Prognostic Value of Cutaneous Disease Severity Estimates for Chronic Graft-Versus-Host Disease

Emily Baumrin, MD, University of Pennsylvania, Philadelphia, Pennsylvania shares research on the value of cutaneous disease severity estimates in predicting survival outcomes in patients with chronic graft-versus-host disease (GVHD).

Transcript:

Hi, everyone. My name is Dr Emily Baumrin, and I am a physician dermatologist at the University of Pennsylvania in the Department of Dermatology with a special interest in treating patients with skin complications of allogeneic hematopoietic cell transplantation, including graft-versus-host-disease. I will be presenting today the results of our study entitled “Prognostic Value of Cutaneous Disease Severity Estimates on Survival Outcomes in Patients with Chronic Graft-Versus-Host-Disease.”

The objective of this study was to identify whether the severity of skin involvement of chronic graft-versus-host-disease is a prognostic marker for poorer outcomes in patients with chronic graft-versus-host-disease. In summary, we found that erythema-type chronic graft-versus-host-disease, which includes all types of rashes that do not involve sclerosis, does predict mortality. We further found that grading severity of the skin involvement by body surface area had better prognostic value compared to the NIH skin score, which is the current gold standard measure used in clinical trials.

To give you a little bit of background, allogeneic hematopoietic cell transplantation is a curative procedure for many hematologic malignancies. However, chronic graft-versus-host-disease limits the success of this lifesaving intervention for up to 70% of transplant recipients. Chronic graft-versus-host-disease is the leading cause of long-term morbidity and non-relapse mortality after allogeneic hematopoietic cell transplantation, and our current therapies are ineffective. 

Chronic immunosuppression can lead to infections, adverse effects, and even loss of graft-versus-tumor effect. We therefore need to identify clinical indicators of poor outcomes so that we can better target prophylactic and preemptive treatments for patients with high-risk chronic graft-versus-host-disease.

The objective of this study was to identify whether early skin involvement could serve as one of these clinical indicators of poor outcomes. In order to answer this question, we use the chronic GVHD consortium, which is a longitudinal cohort of patients with chronic graft-versus-host-disease from 9 US academic medical centers. 

Patients were enrolled if they had chronic GVHD by NIH criteria and required systemic immunosuppression, therefore, were on the moderate to severe end of the chronic GvHD spectrum. We included patients if they had skin involvement with their chronic GvHD. We followed these patients over time who had severity by both NIH skin scoring and body surface area, and we followed them for outcomes including overall survival, non-relapse mortality, and failure-free survival, which is a composite outcome including relapse mortality and a failure event including switch of immunosuppression for chronic GVHD.

In total, we had 267 patients with cutaneous chronic graft-versus-host-disease in the cohort. We first sought to describe the natural history of skin graft-versus-host-disease to determine whether erythema-type disease—again, all rashes that do not have sclerosis—or sclerosis including superficial sclerosis all the way down to deep sclerosis and fasciitis, behaved similarly or different over time. First we found that erythema-type disease occurs much earlier after transplant compared to sclerosis. On average, sclerotic disease occurred twice as far from transplant as erythema-type chronic-graft-versus-host-disease. We also found that most patients who have erythema-type disease do not progress to sclerosis, and patients who present with sclerosis, the majority do not progress to erythema. We therefore concluded that these skin disease subtypes behave quite differently and are not necessarily a natural progression from one to the other, which is a prior thought in the field.

We also found that erythema-type disease was much more responsive to treatment by plotting their NIH skin score and body surface area involved over time. Erythema-type based patients on systemic immunosuppression had a great response, whereas sclerosis-type patients did not have an adequate response by NIH skin scoring and body surface area. From our descriptive statistics describing the natural history of chronic skin GVHD, we found that erythema and sclerotic-type disease are quite different in how they behave.

Building upon that, we looked at survival associations of erythema and sclerosis. Sclerotic-type disease was actually not associated with overall survival, non-relapse mortality or failure-free survival after adjusting for confounders associated with mortality. This is an interesting finding because sclerosis carries significant morbidity over time. Erythema-type disease, however, was associated with mortality, overall survival, non-relapse mortality and failure-free survival in multi-variated adjusted models. 

We took this finding one step further to identify which clinical severity assessment had the highest prognostic value in predicting mortality for our erythema-type patients. In summary, we found that body surface area involvement of the erythema had the highest prognostic value in predicting mortality. This outperformed the NIH skin score, which is the current gold standard assessment.

Moving on to the implications of this study, and for next steps, we found that erythema and sclerotic-type chronic GVHD are quite different in their natural histories and in their prognostic abilities in predicting poor outcomes. Erythema-type disease is a clinical indicator of increased risk of death, while sclerotic-type chronic GVHD is not. 

We found that within erythema-type disease, body surface area was the best way of measuring clinical severity to predict poor outcomes. This would require a shift in the current standard, in that we do not currently recommend a detailed assessment of body surface area, in that the NIH skin score is just a categorical assessment in 3 large buckets. We would therefore propose that patients presenting with erythema-type chronic graft-versus-host-disease should have a detailed body surface area assessment, and those with higher body surface area involved have higher risk of poor outcomes including death.

The last thing to note of our study is that the current assessment tools, including the body surface area involvement and the NIH skin score, are not able to identify patients with sclerosis at high risk for poor outcomes including death. This highlights the need for better measurement tools for this type of chronic cutaneous graft-versus-host-disease.

In summary, patients with skin involvement, and particularly erythema, are at risk for poor outcomes. These patients might benefit from early aggressive treatment or increased targeted preemptive and prophylactic strategies. Future studies include employing these strategies in a prospective randomized control matter. Thank you very much.

Source:

Baumrin E, Baker LX, Byrne M, et al. Prognostic value of cutaneous disease severity estimates on survival outcomes in patients with chronic graft-vs-host disease. JAMA Dermatol. Published online March 8, 2023. doi:10.1001/jamadermatol.2022.6624