Pegfilgrastim (Neulasta) is a recombinant human granulocyte-colony stimulating factor (G-CSF) agent used for primary or secondary prophylaxis of febr ile neutropenia in patients receiving myelosuppressive chemotherapy. It is administered as a single 6 mg subcutaneous injection 24–72 hours after completion of a chemotherapy cycle. Pain is a common side effect and, although G-CSF-associated bone pain is usually mild to moderate, severe pain can occur (Gregory et al, 2010; Bondarenko et al, 2016; Xu et al, 2016). The exact mechanisms of bone pain related to G-CSF have yet to be elucidated, but a younger age and a history of bone pain have been identified as risk factors (Xu et al, 2016). In addition, it is reported that administering G-CSF at 14:00 (as opposed to 10:00) results in significantly less pain for patients (Özkaraman et al, 2017).
Pain medications commonly administered for bone pain include acetaminophen and nonsteroidal anti-inflammatory drugs (NSAIDs). However, some patients are unable to take these medications or find them ineffective. Bone pain can often be so severe that it results in G-CSF being discontinued or delayed, which renders patients more susceptible to febrile neutropenia, which may require hospital admission.
Histamine has been suggested as a chemical mediator that causes pain associated with bone marrow oedema. Although the role of G-CSF in histamine release is unclear, this concept has been hypothesised as the basis for anecdotal use of antihistamines to treat G-CSF-induced bone pain (Romeo et al, 2015). Loratadine (Clarityn) is a long-acting, non-sedating antihistamine that is used for the treatment of allergies (Casale et al, 2003). Histamine is released from mast cells (where it is stored) and attaches to other cells that have receptors for histamine. It then stimulates these cells to release chemicals that cause the signs and symptoms of allergy. Loratadine acts by blocking a histamine receptor—the H1 receptor. An advantage of loratadine is that it does not to cross the blood–brain barrier, so is unlikely to cause drowsiness when taken at a recommended dose (Casale et al, 2003).
Some recent papers have reported on the role of antihistamines in the management of G-CSF-induced bone pain (Pawloski et al, 2015; Romeo et al, 2015; Moukharskaya et al, 2016; Gavioli and Abrams, 2017; Moore and Haroz, 2017). However, evidence supporting the use of loratadine remains sparse (Moore and Pellegrino, 2017). The strongest evidence in support of loratadine for pain relief with G-CSF is reported in a recent randomised multicentre study (NOLAN: Naproxen Or Loratadine And Neulasta; ClinicalTrials.gov identifier: NCT01712009), which evaluated the effect of prophylactic naproxen, prophylactic loratadine or no prophylactic treatment on bone pain in patients with breast cancer receiving chemotherapy and pegfilgrastim (Kirshner et al, 2018). The study was conducted with 600 patients attending 83 centres in the US: loratadine group n=202; naproxen group n= 200; no prophylaxis group n=198. A total of 391 patients (97.3%) received at least one dose of naproxen or loratadine (n=198 received loratadine). The researchers found bone pain was lower in the naproxen and loratadine groups than in the group that received no prophylaxis. Moreover, the patients who were randomised to the loratadine group reported fewer treatment-related adverse events than those in the naproxen group.
Methodology
The primary aim of this pilot study was to ascertain if bone pain in a heterogeneous group of patients with cancer induced by G-CSFs could be alleviated or eliminated by loratadine. A secondary aim was to determine if there was any association between patients' vitamin D and magnesium levels and pain. It has been reported that vitamin D deficiency increases musculoskeletal pain (Shipton and Shipton, 2015; Wintermeyer et al, 2016), and it is postulated that magnesium deficiency can affect pain management as magnesium supplementation has been shown to reduce pain (Dabbagh et al, 2009; Bujalska-Zadrożny et al, 2017).
The study was undertaken at a nurse-led satellite oncology unit in a rural area in Ireland. Ethical approval to undertake the study was granted by the regional university hospital's ethics.
A non-experimental prospective design was used. All patients were screened by the first author (an advanced nurse practitioner in oncology) at their education session before they started chemotherapy between January 2017 and August 2017 to determine if they met the inclusion criteria, which were that they were undergoing chemotherapy for a diagnosis of breast, colorectal or genitourinary cancer and would be receiving G-CSF as part of their regimen. Exclusion criteria were: taking daily antihistamines for allergies, asthma or other indications not including bone pain; taking daily NSAIDs, with the exception of aspirin, for long-term conditions; or receiving bone-modifying agents for bone pain associated with metastatic disease or other long-term conditions. Patients with conditions that meant loratadine was contraindicated were also excluded. These conditions included severe liver impairment, hereditary galactose intolerance, prostatic hypertrophy, urinary retention and glaucoma. Three potential participants who fitted the study's inclusion criteria were later excluded as they had independently taken oral loratadine after being informed about the study before starting cycle 1 of chemotherapy.
Written consent was obtained from all patients before they were included in the study. The study questionnaire consisted of demographic questions and biophysical data, including vitamin D and magnesium levels.
Pain was measured using the Brief Pain Inventory (BPI) (Cleeland and Ryan, 1994). The BPI was self-administered by each patient with a nurse available for assistance or clarification if necessary. The BPI was completed before the first cycle of chemotherapy, again on day 7 after the first cycle and on the first day of each cycle thereafter.
The worst pain (WPS Q3) and average pain scale (APS Q5) questions of the BPI assess the rates of worst or average pain respectively on a scale of 0–10, with a higher score reflecting higher pain severity. Sites of pain were also used to identify the main areas of pain cited by patients over the different cycles of chemotherapy. The BPI includes diagrams showing the front and the back of the entire body, and patients are asked to shade the areas of pain.
Oral loratadine was started on day 1 of cycle 2 if the patient's pain was ≥3 at its worst (Q3 on BPI). This level of pain was chosen because pain <3 is mild pain. The patient was instructed to take loratadine 10 mg for each of the remaining cycles of chemotherapy thereafter (the day of their chemotherapy and the following six consecutive days). The patient's BPI was recorded at each cycle of chemotherapy thereafter.
Data analysis
Data were analysed using the statistical package for the social sciences IBM SPSS version 24. Descriptive statistics were used to compare variables and displayed in tables and scatterplots using frequencies, means, medians and standard deviations. The WPS Q3 and APS Q5 of the BPI were used to compare pain with other study variables. Inferential statistics using parametric two-tailed t-tests and Pearson's correlations were carried out to explore any potential relationship between pain and magnesium and vitamin D levels among participants.
Results
Participants and biophysical data
Twelve patients took part in the study (2 men and 10 women). The main diagnoses were breast (n=7), genitourinary (n=4) and colorectal cancer (n=1). Patients' sociodemographic characteristics are summarised in Table 1. They were in the 29-72 years age range, with a mean age (SD) of 54.58 (11.65) years overall, 47 (19) for men (n=2) and 56 (8) for women (n=10). Participants' BMI was in the 20.20–38.78 kg/m2 range, with the mean BMI (SD) being 27.84 kg/m2 (6.24 kg/m2), which falls into the overweight category (25–30 kg/m2). Women had a slightly higher BMI of 29.10 kg/m2 (5.84 kg/m2) than men at 21.60 kg/m2 (0.10 kg/m2). However, there were only two men in the study.
Table 1. Sociodemographic characteristics of study participants
| Variable | Participant (n=12) |
|---|---|
| Sex | n |
| Male | 2 |
| Female | 10 |
| Age | Mean (SD) |
| All | 54.58 (11.65) |
| Male | 47 (19) |
| Female | 56 (8) |
| BMI | (kg/m2) mean (SD) |
| All | 27.84 (6.24) |
| Male | 21.60 (0.10) |
| Female | 29.10 (5.84) |
| Smoking status | n (%) |
| Smoker | 0 (0) |
| Non-smoker | 12 (100) |
The participants' vitamin D levels were in the range of 14–106 ng/ml, with the overall mean baseline vitamin D level (SD) being 48.45 ng/ml (25.39 ng/ml), which is within normal vitamin D level parameters (20–50 ng/ml). Four participants were taking vitamin D supplements for low vitamin D levels before the study began. It should be noted that one participant had a very high vitamin D level of 106 ng/ml and was not taking vitamin D supplementation.
Magnesium levels across the sample were in the 0.66–0.9 mmol/l range, with the overall mean (SD) baseline magnesium level being 0.773 mmol/l (0.883 mmol/l), which is within the normal magnesium level parameters (0.7–1.10 mmol/l) (Williamson et al, 2011). None of the patients took magnesium supplements.
All participants were fully active and had an ECOG score of 0 (ECOG-ACRIN Cancer Research Group, 2019) and were non-smokers.
The patients' clinical characteristics are summarised in Tables 2 and 3. The number of cycles of chemotherapy varied between group members, ranging from a minimum of two cycles (n=1; 8%) to a maximum of five cycles (n=3; 17%). Half of the group (n=6) underwent four cycles of chemotherapy.
Table 2. Clinical characteristics of study participants (n=12)
| Variable | n | Percentage |
|---|---|---|
| Cancer diagnosis | ||
| Breast | 7 | 58 |
| Genitourinary | 4 | 33 |
| Colorectal | 1 | 8 |
| Chemotherapy regimen | ||
| Paclitaxel/carboplatin | 3 | 25 |
| Docetaxel (Taxotere), cyclophosphamide and trastuzumab (Herceptin) (TCH) | 4 | 33 |
| Doxorubicin (Adriamycin) and cyclophosphamide four cycles, and paclitaxel (Taxol) four cycles x 2/52 | 2 | 17 |
| Doxorubicin (Adriamycin) and cyclophosphamide | 1 | 8 |
| Folfox (folinic acid, calcium folinate, fluorouraci and oxaliplatin) | 1 | 8 |
| Paclitaxel (Taxol) | 1 | 8 |
| Chemo cycles completed | ||
| 1 | 0 | 0 |
| 2 | 1 | 8 |
| 3 | 2 | 17 |
| 4 | 6 | 50 |
| 5 | 3 | 25 |
| Previous medical conditions | ||
| Past psychiatric history | 1 | 8 |
| Type 2 diabetes, arthritis and pre-existing chronic pain | 1 | 8 |
| None | 10 | 83 |
| ECOG score | 12 | 100 |
Table 3. Haematological results and supplements
| Variable | Number (percentage) of participants (n=12) |
|---|---|
| Vitamin D levels | ng/ml: mean (SD) |
| All | 48.45 (25.39) |
| Men | 50 (1) |
| Women | 48 (27) |
| Vitamin D supplement | n (%) |
| Yes | 3 (25) |
| No | 9 (75) |
| Magnesium levels | mmol/l: mean (SD) |
| All | 0.773 (0.883) |
| Men | 0.87(0.03) |
| Women | 0.75 (0.08) |
| Magnesium supplement | n (%) |
| Yes | 0 (0) |
| No | 0 (0) |
Areas of reported pain
The most common area of pain reported across all cycles was lower front and back pain on both sides (n=15; 48.8%), which accounted for almost half of all the total reported by all participants. The front and back right upper body was the second cited area of pain (n=4; 12.9%). Other areas of pain reported included front right (n=3; 9.7%), back both sides (n=3; 9.7%), lower back both sides (n=2; 6.5%) and skull and neck (n=2; 6.5%)
Pain medication
Eight patients were taking no pain medication before they started their first cycle of chemotherapy. All participants were started on oral loratadine on cycle 2, with three participants also taking additional pain medication concurrently; these were ibuprofen (n=1) or tramadol (n=2). The amount of additional pain medication taken towards the later cycles decreased after loratadine was started in cycle 2, with the exception of one participant who took tramadol as needed, including in cycle 3 (Table 4). It is important to note that this patient had a BMI of 33.45, a very high baseline vitamin D level of 106 ng/ml, a particularly low magnesium level of 0.67 mmol/l and was taking pain relief medications as needed after cycles 1, 2 and 3. Magnesium and vitamin D levels may have affected her overall reported pain levels. Two patients did not report pain relief from loratadine from commencement in cycle 2. One of these patients had a very low vitamin D level of 14 ng/ml and completed two cycles of chemotherapy.
Table 4. Pain medication and treatment at each chemotherapy cycle
| Pain medication | Before cycle 1 | After cycle 1 | Cycle 2 | Cycle 3 | Cycle 4 | Cycle 5 |
|---|---|---|---|---|---|---|
| None n (%) | 8 (67%) | 6 (50%) | 0 | 0 | 0 | 0 |
| Loratadine 10 mg n (%) | 0 | 0 | 9 (75%) | 9 (75%) | 9 (75%) | 3 (25%) |
| Loratadine and Ibuprofen as needed n (%) | 0 | 0 | 1 (8%) | 0 | 0 | 0 |
| Loratadine & Tramadol as needed n (%) | 0 | 0 | 2 (17%) | 1 (8%) | 0 | 0 |
| Ibuprofen as needed n (%) | 1 (8%) | 1 (8%) | 0 | 0 | 0 | 0 |
| Tramadol as needed n (%) | 2 (17%) | 5 (42%) | 0 | 0 | 0 | 0 |
| Difene 75 mg as needed n (%) | 1 | 0 | 0 | 0 | 0 | 0 |
Daily pain initially rose between before cycle 1 started and the period after it, decreasing thereafter (Figure 1). Mean average pain diminished over time. Participants reported increasing relief from pain up to cycle 3 but this reduced in cycle 4.
Average pain (before cycle 1) and vitamin D levels appeared to be negatively correlated. Pearson's correlation was–0.677 and this is statistically significant (p=0.045). However, it should be noted that this parametric testing was carried out on a small sample size so should be interpreted with caution.
A similar examination of the effect of magnesium levels on pain found no evidence that this was a factor before cycle one, but there was a statistically significant negative correlation evident at cycles 2 (r=–0.642; p=0.024) and 4 (r=–0.822; p=0.012) (Table 5).
Table 5. Correlations between study variables
| Magnesium level | Vitamin D levels | Average pain before cycle 1 | Average pain after cycle 1 | Average pain at cycle 2 | Average pain at cycle 3 | Average pain at cycle 4 | Average pain at cycle 5 | |
|---|---|---|---|---|---|---|---|---|
| BMI | r=–0.143p=0.658n=12 | 0.3930.2311 | –0.2940.4110 | 0.0360.91711 | –0.2830.37312 | –0.4290.21610 | –0.2630.5298 | –0.4760.5244 |
| Magnesium level | 0.0860.80211 | –0.0550.87910 | –0.5220.09911 | –0.642▪0.02412 | –0.3070.38810 | –0.822▪0.0128 | –0.9140.0864 | |
| Vitamin D levels | –0.677▪0.0459 | –0.1990.58110 | 0.0470.8911 | 0.1480.7049 | 0.899▪▪0.0067 | 03 | ||
| Average pain before cycle 1 | 0.751▪0.029 | 0.2890.41810 | 0.1410.748 | 0.4480.3736 | 0.8160.1844 | |||
| Average pain after cycle 1 | 0.674▪0.02311 | 0.4520.2229 | 0.777▪0.047 | 0.7860.4253 | ||||
| Average pain at cycle 2 | 0.854▪▪0.00210 | 0.8930.0038 | 0.9050.0954 | |||||
| Average pain at cycle 3 | 0.5390.1688 | 0.3330.6674 | ||||||
| Average pain at cycle 4 | 1.000▪▪04 |
Discussion
This study provides evidence that supports the use of loratadine in the management of G-CSF-induced bone pain in patients receiving chemotherapy for cancer. Support for the use of loratadine has been shown in two case studies (Romeo et al, 2015; Moore and Haroz, 2017), in a recent retrospective cohort study (Gavioli and Abrams, 2017) and in the large randomised study with breast cancer patients described above (Kirshner et al, 2018).
However, these findings are not supported by a recent trial on the use of loratadine to manage G-CSF pain (Moukharskaya et al, 2016). In this trial, out of a total of 213 patients, 46 with severe pain on the BPI were divided into those receiving loratadine (n=22) and a placebo. While the researchers concluded that prophylactic loratadine did not decrease the incidence of severe bone pain, they pointed out that an exploratory analysis had demonstrated that patients receiving taxane-based chemotherapy benefited from loratadine. This may explain the findings reported here as most patients in the present study had breast cancer and were on a taxol-based regimen, and it is widely documented that taxane therapy is associated with arthralgias and myalgias. Moreover, the findings of the present study could also be explained by the measurement of the study participants' pain up to cycle 5. Bone pain associated with G-CSF improves with subsequent cycles (Gregory et al, 2010) and the trial reported by Moukharskaya et al (2016) measured pain only on days 1 and 8.
Of note also is that participants in this study reported that pain relief increased up to cycle 3 then reduced in cycle 4. This reduction may be because one participant who completed cycle 4 had a pre-existing chronic pain condition, which would have had a greater influence on the findings than would happen with a large sample size. It is important to note that one of the patients who reported no pain relief from loratadine had a BMI of 33.45, a very high baseline vitamin D level of 106 ng/ml, a particularly low magnesium level of 0.67 mmol/l and was taking pain relief medications as needed after cycles 1, 2 and 3, which may have affected her overall reported pain levels. Two patients did not report relief of pain from loratadine from when it was started in cycle 2; one of these had a very low vitamin D level of 14 ng/ml and completed two cycles of chemotherapy.
Limitations
This was a pilot study with a number of limitations. The sample size was small so parametric testing on any potential relationship between vitamin D and magnesium levels should be interpreted with caution. The pilot study site was a satellite centre so the study was constrained by the numbers of patients attending for treatment within the planned time frame. Also, three potential participants who met the study's inclusion criteria could not be included as they had independently taken loratadine after being informed about the study before starting cycle 1 of chemotherapy.
The study protocol was to measure pain using the BPI on day 7 after the first course of chemotherapy. However, not all patients could return to the unit to meet with the researcher (first author) and have their BPI measured on day 7 after the first course of chemotherapy. These patients' BPI was recorded retrospectively, which relied on the their recall of their pain.
Conclusions
This pilot study has shown that oral loratadine can provide effective pain management in patients with cancer who develop pain after their first dose of G-CSF. Treatment with oral loratadine once daily for five days in each chemotherapy cycle for patients with cancer receiving chemotherapy and pegfilgrastim is advised (Kirshner et al, 2018).
Future studies are also needed to determine if patients on taxane-based regimens benefit more from the pain-relieving properties of loratadine than those on other regimens. Moreover, future studies are needed to determine if vitamin D and magnesium levels play a role in affecting the pain-relieving properties of loratadine in patients with cancer.
Key Points
- Bone pain associated with G-CSF (granulocyte-colony stimulating factor) chemotherapy agents such as pegfilgrastim is a common side effect experienced by patients with cancer
- Bone pain can often be so severe that it results in G-CSF being discontinued or delayed, which renders patients more susceptible to febrile neutropenia, which may require hospital admission
- An increasing number of studies have reported on the role of antihistamines in the management of G-CSF-induced bone pain. However, evidence supporting the use of loratadine remains sparse
- This study provides further evidence in support for the use of loratadine in the management of G-CSF-induced bone pain among patients receiving chemotherapy for cancer
CPD reflective questions
- Before patients start chemotherapy, do you ask them if they are taking Vitamin D supplementation?
- How do you manage pain in your patients receiving granulocyte-colony stimulating factor (G-CSF) chemotherapy?
- Do you check patients' magnesium levels before they start G-CSF?