up time was defined as time in months from completion of chemoradiation. Failure and survival were calculated using the Kaplan–Meier method with log rank statistics used to compare groups. For failure analyses, patients were censored at the time of death or last follow-up. An alpha value ≤0.05 was considered significant. All statistical analyses were completed using SPSS version 22 (IBM Corporation, Armonk, NY, USA). Survival curves for time to event endpoint were estimated using the Kaplan–Meier method.

Results

A total of 77 patients with 206 lymph nodes were identified. The median age of patients was 59 years with a median Karnofsky Performance Status (KPS) of 90. The median stage at presentation was FIGO IIB. The majority, 57 patients (74%), had positive pelvic nodes only, 19 (25%) had positive pelvic and para-aortic nodes and 1 (1%) had positive para-aortic nodes without pelvic lymphadenopathy. A total of 26 patients (34%) had at least one lymph node ≥2 cm and 51 (66%) had only lymph nodes <2 cm. Baseline patient characteristics are summarized in Table 1.

Table 1

Baseline characteristics for all patients (n=77)

Treatment details are summarized in Table 2. Thirteen patients were treated with definitive surgical resection followed by adjuvant radiation, of which 12 were treated to doses ≤5040 (range 2700–5940) cGy. Sixty-four patients were treated with definitive chemoradiation, of which 42 patients (66%) received ≤5040 (range 4500–5040) cGy and 22 (34%) received >5040 (range 5300–6640) cGy. Intensity-modulated radiation therapy was utilized in 59 patients (77%).

Table 2

Treatment-related details for all patients (n=77)

Median follow-up from completion of radiation was 16.7 (range 1.5–98.4) months. Overall survival at 12 months was 95%. Local, regional, and distant control at 12 months was 84%, 78%, and 88%, respectively. Rate of individual lymph node control at 12 months was 92%. At final follow-up, 24 patients had experienced regional failure. Of these, eight patients had lymph node failure at sites involved at the time of treatment, while 16 had failure at nodal sites that were not radiographically involved at the time of treatment.

Initial choice of imaging study was not significantly related to rate of failure. Rates of individual lymph node control for patients staged by PET (n=52), CT (n=5), or both (n=20) at 12 months were 89%, 100%, and 93%, respectively (p=0.63).

Radiation dose >5040 cGy was not associated with improved individual lymph node control compared with ≤5040 cGy when analyzing all patients (12 months, 100% vs 89%, p=0.11). Patients treated with pre-chemoradiation lymph nodes ≥2 cm had inferior individual lymph node control compared with patients with pre-chemoradiation lymph nodes <2 cm (12 months, 77% vs 100%, p<0.01) as shown in Figure 1. This was also seen in the analysis of patients who received definitive chemoradiation (n=64), with inferior individual lymph node control in lymph nodes ≥2 cm versus <2 cm (12 months, 70% vs 100%, p<0.01). Radiation dose >5040 cGy was associated with improved individual lymph node control in patients with pre-chemoradiation lymph nodes ≥2 cm (12 months, 100% vs 60%, p=0.02) as shown in Figure 2. Radiation dose >5040 cGy was also associated with improved lymph node control in the analysis of definitive chemoradiation patients with pre-chemoradiation lymph nodes ≥2 cm (12 months, 100% vs 53%, p<0.01).

Figure 1

Individual lymph node control in patients with lymph nodes ≥2 cm versus <2 cm. Lymph nodes ≥2 cm had lower rates of control at 12 months than lymph nodes <2 cm (77% vs 100%). RT, radiation therapy.

Figure 2

Individual lymph node control in patients with lymph nodes ≥2 cm. Control was significantly higher at 12 months with dose escalation >5040 cGy than those treated with ≤5040 cGy (100% vs 60%).

Local failure of the primary cervical tumor was highly predictive of individual lymph node failure. Patients who experienced failure of their primary cervical cancer (n=13) had lower rates of individual lymph node control than patients who had controlled primaries (n=64) (12 months, 59% vs 98%, p<0.01). See Figure 3 for follow-up imaging results. Median time to first follow-up imaging after completion of chemoradiation was 2.7 (range 0.2–2.7) months, consisting of 49 (71%) PET scans, 12 (17%) CTs, and 8 (12%) MRI. A total of 42 (61%) scans at first follow-up showed a complete response of lymphadenopathy, 22 (32%) showed a partial response, and 5 (7%) showed no response. Median time to second follow-up imaging after completion of chemoradiation was 6.7 (range 3.1–17.6) months, and median time to third follow-up imaging was 13.4 (range 4.3–47.6) months.

Figure 3

Results from initial post-treatment imaging. There were no individual lymph node failures in patients with a complete response on first follow-up scan, while failure rates increased in partial and non-responders. CR, complete response; NR, no response; PR, partial response.

Of the 42 patients with a complete response on first scan, there were nine regional failures on further follow-up. None of these patients had individual lymph node failure of nodes that were present at the time of chemoradiation, but instead had new pelvic lymph nodes that appeared on later scans. A total of 22 patients had a partial response on first scan, all of whom were initially observed. Of these, ten had a complete response on second scan, three had a partial response on second scan with subsequent complete response on later scans, and three had individual lymph node failure of a previously treated node. Six were not re-scanned, of whom one had a local recurrence of the primary at first scan, three had distant recurrence at first scan, and two were lost to follow-up. Of the five patients with no response on first scan, one developed a complete response on subsequent scan and one had stable disease on all subsequent scans. The other three patients were considered regional failures and received salvage chemotherapy. All three of these patients had individual lymph node failure of nodes that had been present at the time of chemoradiation.

Patients who had follow-up imaging showing stable or increased size of lymph nodes at least 3 months after completion of chemoradiation had significantly worse individual lymph node control than patients with follow-up imaging showing resolution of or diminished size of lymph nodes 3 months after completion of chemoradiation (17% vs 97%, p<0.01). In total with long-term follow-up, there were 24 patients with regional failure, including eight individual lymph node failures of nodes present at the time of chemoradiation and 16 patients with appearance of new pelvic lymph nodes not present at the time of chemoradiation. Of the eight individual lymph nodes present at the time of chemoradiation which eventually failed, four were treated to 4500 cGy: one to 4950 cGy, one to 5040 cGy, one to 5400 cGy, and one to 5500 cGy.

Gastrointestinal toxicity data are summarized in Table 3. Only 14 patients (18%) experienced acute grade ≥2 gastrointestinal toxicity. Acute grade ≥2 gastrointestinal toxicity was similar when comparing patients treated to dose >5040 cGy versus ≤5040 cGy (13% (n=3/23) vs 20% (n=11/54), p=0.424). Additionally, only two patients (2%) experienced chronic grade ≥2 gastrointestinal toxicity, both of whom were treated to ≤5040 cGy.

Table 3

Gastrointestinal toxicity data for all patients (n=77)

Discussion

Improved diagnostic imaging tools and treatment delivery techniques have facilitated consideration of dose escalation in managing patients with lymph node-positive cervical cancer. In an era where definitive treatment has been shown to improve survival in the setting of node-positive cervical cancer, optimizing the dose and fields of radiation becomes paramount.14 The standard dose range to involved nodal sites in patients who present with gross lymphadenopathy remains controversial, particularly in patients with bulky lymph nodes who may require dose escalation. Early trials proving the benefit of chemoradiation such as Radiation Therapy Oncology Group (RTOG) 90–01 prescribed 4500 cGy of external beam radiation,2 6 while more recent studies have suggested that higher doses can be prescribed utilizing intensity-modulated radiation techniques.13

We found that in patients with pelvic lymph nodes <2 cm, dose escalation did not impact control. However, in patients with bulky pelvic lymphadenopathy which we defined as nodes ≥2 cm, dose escalation >5040 cGy improved control rates (100% vs 60%). We also found that patients who experienced local failure of the primary cervical tumor had higher rates of individual lymph node failure. We believe this suggests an inherent difference in tumor biology and radioresistance in this subset of patients, as opposed to tumor cells recurring at the cervix and eventually metastasizing to pelvic lymph nodes through lymphatic drainage.

Equally important, dose escalation did not increase the risk of gastrointestinal toxicity. Our findings correlate with data from other retrospective analyses suggesting a benefit to dose escalation in larger lymph nodes.9 15–17 This supports the consensus that locally advanced cervical cancer requires dose escalation, reflected in the American College of Radiology expert panel recommendations published by Gaffney et al.18

Our data are also important in understanding how to interpret follow-up imaging after definitive treatment. Findings from the first follow-up scan after completion of treatment were highly predictive of long-term lymph node control. No patient with a complete response on first scan after chemoradiation had a recurrence of lymph nodes that were present at the time of treatment. In patients with partial response on first scan, the majority will have a complete response on subsequent scans. Of the 23 patients with partial response on first scan, only three were felt to have regional failure. Our results are consistent with published data by Schwarz et al, which found that the results of first follow-up imaging after completion of treatment are highly predictive of long-term control.19

Data from other subsites such as head/neck suggests a PET response at 3-month follow-up correlates with control outcomes.20 21 Based on this, we analyzed our patient cohort to assess if there was an association between 3-month PET nodal response and nodal control. We found that follow-up imaging 3 months after completion of chemoradiation was highly predictive of lymph node control for all patients.

One weakness of our study was the retrospective nature of the analysis. Given the lack of prospective trials addressing this subject, we feel that this study still provides valuable information about the effects of dose escalation on control rates and gastrointestinal toxicity in node-positive cervical cancer.

Conclusions

We found that patients with node-positive cervical cancer ≥2 cm were more likely to develop lymph node failure; and in this subgroup, dose escalation beyond 5040 cGy is associated with an improvement in control without an increase in gastrointestinal toxicity. We do not have sufficient data to suggest dose escalation for patients with nodal disease <2 cm in size. In patients with a partial response on initial post-treatment scan, repeat imaging should be followed as the majority of patients will obtain long-term control without further intervention. Based on these data, we consider dose escalation above 5040 cGy for any patient with bulky lymphadenopathy undergoing definitive chemoradiation therapy for cervical cancer.

References

Footnotes

  • Twitter @emilegogineni, @jeanninevillel

  • Presented at This data was presented in abstract form at ASTRO 2018.

  • Contributors EG was involved with chart mining, data analysis, and writing. FDM was involved with dosimetric data extraction. BB, JV, and AG treated the majority of the patients, and provided mentorship and oversight.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data availability statement Data are available upon reasonable request. De-identified participant data are available upon request.