Melanoma costs: A dynamic model comparing estimated overall costs of various clinical stages
University of California, San Diego. firstname.lastname@example.org
BACKGROUND: The rapidly increasing incidence of melanoma occurs at the same time as an increase in general healthcare costs, particularly the expenses associated with cancer care. Previous cost estimates in melanoma have not utilized a dynamic model considering the evolution of the disease and have not integrated the multiple costs associated with different aspects of medical interventions and patient-related factors. Futhermore, previous calculations have not been updated to reflect the modern tendencies in healthcare costs. METHODS: We designed a comprehensive model of expenses in melanoma that considers the dynamic costs generated by the natural progression of the disease, which produces costs associated with treatment, surveillance, loss of income, and terminal care. The complete range of initial clinical (TNM) stages of the disease and initial tumor stages were analyzed in this model and the total healthcare costs for the five years following melanoma presentation at each particular stage were calculated. RESULTS: We have observed dramatic incremental total costs associated with progressively higher initial stages of the disease, ranging from a total of $4,648.48 for in situ tumors to $159,808.17 for Stage IV melanoma. By stage, early lesions associate 30-55 percent of their costs for the treatment of the primary tumor, due to a low rate of recurrence (local, regional, or distant), which limits the need for additional interventions. For in situ melanoma, T1a, and T1b, surveillance is an important contributor to the medical costs, accounting for more than 25 percent of the total cost over 5 years. In contrast, late lesions incur a much larger proportion of their associated costs (up to 80-85%) from the diagnosis and treatment of metastatic disease because of the increased propensity of those lesions to disseminate. This cost increases with increasing tumor stage (from $2,442.17 for T1a to $6,678.00 for T4b). The most expensive items in the medical care of patients with melanoma consist of adjuvant treatment with IFN-alpha ($75,955.18), palliative care ($14,500), and administration of chemotherapy ($1,967.10 for a triple combination of agents); there are even higher costs for biochemotherapy, the new tyrosine kinase and antiangiogenic drugs, and hospital treatment of neutropenic fever ($1,535.00 to $1,800.00/day). CONCLUSIONS: There is a significant cost decrement when melanoma is diagnosed at an earlier stage, with a T4b lesion being approximately 2200 percent more expensive to diagnose and treat than an early in situ melanoma and 1000 percent more expensive than a stage T1a tumor. Although a direct comparison with other cancers would require the use of the same dynamic model, it is apparent that the high costs of melanoma care places it at the top of the most expensive cancers to diagnose, follow, and treat. These high costs for advanced-stage melanoma warrant an increased emphasis on developing effective strategies for its early diagnosis and treatment.
Melanoma incidence is increasing rapidly, with an annual percentage change (APC) of 2.8 percent. Early detection of melanoma is reflected in an increase of thin (<1 mm) lesions by an annual percentage change (APC) of 2.3 in males and 3.3 percent in females. As such, the incidence of melanoma increased over the course of a decade from 16.38/100,000 in 1995 to 21.25/100,000 in 2005 [1, 2]. However, due to the unique nature of the disease and problems surrounding diagnosis, treatment and documentation, the reported incidence of melanoma may be an underestimate.
The lifetime individual risk of developing melanoma is predicted to be 1 in 58 [1, 3]. Consequently in 2009, 68,720 new cases of melanoma are projected to arise and 8,650 deaths are expected to occur in the United States . In addition to these cases, it is estimated that 48,290 cases of in situ melanoma will be diagnosed; this is in addition to the 59,940 cases mentioned above.
This formidable and yet increasing burden of melanoma is likely to create a major economic impact on the public and private healthcare systems in the US. Our analysis was designed to demonstrate the economic consequences of detecting earlier melanoma lesions by showing the cost in differences between progressive tumor and TNM stages. As screening for melanoma may improve survival and lead to detection of earlier lesions, we sought to translate this stage migration produced by early detection into an economical value associated with the intervention, which may consequently produce an additional stimulus for the implementation of such programs. Although the improvement in patient survival is logically the most important endpoint of a medical intervention directed against an aggressive disease such as melanoma, the economic burden of any intervention needs to be carefully balanced against its costs for the society .
The economic burden of melanoma translates into costs of screening, diagnostic procedures, adjuvant, curative, palliative and end-of-life treatment, medical follow-up through physician visit, and non-invasive and invasive procedures. Equally significant are the costs associated with years of life lost and loss of productivity. The purpose of our analysis is best served through a model which is dynamic, in that is starts with a given lesion and summarizes the impact of projected natural outcomes and the standard responses of the medical system triggered by these evolutions. The contribution of false-positive procedures used in the diagnosis and follow-up of melanoma is also taken into consideration. However, the costs of instances in which multiple biopsies were taken to reach the diagnosis was not added, because this parameter cannot be quantified across different practices and institutional settings. Further, it is likely that the differences in expense associated with different clinical and pathological stages of melanoma will remain unaltered by adding a fixed initial cost of diagnosis to the leading variable (the overall cost derived from an increased biological aggressiveness of a more advanced primary lesion), which necessitates additional diagnostic, surveillance, and treatment expenses. Although no model can be overly inclusive and account for differences in practice, regional expenses, and other socioeconomic and cultural factors, our schema of costs defines the characteristic costs of each stage of melanoma; therefore, it highlights the economic impact of early detection of this disease. A previous estimate of the overall costs of melanoma treatment was made by Tsao et al. . This paper will update and expand previous estimates of the direct costs of treating malignant melanoma in a novel dynamic model, which includes multiple costs and takes into consideration the evolution of the disease that generates incremental costs.
Materials and methods
The costs incorporated in our model are those resulting from usual diagnostic procedures and administration of standard treatments for newly diagnosed malignant melanomas. Medicare reimbursements for the Mid Atlantic region were used according to the DSM classification. Other costs for chemotherapy, biological treatments, and clinical trials were averaged from those encountered in different medical settings in the Mid Atlantic area.
We have built a model of cumulative costs depending on the initial stage and subset stage migration due to disease progression. This is a dynamic model, assessing the newly diagnosed melanoma patients over a standard period of 5 years, when lymph node, in-transit, or distant metastases may most likely occur, mandating diagnostic procedures and administration of additional treatment.
In brief, our costs model comprises the first 5-years after the diagnosis of malignant melanoma. The main characteristics of this model are: (1) It is dynamic, accounting for future development of relapses, metastases, morbidity of cancer, and toxicity/complications of treatment. (2) It is based on 2008 Mid Atlantic Medicare costs (Table 1 and Table 2). (3) It contains the costs associated with melanoma in situ (Tis). (4) It includes the costs for diagnosis and surveillance (routine follow-up). (5) It accounts for false positive (FP) results of imaging tests. (6) It includes the costs for fatal and non-fatal recurrences. (7) It accounts for collateral costs such as loss of days of work and income. (8) It accounts for the terminal care of patients. (9) For the purpose of comparing costs stage-by-stage, all patients were assumed to have complied and completed the diagnostic and therapeutic procedures.
This model is not an epidemiologic one (e.g., looking at total costs for large patient populations), but accounts for costs for individual patients.
We have included the costs of palliative terminal care, for which estimates in the literature range widely between $27,000 , $14,500 , and $6,500 . We have considered an average cost of $14,500/patient. In addition, we have considered the loss of income, at an average rate of $16/hour. We have not included the cost for all nevi biopsied in order to find a melanoma. The average ratio of nevi biopsied to find one melanoma may be in the range of 30:1. However, this ratio varies, according to different sources and practice settings, between 5:1 and 100:1. We have also not included the psychological distress to patients, which is not a readily quantifiable parameter in monetary value.
The medical and patient-related costs in melanoma have been organized in a diagnosis, treatment, and follow-up schema (Fig. 1), which is in accordance to the standard of care in management of melanoma. The main clusters of melanoma-related tests and procedures are divided into: (1) Melanoma Diagnosis, (2) Local Treatment, (3) Lymphatic System Staging, (4) Initial Systemic Staging, (5) Local and Regional Follow-up, (6) Systemic Follow-up, (7) Late workup for recurrences (local, in-transit, nodal, and systemic recurrences). Where complete agreement is not present, we resorted to the most commonly employed practices in the care of melanoma. The non-standard items where general agreement has not been completely reached include the use of diagnostic imaging for staging of melanoma and the optimal schedule of follow-up. In these areas we have used an estimate of the most commonly utilized medical practices in a calculation that included university, hospital-based, and private office settings.
Common work frames, which influence the staging workup and management plans and, implicitly, the costs in melanoma care are:
1. The initial workup (Fig. 1, Table 3) and subsequent follow-up of melanoma (Table 4 and Table 5) are conducted according to the NCCN guidelines , other published guidelines [10, 11], and the common medical practices. Management of a primary cutaneous melanoma consists of a biopsy, surgical wide local excision of the lesion (WLE) (with primary closure or additional use of flaps or grafts depending on the location, size of the excision defect, and other local or systemic factors), pathology examination, dermatology and surgery initial consultation, and follow-up visits. Starting with Stage IIB, a medical oncology visit is required for evaluation in view of the consideration of adjuvant treatment with IFN-alpha or other experimental approaches and for coordination of an initial or follow-up work-up, which searches for systemic involvement. Occasionally, a medical oncology visit may be required for earlier stages in cases of suspicious findings or at the patient's request.
2. A wide local excision (WLE) is done for all melanoma cutaneous lesions. In addition, defect-covering procedures such as flaps and grafting are used in different proportions, as indicated for different melanoma clinical stages (Item 1 of Tables 6 - 15).
3. Lymphatic staging is performed routinely as a sentinel lymph node (SLN) sampling for cutaneous lesions of Breslow's depth >1 mm, or <1 mm if positive ulceration, Clark level IV or V, presence of regression, or a mitotic rate >1 per 10 high-powered fields. SLN is therefore indicated for Stage IB (T1b and T2a) and up. A WLND (whole lymph node dissection) is being done for CLN (clinically LN) positive and for patients with positive SLN biopsy. An initial surgical visit for evaluation, operating room appointments for SLN and WLND, and subsequent surgical follow-up visits are included.
4. Systemic work-up for melanoma consists of a physical exam, blood tests (CBC, LDH, ALP, LFTs), and imaging studies (CXR, CT head, chest, abdomen, pelvis), MRI, PET scans. As represented in Figure 1, the use of the blood chemistry analyses, LDH and ALP, starts at Stage IB, whereas the use of a complete blood count (CBC) starts at Stage IIIB. The indications for imaging studies are not consistent among references. The most commonly recommended imaging tests are CXR, starting at Stage IB. Utilization of Computed Tomography (CT), Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET) start at Stage IIIB, but earlier use is indicated in the presence of symptoms or other suspicious clinical signs . Further systemic work-up is indicated in case of suspected recurrence. Although different management options are available, a summary of the common "standard" diagnostic and therapeutic actions is summarized in Figure 1 for the local recurrences (LR), in-transit metastases, lymphatic (LN), and systemic recurrences.
5. The management of findings resulting from false-positive radiological studies has been incorporated in our schema of costs, because they result in significant expenses. The rates of false-positivity are shown in Table 3 and Appendix B. Expenses are calculated based on the cost of the confirmatory tests/interventions multiplied by the percentage of cases for which these second procedures are utilized.
6. Cost of treatment of Stage IV was computed by averaging the costs of chemotherapy monotherapy (DTIC, the least expensive, and temozolomide), combination chemotherapy (CVD), biochemotherapy (CVD plus Interferon-alpha plus Interleukin-2), and the mean cost of a clinical trial involving a combination of chemotherapy with biological modifying agents. The proportions of these treatments are those shown by a recent pool on treatment practices of US oncologists .
7. Costs for terminal care were obtained from the literature, with different amounts being estimated in various publications: $27,000 published in year 1998 , $6,500 published in year 2000 , $14,500 in 2008 .
"Pathway costs" by group interventions were defined as the sum of common interventions, which result from the decision to assign the patient to a certain stage/risk category. These pathway costs are useful in computing the total future actions resultant from disease relapse, complications of therapy, or the consequences of false-positive diagnostic tests. Examples of such pathways include all the actions pertaining to the diagnosis of the primary cutaneous lesion, lymphatic staging, administration of adjuvant treatment, the work-up and treatment of distant metastatic lesions, local recurrences, in-transit metastases, and follow-up visits. A summary of the costs for specifically defined pathways is provided in Table 2. At the top of pathway expenses are the treatment of a metastatic tumor ($116,294.18), adjuvant treatment with IFN-alpha administered for high-risk melanomas ($75,955.18), cost of palliative care ($14,500), and management of in-transit metastases ($8,775.20).
Clinical stages and risk of recurrence per clinical stage
Clinical stages have been classified according to the AJCC criteria  and the NCCN schema : Stage 0: TIS N0 M0, Stage 1A: T1a N0 M0, Stage IB: T1b N0 M0 and T2a N0 M0, Stage IIA: T2b N0 M0 and T3A N0 M0, Stage IIB: T3B N0 M0 and T4a N0 M0, Stage IIC: T4b N0 M0, Stage IIIA: T1-4a N1a or N2a M0, Stage IIIB: T1-4b N1a-2a M0 and T1-4a N1b-2b M0 and T1-4a/b N2c, Stage IIIC: T1-4b N1b-2b M0 and Any T N3 M0, Stage IV: Any T Any N M0 [NCCN 2008].
Reference work by Balch et al. was used to provide the percentage of patients in each tumor stage in each clinical stage .
The incidence of melanoma relapses (local, regional, and distant) for each clinical stage is shown in Appendix A. LN recurrence rate was calculated according to the incidence found by Rousseau et al. , whereas palpable LN recurrence was calculated using data published by Balch et al. . Because LN recurrence is responsible for the most recurrences, an increased incidence is observed through Stage IIC. The risk of LN recurrence is significantly different at each substage, varying from 2 percent in Stage IA, 9 percent for Stage IB, an up to 53 percent in Stage IIC (Appendix A).
The incidence of local, satellite/in-transit, and metastatic recurrences for Stage III (A-C) was calculated from a study of 1001 consecutive Stage III melanoma patients of Calabro et al. , and was averaged with the results of Fincher et al.  for Stage IIIA local recurrences. Interestingly, Zaludek et al.  showed a mean 5-year local recurrence rate of 6.8 percent for surgically removed in situ melanoma (TIS); this is higher than the 5-year local recurrence rate for Stages IA, IB, and IIA (0.45%, 1.48%, and 4.86%, respectively), as shown in Appendix A, and it is reflected in our calculations of the costs associated with tumor relapses. Once a positive LN involvement is diagnosed, the risk of recurrence was determined to be 14 percent after SLN, and 28 percent after total lymph-node dissection (TLND) . Overall for the totality of melanoma patients who undergo a sentinel lymph node procedure, a negative SLN is associated with metastatic recurrence in 6.8 percent, local recurrence in 2.5 percent, and lymph node recurrence in 0.6 percent . Overall a positive SLN is associated with metastatic recurrence in 10.5 percent and local recurrence in 15.8 percent of cases .
Tumor stages and risk of recurrence per tumor stage
Tumor stages have been classified in the present work according to the AJCC criteria  and the NCCN schema : Tis: melanoma in situ, T1a: <1 mm thickness and level II or III, no ulceration, T1b: <1 mm thickness and level IV or V with ulceration, T2a: 1.01-2.0 mm thickness no ulceration, T2b: 1.01-2.0 mm thickness with ulceration, T3a: 2.01-4.0 mm thickness no ulceration, T3b: 2.01-4.0 mm thickness with ulceration, T4a: >4 mm thickness no ulceration, T4b: >4 mm thickness with ulceration.
The incidence of local recurrences per tumor stage was calculated according to Urist et al.  and Schmid-Wendtner et al.  as 2.1 percent for T2, 6.4 percent for T3, and 13.2 percent for T4 lesions. Considering T1, as an example, the incidences obtained from the two sources was averaged, yielding an incidence of 0.45 percent which was used for this model. The incidence of LN recurrence was determined according to the tumor stage from a number of different sources [13-15, 21-24].
The incidence of satellite/in-transit metastases was calculated according to Kang et al. as 1.59 percent for T1, 3.82 percent for all T2, 9.02 percent for all T3, and 8.95 percent for all T4 lesions . Finally, the estimated incidence of metastases was found to be by Balch et al.  Eight percent for T2 (5-year estimated), 15 percent for T3 (5-year estimated), and 72 percent for T4 (5-year estimated). For T1, in order to increase the accuracy of the estimate, the 5-year recurrence rate was averaged from multiple references [15, 21] because of the high variability of the rate given by these sources (1.2% to 3%), yielding an incidence of 2.1 percent, which was used for this model.
Adjuvant radiation therapy use for patients with positive lymph nodes
Use of adjuvant radiation (RT) for macroscopic positive LN after LN dissection of melanoma increases the costs of medical care. In a study of radiation therapy (RT) given at John Wayne Cancer Center to 21/217 patients, RT was needed in 9.7 percent of cases after LN dissection for melanoma with macroscopic LN , and its utilization was included in our model for the cases of LN recurrences after primary surgical treatment. For microscopic LN involvement, a study of 23/40 patients with ≥4 LN being positive for melanoma found that RT was need in 0.7 percent of cases detected pathologically after LN dissection .
False positive (FP) radiological scans
A study by Tyler et al. showed the FP rate with Positron Emission Tomography (PET) to be 7.7 percent for diagnosing malignant melanoma . In 106 whole-body PET scans, 234 areas of increased activity were identified. Twenty-one areas of melanoma were missed. Of the 234 lesions identified, 39 were FP, meaning that they were not associated with a malignancy. Of the 39 FP, 18 were not otherwise attributable, resulting in a "true" FP rate of 7.7 percent.
The rate of false positivity for Computerized tomography (CT) was calculated by pooling the results of 3 studies, given the smaller sizes of these analyses. An average of approximately 14 percent FP was found for CT in staging asymptomatic patients with Stage III melanoma (22%, 8.4%, and 12%, respectively) [28-30]. The cost of FP radiological scans is detailed in Appendix B, based on the assumption that a false-positive CT is being followed with a PET, and in one-third of cases with a PET and a guided biopsy; a false positive PET is followed also in one third of cases with a CT scan and in one third with a biopsy; a combined PET/CT is followed in one third of cases with a guided biopsy, and one third of these cases eventually require an oncology consult.
Patient follow-up schedule
Follow-up surveillance tests are one of the areas of controversy in the care of melanoma. We have calculated the costs of a 5-year follow-up schedule, which considers both the academic and community common practices, according to the frequencies of care in both settings (Table 4). Our estimates (Table 5) reflect the mostly frequently made recommendations in the literature and the most common clinical practices, although we acknowledge great variations in the medical decisions in the follow-up of melanoma. The patterns of follow-up of academic versus private practitioners are inferred from the NCCN guidelines and the projected compliance in different practice settings.
Follow-up for melanoma of Breslow thickness 0-1 mm was recommended to consist of 7 visits over the course of 5 years, which would include physical exams , whereas all Stage I tumors over 0.4 mm were recommended by other authors to have twice a year clinical exams, CXR and abdominal U/S . Prior to 1984, patients with thin melanomas also received a CT of the chest and blood tests . A follow-up schedule developed in Europe includes 4 clinical and laboratory evaluations per year for 5 years, LN ultrasound, CXR, and abdominal U/S once/year . After 1984, melanomas > 1.5 mm were prescribed an abdominal CT at the first visit, and brain + chest CT at the second yearly visit . The cumulative frequency of follow-up visits for this diagnostic category was established for a total of 14 visits in 5 years . The highest frequency of follow-up visits pertains to Stage II melanomas over 2 mm, for which 16 visits have been advocated for the first 5 years of follow-up (4 visits during the first two years, 3 visits during the 3rd year, and 2 visits during years 4 and 5) . Stage II melanomas were assigned in Europe a follow-up schedule consisting of clinical and laboratory evaluations 4 times/year, LN ultrasound, CXR, and abdominal U/S 2 times/year. A recent quantification of the frequency of different procedures used for follow-up of melanoma demonstrates that a more advanced stage is associated with a higher use of resources (expressed as a procedure/patient ratio during 5 years of follow-up), such as the clinical exam, LN U/S, lab tests, and imaging studies (CT/MRI). The frequency for each of these procedures during the 5 year follow-up for melanoma was 18.8, 3.5, 13.2, 0.01 for Stage I; 14.1, 11.1, 19.3, 0.67 for Stage II; and 14.3, 12.1, 18.1, and 1.6 for Stage III, respectively . Although the CT, MRI, bone scans, CXR and abdominal U/S become more cost effective and of increased importance in Stage III disease, their use represents common practice even in earlier stages. Less specific guidelines for follow-up have been provided by NCCN, where optional CXRs and blood tests are stipulated every 3-12 months for Stage IB and up, whereas CT scans would need to be considered to screen Stage IIB and higher for recurrent/metastatic disease .
Master schema of costs
A master schema of costs was used o compute the total expenses associated with each particular tumor stage of melanoma. The individual components begin with the workup and treatment of the primary tumor, followed by systemic workup for detection of metastases, and end with the treatment of such events in case of occurrence:
1. Primary tumor treatment including (triple/double/single) medical (dermatology, oncology, surgical) visits (initial visit).
2. Initial systemic workup.
3. Workup and treatment costs for SLN including LN recurrence (systemic recurrence included in #8 as rate of distant metastases).
4. Workup and treatment for palpable lymph node including LN recurrence (systemic recurrence included in #8 in rate of distant metastases). Costs for radiation therapy (RT) after SLN procedure (needed in 0.7% of cases) and after TLND (needed in 9.7%) were included.
5. Workup and treatment for local recurrence (LR), which is similar to the one for performed for primary lesion, including medical visits.
6. Workup and treatment for in-transit metastases, which is also similar to the one performed for the primary tumor, including medical visits, in addition to 2 (median number) resections, and the performance of a SLN procedure.
7. Workup and treatment for metastatic disease
8. Five years of surveillance (include follow-up consults and tests, costs for FP radiological tests)
9. Patient time costs
10. Costs for terminal care
Assumptions used in our analysis
In addition to the estimated frequencies and costs for medical procedures and physician visits detailed above, we have relied on the following assumptions during the process of integrating all costs: (1) All patients receive the recommended treatment (since the purpose is to compare the cost of different stages, equal adherence is assumed across all strata). (2) No patients are lost for follow-up or decide to voluntarily interrupt the treatment. (3) The possibility of early deaths for other causes was not included in the model because it is expected to only occur in a minority of patients and be similar among different clinical or tumor stage groups. (4) A general population of patients and the availability of the routine "standard" diagnostic and therapeutic procedures is assumed. Relevant differences in local standards of care and patient cultural models may cause variations in the medical care and associated costs across geographical regions. Similarly, various areas may receive different Medicare payment schedules. Although Medicare costs were used as a baseline estimate of healthcare costs, variations in reimbursement across other insurers can reach significant differences. (5) Finally, we assumed that all the patients diagnosed with metastatic disease died within the follow-up period of 5 years, which represents a close approximation of the dismal actuarial 1 percent patient survival rate.
Assignment of costs reflecting Mid Atlantic Medicare reimbursements is presented in Table 1. Among the most expensive items are performing an MRI of the brain ($4,066.00), palliative radiation therapy ($15,999.000, hospitalization for neutropenic fever ($1,535.00-$1,800.00/day), and primary excision of a skin tumor (total costs $7,149.92). A combination of these individual items was made according to the Costs Per Grouped Interventions (Table 2), or "pathway costs", which, as detailed in the Materials and Methods section were defined as the common sum of interventions necessary when the patient is placed in a certain disease stage or risk group (as a resultant of the initial disease staging work-up or as a consequence of disease progression). The most expensive cumulative "pathway costs" are the treatment of a metastatic tumor ($116,294.18), adjuvant treatment with IFN-alpha administered for high-risk melanomas ($75,955.18), palliative care ($14,500), and management of in-transit metastases ($8,775.20).
Treatment of primary tumor per stage (Tables 6 - 15, Section 1) reflects the local treatments for the skin lesion only. The initial systemic workup per stage is also presented in Tables 6 - 15, Section 2. Both primary tumor and initial systemic workup costs manifest an increment with progression of the melanoma stage. For example, the treatment for a primary lesion, which includes the costs for surgical reconstruction in a certain percentage of patients, increases from $2,434.88 for In-situ melanoma, to $3,711.26 for a primary cutaneous tumor in Stage IIIC. Corresponding staging costs (systemic workup) also present a progressive increment across increasing clinical stages, from $0 for in situ melanoma up to $13,836.00 in Stage IIIC. Staging costs are comprised of radiological and various blood tests (Table 3).
Follow-up (surveillance) costs per clinical stages (Table 5), are based on current recommendations/guidelines from [9, 10, 11] (Table 4). For in situ melanoma, the only surveillance consists of dermatology office visits. For Stages IA-IIB, surveillance includes CMP, LDH, chest x-rays, oncology visits, and dermatology visits. An increase in surveillance costs from $3,759.00 to $50,566.44 is seen from in situ melanoma to clinical Stage IIIC. As shown in Figure 2, surveillance costs rise dramatically in Stages IIIA-IIIC; this is largely due to the addition of radiological scans (CT and PET) to the surveillance protocol. However, in Stage IV, there are no surveillance costs because the patient is receiving active treatment. For a detailed breakdown of surveillance costs in each clinical stage, please see Table 5; the total surveillance costs per tumor stage are detailed in Table 17.
Total costs by clinical stage (Tables 6 - 15)
Total cost for each clinical stage was calculated by adding up the estimates for each individual component of cost, as shown in Table 1. Rising costs observed across Stage II (IIA-IIC) are largely due to the differences in the incidence of metastatic disease among these substages: 12.49 percent for IIA, 30.37 percent for IIB, and 72 percent for IIC, respectively (Fig. 2). The cost for Stage IIIC is similar to cost of Stage IV melanoma, but the components of cost are different, as follows: Stage IV has higher costs for the workup and treatment of metastatic disease; 100 percent of patients are diagnosed with metastatic disease compared with 76.7 percent in Stage IIIC; Stage IIIC disease incurs surveillance costs ($55,043), whereas Stage IV does not; Stage IIIC disease is associated with high expenses largely because this stage requires additional tests to assess the systemic extent of the disease. Of note, a significant component of these costs results from the FP radiological scans. A breakdown distribution of costs per clinical stage demonstrated a significant decrease in the percentage contribution of the treatment of the primary tumor with the increase in clinical stage, mainly because other significant costs offset the balance in their favor (e.g., the workup and treatment of metastatic disease, which becomes a more likely occurrence as the disease manifests at a more advanced initial stage, and the costs associated with surveillance) (Fig. 3).
Stage III and IV patients, who account for 15.2 percent of all patients diagnosed with melanoma account for approximately half of melanoma costs (48%), equal to the expenses incurred by the rest of 84.5 percent of patients in Stages I and II (52% of costs) (Fig. 2).
Total costs by tumor stage (Table 17)
The total costs for each tumor stage result from the addition of the following categories: primary tumor; initial systemic workup; workup and treatment for SLN, palpable LN, local recurrence, in-transit metastases, metastatic disease; surveillance costs; patient time costs (Table 17, Fig. 4). Of note, considering in situ melanoma, T1a, and T1b, surveillance accounts for more than 25 percent of the total cost over 5 years. As shown in Figure 4, the absolute cost of each category increases with increasing stage, from $5,043.57 for in situ lesions to $110,149.88 for T4b melanomas. However, the workup and treatment for local recurrence is greater for in situ melanomas than for all other tumor stages except T4a and T4b. One important component that influences the total cost of melanoma is the cost attributed to the workup and treatment for metastatic disease, which increases with increasing tumor stage (from $2,442.17 for T1a to $6,678.00 for T4b).
Most expensive interventions
The most expensive items in the medical care of patients with melanoma consist of adjuvant treatment with IFN-alpha administered for high-risk melanomas ($75,955.18), cost of palliative care ($14,500), administration of chemotherapy ($1,967.10 for a triple combination of chemotherapy agents, and even higher costs for biochemotherapy, the new tyrosine kinase and antiangiogenic drugs, as well as other investigational clinical drugs), and hospital treatment of neutropenic fever ($1,535.00 to $1,800.00/day). By stage, early tumors are associated with 30-55 percent of their costs for the treatment of the primary lesion, due to a low rate of recurrence (local, regional, or distant), which limits the need for additional interventions and also decreases the need for surveillance compared to the more advanced stages. In contrast, late tumors incur a much larger proportion of their associated costs (up to 80-85%) from the diagnosis and treatment of metastatic disease which results from an increased propensity of those lesions to disseminate (Fig. 3).
Our work is intended to provide an accurate estimate of the cost differences between different tumor and TNM stages of melanoma, which reflects the economical impact of diagnosing the disease in an early vs. late stage.
This study updates and expands the economic landscape in melanoma, which was previously presented in 1998 in a study by Tsao et al., by incorporating the more recent data and technologies . Our analysis includes all components of cost, including the initial patient visits, costs related to lost patient time, and migration of melanoma to higher stages because of natural disease progression. It also includes components of cost excluded from previous models, such the one of Tsao et al. , such as the cost of surveillance for 5 years and costs of terminal care. As opposed to prior models , we employed a dynamic model, which starts with the patient presentation to the physician at a certain stage and includes the projected future progression to more advanced stages, which generates consequent costs. It further considers the costs associated with toxicities of treatment, hospitalization, and costs incurred by false positive results of radiological tests. Moreover, the cost of in situ melanoma is also included. Thus, the total costs covering diagnosis, 5-year follow-up, and workup/treatment of relapses (local, regional, and systemic) have been calculated based upon melanoma initial clinical stage at presentation.
Caveats of such a model include that the cost assumptions are based only on Medicare reimbursements. However, this represents a good approximate of the reimbursement provided by other insurers, and nevertheless they are likely to reflect proportionally across evolving melanoma stages, because a stage-by-stage comparison is in fact the purpose of this work. Because costs are based on those in a hospital setting, there is a facility fee included in many components of cost in this model, which may be variable between different institutions and between different treatment settings (primary vs. tertiary referral academic centers vs. private practices). Some difficult to quantify patient-related factors are also come into play. One example would be the fact that patients with poorer adjustment to their illness had greater healthcare expenditures , as shown by a Canadian study including patients with breast, lung, and prostate cancer. Another caveat results from the fact that all costs are derived for the Mid Atlantic Region only, whereas actual costs may vary from region to region. Although the relative costs for the care of melanoma are expected to vary in parallel with different stages, as a result of these regional differences, the absolute cost may nevertheless vary substantially.
Most increased costs resulted from the medical actions directed in response to finding lymph node metastases associated with more advanced T stage primary skin tumors, followed by the costs of the medical actions required by the development of distant metastases.
Our higher than previously estimated costs of melanoma have multiple roots. They are the result of: (1) a comprehensive consideration of all the costs after diagnosis, including expensive hospitalization and newer treatment strategies, complex medications, as well as complications of the baseline disease; (2) consideration of a follow-up plan, which includes expensive tests used in general practice even if current evidence points against the cost-effectiveness of their use; (3) inclusion of false-positive results and the economic consequences of the follow-up tests and procedures incurred to reach a definitive evaluation of the finding; (4) the effects of the medical inflation and new technologies and treatments; (5) consideration of a 5-year time interval for the evolution and treatment of the disease, which ensures homogeneity of the estimate and includes the vast majority (over 90%) of the events and expenses associated after the diagnosis of melanoma; (6) estimates are billed costs, and not actual reimbursement because the exact reimbursement cannot be known with certainty at a national scale; [A 60-65 percent return rate is a reasonable approximation. In addition, the percent reimbursements are probably similar across investigations, procedures, and treatments, although significant variations are possible. Because our goal was to show total costs per stage of melanoma at diagnosis, we believe that at a global scale the percent return for different melanoma stages is probably very similar.]; (7) the expenses that resulted from loss of time and productivity during the evaluation, medical procedures, and treatment of the complications that resulted iatrogenically were calculated and integrated. [These related to false-positive tests and the ones resultant from medical complications of the baseline cancer.]
An underestimate of costs would result from not considering all these factors. Along with the tremendous global increase in health costs over the past decade, these factors explain why, an estimate of the costs for melanoma in 1997/1998, for example, rendered much lower values of $40K-60K for Stage III and IV patients than our current calculations .
The costs of melanoma care appear to at least parallel, if not exceed the general escalation of the National Health Expenditures (NHE), which from $714.0B in 1990 escalated up to $2,105.5B in 2006, in the presence of a population increase of only 18 percent . A marked increase was also recorded in the US for the average chemotherapy payments. Thus, from 1995 to 2004, the average costs of treating cancer increased by 75 percent. In particular, tremendous increments in costs are associated with the use of new agents. For example, the costs of the new agents cetuximab and bevacizumab were estimated to represent approximately $161K for a one-year course of treatment . As more complicated diagnostic tests and therapeutic interventions are implemented, it is to be expected that they would involve more patient-time. Indeed, in 2005, the total patient-time cost associated with the initial phase of cancer care in patients aged 65 or older in the United States was estimated to be $2.3 billion .
The costs reported in the present article are largely predictable if the general rate of increase in medical costs is applied to costs reported in the previous years. Thus, the cost of an episode of surgical care increased over the past decade from $410.50 to $1694.70 . Similarly, single-agent chemotherapy costs are currently high (Temozolomide $7,440.00, and DTIC $3,900.00 for a 4 cycles course, which includes pharmacy costs, costs for the drug, and lost wages ). Even more dramatic are the costs of newer medications, including the antiangiogenesis agents and vaccine-based treatments, which make the cost of clinical trials to total between $75K to $200K/patient. A poll for treatments used by US oncologists for advanced/metastatic melanoma reveals that the most commonly employed treatments are clinical trials (anti-CTLA4 plus peptide vaccine) in 31 percent, single-agent chemotherapy or high-dose interleukin-2 each in 17 percent, DTIC plus sorafenib in 12 percent, DTIC plus oblimersen in 11 percent, and paclitaxel/carboplatin plus sorafenib in 8 percent of cases . Cost for IFN-alpha in 2007, consisting of the expenses related to the drug, administration, and associated costs, total $51,329.00, to which are further added the observation costs $8,280.00 . IFN-alpha is the most expensive item.
A comparison with costs calculated for other cancers allows us to conclude that melanoma is one of the most expensive neoplastic disorders to diagnose, treat, and follow-up. For example, the costs for breast cancer were found to total $91K for the primary tumor and $100.9K for a recurrence, as assessed using year 2007 fees. The total amount for 5-years of care at a rate of $46.5K/year totalize $232.5K . Similar results were reported in 2008 (using rates at the level of year 2004) using an all-inclusive rate for breast cancer expenses consisting of outpatient office, skilled nursing, pharmacy, inpatient, and ER-related fees of $4,421/month represent a total of $53.1K/year, or $265.2K for 5 years (Baron JJ 2008). In another breast cancer analysis, the costs for managing this disease were $39,000 in the first year after diagnosis and $48,900 after a local recurrence, followed by $13,000 annual costs thereafter. For metastatic breast cancer, the yearly costs were $46,500. Translated in 5-year costs, the total amount reaches $91,000 for primary tumors, $100,900 for recurrences and $232,000 for metastatic disease, with higher totals if the evolution from primary to metastatic disease is compounded .
Colorectal patients incurred excess costs of $33,500 in the initial phase, $4500/y in the continuing phase, and $14,500 in the terminal phase . The cost of treating early prostate cancer in the first 6 months after diagnosis in 1999 ranged from $13,569 for men treated with radical prostatectomy, $10,996 for men treated with external beam radiotherapy, and $12,699 for men treated with brachytherapy . The decrease in 6-month costs for prostate cancer treatments from 1991 to 1999 is thought to be due to increased PSA screening . Another study found a similar cost for treatment during the first 6 months after prostrate cancer diagnosis ($11,495 with a cumulative cost of $42,570 over 5.5 years) .
References1. http://seer.cancer.gov/, accessed October 2008.
2. Purdue MP, Freeman LE, Anderson WF, Tucker MA. Recent trends in incidence of cutaneous melanoma among US Caucasian young adults. J Invest Dermatol. 2008;128:2905-8 [PubMed]
3. Ries LAG, Melbert D, Krapcho M, Mariotto A, Miller BA, Feuer EJ, Clegg L, Horner MJ, Howlader N, Eisner MP, Reichman M, Edwards BK (eds). SEER Cancer Statistics Review, 1975-2004, National Cancer Institute. Bethesda, MD [UI:]
4. American Cancer Society. Cancer Reference Information. Skin Cancer-Melanoma 2009. www.cancer.org, accessed November 2009.
5. Porzsolt F, Kirner A, Kaplan RM. Predictors of successful cancer prevention programs. Recent Results Cancer Res. 2009;181:19-31 [PubMed]
6. Tsao H, Rogers GS, Sober AJ. An estimate of the annual direct cost of treating cutaneous melanoma. J Am Acad Dermatol. 1998;38:669-680 [PubMed]
7. Lang K, Lines LM, Lee DW, Korn JR, Earle CC, Menzin J. Lifetime and Treatment-Phase Costs Associated With Colorectal Cancer: Evidence from SEER-Medicare Data. Clin Gastroenterol Hepatol. 2009;7:198-204 [PubMed]
8. González-Larriba JL, Serrano S, Alvarez-Mon M, et al. Cost-effectiveness analysis of interferon as adjuvant therapy in high-risk melanoma patients in Spain. Eur J Cancer. 2000;36:2344-52 [PubMed]
9. www.nccn.com, accessed January 2009.
10. Romero JB, Stefanato CM, Kopf AW, Bart RS. Follow-up recommendations for patients with stage I malignant melanoma. J Dermatol Surg Oncol. 1994;20:175-8 [PubMed]
11. Govindan R. Melanoma, in DeVita, Hellman, and Rosenberg's Cancer: Principles and Practice of Oncology Review. Lippincott Williams & Wilkins 2005 ISBN 0-7817-5278-7.
12. Linette GP. Emerging Therapies for Melanoma. The American School of Oncology. www. asoncology.com, accessed July 2008.
13. Balch CM, Buzaid AC, Soong SJ, et al. Final version of the American Joint Committee on Cancer staging system for cutaneous melanoma. J Clin Oncol. 2001;19:3635-3648 [PubMed]
14. Rousseau DL Jr, Ross MI, Johnson MM, et al. Revised American Joint Committee on Cancer staging criteria accurately predict sentinel lymph node positivity in clinically node-negative melanoma patients. Ann Surg Oncol. 2003;10:569-574 [PubMed]
15. Balch CM. Surgical management of regional lymph nodes in cutaneous melanoma. J Am Acad Dermatol. 1980;3:511-524 [PubMed]
16. Calabro A, Singletary SE, Balch CM. Patterns of relapse in 1001 consecutive patients with melanoma nodal metastases. Arch Surg. 1989;124:1051-1055 [PubMed]
17. Fincher TR, McCarty TM, Fisher TL, et al. Patterns of recurrence after sentinel lymph node biopsy for cutaneous melanoma. Am J Surg. 2003;186:675-681 [PubMed]
18. Zalaudek I, Horn M, Richtig E, Hodl S, Kerl H, Smolle J. Local recurrence in melanoma in situ: influence of sex, age, site of involvement and therapeutic modalities. Br J Dermatol. 2003;148:703-708 [PubMed]
19. Pidhorecky I, Lee RJ, Proulx G, et al. Risk factors for nodal recurrence after lymphadenectomy for melanoma. Ann Surg Oncol. 2001;8:109-115 [PubMed]
20. Urist MM, Balch CM, Soong S, Shaw HM, Milton GW, Maddox WA. The influence of surgical margins and prognostic factors predicting the risk of local recurrence in 3445 patients with primary cutaneous melanoma. Cancer. 1985;55:1398-1402 [PubMed]
21. Schmid-Wendtner MH, Baumert J, Eberle J, Plewig G, Volkenandt M, Sander CA. Disease progression in patients with thin cutaneous melanomas (tumour thickness < or = 0.75 mm): clinical and epidemiological data from the Tumour Center Munich 1977-98. Br J Dermatol. 2003;149:788-793 [PubMed]
22. Lens MB, Dawes M, Newton-Bishop JA, Goodacre T. Tumour thickness as a predictor of occult lymph node metastases in patients with stage I and II melanoma undergoing sentinel lymph node biopsy. Br J Surg. 2002;89:1223-1227 [PubMed]
23. Wagner JD, Ranieri J, Evdokimow DZ, et al. Patterns of initial recurrence and prognosis after sentinel lymph node biopsy and selective lymphadenectomy for melanoma. Plast Reconstr Surg. 2003;112:486-497 [PubMed]
24. Kruper LL, Spitz FR, Czerniecki BJ, et al. Predicting sentinel node status in AJCC stage I/II primary cutaneous melanoma. Cancer. 2006;107:2436-2445 [PubMed]
25. Kang JC, Wanek LA, Essner R, Faries MB, Foshag LJ, Morton DL. Sentinel lymphadenectomy does not increase the incidence of in-transit metastases in primary melanoma. J Clin Oncol. 2005;23:4764-4770 [PubMed]
26. Shen P, Wanek LA, Morton DL. Is adjuvant radiotherapy necessary after positive lymph node dissection in head and neck melanomas? Ann Surg Oncol. 2000;7:554-559 [PubMed]
27. Tyler DS, Onaitis M, Kherani A, et al. Positron emission tomography scanning in malignant melanoma. Cancer. 2000;89:1019-1025 [PubMed]
28. Buzaid AC, Tinoco L, Ross MI, Legha SS, Benjamin RS. Role of computed tomography in the staging of patients with local-regional metastases of melanoma. J Clin Oncol. 1995;13:2104-2108 [PubMed]
29. Kuvshinoff BW, Kurtz C, Coit DG. Computed tomography in evaluation of patients with stage III melanoma. Ann Surg Oncol. 1997;4:252-8 [PubMed]
30. Johnson TM, Fader DJ, Chang AE, et al. Computed tomography in staging of patients with melanoma metastatic to the regional nodes. Ann Surg Oncol. 1997;4:396-402 [PubMed]
31. Bassères N, Grob JJ, Richard MA, et al. Cost-effectiveness of surveillance of stage I melanoma. A retrospective appraisal based on a 10-year experience in a dermatology department in France. Dermatology. 1995;191:199-203 [PubMed]
32. Hengge UR, Wallerand A, Stutzki A, Kockel N. Cost-effectiveness of reduced follow-up in malignant melanoma. J Dtsch Dermatol Ges. 2007;5:898-907 [PubMed]
33. Butler L, Downe-Wamboldt B, Melanson P, et al. Prevalence, correlates, and costs of patients with poor adjustment to mixed cancers. Cancer Nurs. 2006;29:9-16 [PubMed]
34. Crott R. Cost effectiveness and cost utility of adjuvant interferon alpha in cutaneous melanoma: a review. Pharmacoeconomics. 2004;22:569-80 [PubMed]
35. Catlin A, Cowan C, Hartman M, Heffler S; National Health Expenditure Accounts Team. National health spending in 2006: a year of change for prescription drugs. Health Aff (Millwood). 2008 Jan-Feb;27(1):14-29. [PubMed]
36. Warren JL, Yabroff KR, Meekins A, Topor M, Lamont EB, Brown ML. Evaluation of trends in the cost of initial cancer treatment. Natl Cancer Inst. 2008;100:888-97. [PubMed]
37. Yabroff KR, Davis WW, Lamont EB, et al. Patient time costs associated with cancer care. J Natl Cancer Inst. 2007;99:14-23 [PubMed]
38. Hernández Martín A, Núñez Reiz A, Sáiz Martínez M, Rovirosa i Juncosa J. Cost per episode of care in the surgical treatment of skin cancer. Gac Sanit. 2006;20:273-9 [PubMed]
39. Hillner BE, Agarwala S, Middleton MR.Post hoc economic analysis of temozolomide versus dacarbazine in the treatment of advanced metastatic melanoma. J Clin Oncol. 2000;18:1474-80 [PubMed]
40. Cormier JN, Xing Y, Ding M, Cantor SB, Salter KJ, Lee JE, Mansfield PF, Crott R. Cost effectiveness and cost utility of adjuvant interferon alpha in cutaneous melanoma: a review. Pharmacoeconomics. 2004;22:569-80 [PubMed]
41. Lidgren M, Wilking N, Jonsson B, Rehnberg C. Resource use and costs associated with different states of breast cancer. Int J Technol Assess Health Care. 2007;23:223-231 [PubMed]
42. Zeliadt SB, Etzioni R, Ramsey SD, Penson DF, Potosky AL. Trends in treatment costs for localized prostate cancer: the healthy screenee effect. Med Care. 2007;45:154-159 [PubMed]
43. Wilson LS, Tesoro R, Elkin EP, et al. Cumulative cost pattern comparison of prostate cancer treatments. Cancer. 2007;109:518-527. Erratum in: Cancer. 2007;109:2155 [PubMed]
© 2009 Dermatology Online Journal