Acute and persistent post-operative pain following mastectomy: a descriptive study in a tertiary hospital cohort

Breast cancer is the most common cancer affecting women in New Zealand,¹ and the third most common cancer overall on a population basis. Last published data, in 2012, showed patients from the Canterbury Region made up 14.5% of new breast cancer registrations in New Zealand.² Surgical treatment is the most common first intervention for these patients, and mastectomy is the most common surgical procedure performed.² Survival in female breast cancer in New Zealand is improving; 5-year survival in 1998/1999 was 79% compared with 89% for the same metric in 2016/2017.³

Persistent post-surgical pain following mastectomy is a well-described post-operative complication that can have significant negative effects on quality of life and function after mastectomy.⁴ Pain in the ipsilateral chest wall, arm, flank or axilla, often with neuropathic features, are typical complaints, and occur with variable frequency.⁵ Symptoms range from moderate to severe, and literature suggests, once present, they are likely to persist.⁵ The quoted prevalence of this condition in the literature ranges between 20 and 74%;^4–17 such wide variance makes it difficult to understand the true burden of the condition in a particular population.

The variable rates seen are likely to be in part due to lack of a universally accepted definition for post-mastectomy pain syndrome (PMPS).¹⁸ An attempt at a consensus definition of PMPS based on common elements across the existing literature was published in 2016.¹⁸ PMPS was defined as: pain that occurs after any breast surgery; is of at least moderate severity; possesses neuropathic qualities; is located in the ipsilateral breast/chest wall, axilla and/or arm; lasts at least 6 months (although acknowledges that, practically, 3 months is an adequate duration for research); occurs at least 50% of the time; and may be exacerbated by movements of the shoulder girdle.¹⁸ This definition now presents consistent criteria for diagnosing PMPS, allowing research and audit in different populations.

The incidence and severity of persistent pain post-mastectomy is somewhat surprising to many clinicians given the usual in-hospital course for these patients; the procedure appears well tolerated, is typically managed with oral analgesia post-operatively and is associated with a relatively short length of stay.¹⁹ The cause of PMPS is not precisely known, but is almost certainly multifactorial.²⁰ Many risk factors have been identified, including younger age, non-European ethnicity, pre-operative depression, anxiety or catastrophising, radiation therapy, axillary surgery and the presence of pre-operative pain.^4–17,20 Many of these factors appear significantly associated in some studies but are not reproduced in others. There are also reported associations with severe acute pre-operative pain, but studies often rely on recall of acute pain, which can be unreliable.^4,12,17

Peri-operative decision making about surgical and anaesthetic techniques, as well as appropriate counselling of patients, requires local data to quantify the risk of the development of PMPS in a given population. In the New Zealand population data, examining PMPS is sparse; there is one publication only that has studied its prevalence, which was published after our study was commenced²⁰ and did not examine a possible correlation between acute pain after mastectomy and the development of PMPS. This is important, as acute pain may be modifiable through surgical and anaesthetic techniques that could alter the long-term outcome for the patient.⁴

The aims of this study were:

1. To estimate the prevalence of PMPS for a cohort of mastectomy patients in Canterbury.
2. To describe post-operative analgesic requirements for the same cohort.
3. To explore risk factors for development of persistent post-surgical pain, including possible association with severity of acute post-operative pain.

Method

We designed a single-centre, retrospective, cross-sectional observational study, to include all women who had undergone mastectomy in Christchurch, New Zealand, administered by Canterbury District Health Board (CDHB) between August 2017 and August 2018, and met the inclusion criteria of the study. This allowed for at least 1 year between mastectomy and recruitment for the study.

Ethics approval was obtained for both the questionnaire-based component (New Zealand Health and Disabilities Ethics Committee reference 19/CEN/180, locality approval from CDHB Research Office and Te Komiti Whakarite) and for audit of peri-operative analgesia (CDHB Research Office and Te Komiti Whakarite).

Eligible patients included any patient over 18 years of age who underwent mastectomy without immediate reconstruction during the study period, including patients who had sentinel lymph node biopsy (SLNB) or axillary node dissection. Patients were identified using clinical coding data. Patients were excluded if unable to give informed consent to participate; unable to speak English and interpreter unavailable; and those with cancer recurrence, development of new primary cancer or metastatic disease.

Eligible patients were phoned by CDHB breast clinic nurses or study investigators to discuss involvement in the study and consent, and then a written consent form was posted to be signed. Patients who accepted the invitation to participate were allocated a study number, and then underwent a telephone-administered questionnaire based on the proposed consensus definition of PMPS (Appendix 1).¹⁸

Our primary outcome is to describe the prevalence of PMPS at least 1-year post-mastectomy in Christchurch. PMPS is defined as:

• pain that occurs after breast surgery
• is located in the ipsilateral breast/chest wall, axilla and/or arm
• is present at least 3 months after breast surgery
• is of at least moderate severity
• occurs at least 50% of the time
• possesses neuropathic qualities.¹⁸

Individual medical records were reviewed for demographic data, surgical and anaesthetic details, and the presence of possible risk factors for persistent post-surgical pain (age, body mass index [BMI], axillary surgery, radiation or chemotherapy, pre-existing persistent pain or opioid use and severity of acute post-operative pain after mastectomy). Acute pain data were obtained by retrospective chart review and expressed as use of strong (morphine/oxycodone/fentanyl), weak (codeine/tramadol) or no opioids, and the amount used in morphine equivalents.

Descriptive data about ongoing symptoms were also sought to aid better understanding of the range of symptoms in these patients.

Descriptive statistics were used to calculate the prevalence of PMPS. The associations between the putative predictors and post-surgical pain were tested using logistic regression models and are summarised using odds ratios (OR) and 95% confidence intervals (CI). When the observed frequencies were small, e.g., for PMPS, Fisher’s exact tests were used. A two-tailed p-value <0.05 was used to indicate statistical significance.

Results

Of 71 eligible patients, 59 responded to the telephone questionnaire, giving a response rate of 83% (Figure 1).

View Figure 1, Table 1–7.

Of 59 patients who participated in the telephone questionnaire, three (5%) met criteria for PMPS.

Thirty-six patients described persistent symptoms that did not meet the definition of PMPS at the time of questionnaire administration (61% of total). The demographic data of these patients are presented in Table 2.

Post-mastectomy pain syndrome

The description of each of the patients with PMPS is presented below (Table 3). No significant associations between the presence of PMPS and predictive risk factors (Table 4) were found, likely due to small numbers. ORs were unable to be reported for this outcome, similarly due to small numbers.

Any ongoing symptoms

Persistent symptoms of any type were reported by 66% of patients (39 patients). This included the three patients who met criteria for PMPS. Of these, 59% (23 patients) reported a severity of 3 or greater out of 10, and 25% reported severity of 6 or greater out of 10. Thirty-nine percent reported daily symptoms. Seventy-two percent reported neuropathic features such as shooting pain, electric shocks, numbness, tingling or burning. The remainder of patients described their persistent symptoms as either sharp, dull, aching or cramping. Thirteen percent (five patients) reported these symptoms interfering with their daily life.

In patients with any ongoing symptoms, significant associations were seen in groups of younger age (OR for age >65 0.3, 95% CI 0.02–0.94, p=0.035), who underwent axillary surgery (OR 5.6, 95% CI 1.14–27.7, p=0.02), chemotherapy (OR 5.6, 95% CI 1.14–27.7, p=0.02) and adjuvant radiation therapy (OR not calculated, p=0.044) (Table 5). No significant association was observed either with pre-operative chronic pain or opioid use, or regional anaesthesia. Intraoperative local anaesthetic infiltration, and timing of this (pre- or post-incision) was inconsistently documented and not able to be included for analysis. Ethnicity data showed a trend towards an increased risk of ongoing symptoms in the non-European group compared to European as previously shown in the literature,²⁰ but this was not significant. When comparing Māori, NZ European and other groups, this remained non-significant. A non-statistically significant trend towards greater use of strong opioids in the groups with PMPS and persistent post-operative symptoms was observed.

Impact on daily life

Four patients continued to require analgesia for their symptoms. Paracetamol was the single analgesic used in most. One patient who fulfilled criteria for PMPS required multiple analgesic modalities, including paracetamol, non-steroidal anti-inflammatory, gabapentin, amitriptyline, acupuncture, deep tissue massage and review in a chronic pain management clinic.

Five patients stated that their persistent symptoms impacted their daily life, one of whom met the definition for PMPS. All five had spoken to a medical practitioner.

Analgesic requirement in the immediate post-operative period

Acute post-operative analgesic requirements were examined for all patients and included for analysis in the 59 study participants (Table 6). In 50% of patients only weak opioids or simple analgesia was required as an inpatient. Of those who did require strong opioids, total amounts used were generally modest (Table 7).

Discussion

This study is the first to describe the prevalence of persistent post-operative pain in mastectomy patients in Canterbury, New Zealand, and only the second to describe the same in a New Zealand population. It also describes the acute post-operative analgesic requirement of the same cohort and provides descriptive information to characterise risk factors for development of PMPS, assisting us to better inform our patients regarding their post-operative course.

The prevalence of PMPS in this study was only 5%. This is extraordinarily low compared to existing literature, including elsewhere in New Zealand, where it is estimated that 15–25% of patients experience moderate to severe persistent pain after breast cancer treatment.²⁰ There are several possible reasons for this. Our study was small and only detected three patients with PMPS, which decreases the precision of the calculated prevalence value. However, we believe the definition used for PMPS is robust, clinically significant and appropriate to use for comparison.¹⁸ This indicates that despite some reports of very high rates of persistent post-mastectomy pain, the actual prevalence of pain meeting consensus criteria for PMPS might be much lower.

We found that persistent symptoms of lower severity are common (reported in 61%), but unlikely to limit patients’ daily activities (14%). This emphasises the importance of counselling about the risk of post-operative pain, and continuing efforts to determine possible modifiable peri-operative factors.

Factors examined in this study associated with persistent pain symptoms include younger age (<65 years), axillary surgery, adjuvant radiation therapy and chemotherapy. These are consistent with existing literature, particularly younger age, which has frequently been identified as a risk factor.^4,7,9–11 Ethnicity data showed a non-significant trend towards increased risk of ongoing symptoms in non-Europeans, as previously shown,²⁰ but more descriptive hospital demographic ethnicity data would improve accuracy of ethnicity analysis in future studies to identify groups at highest risk. Local anaesthetic infiltration by the surgical team was inconsistently documented in this cohort but would be an important variable to capture reliably in a prospective study. A non-significant trend towards requiring strong opioids post-operatively was seen in those who developed persistent pain. Further larger studies may be useful to determine if this is a modifiable risk factor.

Strengths of our study include the use of acute pain data, which is not reliant on recall, but on documented inpatient analgesic requirements. While this is a surrogate measure of pain and may be influenced by patient comorbidities, analgesic preferences and nursing practices, it is a pragmatic marker of pain severity. The suggested trend towards higher inpatient analgesic requirement in those with persistent pain could at least allow early identification of these patients for provision of support or treatment and is also a potential modifiable risk factor. More research is required in this area to distinguish between association and causation.

Limitations of our study include its small numbers, retrospective observational nature and lack of a standardised definition for PMPS. Using the above criteria for PMPS, we only identified three positive cases, but a larger group of patients with less severe ongoing symptoms. This confirms the challenges in accurately defining this condition, which is not described well by general definitions of chronic pain. Our patient survey was designed to be descriptive and to identify cases of PMPS based on a specific definition.

Conclusion

Acute post-operative pain after mastectomy in Canterbury appears well controlled with routine peri-operative care, with modest analgesic requirements in the majority of patients. Sixty-six percent of patients had some form of ongoing pain symptoms at least 1 year after mastectomy with identifiable but not modifiable risk factors. Only 5% of patients met consensus criteria for the more severe PMPS. The authors recommend that future study in this area consistently distinguishes patients with persistent pain symptoms from PMPS. Both are clinically relevant but may have different antecedents and therapeutic options. This study provides local data to inform patients about their peri-operative journey and prompts clinicians to remain vigilant in identifying and managing pain symptoms long after surgical recovery.

View Appendices.