CORRESPONDENCE

 

Comment on 'Paediatric nephroblastoma at a South African tertiary hospital: A 21-year retrospective analysis'

 

 

To the Editor: I read the article 'Paediatric nephroblastoma at a South African tertiary hospital: A 21-year retrospective analysis'^[1] with interest.

The first striking welcome feature was that the first/corresponding author involved many medical students in publishing this article.

Allow me to make certain queries of the corresponding author.

In the abstract, the authors mention favourable preoperative histology, but do not mention the preoperative use of cutting needle biopsy in the article. The histopathological report and tumour staging are available only postoperatively. Surprisingly, the number of patients undergoing surgery was 180 (Table 1), but histopathology has been mentioned for 207 patients (Table 3). How do the authors explain this discrepancy of 27 patients? Did these patients undergo a preoperative cutting needle biopsy? I think the authors should mention n=180 as the legend of Table 3, and include only the patients who underwent surgery. The numbers of the high, intermediate and low-risk tumours in Table 3 still add up to 179, 1 less than the total number of patients operated on in the study.

Were there any syndromic patients in the study, or did any patient have a single kidney or horseshoe kidney?

There is no mention of lactate dehydrogenase (LDH) levels in the diagnostics of Wilms' tumour (WT) in the world literature^[2] or even Société Internationale d'Oncologie Pédiatrique [International Society of Paediatric Oncology] Paediatric Oncology in Developing Countries (SIOP PODC) protocols to discover the tumour burden. Why was it done in this study?

What do the authors mean when they state, 'negative isotope studies demonstrated skeletal involvement in 143 (69.1%) patients, liver involvement in 128 (61.8%) and clear bone marrow aspirates in 189 (91.3%)'? Isotope studies are not recommended for WT in any of the collaborative group protocols, the only exception being performing pre- and postoperative renal dynamic scans in patients with WT, where the surgeon is contemplating performing nephron-sparing surgery.^[3] This helps to learn the pre- and postoperative function of the salvageable kidney. Similarly, bone marrow aspiration is not an integral part of the diagnostics in WT, so why was it performed? In any case, the yield of this investigation has been very poor in this study and, therefore, it is not cost-effective against the background of poor healthcare budgets available in the region.

Referring to Table 1, the six patients with inferior vena cava (IVC) are not to be counted as having metastases. IVC invasion, even if the tumour thrombus reaches the right atrium, is considered a local spread and not metastasis. How many of the patients had multiple-site metastases? How many patients had non-pulmonary, non-hepatic metastases?

Again, referring to Table 1: surprisingly, a diagnostic delay of >2 weeks had a significant impact on survival in the study. A fortnight's delay is common even in high-income countries. Maybe the authors should use receiver operating characteristic (ROC) curves to decide the cut-off delay that results in lower survival.

Again, referring to Table 1: the staging is mentioned for 198 patients, when the number of patients who have been operated on is 180. How do you explain this discrepancy? Are the patients who are known to have stage IV and stage V disease preoperatively included there? With preoperative imaging, one can only know whether the patient has a localised or metastatic disease, and whether the disease is unilateral or bilateral.

Again, referring to Table 1: what was the indication for radiotherapy in 37 patients? The conclusion section does mention that it was administered to advanced cases. The term 'advanced' is not defined in the 'methods' section. What was the breakup of these patients in terms of the region irradiated? How many had flank radiation, whole abdominal radiation and/or pulmonary radiation? It is important to understand that not all stage IV patients need to have flank radiation in cases where the local stage is I or II. Thus, every patient with metastases has an overall stage IV, but the local stage could be I/II or stage III. Only patients with overall stage IV and local stage III would need flank radiation.

Again, referring to Table 1: how did you define a good response to neo-adjuvant chemotherapy (NACT)? If it was as per the reduction of tumour volume on imaging, it is not mentioned in the methods section. In our recently published study, only half of the patients had good responses to NACT; the good response was defined as >40% reduction in tumour volume.^[3]

Could you please provide data on 'fast complete responders' and 'late slow responders' from the cohort of patients who had pulmonary metastases alone?

Again, referring to Table 1: did the 18 patients who had a relapse during treatment complete the treatment, or abandon the treatment after the relapse? What was done for 18/132 patients who relapsed after completing the treatment? Were they administered second-line chemotherapy/radiotherapy?

Referring to Table 2, could the authors provide the median (interquartile range) for age for unilateral and bilateral tumours separately?

The median delay of symptoms-to-diagnosis time has been mentioned in both abstract and Table 2 as 30 days. Why did the authors then choose to apply statistics to >2 weeks' delay in Table 1?

Does the survival time in Table 2 refer to survival after the completion of treatment?

Referring to the 'outcome' subsection of the results section, what do the authors mean by the median survival rate throughout the study? The overall survivals (OSs) and event-free survivals (EFSs) are mentioned as available after 2 years, 4 years or 5 years.^[1] In the discussion section, 2-year OS has been mentioned as 59.4%. But this is a 21-year study, so the authors could easily mention 5-year and 10-year OSs.

The 'outcome' subsection also mentions that 20 (9.7%) patients defaulted on their treatment and 26 (12.6%) were lost to follow-up, although no significant subsequent impact on survival was observed. The p-value mentioned in Table 1 is 0.019, so it shows a significant impact on survival.

What were the 'other areas' in nine patients where relapse happened?

In the subsection 'factors associated with outcome', it is mentioned that the patients with lung metastasis had an 85% higher hazard of mortality (hazard ratio 1.85; 95% confidence interval 1.15 - 2.95; p=0.010) compared with metastasis to different locations. This seems unusual, and is opposite to the global experience.^[4]

In Fig. 1, the authors have depicted an adjusted Kaplan-Meier survival curve for relapse in paediatric patients with WT for different risk stratifications for 90 months (7.5 years). It would have been better if the authors had shown Kaplan-Meier OS and EFS curves for 20 years for unilateral v. bilateral WT, localised v. metastatic WT and for different stages of WT.

Last but not least, may I request the corresponding author to provide raw data. The confidentiality and privacy of the data would be honoured.

Y K Sarin

Department of Pediatric Surgery, Lady Hardinge Medical College, and associated Kalawati Saran Children's Hospital, New Delhi, India yksarin@gmail.com

 

References

1. Brits E, Gerber E, Iroka I, et al. Paediatric nephroblastoma at a South African tertiary hospital: A 21-year retrospective analysis. S Afr Med J 2025;114 (12):56-64. https://doi.org/10.7196/SAMJ.2024.v114i12.2223

2. Vig A, Rathod, KJ. Laboratory workup. Chapter 9. In: Sarin YK (ed). Wilms' Tumor. 1st ed. Singapore: Springer, 2022:79-81. https://doi.org/10.1007/978-981-19-3428-5_9

3. Sarin YK, Losu PU, Nangia A. Wilms' tumor - an audit. J Indian Assoc Pediatr Surg 2024;29(6):600-607. https://doi.org/10.4103/jiaps.jiaps_87_24

4. Joshi M, Singh UB. Prognosis and outcomes. Chapter 38. In: Sarin, YK (ed). Wilms' Tumor. 1st ed. Singapore: Springer, 2022:343-53. https://doi.org/10.1007/978-981-19-3428-5_38