Page 56 - ARNM-3-1

Advances in Radiotherapy
& Nuclear Medicine The novel index and osteoradionecrosis

established to mitigate any potential bias stemming from in the “Baseline oral examination” section. ORNJ was
pre-existing immune-inflammatory conditions and the diagnosed as per the following clinical and radiological
influence of medication use. criteria: “irradiated bone that fails to heal over 3 months
without evidence of persisting, recurrent, or metastatic
2.2. Baseline oral examination tumor.” 31-35
Before undergoing C-CRT, each patient was examined for
oral and dental pathologies, regardless of symptom status, 2.6. Statistical analysis
per the American Dental Association and US Food and The descriptive characteristics are presented as the patient
Drug Administration guidelines, and our institutional count and corresponding percentages or as the median
standards. Following our institutional protocols, oral and interquartile ranges. Nominal factors were contrasted
31
examinations included panoramic radiographs for using either Fisher’s exact test or Pearson’s Chi-squared
each patient and followed the specifications advised by test. The Mann–Whitney and Kruskal–Wallis tests
the manufacturer. The digital panoramic radiographs were employed to compare continuous variables. This
were obtained using the same X-ray machine (J Morita, retrospective cohort study primarily aimed to identify
Veraviewepocs 2D, Kyoto, Japan). An experienced oral any associations between post-C-CRT ORNJ rates and
and maxillofacial surgeon (ES) performed all clinical and groups formed by merging pre-C-CRT. The combined
radiological assessments. groups were established using a pre-C-CRT PIV cutoff
2.3. Determination of pretreatment PIV and TE of 833 (<833 vs. ≥833) and pre-C-CRT TE groups (<4
29
cutoffs and ≥4 extractions) identified in a previous study.
Briefly, in our previous investigation, receiver operating
We opted not to set new PIV and TE cutoffs as 210 out of characteristic (ROC) curve analysis was employed to
220 patients in this study had also participated in our previous ascertain the optimal cutoff points for pre-C-CRT PIV
research published elsewhere. Therefore, the PIV and TE measures and pre-C-CRT TE numbers that could divide
cutoffs used in this study to categorize patients into two groups the study cohort into two significantly discrete ORNJ
for each factor were set at 833 (<833 vs. ≥833) and 4 (<4 vs. ≥4 risk groups for each variable under investigation. The
extractions), respectively, based on our earlier work. 29 cutoff points for pre-C-CRT PIV measures and pre-
2.4. Treatment details C-CRT TE numbers were 833 (area under the curve,
88.0%; sensitivity, 81.1%; specificity, 78.8%; and Youden
Target volumes were defined using pre-C-CRT computed index, 59.9) and 4 (the median pre-C-CRT value was
tomography (CT), 18F-fluorodeoxyglucose positron determined as a cutoff point), respectively, which resulted
emission tomography-CT, and/or magnetic resonance in the formation of four groups: Groups 1 (PIV < 833
imaging of the entire neck and nasopharyngeal primary, and TE < 4), 2 (PIV < 833 and TE ≥ 4), 3 (PIV ≥ 833
as mandated by our institutional standards for patients and TE < 4), and Group 4 (PIV ≥ 833 and TE ≥ 4). The
diagnosed with LA-NPC. The treatment approach optimal cutoff points for each variable were determined
employed was simultaneous integrated boost IMRT (SIB- by identifying the points at which the J-index reaches its
IMRT), which was consistently applied to all patients. maximum value on the ROC curve.
The methods for target volume definition, delineation,
and prescribed SIB-IMRT doses adhered to those Student’s t-test, Chi-square test, or Spearman correlation
previously documented. The recommended RT doses for tests were used to compare different groups as needed.
low-, moderate-, and high-risk planning target volumes The univariate analyses explored how patient, disease,
were 54, 59.4, and 70 Gy, respectively. These doses were and treatment variables might influence with ORNJ
systematically administered daily over 33 days. The prevalence. The multivariate Cox proportional hazard
concurrent chemotherapy regimen consisted of three cycles model included only the statistically significant factors in
of cisplatin, administered every 21 days, followed by an the univariate analysis. All statistical tests were two-sided,
additional two cycles of a combination regimen involving and significance was considered at P < 0.05. Bonferroni
cisplatin and 5-fluorouracil in the adjuvant setting. corrections were used to correct treatment variables for
Supportive measures, such as nutritional supplements and multiplicity, which reduced the risk of random false-
antiemetic medications, were prescribed as required. positive results when comparing three or more subgroups.
The adjusted P-values obtained from this correction were
2.5. Follow-up oral examination then used to define the significance levels (significance
Dental and oral evaluations were performed during the level for Bonferroni corrected P-value for three possible
follow-up appointments using the same protocol outlined comparisons was at 0.0167 [0.05÷3]).

Volume 3 Issue 1 (2025) 48 doi: 10.36922/arnm.5799