Among women diagnosed with early-stage breast cancer enrolled in the Treatment-Induced Nausea and Vomiting study (N = 287) at the University of Pittsburgh (NR0196556), 63.8 percent underwent adjuvant radiotherapy (n = 183), and the prevalence of RIN was 41.0 percent (n = 75).
Baseline Characteristics: Cases versus Controls
The baseline characteristics of the 183 women included in the current analysis are summarized in Table 1. Compared to controls who reported no RIN, cases reported higher average pain levels (p < 0.0001). The mean pain score among participants who experienced RIN was nearly twice that of participants who did not experience RIN. Cases reported a mean pain score of 3.92/10, while controls reported a mean pain level of 1.89/10. Less than one third of controls reported use of over-the-counter medications to manage pain, and only three reported using prescription medication to manage it. In contrast, nearly two-thirds of cases reported use of pain medication at least once, with 36 using over-the-counter medications and 11 using prescription medications. Overall, cases were significantly younger (p = 0.024), with a mean age of 58.87 years (median: 59.0) compared to controls, who had a mean age of 62.82 years (median: 65.0). Cases reported significantly higher rates of history of nausea, including family history of nausea and vomiting (p = 0.052), history of motion sickness (p = 0.051), and history of postoperative nausea and vomiting (p = 0.048).
Cases and controls did not significantly differ in estrogen receptor positivity, history of nausea during pregnancy, progesterone receptor positivity, surgical side, smoking history, surgery type, history of neoadjuvant therapy, handedness, or race.
Nausea Findings
Among cases, mean nausea severity was 3.27/10, with a median score of 3.00, and mean scores ranging from 0.25 to 10.
Table 1: Participant Characteristics by Radiation Induced Nausea (RIN) Status (N=183)
Characteristic
|
No RIN (n=108)
n (%)
|
RIN (n=75)
n (%)
|
Odds Ratio (95% Confidence Interval)
|
p-value
|
Concurrent Treatment
|
No: 90 (83.3%)
Yes: 18 (16.7%)
|
No: 55 (73.3%)
Yes: 20 (26.7%)
|
1.82
(0.89, 3.73)
|
0.101
|
Estrogen Receptor
|
Negative: 26 (25.0%)
Positive: 78 (75.0%)
|
Negative: 16 (21.9%)
Positive: 57 (78.1%)
|
1.19
(0.58, 2.42)
|
0.635
|
Family History of Nausea and Vomiting
|
No: 87 (80.6%)
Yes: 21 (19.4%)
|
No: 51 (68.0%)
Yes: 24 (32.0%)
|
1.95
(0.99, 3.85)
|
0.052
|
Motion Sickness
|
No: 73 (67.6%)
Yes: 35 (32.4%)
|
No: 40 (53.3%)
Yes: 35 (46.7%)
|
1.83
(0.99, 3.35)
|
0.051
|
Prescription Opioids[1]
|
No: 105 (97.2%)
Yes: 3 (2.8%)
|
No: 64 (85.3%)
Yes: 11 (14.7%)
|
8.86
(1.62, 22.38)
|
0.004
|
Pregnancy Nausea
|
No: 73 (67.6%)
Yes: 35 (32.4%)
|
No: 41 (54.7%)
Yes: 34 (45.3%)
|
1.73
(0.94, 3.18)
|
0.076
|
Postoperative Nausea
|
No: 86 (79.6%)
Yes: 22 (20.4%)
|
No: 50 (66.7%)
Yes: 25 (33.3%)
|
1.95
(1.01, 3.82)
|
0.048
|
Progesterone Receptor
|
Negative: 27 (26.5%)
Positive: 75 (73.5%)
|
Negative: 21 (28.8%)
Positive: 52 (71.2%)
|
0.89
(0.46, 1.74)
|
0.737
|
Surgical Side*
|
Left: 51 (47.2%)
Bilateral: 13 (12.0%)
Right: 44 (40.8%)
|
Left: 37 (49.3%)
Bilateral: 8 (10.7%)
Right: 30 (40.0%)
|
1.09
(0.60, 1.96)
|
0.942
|
Smoking History
|
Nonsmoker: 102 (94.4%)
Smoker: 6 (5.6%)
|
Nonsmoker: 68 (90.7%)
Smoker: 7 (9.3%)
|
1.75
(0.56, 5.43)
|
0.328
|
Mastectomy Type
|
Segmental: 88 (81.5%)
Total: 20 (18.5%)
|
Segmental: 58 (77.3%)
Total: 17 (22.7%)
|
1.29
(0.62, 2.67)
|
0.492
|
Neoadjuvant Therapy
|
No: 82 (75.9%)
Yes: 25 (24.1%)
|
No: 56 (74.7%)
Yes: 19 (25.3%)
|
1.11
(0.56, 2.21)
|
0.760
|
Handedness
|
Right: 86 (79.6%)
Left: 14 (20.4%)
|
Right: 61 (81.3%)
Left: 6 (18.7%)
|
0.60
(0.22, 1.66)
|
0.325
|
Race
|
White: 94 (87.0%)
Nonwhite: 14 (13.0%)
|
White: 67 (89.3%)
Nonwhite: 8 (10.7%)
|
0.80
(0.32, 2.02)
|
0.639
|
|
Mean (SD) – Controls
|
Mean (SD) – Cases
|
|
|
Average Pain Level
(0-10 scale)
|
1.89 (2.05)
|
3.92 (2.83)
|
1.390
(1.211, 1.596)
|
<0.001
|
Age (years)
|
62.82 (11.74)
|
58.87 (10.78)
|
0.970
(0.945, 0.996)
|
0.024
|
*Left breast surgery used as referent category for odds ratio calculation.
Note: “Yes,” “Positive,” “Smoker,” and “Left” were used as the present variables in odds ratio calculations
[1] Fisher’s Exact test used
Table 2: Association of Risk Factors with the Probability of Having RIN - Logistic Regression
Variables
|
Crude Odds Ratio
|
Wald Chi-squared p-value
|
Adjusted Odds Ratio
|
Wald Chi-squared
p-value
|
95% Confidence Interval
|
Lower
|
Upper
|
Pain
|
1.390
|
<0.0001
|
1.536
|
<0.0001
|
1.268
|
1.862
|
Motion Sickness1
|
1.825
|
0.052
|
2.280
|
0.024
|
1.112
|
4.673
|
Age (years)
|
0.970
|
0.024
|
0.965
|
0.032
|
0.933
|
0.997
|
Postoperative Nausea2
|
1.955
|
0.050
|
1.847
|
0.116
|
0.860
|
3.965
|
Prescription Opioids3
|
8.857
|
0.004
|
1.339
|
0.065
|
0.918
|
15.854
|
1 Reference category for history of motion sickness is “No history of motion sickness.”
2 Reference category for history of postoperative nausea is “No history of postoperative nausea.”
3 Reference category for prescription opioids is “No prescription opioids.”
Binary Logistic Regression for RIN
Table 2 summarizes the results of the multivariable binary logistic regression analysis of RIN prevalence predictors. In this model, average pain level, age and having a history of motion sickness were significant predictors of RIN. Adjusting for all other variables in the model, an increase in pain by one unit of measure increased the odds ratio of RIN by 1.54 (OR = 1.536, 95% CI = 1.268–1.862, p < 0.0001). In this model, history of motion sickness doubled the odds of RIN (OR = 2.280, 95% CI = 1.112–4.673, p = 0.024). An increase in age by one year resulted in 0.03 lower odds of reporting RIN (OR = 0.970, 95% CI = 0.933–0.997, p = 0.032). There was a positive relationship between prescription opioid use and RIN prevalence, but this relationship was not significant (p = 0.065).
Linear Regression for RIN Severity
Table 3 reports the results from the multiple linear regression analysis where average nausea severity was the dependent variable. The full multiple linear regression model yielded an adjusted R2 of 0.198. Thus, 19.8 percent of variation in RIN severity can be explained by the predictors in this model. Average pain level and history of postoperative nausea were significantly associated with increased RIN severity. For every one-point increase in average pain level, average RIN severity increased by nearly one-half of a point (p < 0.0001). If history of postoperative nausea and vomiting was reported, average RIN severity increased by 0.138 points (p = 0.042). Older age was significantly associated with decreased RIN severity. For every one-year increase in age, average RIN severity decreased by 0.144 points (p = 0.038).
Table 3: Association of Risk Factors with Average Radiation Induced Nausea Severity - Linear Regression
Variable
|
Unstandardized Regression Coefficient, b
|
Standard Error for b
|
t-value
|
p-value
|
Standardized Regression Coefficient
|
95% Confidence Interval Limits for b
|
|
|
Pain
|
0.344
|
0.057
|
6.045
|
<0.001
|
0.413
|
0.232
|
0.456
|
|
Age
|
-0.026
|
0.012
|
-2.088
|
0.038
|
-0.144
|
-0.050
|
-0.001
|
|
Postoperative Nausea
|
0.655
|
0.320
|
2.045
|
0.042
|
0.138
|
0.023
|
1.286
|
|
Neoadjuvant Chemotherapy
|
-0.614
|
0.343
|
-1.790
|
0.075
|
-0.127
|
-1.291
|
0.063
|
|