2018-19 Section 7-Neoplastic and Inflammatory Diseases of the Head and Neck eBook

Oral Oncology 79 (2018) 9–14

C.L. Barney et al.

Table 3 Multivariate analysis of prognostic factors for clinical outcomes (all patients).

Variable

Clinical outcomes

LRC

DMFS

RFS

OS

HR (95%CI)

p-value

HR (95%CI)

p-value

HR (95%CI)

p-value

HR (95%CI)

p-value

ECOG PS (1 vs. 0)

0.61 (0.28 – 1.33) 0.60 (0.27 – 1.32 0.54 (0.25 – 1.16) 0.45 (0.21 – 0.97)

.21 .20 .11 .04

0.58 (0.28 – 1.21) 0.46 (0.22 – 0.97) 0.38 (0.18 – 0.81) 0.55 (0.26 – 1.14) 0.78 (0.39 – 1.58) 0.26 (0.12 – 0.53)

.15 .04 .01 .11 .49

– –

– –

– –

– –

Smoking (pkyrs. > 10 vs. ≤ 10)

.03 .17 .08

ICON-S (N2-3 vs. N0-N1) T-Stage (T3-4 vs. T1-T2)

0.45 (0.22 – 0.94) 0.61 (0.30 – 1.24) 0.54 (0.27 – 1.08) 0.33 (0.16 – 0.64)

0.60 (0.33 – 1.08) 0.52 (0.29 – 0.93) 0.61 (0.35 – 1.07) 0.21 (0.12 – 0.38)

.09 .03 .08

Age( ≥ 60 vs. <60)

–

–

< .001

.001

< .001

< .001

Cetuximab

0.20 (0.09 – 0.44)

Abbreviations : LRC= locoregional control; DMFS = distant metastasis-free survival; RFS = recurrence-free survival; OS =overall survival; ECOG PS = Eastern Cooperative Oncology Group performance status; ICON-S = International Collaboration on Oropharyngeal cancer Network for Staging.

given. As previously mentioned, the improvement in DMFS with cisplatin compared to cetuximab may be a downstream e ff ect of better LRC in the cisplatin group. A more potent direct action of cisplatin on micro- scopic metastatic disease compared to that of cetuximab may also be a possibility, although, further research is needed to test this hypothesis. Of course, biases inherent in retrospective studies may also contribute to di ff erences in outcomes. Older age and di ff erences in baseline co- morbid conditions may have in fl uenced outcomes between groups; al- ternatively, other in fl uential factors such as lead time bias and in- creased use of AFRT may have acted in favor of the cetuximab cohort. Despite these biases, the magnitude of outcome di ff erences between our treatments groups strongly suggests that the threshold for which to substitute cisplatin with cetuximab should remain high. According to a recent Surveillance Epidemiology and End Results (SEER) database analysis, the use of concurrent cisplatin in CRT for head and neck cancers has been slowly declining since 2006, even in patients with low comorbidity scores, as has the use of other platinum regimens [7] . In context of our results, such a trend can seem concerning, however, further data is needed to understand which factors are most in fl uential in chemotherapy decision-making. Patients with high comorbidity scores, poor performance status, and/or elderly patients treated with cisplatin are at increased risk of developing severe side e ff ects [1,18,19] . However, the use of concurrent cetuximab may also be no better than RT alone in such patients (age > 65 years or KPS < 80) [6] . In patients who are borderline cisplatin eligible, shorter interval dosing of cisplatin (weekly or daily) allows for more prompt regimen adjustments should toxicities arise, potentially mitigating long-term morbidity. A number of prior retrospective studies have suggested that these schedules are equally e ff ective to high-dose cisplatin in non-HPV and HPV-related disease [20 – 22] . However, a recently published ran- domized trial reported high-dose cisplatin (100 mg/m 2 every 3 weeks) to be associated with improved 2-yr LRC (58.5 vs 73.1%; HR 1.76) compared to weekly cisplatin (30 mg/m 2 ), but did not show a statisti- cally signi fi cant di ff erence in median progression-free or overall sur- vival [23] . Given these con fl icting fi ndings, there is still debate on weekly cisplatin dosing. In our analysis, there was no di ff erence in clinical outcomes when 40 mg/m 2 cisplatin was given weekly con- current with RT. Other concerns, such as baseline hearing loss, should not always preclude the use of concurrent cisplatin in the absence of other comorbidities. For example, in the Veterans A ff airs healthcare system, where the majority of patients enter treatment with hearing loss, a comprehensive ototoxicity monitoring program has been estab- lished. This protocol aims to increase access to standard of care ototoxic treatments, such as cisplatin, while minimizing risk of further hearing loss [24,25] . While e ff orts to minimize CRT-related toxicity in p16+ OPC con- tinue to be explored, their impact on disease control and survival is of primary concern. Our study suggests that the threshold for cisplatin ineligibility should be maintained appropriately high, and o ff -trial use

Fig. 2. Overlaid Kaplan Meier plots of overall survival in patient s with p16 positive oropharyngeal carcinoma treated with radiotherapy plus concurrent cisplatin (blue) versus cetuximab (yellow). Recursive partitioning analysis (RPA) was used to stratify cohorts into low-risk (LR-RPA; solid lines) and intermediate-risk (IR-RPA; dashed lines) groups. (For interpretation of the references to colour in this fi gure legend, the reader is referred to the web version of this article.)

Fig. 3. Kaplan Meier plot with log-rank p-value of locoregional control in patients with p16 positive oropharyngeal carcinoma treated with altered fractionation radiotherapy (AFRT; blue) versus standard fraction radiotherapy (SFRT; yellow) plus concurrent ce- tuximab. (For interpretation of the references to colour in this fi gure legend, the reader is referred to the web version of this article.)

97