Combined proton and carbon ion radiotherapy for uterine cervical squamous cell carcinoma: A retrospective analysis

Background: To skip brachytherapy from external radiation and brachytherapy for cervical carcinomas, a combined proton and carbon ion radiotherapy (PCRT) was used. This retrospective study reported the toxicity and short-term outcome. Methods: 16 cervical SCC patients were consecutively treated with PCRT in our center from August 2016 to July 2019. The acute and late toxicities, overall survival (OS), local control rate (LC), progression-free survival (PFS) and distant metastasis-free survival (DMFS) were analyzed. Results: For 16 cervical SCC, the prescribed dose for tumor was 76 Gy to 85.6 Gy (relative biologic ecacy (RBE)). With a median follow-up of 34.3 (range, 20.7-56.8) months, all of the acute toxicity were grade1-2 with 12.5% of hepatic toxicity, 75% of hematological toxicity and 18.8% of gastroenteric toxicity. No acute urinary toxicity occurred. No severe late toxicity was observed during the follow-up period. At the last follow-up visit, 11 patients were alive, 4 patients died and 1 lost for follow-up; 3 patients had local recurrences and 6 patients had distant metastases. OS, LC, PFS, and DMFS at 1-year, 2-year and 3-year were 100%, 100% and 77.9%; 86.7%, 86.7% and 78.0%; 75.0%, 68.8% and 61.9%; and 87.5%, 81.3% and 75.0%, respectively. Conclusions: PCRT was feasible to deliver high dose to cervical SCC, and the outcome was quite good. We thought that brachytherapy had the potential to be skipped when PCRT was used and long term results need to be further investigated.

bladder, rectum, head of femur and pelvis [5]. Therefore, it has more potential to spare normal organs, and meanwhile to deliver higher doses to targets compared to photon irradiation [7].
Carbon ion radiotherapy (CIRT) also has the physical dosimetric advantages. Besides, it possess biological advantages due to the nature of high linear energy transfer (LET) in the Bragg peak, which has strong inactivation effect on tumors with relative biological effect (RBE) of 2-3 [9][10], and low oxygen enhancement ratio (OER). National Institute of Radiological Sciences (NIRS), in Japan has completed several phase I/II or phase II clinical trails and showed high dose CIRT could improve the local control of the uterine cervical carcinoma [11]. CIRT with concurrent chemotherapy was safe for bulky tumors and could improve local control and overall survival [12].
Brachytherapy is a necessary part of irradiation for cervical carcinomas. However, we intended to skip it by combined proton and carbon ion radiotherapy (PCRT). Because there are radiation sensitive organs, intestine and colon adjacent to uterine cervical carcinoma, CIRT only would probably injure them more severe and result in more toxicity than proton. Therefore, we used proton for the primary tumor and pelvic nodes and CIRT boost for the clinical cervical tumors and metastatic nodes. In this report, we retrospectively analyzed the cervical carcinoma patients we treated with PCRT to observe its safety, tolerance and local control and survival. This study was approved by Ethical Committee of our center.

Patient eligibility
The patients selection eligibility criteria were: (1). histologically proven cervical squamous cell carcinoma (cervical SCC); (2). International Federation of Gynecology and Obstetrics (FIGO) stage (2014) IB2-IIIB, without rectum invasion; or FIGO stage IA-IB1, who refused surgery or were contraindication for surgery due to comorbidity diseases; (3). received PCRT with or without chemotherapy for curative intention.
From our center database 16 cervical SCC patients from August 2016 to July 2019 were eligible for the analysis. Their clinical characteristics were shown in Table 1. The median age was 48.5 years (range: 30 to 74). The median size of maximal diameter of tumor was 5.5 cm (range: 2.9-9.5) before treatment. The tumor size decreased after induction chemotherapy in 4 patients. The median size of maximal length of tumor was 7.0 cm (4.7-9.5 cm) in those 4 patients and it decreased to 4.8 cm (3.2-6.0 cm) after induction chemotherapy. Gross tumor volume (GTV) was de ned as the primary lesion in cervix identi ed by MRI or PET. Clinical target volume (CTV) consisted of CTV1, CTV2, CTV3 and CTVln. CTV1 was de ned as the pelvic lymph node regions (common iliac, internal iliac, external iliac, obrurator and presacral node regions) and suspicious lymph nodes (LN). CTV2 was de ned as high-risk clinical region, including CTV3, uterus, cervix not included in GTV, bilateral parametrium, ovaries, subterminal vagina. CTV3 was de ned as GTV expanded with a 5-mm margin. CTVln was de ned as enlarged lymph nodes with a 5-mm margin. Planning target volume (PTV) included CTV plus a 5-mm margin for positioning uncertainty, and more margins at the beam directions.
The prescribed doses were listed in Table 2. Proton of 46 Gy (RBE) in 23 fractions was delivered to CTV1. Various proton plus carbon or pure carbon ion boost doses were given to GTV or CTVln. Biological effect dose (BED) was calculated by Linea-Quadratic model with α/β of 10Gy.
PCRT quality assurance included water phantom dose veri cation before irradiation; KV X-ray position veri cation lms before each irradiation; and CT review weekly and necessary plan modi cation when needed.

Discussion
Cervical carcinoma in late stages were believed the radio-resistant due to the tumor hypoxia, which resulted from the associated anemia. Therefore, to achieve good outcome, high dose of radiotherapy is necessary. The external irradiation followed by brachytherapy irradiation, intracavitary or interstitial implantation, has been the standard irradiation technique. Typically, patients were irradiated with external beam radiotherapy of 45 to 50 Gy, and followed by intracavitary brachytherapy with 3 to 5 fractions of 5 to 7 Gy per fraction (BED10 of 72.6Gy-119.5 Gy). However, intracavitary brachytherapy might not deliver a su cient dose to extensive and bulky tumors in stage III or IVA disease [13][14]. In that case, interstitial brachytherapy could be used [15][16]. However, it was invasive and had limited indications. In an effort to exclude brachytherapy, the dosimetric comparison studies had been done between external irradiation plus brachytherapy and external irradiation alone with stereotactic radiotherapy or IMRT [17][18]. However, the high dose to central tumors by brachytherapy could not be reproduced by stereotactic radiotherapy or IMRT. Moreover, a retrospective analysis showed that curative external irradiation and brachytherapy achieved a higher 5-year cancer speci c survival than external irradiation alone in cervical carcinoma, which implied the important role of brachytherapy [19]. However, intracavitary or interstitial implantation brachytherapy brought patients inconvenience, or procedures was invasive. Therefore, most advanced irradiation techniques were tried to deliver intensive doses in a hope to skip the brachytherapy.
Although our analysis was a retrospective study, we had some basic considerations to implement PCRT for cervical SCC. Our intention was to exclude brachytherapy from the combination of external beam and brachytherapy, and meanwhile to keep a similar outcome as that in the external and brachytherapy. From dosimetric comparative study of cervical carcinoma, intensity-modulated proton radiotherapy (IMPT) had a signi cant reduction for the mean dose of small bowel and functional bone marrow than photon techniques such as IMRT, helical tomotherapy and RapidArc [5][6]. Furthermore, mean dose to the bladder and rectum was decreased by 7-9 Gy with IMPT in patents with pelvic radiation [5]. Volume of pelvic bone marrow receiving 10 Gy-20 Gy or 40Gy was associated with hematologic toxicity [20][21][22][23][24]. Thus, protons offered the best sparing of small bowel, rectum and pelvic bone marrow, and could contribute to a signi cant reduction of acute and late gastrointestinal and hematologic toxicity in cervical carcinoma radiotherapy, especially when combined with concurrent chemotherapy. However, on the other hand, cervical carcinoma often contained a large percentage of radio-resistant hypoxic cells due to the associated anemia. The proton, as low LET beam, would probably not be effective enough to sterilize hypoxic cells. Carbon ion beam is a high LET beam and possesses both physical and biological advantages. From experimental and clinical studies, it had shown strong capabilities to inactivate malignancies with RBE of 2 to 3. Moreover, it could overcome tumor hypoxia with decreased OER of around 2. Actually, CIRT had been used to treat cervical carcinomas and yielded good outcome for locally advanced bulky cervical SCC [11]. Therefore, we irradiated the cervical tumor and pelvic nodes by proton of 46Gy (RBE), and followed by CIRT boost doses to cervical gross tumor and the metastatic nodes.
As proton and CIRT are new radiation techniques for cervical carcinomas, some technical problems we have to deal with, including uterine motion [32], and position variations of bladder and rectum [33]. The above problems would result in dose delivery uncertainty. We applied some measurements to decrease the gas in intestine and colon, to keep the same bladder urine volume and rectum empty. Recently we took CT by in-room CT just before the dose delivery and con rmed the effectiveness of those measurements on keeping anatomy position consistent.
Concurrent platinum-based chemotherapy is a part of the current standard treatment for patients with FIGO stage IB2-IVA. NIRS carried out a dose-escalation study on locally advanced cervical SCC by CIRT alone in 2014, and showed 2-year OS and LC were 64% and 68.2%, respectively [11]. In 2019, NIRS did another phase 1/2 study on CIRT, but with concurrent chemotherapy, and the 2-year OS and LC were 82% and 67%, respectively [31]. The 2-year OS was improved from 64-82% in concurrent chemo-radiation study, which demonstrated again the importance of concurrent chemotherapy in CIRT as the same as in photon irradiation. However, concurrent chemotherapy did not improve the outcome in our study probably due to the small example size.
Univariate analysis in our study showed that only stage and tumor size were related to OS as that reported in literature [31]. If we could make the disease down staged before irradiation, it would be of help. Induction chemotherapy might be a strategy to reduce tumor burden in locally advanced cervical carcinomas. A British phase II trail included 46 patients with FIGO stage Ib2-IVa cervical carcinoma showed a good response rate to induction chemotherapy before CCRT, and the 5-year OS was 67% [34], which was better than the outcome after CCRT alone [35]. An international randomized multicenter phase III trial, INTERLACE (NCT01566240) is going on to compare the induction chemotherapy followed by CCRT and CCRT alone. In our center, we are planning to initiate a prospective clinical trial of induction chemotherapy, and concurrent chemotherapy and PCRT for locally advanced cervical carcinomas to further con rm PCRT feasibility and its e cacy.

Conclusions
In summary, our preliminary result demonstrated that PCRT was feasible to deliver high dose to cervical SCC, and the outcome was quite good. We thought that brachytherapy had the potential to be skipped when PCRT was used. However, as it was a retrospective analysis with small sample size and short follow-up period, the conclusion was very preliminary and long term results need to be further investigated.