Abstract
Background/Aim: The latest developments in oncological therapies for malignant melanoma, and the discovery that complete lymph node dissection offers no survival benefit, are changing the landscape of melanoma surgery. There is a need for more information on health-related quality of life (HRQoL) consequences of melanoma surgery. Patients and Methods: This longitudinal cohort study was carried out from 2004 to 2009 in the Helsinki and Uusimaa Hospital District and patients were followed-up at 6, 12 and 24 months. The patients were asked to fill in the generic 15D questionnaire and the cancer-specific European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-30). In addition, they were asked selected questions from the EORTC Item Library regarding upper and lower limb edema. Results: A total of 169 (64.5%) patients with local or locally advanced melanoma referred for surgical treatment responded, of whom 161 were included in the final analysis. For the whole patient group, distress, depression and emotional function improved over time. Worse HRQoL in some of the dimensions were associated with female sex, skin transplant versus direct wound closure and complications 30 days or more after surgery, but none was associated with worse overall HRQoL. Postoperative complications, type of wound closure or lymph node surgery had no effect on overall HRQoL. Patient-reported limb edema was associated with worse overall HRQoL at baseline and during follow-up by both instruments. Patients reporting limb edema reported worse mobility and more pain throughout the study. Conclusion: Patient-reported limb edema, regardless of the cause, seems to be an important predictor of worse HRQoL among patients with melanoma.
- Cutaneous malignant melanoma
- health-related quality of life
- limb edema
- limb lymphedema
- extremity lymphedema
- cancer surgery
In recent years, there has been a growing interest in health-related quality of life (HRQoL) as an outcome measure in studies assessing the effectiveness of cancer surgery. The discovery that complete lymph node dissection (CLND) offers no survival benefit in comparison to nodal observation (1, 2), and the development of effective oncological treatments are changing the landscape of melanoma surgery, since surgery is no longer the only effective treatment option for melanoma. There is therefore a need for information on the possible consequences of surgical procedures on HRQoL, for clinicians to choose the best treatment option for their patients, as well as for optimal resource allocation in healthcare. While the HRQoL implications of modern oncological treatments have been frequently studied as part of randomized controlled trials (3-5), little previous research exists with a focus on the HRQoL-related consequences of melanoma surgery.
This study assessed the HRQoL and morbidity of patients after surgery for melanoma with one generic and one disease-specific HRQoL instrument, the 15D and the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-30). The 15D is a generic, comprehensive (15-dimensional), self-administered instrument for measuring HRQoL among adults (age 16+ years). The EORTC QLQ-30 is a questionnaire developed to assess the quality of life of patients with cancer.
Patients and Methods
This longitudinal cohort study was carried out from 2004 to 2009 in the Helsinki and Uusimaa Hospital District, which encompasses nearly 30% of the Finnish Population. The baseline questionnaires were carried out between June 2004 and July 2007 and the patients were followed-up for 24 months. The study was approved by the Ethics Committee of the Helsinki University Hospital (approval number § 215/2004) and all patients signed an informed consent before participation. Permission for using selected items added to the EORTC QLQ-30 was obtained from the EORTC Item Bank (now Item Library).
The participants were recruited among the patients referred from primary or secondary healthcare to the Helsinki University Hospital Plastic Surgery Department outpatient clinic for surgical treatment of cutaneous melanoma. All patients with local or locally advanced melanoma referred for either excision with/without sentinel node biopsy (SNB) or CLND were eligible.
The patients were asked to fill in two HRQoL questionnaires: the 15D and the EORTC QLQ-30. In addition, the patients were asked two questions selected from the EORTC Item Bank regarding swelling of the arms or legs. They were asked in separate questions to rate whether during the previous week they had had upper limb edema: “Have you had a swollen arm or hand?” or lower limb edema: “Have you had swelling on one or both legs?” on a scale from 1-4 (1=“not at all”, 2 =“a little”, 3 =“quite a bit”, 4 =“very much”). For analysis, the answer 1=“not at all” to either question was then converted into one response indicating no limb edema and answers from 2-4 (a little, quite a bit, and very much), into a single response indicating limb edema.
The questionnaires were carried out at baseline, and at 6, 12, and 24 months after surgery. During follow-up, patients who simultaneously underwent surgical or oncological treatment for another malignancy were excluded from the study. Clinical data, such as information about surgical technique, oncological treatments, and postoperative complications were gathered retrospectively from the Helsinki and Uusimaa Hospital patient registry and analyzed together with the HRQoL data. Early complications were defined as complications that occurred within 30 days after surgery, while late complications were defined as any complication that occurred 30 days or more after surgery. The patients’ HRQoL according to sex, type of excision wound closure, type of lymph node surgery performed, reported limb edema, presence of metastatic disease, and use of oncological treatments (radiation therapy or medications) were analyzed to detect possible differences in HRQoL between groups.
The statistical analysis was carried out using SPSS version 26 (IBM, Armonk, NY, USA). For statistical analysis, the significance level used was 95%. Statistical differences between groups were calculated using the Mann–Whitney U-test or one-way analysis of variance with Bonferroni correction, as appropriate.
Results
A total of 262 patients were approached and 169 agreed to participate, yielding a response rate of 64.5%, out of which 168 were found to be eligible. During follow-up, seven patients received active surgical or oncological treatment for another malignancy and were therefore excluded from the final analysis. The final analysis included 161 patients; 160 patients answered the baseline EORTC QLQ-30 questionnaire and 159 answered the baseline 15D questionnaire. At 6, 12 and 24 months, the respective response numbers were 149, 143, and 126 patients for the EORTC QLQ-30 questionnaire and 150, 138, and 123 patients for the 15D questionnaire. The losses to follow-up at 6, 12 and 24 months were 7.5%, 11.2%, and 21.7% for EORTC QLQ-30, and 6.8%, 14.3%, and 23.6% for 15D. Between 6 and 12 months, one patient died of melanoma, and between 12 and 24 months, 10 patients died, one of liver cirrhosis and nine of melanoma.
The patient characteristics are presented in Table I. Patients with stage IV disease (1.2%) had either local lymph node metastases or subcutaneous metastases with no evidence of distant metastasis at baseline and were, therefore, treated surgically with local excision or CLND. Ten patients had clinical regional lymph node metastasis or in-transit metastasis at baseline. At 6, 12 and 24 months, the number of patients with distant metastasis were 12, 14, and 12, respectively.
Study population characteristics.
A total of 156 (96.9%) patients were treated with local excision of the primary tumor or skin/subcutaneous metastases, followed by a possible SNB and/or CLND. For the 156 patients treated with local excision, most of the wounds were closed directly. The other surgical techniques used were skin grafts, a local flap, or amputation. Three (1.9%) of the patients presented with lymph node metastasis after a previously excised localized melanoma and were treated with CLND only.
The most frequently used lymph node procedure was SNB (145 patients, 90.1%). Out of the 145 patients, 27 (18.6%) underwent further CLND because of metastasis in the sentinel nodes. This corresponds to 16.8% of the whole study population. Six patients (3.7%) had no lymph node procedure, and ten patients underwent direct CLND because of regional lymph node metastasis. The most frequent procedure was axillary lymphadenectomy, followed by groin/inguinal lymphadenectomy, and neck dissection. Two patients (1.2%) had both axillary and groin CLND. Three patients (1.9%) had already had CLND previously, before the baseline questionnaire, of them two for a previously treated melanoma and one for breast cancer.
Surgical morbidity: SNB, CLND, and limb edema. At baseline, 15 (9.6%) respondents reported upper limb edema. One of these patients had a history of axillary CLND from a previous primary melanoma operation, while others who reported limb edema at baseline had no history of lymph node surgery. Thirty-four (21.3%) patients reported lower limb edema at baseline, of whom none had a history of previous lymph node surgery. Nine patients without previous lymph node surgery reported both lower and upper limb edema. When those with baseline limb edema and those who had had a lymph node procedure before baseline were excluded, 16/29 (55.2%) of respondents who had undergone an inguinal/groin lymph node procedure reported lower limb edema at 6 months, 10/29 (34.5%) at 12 months, and 6/13 (46.2%) at 24 months.
No statistically significant difference was detected in reported limb edema between patients who had SNB and those who had CLND at any of the time points. However, when comparing patients who had CLND to those who did no, those with CLND reported more limb edema at 6 months (55.9% vs. 37.5%, p=0.035).
Early complications. Forty-nine patients (30.4%) had an early complication. The complications and their frequencies are reported in Table II. Patients who had had a CLND directly or after a positive SNB, had statistically significantly (p=0.002) more early complications (51.4%) in comparison to those who had SNB only (24.6%). The most frequent complication was postoperative seroma, followed by wound dehiscence, flap necrosis or other wound healing issues, and surgical site infection. Patients who had had CLND had statistically significantly (p=0.006) more seromas (27.0%) than those who had had SNB only (9.3%). The reoperation rate due to early complications was 6.2%.
Early (≤30 days) and late (>30 days) complications experienced by patients after surgery for melanoma (n=161).
Late complications. Late complications occurred among 23 (14.3%) of the patients and are presented in Table II. Prolonged seroma formation occurred among three (3) patients (1.9%) after SNB, followed by CLND. There were no statistically significant differences in late complications between patients who had CLND in comparison to those who had SNB only. The reoperation rate due to late complications was 0.6%.
Health-related quality of life. For the whole patient cohort, the overall 15D score did not differ statistically between any of the time points nor did the differences reach the minimum important change threshold of ±0.015 (6). In comparison to baseline, at 6, 12, and 24 months, there was a statistically significant improvement in distress (0.020 p=0.001, 0.023, p=0.001 and 0.033, p<0.001, respectively) and, in comparison to baseline, there was also an improvement in depression at 6 months (0.015, p=0.002) and at 24 months (0.011 p=0.008).
For EORTQ-QLQ-30, there was no statistically significant change in global health status/QoL over time. The only statistically significant differences were detected in the emotional functioning dimension, which improved over time. In comparison to baseline, emotional functioning improved from 77.41 to 86.51 at 6 months, to 86.93 at 12 months, and to 87.86 at 24 months (all p<0.001).
For overall 15D, women scored worse at baseline (−0.034, p=0.009). During follow-up, women also scored worse on overall 15D score but the differences were not statistically significant. Between the sexes, women had a statistically significantly worse score in mobility at baseline (−0.077, p=0.002), 6 months (−0.070, p=0.006), 12 months (−0.059, p=0.015), and 24 months (−0.096, p<0.001). In the breathing dimension, women also had a statistically significantly worse score at baseline (−0.057, p=0.020), 6 months (−0.059, p=0.030), and 24 months (−0.058, p=0.045).
No statistically significant differences were found in 15D or EORTC QLQ-30 Global Health/Qol scores between patients with and without metastatic disease at any of the follow-up points. Likewise, there were no differences in overall HRQoL for either of the QoL instruments between patients who received radiation therapy during follow-up in comparison to those who did not, nor in those who received other oncological treatments.
Surgical technique and CLND. For 15D, no statistically significant differences in HRQoL related to different wound-closure techniques were detected in the overall score or in any of the dimensions at baseline or during follow-up. For EORTC-QLQ30, the HRQoL scores were statistically significantly better for direct wound closure versus skin transplant closure in role functioning at 6 months (90.94 vs. 77.50, p=0.019) and 12 months (87.73 vs. 76.98, p=0.049), social functioning at 6 months (93.79 vs. 85.00 p=0.023), and in emotional functioning at 24 months (89.66 vs. 78.33), p=0.029), with worse scores associated with wound closure by skin transplant.
There were no statistically significant differences at any of the timepoints in overall 15D scores between patients who had not undergone a lymph node procedure or SNB only versus patients who had CLND. However, the change in the 15D score between baseline and each time point reached the threshold of 0.015 and was also statistically significant during follow-up. Over time, the 15D score continued to deteriorate in the CLND group in comparison to baseline, while the no CLND group’s 15D scores improved statistically significantly between baseline and follow-up. The change in the CLND group between baseline and each of the different time points (6, 12, and 24 months) were – 0.022 (p<0.01), −0.021 (p<0.05), and −0.022 (p<0.05), respectively.
For EORTC QLQ-30, no statistically significant differences were detected between patients according to CLND status at baseline. In the follow-up, the only statistically significant differences between patients without and with CLND were in appetite loss at 6 months (7.07 vs. 1.71, p<0.05), and at 12 months (16.13 vs. 3.15, p<0.001), and in social function (87.68 vs. 94.95, p<0.01) and financial problems (10.15 vs. 3.78, p<0.05) at 24 months; the patients with CLND had a worse score/more symptoms for each of these dimensions.
For early and late surgical complications, no statistically significant differences were detected for overall HRQoL using either of the instruments. For EORTC QLQ-30, the only statistically significant difference between baseline and follow-up was in financial problems, where those with a late complication had more financial problems than those without a late complication (1.28 vs. 8.54, respectively, p=0.007). For 15D, in comparison to baseline, at 6 months, statistically significant deterioration was found for hearing (−0.018, p=0.011) and breathing (−0.038 p=0.028) and for usual activities (−0.018, p=0.017), whilst there was an improvement in the distress dimension (0.057, p=0.045) at 12 months.
HRQoL and reported limb edema. 15D results: Patients who reported limb edema, had a statistically significantly worse overall 15D score at all time points. At baseline, those who reported limb edema had a statistically significantly worse score in mobility, eating, and distress. At 6 months, those with edema had a significantly worse overall 15D score (−0.05, p=0.001) and a significantly worse score in eight of the dimensions (mobility, vision, breathing, eating, excretion, mental function, discomfort, vitality). Discomfort and mental function continued to score lower among those with limb edema at 12 and 24 months, mobility, breathing, and eating at 12 months, and excretion and vitality at 24 months. The results are presented in Figure 1.
15D Scores according to patient-reported limb edema at baseline (A), and at 6 (B), 12 (B) and 24 (B) months after surgery for melanoma. Significantly different at: *p<0.05, **p<0.01 and ***<p0.001; #above the minimum important change (6) threshold.
EORTC QLQ-30 results: At baseline, those reporting limb edema had a statistically significantly worse global health status, and also scored lower on physical and emotional functioning. In the symptom scales, there were no statistically significant differences between the patients. At 6 months, those reporting limb edema scored lower on global health status, physical and role functioning, as well as on several of the symptom scales with clinically important differences at several of the dimensions (7). At 12 and 24 months, limb edema patients continued to score lower on several of the functional scales and higher on the symptom scales. Those reporting limb edema continued to score higher on pain throughout the study (Figure 2, Figure 3, Figure 4 and Figure 5).
Scores from European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-30) functional scales and global health status (A) and symptom scales (B) at baseline according to patient-reported limb edema. Significantly different at. *p<0.05 and **p<0.01; #slightly clinically meaningful, ##moderately clinically meaningful.
Scores from European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-30) functional scales and global health status (A) and symptom scales (B) at 6 months according to patient-reported limb edema. Significantly different at: *p<0.05, **p<0.01 and ***p<0.001; #slightly clinically meaningful, ##moderately clinically meaningful.
Scores from European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-30) functional scales and global health status (A) and symptom scales (B) at 12 months according to patient-reported limb edema. Significantly different at. *p<0.05, **p<0.01 and ***p<0.001; #slightly clinically meaningful, ##moderately clinically meaningful.
Scores from European Organization for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ-30) functional scales and global health status (A) and symptom scales (B) at 24 months according to patient-reported limb edema. Significantly different at. *p<0.05, **p<0.01 and ***p<0.001; #slightly clinically meaningful, ##moderately clinically meaningful.
Discussion
In this longitudinal cohort study, we assessed the HRQoL of patients after surgical treatment of local or locally advanced cutaneous melanoma. Our aim was to detect possible differences in HRQoL over time and across different patient subgroups. Our main finding was that patient-reported limb edema was associated with a worse HRQoL measured by both instruments. On the other hand, no difference in HRQoL between different wound-closure techniques or type of lymph node surgery was found.
In the past, the majority of research on the consequences of melanoma surgery focused on oncological safety and morbidity (8-10). While some studies have applied HRQoL to measure the effectiveness of melanoma excision (11, 12), few studies have examined the HRQoL consequences of lymph node surgery among patients with melanoma (13, 14). Of the previous studies that assessed the HRQoL-related implications of melanoma surgery, the randomized controlled trial by Moncrieff et al. (15) compared 1 cm versus 2 cm excision margins for excision of primary cutaneous melanoma of Breslow thickness >1 mm. Even though the patients in the 2 cm margin group more often required reconstruction and had a slightly increased wound necrosis rate, the authors found no difference in QoL between the groups after a 12-month follow-up. Another prospective study examining the effects of surgical excision of melanoma on HRQoL among other skin cancers found that patients with melanoma had statistically significant impairment in their health state 1 month after surgery and that continued 1 year after surgery (16). The study did not report the type of surgery nor whether lymph node surgery was performed.
A randomized controlled trial from 2010 compared the HRQoL of patients treated with a narrow (2 cm) versus a wide (4 cm) excision margin at 3, 9, and 15 months (12). No difference in HRQoL was found between the two groups. In our study, instead of comparing resection margins, we compared wound-closure techniques, since the reconstruction technique was anticipated to have more impact on wound healing and scar appearance, and therefore on HRQoL, than the amount of skin tissue removed. Only a few differences in some of the EORTC QLQ-30 functional scales were detected, which suggests that the surgical technique used has a minor effect on HRQoL, as the previous studies (9-12) have also indicated.
Several studies, including the Multicenter Selective Lymphadenectomy Trial-2 (MSLT-2 study) (1) found that CLND was associated with a higher incidence of lymphedema as well as higher morbidity in general (17) in comparison to SNB only. Gjorup et al. in their cross-sectional analysis (14) found that patients with melanoma who had lymphedema after SNB or CLND had a significantly worse HRQoL measured by EORTC QLQ-30 and the QLQ-BR23 body image subscale than patients who did not develop lymphedema. Moreover, based on research on HRQoL after surgical treatment of other cancer type, there is evidence that limb lymphedema after lymph node dissection has a negative effect on HRQoL (18-20).
Despite finding a statistically significant difference in patient-reported limb edema at 6 months between patients who did not undergo CLND compared to those who did, we did not detect any major differences in HRQoL between patients who had CLND versus those who did not at any time point using either of the HRQoL instruments. We did however find an association between patient-reported limb edema and worse HRQoL at all time points. In this study, this effect was detected with both the 15D and the EORTC-QLQ-30. No statistically significant difference between limb edema between patients who had SNB only and those who had CLND was detected at any of the time points. This could be explained by the size of our study sample or patients under- or over-reporting limb edema, which is one of the main limitations of our study. Due to limb edema being assessed by the patients themselves and not by a medical specialist, the severity, type, and cause of edema were not confirmed and might therefore be unrelated to melanoma surgery. Another reason may be that the number of SNs harvested may approach that of CLND (21). However, according to our findings, patient-reported limb edema, regardless of the cause, seems to be an important predictor of worse HRQoL among patients with melanoma.
In a previous retrospective study using the 15D instrument, we found that in the long term, the HRQoL of melanoma survivors compared well with that of the general population (22). In this study, the use of two instruments allowed us to detect possible differences in the use of these two instruments and to assess their applicability in the context of melanoma surgery. Both instruments have their advantages, as 15D allows for a comparison with the general population, while the EORTC QLQ-30, with a larger number of items and cancer-specific questions, allows for a comparison between different cancer types. In this study, both instruments gave similar results, which adds to the validity of our findings. While most of the dimensions that showed a statistically significant difference between groups had a logical association with melanoma treatments and symptoms, some of them, e.g. in vision, seem to be unrelated to cutaneous melanoma. In conclusion, while both instruments seem to be applicable for assessing the HRQoL of patients with melanoma, the EORTC QLQ-30 is perhaps better suited to assessing melanoma patients’ symptoms when a comparison to the general population is not needed.
One of the other limitations of our study was that although our questionnaires were collected prospectively, the data regarding surgical margins, technique and morbidity was gathered retrospectively from the patient registry. Information on clinically diagnosed lymphedema was largely lacking. The clinical data are therefore also subject to reporting bias. Another limitation is that our study sample was gathered in 2004-2009 and is thus relatively old. Our main aim, however, was to assess the effects of melanoma surgery on HRQoL and morbidity. Since resection margins and surgical technique have stayed the same to date and most of our patients did not receive oncological therapies, our results can be applied to today’s setting when assessing the HRQoL after surgery without the effect of modern oncological therapies as confounding factors. Among the limitations of this study is also the loss to follow-up, especially at 24 months, which subjects our study to selection bias. In addition, the effect of different clinical factors on HRQoL in this study was assessed by a one-way analysis. A future study, with further multivariate analysis, considering confounding factors, as well as data on clinically diagnosed lymphedema, will help to confirm our results.
The MSLT-2 and DeCOG-SLT studies demonstrated no survival benefit in CLND versus nodal observation (1, 2), and as a result, the use of CLND has diminished significantly in recent years (23). Our findings support the notion that CLND should be aimed at patients who are likely to benefit from it, and the possible lymphedema resulting from either SNB or CLND should be treated appropriately, as limb edema not only causes morbidity but seems to translate into worse HRQoL. Finally, melanoma-specific HRQoL instruments might benefit from adding specific questions regarding swelling of the extremities.
Acknowledgements
The Authors would like to thank Marja Nurmi and Anu Roine for their technical assistance in the data collection.
Footnotes
Authors’ Contributions
Tiina Jahkola, Pirjo Räsänen, Minna-Liisa Luoma and Risto P. Roine designed the study. Pia Mylläri, Tiina Jahkola, Pirjo Räsänen and Risto P. Roine collected the data. Pia Heino wrote this article and analyzed the data with the assistance of Risto P. Roine.
Conflicts of Interest
The Authors report there are no competing interests to declare.
- Received August 3, 2022.
- Revision received August 31, 2022.
- Accepted September 9, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.












