br Statistics br Descriptive statistics for patients charact
Descriptive statistics for patients' characteristics, Background pain and BTcP have been provided in terms of mean values and frequencies according with H&N cancer occurrence. Comparability of groups, af-fected or not aﬀected by H&N cancer, have been checked performing a multivariate Hotelling T2 test, setting a first type error at 5%, followed by pairwise comparisons with Bonferroni's correction for familywise error. Gender and age have been tested for confounding. A logistic re-gression provided diﬀerent probabilities of H&N cancer occurrence conditioning on gender. In the following analysis confounding has been controlled by stratification. χ2 tests have been used to assess the sta-tistical association between H&N cancer and Breakthrough pain char-acteristics. Continuous variables pair-wise comparisons have been performed using t-tests, coping with the lack of homoscedasticity be-tween groups by the Satterthwaite correction. The statistical analysis Oral Oncology 95 (2019) 87–90
Characteristics of patients with H&N cancer and other primary tumors.
Age (yrs, mean
has been carried out using the statistical software STATA (StataCorp, College Station, Texas, version 14).
In comparison with patients with other primary diagnoses recorded in the original study, no significant diﬀerences were found in pain mechanism (p = 0.968), background pain intensity (p = 0.932), the use of nonsteroidal anti-inflammatory drugs and paracetamol (p = 0.962 and p = 0.101, respectively) (see below).
S. Mercadante, et al.
The mean time to meaningful pain relief after taking a BTcP med-ication was 15.3 min (SD = 10.04). No statistical diﬀerences with other types of tumor were found (p = 0.376).
Analgesics used for background pain
Opioids used for background pain in patients with and without H&N cancer (some of them also in combination).
Opioids used for BTcP in patients with and without H&N cancer.
Analgesics used for BTcP
Opioids used for BTcP in patients with and without H&N cancer AZD 2281 reported in Table 3.
Nasal fentanyl preparations were more frequently used in patients H &N cancer, while sublingual fentanyl (SLF) was less frequently used (Table 3).
Level of mucositis and adverse eﬀects
Adverse eﬀects attributable to BTcP medications were found in 4 patients (2.8%), No diﬀerences with other tumors were found (3.7%, p = 0.551). No adverse eﬀects of severe intensity were reported (3 of mild intensity, 1 of moderate intensity)
The findings of this study provided relevant information regarding background pain and a phenomenon, BTcP, that has never been de-scribed in patients with H&N cancer. First, in comparison with the general population aﬀected by other primary tumors, H&N cancer pa-tients exhibited more episodes of BTcP, which were more predictable, particularly with the ingestion of food. This finding can be explained by the local characteristics of H&N cancer, which also suggest a more in-tensive treatment of background pain . Pain is a significant mor-bidity characterized by multifactorial mechanisms in patients with H&N cancer, also resulting from the treatments directed against the tumor. Pain should be managed with multiple medications in a multimodal approach . Of interest, corticosteroids were more frequently given in patients with H&N cancer. Transdermal drugs, particularly fentanyl, were more frequently prescribed in H&N cancer patients. Transdermal fentanyl may be an important option, thanks to the modality of ad-ministration, the good safety, and tolerability profile to control baseline pain, although head-to-head studies of fentanyl versus other strong opioids are lacking . In the only existing comparative study of patients with H&N cancer having a neuropathic component, methadone appeared more eﬀective than transdermal fentanyl . The use of methadone, which appeared more frequently prescribed in women,
S. Mercadante, et al.
may have a potential in neuropathic pain conditions, but in this study there is no apparent explanation for this finding, considering that this substance is really rarely prescribed in this country .
Secondly, a large number of patients had oral mucositis of severe intensity. This finding corresponds to the clinical condition of patients who received many treatments with high local toxicity potential, in-cluding surgery, chemotherapy, and radiotherapy. The presence of high levels of mucositis could have influenced the choice of drugs for both background pain (transdermal preparations) and above all BTcP, with nasal fentanyl preparations being more frequently prescribed. This choice could be explained by the diﬃculties in giving oral medications for background pain and oral transmucosal fentanyl preparations for BTcP. In fact, patients with H&N cancer are likely to have a serious mucosal damage making local absorption unpredictable. There are no existing studies regarding BTcP in patients with H&N cancer for com-parison.